National Institute on Drug Abuse
RESEARCH
MONOGRAPH SERIES
The Context of
HIV Risk Among
Drug Users and
Their Sexual
Partners
143
U.S. Department of Health and Human Services • Public Health Service • National Institutes of Health
The Context of HIV Risk Among
Drug Users and Their Sexual
Partners
Editors:
Robert J. Battjes, D.S.W.
Zili Sloboda, Sc.D.
William C. Grace, Ph.D.
NIDA Research Monograph 143
1994
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
Public Health Service
National Institutes of Health
National Institute on Drug Abuse
5600 Fishers Lane
Rockville, MD 20857
ACKNOWLEDGMENT
This monograph is based on the papers from a technical review on “The
Context of HIV Risk Among Drug Users and Their Sexual Partners”
held on April 22-23, 1993. The review meeting was sponsored by the
National Institute on Drug Abuse.
COPYRIGHT STATUS
The National Institute on Drug Abuse has obtained permission from the
copyright holders to reproduce certain previously published material as
noted in the text. Further reproduction of this copyrighted material is
permitted only as part of a reprinting of the entire publication or chapter.
For any other use, the copyright holder’s permission is required. All other
material in this volume except quoted passages from copyrighted sources is
in the public domain and may be used or reproduced without permission
from the Institute or the authors. Citation of the source is appreciated.
Opinions expressed in this volume are those of the authors and do not
necessarily reflect the opinions or official policy of the National Institute on
Drug Abuse or any other part of the U.S. Department of Health and Human
Services.
The U.S. Government does not endorse or favor any specific commercial
product or company. Trade, proprietary, or company names appearing in
this publication are used only because they are considered essential in the
context of the studies reported herein.
National Institute on Drug Abuse
NIH Publication No. 94-3750
Printed 1994
NIDA Research Monographs are indexed in the Index Medicus. They are
selectively included in the coverage of American Statistics Index,
BioSciences Information Service, Chemical Abstracts, Current Contents,
Psychological Abstracts, and Psychopharmacology Abstracts.
Contents
Page
A Contextual Perspective on HIV Risk
Robert J. Battjes, Zili Sloboda, and William C. Grace . . . . . . . . . 1
I.
Panel I - Heterosexual Males
HIV Risk Behaviors of Heterosexual Male Drug Users . . . . . . . . 5
Richard H. Needle
Injection and Sexual Risk Behaviors of Male Heterosexual
Injecting Drug Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Richard C. Stephens and Sonia A. Alemagno
9
HIV/AIDS Risks Among Male, Heterosexual Noninjecting
Drug Users Who Exchange Crack for Sex . . . . . . . . . . . . . . . . . 26
James A. Inciardi
II. Panel II - Women
Context of HIV Risk Behavior Among Female Injecting
Drug Users and Female Sexual Partners of Injecting
Drug Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Diana Hartel
Female Drug Abusers and the Context of Their HIV
Transmission Risk Behaviors . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Karen Allen
Factors Associated with Sexual Risk of AIDS in Women . . . . . . 64
Ann O’Leary
Ill. Panel Ill - Men Who Have Sex With Men
Drug Use and HIV Risk Among Gay and Bisexual Men:
An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Robert J. Battjes
82
Substance Use and HIV-Transmitting Behaviors Among
Gay and Bisexual Men . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
David G. Ostrow
iii
Drug Use and HIV Risk Among Male Sex Workers: Results of
Two Samples in San Francisco . . . . . . . . . . . . . . . . . . . . . . . . . 114
Dan Waldorf
IV. Panel IV - Adolescents
HIV Risk in Drug-Using Adolescents . . . . . . . . . . . . . . . . . . . . 132
Vincent L. Smeriglio
HIV Risk in Adolescents: The Role of Sexual Activity and
Substance Use Behaviors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cherrie B. Boyer and Jonathan M. Ellen
Going Nowhere Fast: Methamphetamine Use and
HIV Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mary Jane Rotheram-Borus, G. Cajetan Luna,
Toby Marotta, and Hilarie Kelly
135
155
V. Panel V - Measurement Issues
The Context of Risk: Methodological Issues . . . . . . . . . . . . . . 183
Zili Sloboda
Bringing the Context in From the Cold: Substantive,
Technical, and Statistical Issues for AIDS Research in the
Second Decade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ann F. Brunswick
187
The Context of Risk: Ethnographic Contributions to the
Study of Drug Use and HIV . . . . . . . . . . . . . . . . . . . . . . . . . .
Stephen K. Koester
202
Assessing the Reliability and Validity of Self-Reported
Risk Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
David R. Gibson and Martin Young
Future Directions for Studies on the Context of
HIV Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
William C. Grace, Robert J. Battjes, and Zili Sloboda
iv
A Contextual Perspective on HIV
Risk
Robert J. Battjes, Zili Sloboda, and William C. Grace
For over a decade, acquired immunodeficiency syndrome (AIDS) has
been identified as a major public health epidemic and a research priority
of the National Institute on Drug Abuse (NIDA), the National Institutes
of Health (NIH), and other components of the U.S. Public Health Service.
Since the early years of the epidemic, needle and syringe sharing among
injecting drug users (IDUs) has been recognized as a major vector for
spread of the human immunodeficiency virus (HIV). Sexual transmission
among IDUs, and from IDUs to non-IDU sexual partners, also has been a
longstanding concern, as has perinatal transmission to their children. The
number of IDUs diagnosed with AIDS continues to expand in the second
decade of the epidemic, as does the number of related cases in IDUs’
sexual partners and their children. Approximately one-third of AIDS
cases are associated with injecting drug use (Centers for Disease Control
and Prevention 1993).
The potential contribution of non-IDUs to HIV transmission and disease
progression also has been of concern. For example, drug intoxication
may lessen inhibition and facilitate engagement in sexual risk behaviors.
Involvement in drug-using social networks may promote sexual contact
between noninjectors and IDUs, increasing risk of HIV exposure. Some
non-IDUs also engage in sex to obtain drugs or money to buy drugs.
Recently, it has been recognized that crack cocaine users who engage in
sex to obtain their drugs may be at especially high risk for HIV infection
(Inciardi et al. 1992, 1993; Ratner 1993).
Considerable progress has been made in identifying risk behaviors that
contribute to HIV transmission and in enumerating the extent of injecting
and sexual risk behaviors, especially for IDUs. However, much less
progress has been made in clarifying the factors that may explain risk
behaviors and barriers to risk reduction.
To review research on drug-using and sexual behaviors of drug users that
are associated with HIV transmission and to develop a research agenda
for future research on risk behaviors, NIDA convened a technical review
entitled The Context of HIV Risk Among Drug Users and Their Sexual
1
Partners. The review was held April 22 and 23, 1993, in Bethesda,
Maryland. This monograph contains the papers presented at the technical
review and recommendations emanating from it.
The technical review on HIV risk behaviors evolved from an earlier
meeting that NIDA convened in January 1992 as part of a planning
process to develop NIDA’s 5-year plan for AIDS research. A number of
areas for future research emerged from this planning meeting, with three
priority areas identified for emphasis:
Impact of the social context on drug injecting and drug-related sexual
behaviors;
Situational and temporal variations in injecting behaviors, especially
considering the continuum of injection initiation, maintenance, risk
reduction, and relapse; and
Sexual behaviors and changes in sexual relationships in relation to
developmental stages over time and in diverse interpersonal
situations.
These recommendations provided the foundation and framework for the
technical review. The technical review considered injection-related risk
behaviors of IDUs, sexual risk behaviors of IDUs and their sexual
partners, and sexual risk behaviors associated with drug use in non-IDUs.
The intent was to focus on the context of risk behaviors in order to clarify
factors that contribute to or protect against risk and to thereby advance
the knowledge base for future HIV-prevention strategies.
In addressing the context of risk behaviors, authors were asked to
consider situational, temporal, and developmental variation in risk
behaviors, to consider differences based on level of risk (i.e., differences
among individuals based on high, moderate, and low-risk behaviors), and
the effects of competing risks (e.g., risks of withdrawal and overdose
versus risk of HIV infection). Authors also were asked to consider
psychological status, the dyad or social grouping in which behavior
occurs, social networks, community context, culture, economic
conditions, individual’s serostatus, and community AIDS prevalence.
The monograph is organized into five sections. The first section,
introduced by Needle, contains chapters on HIV risk behaviors of
2
heterosexual male drug users, with Stephens and Alemagno addressing
IDUs and Inciardi addressing non-IDUs.
The second section, introduced by Hartel, contains chapters on HIV risk
behaviors among women. Allen addresses drug-using women, and
O’Leary addresses sexual partners of IDUs.
The third section, introduced by Battjes, includes chapters on homosexual
and bisexual male drug users. Ostrow addresses drug use associated with
sexual risk behaviors, and Waldorf addresses male prostitution and drug
use.
The fourth section, introduced by Smeriglio, focuses on adolescent drug
users. Boyer and Ellen address HIV risk behaviors among mainstream
adolescents (i.e., those who can be accessed through schools), while
Rotheram-Borus and colleagues address homosexual and bisexual
adolescent males who are-involved in a range of drug-using and sexual
risk behaviors, including injecting drug use and male prostitution.
The fifth section, introduced by Sloboda, concentrates on measurement
issues in HIV risk behavior research. Brunswick sets forth a
multidimensional conceptual framework to guide research in the area.
Koester discusses the importance of qualitative research in gaining a
better understanding of both the micro and macro contexts of these
behaviors of interest. Gibson and Young examine the reliability and
validity of self-reports of these risk behaviors.
During the technical review, Agar, Des Jarlais, and Robles served as
reactors to the formal presentations and, thereby, focused group
discussion and the formulation of research recommendations. The final
monograph chapter contains recommendations for future research based
on the formal papers and discussion.
REFERENCES
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Vol. 5, No. 3, 1993.
Inciardi, J.A.; Chitwood, D.D.; and McCoy, C.B. Special risks for the
acquisition and transmission of HIV infection during sex in crack
houses. J Acquir Immune Defic Syndr 5:951-952, 1992.
3
Inciardi, J.A.; Lockwood, D.; and Pottieger, A.E. Women and CrackCocaine. New York: Macmillan, 1993.
Ratner, M.S., ed. Crack Pipe as Pimp: An Ethnographic Investigation of
Sex-for-Crack Exchanges. New York: Lexington Books, 1993.
AUTHORS
Robert J. Battjes, D.S.W.
Zili Sloboda, Sc.D.
William C. Grace, Ph.D.
Division of Clinical Research
National Institute on Drug Abuse
Parklawn Building, Room 10A5600 Fishers Lane
Rockville, MD 20857
4
HIV Risk Behaviors of
Heterosexual Male Drug Users
Richard H. Needle
Since reporting of acquired immunodeficiency syndrome (AIDS) cases
began in the United States, over 325,000 persons have been reported as
having been diagnosed with AIDS, and more than 200,000 deaths have
occurred. Approximately 30 percent of AIDS cases in the United States
have been injecting drug users (IDUs), and most of these AIDS classified
cases (80 percent) are among heterosexual IDUs. More AIDS cases have
been reported among heterosexual male IDUs (about 75 percent) than
among female IDUs (Centers for Disease Control and Prevention 1993).
Available surveillance data on AIDS cases do not permit tracking of
trends in heterosexual male or female crack users who have become
human immunodeficiency virus (HIV) infected or acquired AIDS through
sexual transmission.
In this monograph, Stephens and Alemagno review the injection and
sexual risk behaviors of male heterosexual IDUs, and Inciardi reports on
HIV/AIDS risks among male heterosexual non-IDUs who exchange
crack for sex. Stephens and Alemagno focus on HIV risk-taking
behaviors of IDUs related to the multiperson use of contaminated needles
and other drug injection paraphernalia, their sexual risk practices, and the
potential for acquiring or transmitting HIV. Their review reveals that
individual behavioral patterns and risks of HIV exposure depend on a
number of factors, many of which are beginning to be recognized as
contextual. These factors include community prevalence, availability of
syringes and paraphernalia, availability of drugs, presence of shooting
galleries and needle exchanges, and availability of condoms (both female
and male). They also review some newly emerging data on indirect
needle-sharing practices, such as frontloading and backloading, and on
the use and nonuse of condoms, sexual behaviors, and other sexual risk
practices. While their review focuses on heterosexual male IDUs and
emphasizes sharing of needles and other injection paraphernalia among
IDUs as the major mode of HIV transmission, that practice is clearly not
the only mode for acquiring or transmitting HIV infection. Stephens and
Alemagno’s review of sexual practices of male heterosexual IDUs
5
highlights the critical importance of researchers’ awareness of the
increasing percentage of AIDS cases among women who have had sex
with IDUs.
Stephens and Alemagno also reviewed the growing literature on the
effectiveness of intervention programs targeted at changing behaviors
related to HIV/AIDS. They review data from the National AIDS
Demonstration Research (NADR) program sponsored by the National
Institute on Drug Abuse (NIDA) and discuss changes in HIV risk-taking
behaviors of the IDU population. While many drug users reported
reducing their HIV risk behaviors, others continued to practice high-risk
behaviors. NADR data, and more recently the Cooperative Agreement
(CA) for AIDS Community-Based Outreach/Intervention Research
Program data, suggest that it is very important to begin to focus more on
IDUs’ sexual risk practices to help prevent the spread of HIV.
Far less is understood about HIV risk-taking behaviors of male
heterosexual non-IDUs, particularly crack users who engage in sexual
practices that put them at risk for acquiring or transmitting HIV infection.
The crack epidemic emerged during the mid-1980s, years after the first
AIDS case was reported to the Centers for Disease Control and
Prevention (CDC). Most of the HIV epidemiology focused on the modes
of transmission in the high prevalence groups of gay/bisexual men and
male and female IDUs. To date, only 7 percent of AIDS cases are
attributable to heterosexual exposure, and the percentage of AIDS cases
among male heterosexual non-IDUs is very low (Centers for Disease
Control and Prevention 1993). There is no separate CDC category to
classify men or women who are noninjecting crack users, have had sex
with HIV-infected men or women, and attribute their HIV/AIDS status to
sexual transmission and noninjection drug use. These cases would be
classified in the exposure category of heterosexual contact, and more
specifically classified as sex with an HIV-infected person, risk not
specified (Centers for Disease Control and Prevention 1993).
Inciardi developed his chapter on crack, sex, and secondary spread of
HIV by relying on literature related to the efficiency of female-to-male
transmission; cofactors in female-to-male transmission of HIV; and the
small, but rapidly growing, ethnographic literature on bartering of sex for
crack cocaine. The chapter reviews explanations for the documented
higher rates of female-to-male infections in Africa and Europe than in the
United States. While the epidemiologic data are not adequate to permit
estimates of cases attributable to crack-using heterosexuals who have sex
6
with an HIV-infected individual, it is clear that the context of crack use
greatly increases the likelihood of sexual risk behaviors and of HIV
infection.
Relying on ethnographic studies of crack, sex, and HIV, Inciardi
describes the context of exchanges and clearly illustrates the association
of crack and HIV. Though there are few studies that separate out crack
users without a history of injection drug use, NIDA data from the CA
research program has found about a 6-percent seroprevalence rate among
crack users without any history of injection. For male crack users
without a history of injection drug use, a 5-percent seroprevalence rate
has been found (National Institute on Drug Abuse 1994).
Both chapters in this section have contributed to the understanding of
contextual factors influencing HIV risk behaviors and the relationship
between risk behaviors and the likelihood of acquiring and transmitting
HIV infection. The context or setting where behaviors occur, such as
crack houses, shooting galleries, or even public places, affects the risks of
practicing behaviors that expose individuals to HIV infection. The
presentations and discussion suggest that understanding about the future
of this epidemic will improve dramatically if researchers begin to shift
their thinking from examining risk behaviors from an individual
perspective to examining risk behaviors as behavioral transactions
between individuals (dyads), triads, and groups, focusing on the context
in which multiperson use of equipment and sexual risk-taking behaviors
occur. Both chapters raise new research questions that require attention if
researchers are to respond effectively and prevent further tragic
consequences of the AIDS epidemic.
REFERENCES
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Vol. 5, No. 3, 1993.
National Institute on Drug Abuse. Cooperative Agreement for AIDS
Community-Rased Outreach/Intervention Research Program:
Quarterly Database Fact Sheet. Rockville, MD: National Institute on
Drug Abuse, 1994.
7
AUTHOR
Richard H. Needle, Ph.D., M.P.H.
Chief, Community Research Branch
Division of Clinical Research
National Institute on Drug Abuse
Parklawn Building, Room 9A-30
5600 Fishers Lane
Rockville, MD 20857
8
Injection and Sexual Risk
Behaviors of Male Heterosexual
Injecting Drug Users
Richard C. Stephens and Sonia A. Alemagno
INTRODUCTION
For male, heterosexual injecting drug users (IDUs), the greatest risk for
human immunodeficiency virus (HIV) infection is associated with
injection risks such as sharing contaminated needles. Though difficult to
quantify, the risk that male heterosexual IDUs incur through unsafe
sexual practices—even with other IDUs—is almost certainly secondary.
This chapter focuses on studies published in the last 5 years or so in the
areas of injection and sexual risk behaviors. While there is a rather large,
but not too detailed, literature on needle-related behaviors (e.g., Agar
1973; Hanson et al. 1985), many of these studies were conducted before
the age of acquired immunodeficiency syndrome (AIDS). Very little can
be found on the sexual behavior of male IDUs in the early literature; only
since the spread of AIDS has this topic drawn the attention of drug abuse
researchers.
The chapter presents a representative, although not exhaustive, review of
recent topics that could lead to the development of future research
agendas regarding HIV prevention strategies for male heterosexual IDUs.
While some studies reported here have focused exclusively on this target
population, other studies reviewed have presented analyses on broader
IDU samples that control for sexual orientation and are, therefore,
relevant to this review. Several studies focus on relapse to risk behaviors
among homosexual males.
Sharing of needles and other drug injection paraphernalia among IDUs is
well established as the major mode of HIV transmission (Des Jarlais and
Friedman 1987; Nemoto et al. 1990). The epidemiology of sexual
transmission of HIV associated with injecting drug use has focused
largely on transmission from IDUs to their noninjecting sexual partners.
The extent to which HIV is spread sexually among heterosexual IDUs
remains unconfirmed (Haverkos and Edelman 1988).
9
Blower and colleagues (1991) have presented a data-based mathematical
model formulated to assess consequences of heterosexual, injecting drug
use in order to provide qualitative and quantitative insights into the HIV
epidemic in New York City. The results demonstrated the significance of
the dynamic interaction of heterosexual and IDU transmission. In the
early stages of the epidemic, IDU transmission is often more important
than heterosexual transmission; however, the relative importance of
heterosexual transmission increases as the epidemic spreads from the IDU
community to the heterosexual, non-IDU community. Results also
indicated that the effect of the heterosexual transmission risk factor on
increasing the risk of HIV infection depends on the level of injecting drug
use risk. Results demonstrate that the addition of the heterosexual
transmission risk factor does not increase an individual’s risk of HIV
infection among individuals with very high risk IDU activity (such as
needle sharing with strangers). However, the addition of the same
heterosexual risk factor for an individual with lower risk activity (such as
sharing needles with a buddy) can significantly increase the individual’s
risk for infection. Confidence intervals on the prediction estimates of
future cumulative number of AIDS cases are extremely wide. The
authors suggest that long-term precise estimates of the future number of
AIDS cases will only be possible once the values of key variables such as
gender, needle sharing, and sexual behaviors have been evaluated
accurately.
Battjes and colleagues (1990) have explored heterosexual transmission by
comparing HIV seropositivity among IDUs who have IDU sexual
partners with rates of seropositivity among IDUs with no IDU sexual
partners. The hypothesis for this analysis was that if sexual transmission
is contributing substantially to the spread of HIV among heterosexual
IDUs, then prevalence of HIV infection should be greater among IDUs
with IDU sexual partners than among IDUs with no IDU sexual partners.
In bivariate analysis, results indicated that heterosexual IDUs who
reported having IDU sexual partners were more likely to be seropositive
than IDUs who reported no IDU sexual partners. Yet, IDUs with IDU
sexual partners also reported more drug-using practices that placed them
at greater risk for HIV infection. Controlling for drug use and
demographic characteristics, having IDU sexual partners was no longer
associated with increased risk for HIV infection. Thus, the apparent
increased risk for HIV infection found in bivariate analysis may have
been due to differences in drug use practices, not to additional risks
resulting from sexual HIV transmission among heterosexuals.
10
Battjes and colleagues’ (1990) finding is contrary to the findings of
Schoenbaum and colleagues (1989), who found heterosexual contact with
other IDUs to be an independent risk factor for HIV infection when drug
use and demographic factors were controlled statistically. Other
predictors of seropositivity included the number of injections per month,
the percentage of injections with used needles, the average number of
cocaine injections per month, and the percentage of injections with
needles shared with strangers. These studies may have divergent findings
as a result of differences in target groups studied. Battjes and colleagues
(1990) studied recent drug injectors on admission to methadone
treatment; Schoenbaum and colleagues (1989) studied a methadone intreatment sample, some of whom had stopped injecting some time earlier.
Schoenbaum found sexual risk stronger among former injectors
compared with current injectors.
Battjes and colleagues point out that the heterosexual transmission of
HIV may be relatively more important for younger IDUs who are more
sexually active and engage in lower levels of drug-using HIV risk
behaviors. The relative roles of needle sharing and sexual transmission
may differ depending on the seropositivity level of the community. As
IDUs increasingly adopt safer needle-using practices, sexual transmission
can be expected to play an increasingly more important role in the spread
of HIV among heterosexual IDUs, unless corresponding changes in highrisk sexual behaviors are also made.
A recent study in Sydney conducted by Ross and colleagues (1992) has
reported significant differences across sexual orientation in HIV
serostatus for IDU males, with homosexual men having the highest HIV
seroprevalence rate (35 percent), bisexual men intermediate (12 percent),
and heterosexual men lowest (3 percent).
INJECTION RISKS
Currently, injecting drug use continues to be the second most common
risk behavior associated with AIDS in the United States (CDC 1993; Des
Jarlais and Friedman 1988; Lewis et al. 1990). Many studies have
reported the increased risk for infection for IDUs who share needles and
syringes (Chaisson et al. 1987; Chitwood et al. 1990; Hopkins 1988;
Magura et al. 1989; Schoenbaum et al. 1989).
11
Even in areas with low seroprevalence rates, needle sharing can be
common (Metzger et al. 1991). It is, however, by no means a universal
practice of IDUs. Not all IDUs share needles, and some IDUs have
begun to seek new needles (Guydish et al. 1990). Nevertheless, it may be
possible to identify those variables that increase the risk of HIV through
injection behaviors.
Many studies have described the situational nature of needle sharing. An
addict who has the drugs and his or her own outfit will not share. Needle
sharing is related to the supply of needles available. Many studies have
reported that the availability of needles and the problems associated with
procuring them play a major role in needle sharing (Murphy 1987;
Murphy and Waldorf 1991; Power et al. 1988; Waldorf and Murphy
1989; Watters 1988). Qualitative work has pointed to the issue of
availability of clean needles as the main factor for needle sharing. It will
be interesting to determine the impact of needle exchange programs on
needle sharing.
Dolan and colleagues (1987) have identified several variables that
discriminated needle sharing among drug abusers admitted to a 30-day
inpatient drug treatment program. Compared with other injecting drug
abusers, needle sharers were more likely to engage in polydrug use, were
more likely to use a shooting gallery, and had higher scores on a drug use
severity test. No demographic or personality variables, such as age, race,
education, or any of the 24 Minnesota Multiphasic Personality Inventory
(MMPI) scores, discriminated needle sharers from nonsharers. Guydish
and colleagues (1990) have found that needle sharing is predicted by
earlier time of admission to drug treatment, cocaine use, and being
younger in age. This is supported by a study performed by Kleinman and
colleagues (1990), who report that new drug users (persons who had been
using drugs for only 1 or 2 years) are significantly less likely than others
to practice risk-reduction measures and are less likely to have salient
knowledge about AIDS transmission.
Magura and colleagues (1989) reported that needle sharing is related to
peers’ injecting drug use, economic motivation to share, not owning
injection equipment, and fatalism about developing AIDS. Factors that
did not predict needle sharing are also of interest. Knowledge of AIDS
risks, knowing someone with AIDS, gender, age, ethnicity, marital status,
and time in methadone treatment were not associated with sharing. In
this sample, in which knowledge of AIDS risks was high, needle sharing
could not be attributed to ignorance about the consequences of such
12
behavior or how to protect oneself and others against infection. Needle
sharers did not deny the dangers of their behavior; most claimed they
borrowed injection equipment only in what they considered emergencies.
Addicts were more likely to share in order to avoid withdrawal symptoms
or if they believed their friends would be insulted if they refused to share.
This study strongly supports the idea that peer behavior heavily
influences needle-sharing behavior.
This peer dimension has been explored by other researchers who have
studied the extensive network of IDUs. Siegal (1990) described how
IDUs in the Dayton and Columbus, Ohio, areas, where shooting galleries
are not the typical location for drug use, interact primarily within their
own network. The usual sites for drug use for these IDUs are private
residences where only persons well-known to each other are present. It is
rare for users to violate these network boundaries, which are constructed
along social and sociogeographic lines. However, preliminary
ethnographic reports suggest that IDUs who are very frequent crack users
are transcending these boundaries.
Metzger and colleagues (1991) performed a comparison of methadone
clients who continued to share needles, who had injected drugs but not
shared needles, and who had not injected drugs in the preceding 6
months. Those who continued to share reported greater difficulty in
acquiring new needles, more legal difficulties, more severe drug
problems, and a higher level of psychiatric symptoms as measured by the
Beck Depression Inventory and the SCL-90 scales such as scales of
somatization, interpersonal sensitivity, depression, hostility, and anxiety.
Contrary to expectations, fear of AIDS was not associated with safer
injecting practices. Knowledge of methods of self-protection from
infection was not found to relate to needle sharing, although such
knowledge was nearly universal among the study group. The authors
focused on the importance of the association between psychological
distress and recent needle sharing and reported that reduction in
psychiatric symptoms may play a role in reducing high-risk behavior.
Injection risks may also be related to injection practices. Inciardi and
Page (1991) and Kaplan (1983) have described the process of skinpopping—the intramuscular (into the muscle) or subcutaneous (under the
skin) injecting of drugs. Skin-popping is a common method of heroin
use by experimenters and “tasters” who mistakenly believe that addiction
cannot occur by this practice. Another process, known as booting, has
been described by Des Jarlais and Hunt (1988). The booting process
13
involves using a syringe to draw blood from the user’s arm, mixing the
drawn blood with the drug already taken into the syringe, and injection of
the blood/drug mixture into the vein. Many IDUs believe that this
practice potentiates a drug’s effects; however, this procedure is reported
to leave traces of blood in the needle and the syringe, thus placing
subsequent users of the injection equipment at higher risk. Inciardi and
Page (1991) have studied frontloading and backloading practices in
Miami-procedures for making the speedball solution (heroin and
cocaine mixed together) in which two needles and syringes are used.
This practice potentially doubles the risk of HIV transmission.
There is preliminary evidence that suggests that the type of drug used
(heroin, cocaine, etc.) may increase the likelihood of needle sharing
(Turner et al. 1990). Due to cocaine’s short effect time, cocaine injectors
require more injections to maintain their high than heroin injectors. One
study found amphetamine users were more likely to share needles than
heroin users and were less likely to be in contact with formal agencies
and consequently HIV prevention services (Baxter and Schlecht 1990).
Koester and colleagues (1990) studied IDUs in Denver and detailed how
IDUs may inadvertently transmit HIV. They found that drug preparation
and injection involves a complex series of steps that may be influenced
by any number of variables, including the type of drug being injected, the
beliefs and customs of the users, and socioeconomic and psychosocial
factors. Page and colleagues (1990) have identified some of the customs
practiced in settings where injecting drug use takes place and have found
these customs may shift as conditions and interpersonal relations change
in a given scene.
Several studies have indicated the importance of not only needle- and
syringe-sharing practices in HIV transmission but the sharing of drug
containers, cotton, and other injection paraphernalia (Inciardi 1990;
Inciardi and Page 1991; Page et al. 1990). Furthermore, the use of
contaminated syringes may not entail sharing, in the sense of the social
act of passing a recently used needle to a waiting partner, but rather
pooling of used needles (Chitwood et al. 1990).
Virological studies have indicated that HIV can survive in ordinary tap
water for extended periods of time (Resnick et al. 1986); thus, shared
rinse water represents a considerable potential transmission risk. Koester
and colleagues (1990) have described how IDUs often clean their works
from the same cup of water that others are using to prepare their heroin
14
after they shoot drugs. Thus, the same water is used for rinsing of
syringes and mixing of drugs.
Reports that the use of shooting galleries is associated with exposure to
HIV infection are consistent in the literature. Murphy and Waldorf
(1991) have described shooting galleries that provide a relatively private
place to inject drugs, provide syringes for rent, and offer other materials
to prepare the drugs for injection—such as water, cotton, matches, bottle
caps, and syringes for sale. While Des Jarlais and Friedman (1987)
described shooting galleries as an integral part of the addict lifestyle in
New York, Murphy and Waldorf described shooting galleries as integral
for impoverished addicts in the San Francisco Bay area.
Shared injection equipment can transmit HIV when residual contaminated
blood remains in previously used syringes and needles. Blood residue is
often present in the syringe because of aspiration of venous blood into the
syringe. Chitwood and colleagues (1990) reported that shooting gallery
syringes that contain visible blood and those that appear visibly clean
have a significant risk of carrying HIV. Ten percent of the needlesyringe combinations that were tested from three shooting galleries in
South Florida were positive for antibodies to HIV type 1 (HIV-1). Data
from this study suggest that the selection of visibly clean injection
equipment does not eliminate the possibility of using an HIV-infected
needle.
Heterosexual IDUs appear to be less likely to engage in persistent
shooting gallery use as compared with homosexual and bisexual men.
Celentano and colleagues (1991) reported that in a sample of 2,615 active
IDUs in Baltimore, lifetime shooting gallery use was associated with
heavier drug involvement and with being male, homosexual or bisexual,
of low socioeconomic status, and African American.
The understanding of race/ethnicity as a predictor of HIV risk behavior is
still in very early stages. Koblin and colleagues (1990) examined
differences in HIV seroprevalence and patterns of drug use and sexual
behaviors among Hispanic, African-American, and white IDUs in a
sample recruited at multiple sites in Worcester, Massachusetts. After
adjustment for demographic differences, African-American males were
significantly less likely to report risky drug use behaviors (such as ever
sharing needles and recently visiting shooting galleries) compared with
white males. However, Hispanic males were significantly more likely to
report recent risky drug use behaviors. Both African Americans and
15
Hispanics were less likely to report risky sexual behaviors compared with
whites. Odds ratios for HIV seropositivity were significantly higher for
Hispanics compared with whites and maintained marginal significance
for African Americans compared with whites when adjusted for risk
behaviors and demographic variables. These results, indicating variations
in risk behaviors across race/ethnicity groups, are supported by other
studies. Guydish and colleagues (1990) reported that African-American
IDUs in a San Francisco sample were less likely to share needles than
whites.
Studies of HIV prevalence among IDUs have demonstrated varying
racial/ethnic patterns dependent on the geographic area and
subpopulation sampled. It is likely that risk behaviors also vary by
geographic area and subpopulation. The finding that Hispanics and nonHispanic African Americans in the United States share a disproportionate
burden of HIV infection is perhaps due to the earlier onset of the HIV
epidemic in African-American and Hispanic IDU populations. That is,
while risk behaviors observed across race/ethnicity groups may be
comparable, the increased probability of contact with an infected person
may be an independent risk factor for African Americans and Hispanics.
Behavioral, cultural, and socioeconomic differences within racial/ethnic
groups need to be examined in further studies. Serrano (1990) has
described Puerto Rican IDUs. The impact of HIV at the neighborhood
level is described as devastating. Preventing AIDS within minorities is
proposed within the context of several well-defined communities, such as
the family, and entails the development of community resources to deal
with many complex problems. Marin (1990) described non-Puerto Rican
Hispanics and AIDS prevention efforts that use culturally appropriate
interventions.
SEXUAL RISK BEHAVIORS
Feucht and colleagues (1990) described the sexual behavior of IDUs.
Most male IDUs are sexually active and heterosexual, and significant
proportions have multiple female partners. In this sample, while white
males were about as likely to have an IDU partner as a non-IDU partner,
only a third of the African-American males reported having a female IDU
partner during the preceding year, while 85 percent reported having a
female non-IDU partner. African-American males were more likely than
white males to have sex with a non-IDU female and were more likely
16
than whites to have multiple non-IDU female partners. White males were
more likely to have multiple IDU female partners.
Several studies have reported the low use of condoms among
heterosexual male IDUs. Ross and colleagues (1992) compared IDUs
across sexual orientation groups and reported sexual risk related to
condom nonuse was lowest for homosexual men, intermediate for
bisexual men, and highest for heterosexual men; i.e., heterosexual men
were least likely to use condoms. Heterosexual men were, however, least
likely to have multiple partners and to have anal sex. Reported rates of
condom use vary by study; however, most report nonuse at more than
two-thirds.
Factors that lead to condom use remain unclear. Chapman and
colleagues (1990) found that in a general heterosexual sample, three
conceptually coherent factors (condom use as a positive action, condom
use as a cue to embarrassment, and condom use as antithetical to good
sex) discriminated between users and nonusers. Condom use has been
reported to be associated with voluntary testing for HIV, average or more
than average fear of sexually transmitted diseases (STDs), and knowledge
of HIV carrier status in personal relations (Moatti et al. 1991). This study
included 1,088 heterosexuals with multiple partners (IDU status
unknown).
Watkins and colleagues (1992) compared in- and out-of-treatment IDUs
on their sexual risk behaviors. Out-of-treatment IDUs reported
significantly more partners than in-treatment IDUs and more often
exchanged sex for money or drugs. Alcabes and colleagues (1992) also
compared in-treatment to out-of-treatment IDU samples and found that
the out-of-treatment IDUs tended to be younger, male, and African
American. However, associations between HIV-1 seropositivity and a
series of demographic and drug-using characteristics were similar in
direction and magnitude among subjects currently in treatment and those
not in treatment. Lewis and Watters (1991) reported that sexual risktaking behavior in a sample of IDUs was associated with recent increases
in both injecting and smoking cocaine.
Clearly the concept of risk perception is important. Connors (1992)
described the overlap of the risks associated with AIDS transmission with
other risks common in the use of intravenous (IV) drugs (risks associated
with stealing, dealing drugs, carrying a needle, copping drugs, and
borrowing needles). On a daily basis, IDUs risk arrest, overdose, and
17
victimization through theft, violence, and illnesses related to their drug
use. A person may borrow a needle from someone because the borrower
does not want to assume the risk for illegal possession of a syringe.
Similarly, the probability of a fatal overdose is reduced if someone else
with whom they share a needle is present. The immediate costs of arrest,
overdose, or dope sickness are usually more salient than the long-term
risks to one’s health (hepatitis, endocarditis, AIDS). Risk taking and risk
perception are also likely to change over time, as a person’s dependency
on a substance increases and acculturation to the IDU community
increases.
CHANGES IN AIDS RISK BEHAVIORS
There is a growing literature on the effectiveness of intervention
programs whose goals are to impact on the AIDS risk behaviors of IDUs
and their sexual partners. While it is beyond the scope of this chapter to
review this literature, it is nevertheless instructive to review the findings
of a study that summarizes the impact of the National AIDS
Demonstration Research (NADR) projects (Stephens et al. 1993). These
projects targeted IDUs and their sexual partners with a variety of AIDS
prevention messages delivered in a variety of innovative educational
programs.
Twenty-eight NADR programs contributed longitudinal data (measured
by standardized interviews at baseline and 6-month followup) on 13,475
IDUs and 1,637 sexual partners. Analysis indicated significant changes
over the 6-month period in HIV-related risk behaviors among IDUs,
including frequency of injecting drugs, the use of noninjected drugs, the
use of borrowed injection equipment, and the number of sexual partners
reported by the subject. Significant increases were observed in the use of
new (rather than previously used) needles, the use of bleach to clean
injection equipment between uses, and the use of condoms. The
magnitude of change in many of the risk behaviors was very large. For
example, at baseline, 54 percent of the total sample indicated that they
had shared needles with two or more different persons during the
preceding 6 months; at followup, the percentage had dropped to 23
percent. In particular, changes in risky needle-related behaviors were
more dramatic than the changes in the risky sexual behaviors.
Nevertheless, there were significant numbers of subjects who did not
reduce their AIDS risk behaviors.
18
A number of variables linked to level of risk were identified in a series of
analysis of covariance models. At followup, Hispanic IDUs, those who
were unemployed, those who had previously been in drug treatment,
those whose primary injection drug was heroin, and those who had longer
histories of drug injection reported less reduction in the use of injected
drugs than did other IDUs. At followup, overall needle-related risk was
reduced most significantly among African-American IDUs. Overall
sexual risk at followup was lowest among males and nonwhite IDUs.
SUGGESTIONS FOR FUTURE RESEARCH
These results clearly indicate that IDUs are willing and able to reduce
their risks of contracting AIDS. Comparing IDUs receiving drug
treatment with those who have not, drug treatment does not ensure
reduced risk. Studies consistently report high AIDS knowledge among
IDUs, yet knowledge does not necessarily lead to behavior change.
The public and individual health issues of male heterosexual IDUs pose
challenges for prevention, intervention, politics, and research. A review
of the literature draws attention to some of the factors that place male
heterosexual IDUs at highest risk. These factors include high levels of
injecting drug use (which may lead to a hierarchy of risks that places
long-range health risks below the risk of arrest or withdrawal), being
younger and therefore more sexually active, having multiple sexual
partners (particularly IDU sexual partners), being of African-American or
Hispanic race/ethnicity, living in high seroprevalence areas, having
negative attitudes regarding condom use for various cultural or
psychological reasons, being out of drug treatment, and participating in
social networks that do not value AIDS preventive behavior.
Epidemiological efforts need to explore further the interaction of
heterosexual and IDU transmission. What specific heterosexual risk
behaviors pose risks independent of IDU risks? How do these risks vary
across sexual orientation?
It is obvious that much more needs to be known about the circumstances
of needle sharing. While the existent literature addresses the larger
questions relating variables like ethnicity, level of AIDS knowledge, and
other variables to needle-sharing practices, little is known about the
detailed circumstances of this phenomenon. Research questions that beg
for further answers include: What are the various dimensions of needle
19
sharing? When are needles deliberately shared with others versus used in
unknown contexts such as in shooting galleries? Why do IDUs share
needles? What is the role of the heterosexual IDU partner? Does
sharing/not sharing affect the relationship? What factors are associated
with situational variation in behaviors? What is the composition of a
needle-sharing network?
Further efforts should be focused on the factors that predict persistent
high risk. Under what circumstances do low-risk individuals move to
high risk? In the context of simultaneous risks, how is the hierarchy of
risks constructed and how does it impact the decision to share needles and
other injection equipment?
Contextual and social network variables should be examined using
intensive ethnographic observation and in-depth interviewing. The
unique aspects of heterosexual relationships (steady partnership, multiple
partnership, cultural values, family values, smooth social relations) need
to be examined in the context of IDU network relationships such as those
to running buddies. AIDS often does not pose an immediate risk;
however, the introduction and use of condoms, or the suggestion of
cleaning needles shared with a sexual partner, may pose an immediate
interpersonal risk.
Research on HIV transmission from IDUs to non-IDU sexual partners is
today the subject of increased attention due to the confirmed risk of
heterosexual transmission. However, the importance of HIV transmission
among IDUs who engage in sexual relations with other IDUs should not
be ignored. Heterosexual IDUs should be alerted not only to the risk of
sexual HIV transmission to their non-IDU sexual partners but also to the
risks of contracting and transmitting HIV through heterosexual
intercourse with other IDUs.
Intervention efforts may need more realistic goals. Changing HIV risk
behavior is complex since these risk behaviors are based on interpersonal
relations. Further efforts need to be directed at how risk behaviors vary
across sexual orientation (homosexual, bisexual, and heterosexual) and
whether factors that appear to be related to relapse to risk behaviors in
homosexual men are similar for bisexual and heterosexual men. Which
components of interventions are most likely to lead to risk reductions?
Under what legal codes is change in risk accomplished; for example, do
researchers need additional studies of the impact of needle exchange
programs?
20
Finally, additional efforts should be directed at considering special
treatment programs for IDUs with AIDS. There is evidence that
heterosexual seropositive IDUs continue to engage in unprotected sex.
It is clear that intensive interviewing and detailed ethnographic research
will be able to answer those questions satisfactorily.
REFERENCES
Agar, M. Ripping and Running. New York: Seminar Press, 1973.
Alcabes, P.; Vlahov, D.; and Anthony, J.C. Correlates of human
immunodeficiency virus infection in intravenous drug users: Are
treatment-program samples misleading? Br J Addict 87:47-54, 1992.
Battjes, R.J.; Pickens, R.W.; Amsel, Z.; and Brown, L.S. Heterosexual
transmission of human immunodeficiency virus among intravenous
drug users. J Infect Dis 162:1007-1011, 1990.
Baxter, D.N., and Schlecht, B. Patterns of behavior amongst injection
drug-users-Implications for HIV. Public Health 104(5):321-325,
1990.
Blower, S.M.; Hartel, D.; Dowlatabadi, H.; Anderson, R.M.; and May,
R.M. Drugs, sex and HIV: A mathematical model for New York City.
Philos Trans R Soc Lond [Biol]: 171-187, 1991.
Celentano, D.D.; Vlahov, D.; Cohn, S.; Anthony, J.C.; Solomon, L.; and
Nelson, K.E. Risk factors for shooting gallery use and cessation
among intravenous drug users. Am J Public Health 8l(10):1291-1295,
1991.
Centers for Disease Control. AIDS/HIV Surveillance, National
Commission on AIDS Report, 1993.
Chaisson, R.E.; Moss, A.R.; Onishi, R.; Osmond, D.; and Carlson, J.R.
Human Immunodeficiency Virus infection in heterosexual intravenous
drug users in San Francisco. Am J Public Health 77(2):169-172, 1987.
Chapman, S.; Stoker, L.; Ward, M.; Porritt, D.; and Fahey, P.
Discriminant attitudes and beliefs about condoms in young, multipartner heterosexuals. Int J STD AIDS 1(6):422-428, 1990.
Chitwood, D.D.; McCoy, C.B.; and Comerford, M. Risk behavior of
intravenous users: Implications of intervention. In: Leukefeld, C.;
Battjes, R.; and Amsel, Z. eds. AIDS and Intravenous Drug Use:
Future Directions for Community-Based Prevention Research.
National Institute on Drug Abuse Research Monograph 93.
Washington, DC: Supt. of Docs., U.S. Govt. Print. Off., 1990.
21
Chitwood, D.D.; McCoy, C.B.; Inciardi, J.A.; McBride, D.C.;
Comerford, M.; Trapido, E.; McCoy, V.; Page, J.B.; Griffin, J.;
Fletcher, M.A.; and Ashman, M.A. HIV seropositivity of needles from
shooting galleries in South Florida. Am J Public Health 80(2):150152, 1990.
Connors, M.M. Risk perception, risk taking and risk management among
intravenous drug users: Implications for AIDS prevention. Soc Sci
Med 34(6):591-601, 1992.
Des Jarlais, D.C., and Friedman, S.R. HIV infection among intravenous
drug users: Epidemiology and risk reduction. AIDS 1(2):67-76, 1987.
Des Jarlais, D.C., and Friedman, S.R. The psychology of preventing
AIDS among intravenous drug users: A social learning
conceptualization. Am Psychol 43(11):865-870, 1988.
Des Jarlais, D.C., and Hunt, D. AIDS and Intravenous Drug Use.
National Institute of Justice AIDS Bulletin, 1988.
Dolan, M.P.; Black, J.L.; Deford, H.A.; Skinner, J.R.; and Robinowitz, R.
Characteristics of drug abusers that discriminate needle-sharers. Public
Health Rep 102(4):395-398, 1987.
Feucht, T.E.; Stephens, R.C.; and Roman, S.W. The sexual behavior of
intravenous drug users: Assessing the risk of sexual transmission of
HIV. J Drug Issues 20(2):195-213, 1990.
Guydish, J.R.; Abramowitz, A.; Woods, W.; Black, D.M.; and Sorensen,
J.L. Changes in needle sharing behavior among intravenous drug
users: San Francisco, 1986-88. Am J Public Health 80(8):995-997,
1990.
Hanson, B.; Beschner, G.; Walter, J. M.; and Bovelle, E. Life With
Heroin. Lexington, Mass.: Lexington Books, 1985.
Haverkos, H.W., and Edelman, R. The epidemiology of acquired
immunodeficiency syndrome among heterosexuals. JAMA 260: 19221929, 1988.
Hopkins, W. Needle sharing and street behavior in response to AIDS in
New York City. In: Battjes, R., and Pickens, R., eds. Needle Sharing
among intravenous Drug Abusers: National and International
Perspectives. National Institute on Drug Abuse Research Monograph
No. 80. DHHS Pub. No. (ADM)88-1567. Washington, DC: Supt. of
Docs., U.S. Govt. Print. Off., 1988.
Inciardi, J. HIV, AIDS, and intravenous drug use: Some considerations.
J Drug Issues 20(2):181-194, 1990.
Inciardi, J., and Page, J. Drug sharing among intravenous drug users.
AIDS 5(6):772-773, 1991.
Kaplan, J. HIV, heroin, and heterosexual relations. Soc Sci Med
32(9):1037-1050, 1983.
22
Kleinman, P.H.; Goldsmith, D.S.; Friedman, S.R.; Hopkins, W.; and Des
Jarlais, D.C. Knowledge about and behaviors affecting the spread of
AIDS: A street survey of intravenous drug users and their associates in
New York City. Int J Addict 25(4):345-361, 1990.
Koblin, B.A.; McCusker, J.; Lewis, B.F.; and Sullivan, J.L. Racial/ethnic
differences in HIV-1 seroprevalence and risky behaviors among
intravenous drug users in a multisite study. Am J Epidemiol
132(5):837-846, 1990.
Koester, S.; Booth, R.; and Wayne, W. The risk of transmission from
sharing water, drug mixing containers, and cotton filters among
intravenous drug users. Int J Drug Policy 1(6):28-30, 1990.
Lewis, D.K.; Watters, J.K.; and Case, P. The prevalence of high-risk
sexual behavior in male intravenous drug users with steady female
partners. Am J Public Health 80:465-466, 1990.
Lewis, D.K., and Watters, J.K. Sexual risk behavior among heterosexual
intravenous drug users: Ethnic and gender variations. AIDS 5:77-83,
1991.
Magura, S.; Grossman, J.I.; Lipton, D.S.; Siddiqi, Q.; Shapiro, J.; Marion,
I.; and Amann, K.R. Determinants of needle sharing among
intravenous drug users. Am J Public Health 79(4):459-462, 1989.
Marin, B.V. AIDS prevention for non-Puerto Rican Hispanics. In:
Leukefeld, C.; Battjes, R.; and Amsel, Z., eds. AIDS and Intravenous
Drug Use: Future Directions for Community-Based Prevention
Research. National Institute on Drug Abuse Research Monograph No.
93. DHHS. Pub. No. (ADM)90-1627. Supt. of Docs, U.S. Govt. Print.
Off., 1990.
Metzger, D.; Woody, G.; De Phillips, D.; McLellan, A.T.; O’Brien, C.P.;
and Platt, J.J. Risk factors for needle sharing among methadonetreated patients. Am J Psychiatry 148(5):636-640, May 1991.
Moatti, J.P.; Bajos, N.; and Durbec, J.P. Determinant of condom use
among French heterosexuals with multiple partners Am J Public
Health 81(1):106-109, 1991.
Murphy, S. Intravenous drug use and AIDS: Notes on the social economy
of needle sharing. Contemp Drug Probl 14(3):373-410, 1987.
Murphy, S., and Waldorf, D. Kicking down to the street doc: Shooting
galleries in the San Francisco Bay Area. Contemp Drug Probl
28(1):9-29, 1991.
Nemoto, T.; Brown Jr., L.S.; Foster, K.; and Chu, A. Behavioral risk
factors of human immunodeficiency virus infection among
intravenous drug users and implications for preventive interventions.
AIDS Educ Prev 2(2):116-126, 1990.
23
Page, J.B.; Smith, P.C.; and Kane, N. Shooting galleries, their
proprietors, and implications for prevention of AIDS. Drugs Soc
5(1):69-85, 1990.
Power, R.; Hartnoll, R.; and Daviadud, E. Drug injecting, AIDS, and risk
behavior: Potential for change and intervention strategies. Br J Addict
83:649-654, 1988.
Resnick, L.K.; Veren, S.Z.; Salahuddin, S.; Tondreau, S.; and Karkham,
P.D. Stability and inactivation of HTLV-III/LAV under clinical and
laboratory environments. JAMA 255(14):1887-1891, 1986.
Ross, M.W.; Wodak, A.G.; Gold, J.; and Miller, M.E. Differences across
sexual orientation on HIV risk behaviors in injecting drug users. AIDS
Care 4(2):139-148, 1992.
Schoenbaum, E.E.; Hartel, D.; Selwyn, P.A.; Klein, R.S.; Davenny, K.;
Rogers, M.; Feiner, C.; and Friedland, G. Risk factors for human
immunodeficiency virus infection in intravenous drug users. N Engl J
Med 321:874-879, 1989.
Serrano, Y, The Puerto Rican Intravenous Drug User. In: Leukefeld, C.;
Battjes, R.; and Amsel, Z., eds. AIDS and Intravenous Drug Use:
Future Directions for Community-Based Prevention Research.
National Institute on Drug Abuse Research Monograph No. 93.
DHHS Pub. No. (ADM)90-1627. Washington, DC: Supt. of Docs.,
U.S. Govt. Print. Off., 1990.
Siegal, H.A. Intravenous drug abuse and the HIV epidemic in two
midwestem cities: A preliminary report. J Drug Issues 20(2):281-290,
1990.
Stephens, R.C.; Simpson, D.D.; Coyle, S.L.; McCoy, C.B.; and the
National AIDS Research Consortium. In: Brown, B.S., and Beschner,
G., eds. Handbook in AIDS and the Injection Drug User. New York:
Greenwood Press, 1993.
Turner, C.; Miller, H.; and Moses, L., eds. AIDS, Sexual Behavior and
Intravenous Drug Use. Washington, D.C.: National Academy Press,
1990.
Waldorf, D., and Murphy, S. IV drug use and syringe sharing practices in
call men and hustlers. In: Plant, N., ed. AIDS, Drugs and Prostitution.
London, England: Routledge, 1989.
Watkins, K.E.; Metzger, D.; Woody, G.; and McLellan, A.T. High-risk
sexual behaviors of intravenous drug users in- and out-of-treatment:
Implications for the spread of HIV infection. Am J Drug Alcohol
Abuse 18(4):389-398, 1992.
Watters, J.K. Meaning and context: The social facts of intravenous drug
use and HIV transmission in the inner city. J Psychoactive Drugs
20(2):173-177, 1988.
24
AUTHORS
Richard C. Stephens, Ph.D.
Professor
Department of Sociology
University of Akron
Olin Hall
Akron, OH 44325-0602
Sonia A. Alemagno, Ph.D.
Program Manager
Health Care Administration
Cleveland State University
Fenn Tower 1306
Cleveland, OH 44115
25
HIV/AIDS Risks Among Male,
Heterosexual Noninjecting Drug
Users Who Exchange Crack for
Sex
James A. Inciardi
INTRODUCTION
Of the 249, 199 cumulative adult/adolescent acquired immunodeficiency
syndrome (AIDS) cases reported to the Centers for Disease Control and
Prevention (CDC) through December 1992, only 7 percent (N = 16,254)
were the result of heterosexual contact, and of these, some 39 percent
(N = 6,419) occurred among men (CDC 1993). As such, heterosexual
contact accounts for only 3 percent of all reported AIDS cases among
men in the United States.
At present, the biological variables that determine human
immunodeficiency virus (HIV) infectivity (the tendency to spread from
host to host) and susceptibility (the tendency for a host to become
infected) are incompletely understood. HIV has been isolated from the
semen of infected men, and it appears that it may be harbored in the cells
of preejaculated fluids or sequestered in inflammatory lesions (Fischl
1988). Furthermore, there is evidence that women can harbor HIV in
vaginal and cervical secretions at varying times during the menstrual
cycle (Vogt et al. 1986, 1987; Wofsy 1986).
The probability of sexual transmission of HIV among gay and bisexual
men through anal intercourse and to women through vaginal intercourse
has been well documented (Kasalow and Francis 1989; Ma and
Armstrong 1989). However, although there is the potential for viral
transmission from female secretions, the absolute amounts of virus in
these secretions appear to be relatively low. The efficiency of
transmission of male-to-female versus female-to-male is likely affected
by the relative infectivity of these different secretions, as well as sex
during menses, specific sexual practices, the relative integrity of skin and
mucosal surfaces involved, and the presence of other sexually transmitted
diseases (STDs). Given this, there are a number of issues to be examined
when considering the heterosexual transmission of HIV among male
26
non-injecting drug users (non-IDUs). Of particular interest are the
biological variables, risk factors and cofactors, and particular sexual
practices that relate to the target population of this review—specifically,
men who exchange crack for sex with women crack users.
THE EFFICIENCY OF FEMALE-TO-MALE TRANSMISSION
The likelihood of female-to-male sexual transmission of HIV is supported
by biological plausibility, equal numbers of male and female AIDS cases
in some African countries, case reports of males with no risk factors other
than heterosexual intercourse, and seroconversion of male sex partners of
infected women that occurred while the couples were being studied
prospectively. In terms of the biological plausibility of female-to-male
transmission, it has been argued that since other STDs are bidirectional in
nature, it is not unreasonable to assume that HIV can spread in the same
manner. A number of studies have documented that African and Indian
men who have multiple female sex partners or sexual contact with
prostitutes are at high risk for becoming infected with HIV (Cameron et
al. 1987; Carswell et al. 1989; Clumeck et al. 1985; Kreiss et al. 1986;
Singh et al. 1993). The most persuasive case reports of female-to-male
transmission have been those in which the female acquired the infection
from a transfusion or organ transplant and her male partner (without other
known risk factors) subsequently seroconverted (L’Age-Stehr et al.
1985), and those in which a sequential chain of male-to-female-to-male
transmission was observed (Calabrese and Gopalakrishna 1986).
Although significant numbers of female-to-male infections have been
documented in Africa (Barnett and Blaikie 1992; Panos Institute 1988;
Shannon et al. 1991), such a mode of transmission has been reported only
infrequently in the United States, and the majority of the more recent case
reports have come from investigators in Europe (Beck et al. 1989;
European Study Group 1992; Johnson et al. 1989; Lefrere et al. 1988).
Several explanations have been offered for the differences in female-tomale transmission rates between U.S. and African experiences. A
number of researchers have suggested that the infrequent documentation
of heterosexual transmission from women to men in the United States
may be a function of the history of the epidemic. They suggest that the
initial phase was largely confined to male homosexuals and intravenous
drug users (IVDUs). As a result, the number of infected women was low
27
during that time, and the possibility of female-to-male transmission was
small.
Because the majority of AIDS cases occurring today reflect infections
that were acquired during the early years of the epidemic, most
heterosexually acquired infections among men still may be in the
asymptomatic or latent stage (Friedland and Klein 1987; Osmond 1990).
The infectivity of an HIV carrier increases over time, suggesting that this
additional factor in the natural history of HIV infection may magnify the
effects alluded to above. Following this line of reasoning, it has been
argued that virus concentration in genital secretions also may increase
over the course of the infection (Burke and Redfield 1988). On the other
hand, Haverkos and Battjes (1992) recently have argued that the relative
frequency of female-to-male transmission has been underestimated
(primarily as the result of the way that cases are classified), suggesting
that it represents a more significant public health concern than is
generally believed.
COFACTORS IN FEMALE-TO-MALE TRANSMISSION OF HIV
Of additional significance here are the relationships among prostitution,
untreated STDs, condom use, and HIV infection. In 1988, Cohen and
colleagues argued that:
If prostitutes [in the United States] are effectively
transmitting the AIDS virus to their customers, there
would be far more cases of white, heterosexual males
diagnosed with AIDS than are reflected in the current
statistics, because some IVDUs in New York, including
some prostitutes, have been infected with the AIDS virus
since at least 1978. The average street prostitute sees
1,500 customers a year. If even five percent of female
street prostitutes in New York City were infected by
1981, the year AIDS was first identified, even
moderately efficient transmission of the virus from
prostitutes to clients would have resulted in the diagnosis
of at least 100,000 white, heterosexual men by now.
(p. 18)
28
In 1992, at the National Insitute on Drug Abuse (NIDA) AIDS research
planning meeting, Dr. Cohen reiterated her argument:
Direct information on customers becoming infected by
sex workers is very limited, but does not support more
than rare instances of transmission. CDC data report
very low rates of “heterosexual transmission” or “no
identified risk” AIDS cases among men; given the
number of men who purchase sex in this country, these
low rates argue against anything but rare transmission.
(p. 1)
As already noted, in Africa, by contrast, there is considerable evidence of
viral transmission by HIV-infected prostitutes to their male customers
(Barnett and Blaikie 1992; Carswell et al. 1989; Shannon et al. 1991). A
number of factors contribute to this difference. Among African
prostitutes and their customers, there appear to be significant proportions
with untreated STDs, including genital ulcers, and these appear to
increase men’s susceptibility to HIV (Cameron et al. 1987; Johnson and
Laga 1988; Kreiss et al. 1989; O’Farrell 1989; Piot et al. 1987; Plummer
et al. 1991; Simonsen et al. 1988). In addition, several studies have noted
that sex workers in the United States are more conscious of STDs and are
more likely to use safer sexual practices (e.g., vaginal and anal sex with
condoms) with customers than those in Africa (Cohen 1992). For
example, a number of reports have noted that there is a socialization
process associated with becoming a prostitute in the United States.
Would-be and neophyte prostitutes learn the appropriate techniques and
safeguards through apprenticeships with pimps or more experienced
prostitutes (Carmen and Moody 1985; Evans 1979; Goldstein 1979;
Miller 1986; Rosenbaum 1981; Winick and Kinsie 1971). In some cases
there is informal or even formal training on how to protect oneself from
theft, violence, or disease. For example, in one sociological analysis of
prostitution as an occupation, it was found that the recognition of STDs
was a specific topic of instruction for neophyte house prostitutes (Heyl
1979). Furthermore, those who work the streets or in organized houses of
prostitution in the United States have friendships and peer relationships,
however loose, unstructured, and transitory they may often be, through
which experiences are shared, techniques are traded, warnings are
communicated, and knowledge is reinforced.
29
All of these issues and considerations have relevance to the examination
of HIV risks among heterosexual male non-IDUs who exchange crack for
sex.
THE BARTERING OF SEX FOR CRACK COCAINE
Although the bartering of sex for crack had been mentioned in the
popular media at the very beginnings of the crack epidemic (Gross 1985;
Lamar 1986; Lawlor 1986), the first empirical study of the phenomenon
did not appear in the scientific literature until 1989. In that analysis,
drawn from a larger study of drug use and street crime among serious
delinquents in Miami, the potential for HIV acquisition and transmission
through sex-for-crack exchanges was addressed (Inciardi 1989). Of 100
girls in the 14- to l7-year age range, 27 had bartered sex for crack during
the 1-year period prior to interview. Of these, 11 had traded sex for drugs
on fewer than 6 occasions but had nevertheless exchanged sex for money
on an aggregate of 6,850 occasions. By contrast, there were others in the
sample who had bartered sex hundreds and even thousands of times.
At about the same time that this research was being reported, others
began to notice rising rates of syphilis and other STDs among crack users
(Bowser 1989; Fullilove and Fullilove 1989; Fullilove et al. 1990; Kerr
1989; Knopf 1989a, b). Shortly thereafter, sex-for-crack exchanges were
targeted for systematic study, but to date, only a few reports have
appeared in the literature. Moreover, one of the difficulties in assessing
the nature of HIV risks associated with crack use, particularly among
heterosexual male non-IDUs, is the fact that most crack users engage in
multiple risk behaviors.
A study of risk factors for HIV-1 infection was conducted at an STD
clinic in an area in New York City where the cumulative incidence of
AIDS in adults through mid-1990 was 9.1 per 1,000 population and
where the use of illicit drugs, including crack smoking, was common
(Chiasson et al. 1991). The overall seroprevalenee among the 3,084
volunteer subjects was 12 percent, with 80 percent of these seropositives
reporting risk behaviors associated with HIV-1 infection, including maleto-male sexual contact, IV drug use, and heterosexual contact with an
IDU. The seroprevalence in individuals denying these risks was 3.6
percent in men (50 of 1,389) and 4.2 percent in women (22 of 522).
Among these individuals, the behaviors associated with infection were
30
prostitution and use of crack in women and a history of syphilis, crack
use, and sexual contact with a crack-using prostitute in men.
The potential for a male in sex-for-crack exchanges to come into contact
with an HIV positive female partner was demonstrated in a study of 87
New York City women who had been admitted to a municipal hospital
with a diagnosis of pelvic inflammatory disease (PID) (Des Jarlais et al.
1991). Crack use was reported by 56 percent of the subjects (N = 49),
and of these, 20 percent were HIV seropositive. Crack use was
significantly related to both traditional AIDS risk behaviors (injecting
drugs and sex with an IDU) and other unsafe sexual behaviors
(exchanging sex for money or drugs and having casual sex partners).
In 1989, given the potential of sex-for-crack exchanges for spreading
HIV to new populations, NIDA supported ethnographic studies of the
phenomenon in eight cities: Chicago, Denver, Los Angeles, Miami,
Newark, New York, Philadelphia, and San Francisco (Ratner 1993). A
total of 340 crack users were interviewed in depth, 69 percent of whom
were women. Of the 233 women, 108 had participated in sex-for-crack
exchanges, as had 69 of the men. HIV testing was done with 168 of the
subjects, and a total of 14 percent were found to be positive for the HIV
antibody. Of the 24 males who were non-IDUs and who had engaged in
heterosexual sex-for-crack exchanges, 12 percent were HIV positive.
CRACK, SEX, AND THE SECONDARY SPREAD OF HIV
The potential for transmission of HIV from women to heterosexual male
non-IDUs within the context of sex-for-crack exchanges is related to a
number of considerations, including two important independent risk
factors. The first is the cocaine/sexuality connection. Cocaine has long
had a reputation as an aphrodisiac, although sexuality is notoriously a
playground of legend, exaggeration, and rumor. In all likelihood, much
of cocaine’s reputation may be from the mental exhilaration and
disinhibition it engenders, thus bringing about some heightened sexual
pleasure during the early stages of use. At the same time, cocaine users
have consistently reported that the drug tends to delay the sexual climax,
and that after prolonged stimulation, an explosive orgasm occurs. Users
also report that chronic use of the drug results in sexual dysfunction, with
impotence and the inability to ejaculate the common complaints of male
users, the inability to climax among females, and decreased desire for sex
31
becoming the norm for both male and female users (Grinspoon and
Bakalar 1976; Weiss and Mirin 1987).
What applies to powder cocaine with regard to sexual stimulation and
functioning also would apply to crack cocaine. Curiously, however, there
is the rather contradictory evidence that crack appears to engender what
has been referred to as “hypersexual” behavior among many users. This
has been observed in a number of ethnographic studies (Inciardi et al.
1993; Ratner 1993). Many crack addicted women and men engage in any
manner of sexual activity, under any circumstances, in private or in
public, and with multiple partners of either sex (or both sexes
simultaneously). Indeed, the tendency of some crack users to engage in
high-frequency sex with numerous anonymous partners is a feature of
crack dependence and crack house life in many locales, Furthermore,
sex-for-drugs exchanges seem to be far more common among female
crack addicts now than they ever were among female narcotics addicts,
even at the height of the 1967 to 1974 heroin epidemics (Ball and
Chambers 1970; Rosenbaum 1981).
The pharmacological effects of crack (the rapid onset, extreme euphoria,
and short duration), as well as the economic need to pay for the drug,
have a special impact on women. Because crack makes its users ecstatic
and yet is so short-acting, it has an extremely high addiction potential.
Use rapidly becomes compulsive use. Crack acquisition thus becomes
enormously more important than family, work, social responsibility,
health, values, modesty, morality, or self-respect. This makes sex-forcrack exchanges psychologically tolerable as an economic necessity.
Further, the disinhibiting effects of crack enable users to engage in sexual
acts they otherwise might not even consider. Crack-using male
customers involved in sex-for-crack exchanges, although they
consistently report difficulties in maintaining an erection and ejaculating
under the influence of the drug, also state that sex while smoking not only
enhances the drug’s effects but also gives them a sense of power and
control that they typically do not have in other aspects of their lives
(Inciardi et al. 1993).
A second independent risk factor relates to the impact of crack use on
physical health and hygiene. Because of the pharmacology and addiction
potential of crack, it is rare that smokers take only a single hit of the drug.
More likely they spend $50 to $500 during what they call a mission—a
3- or 4-day binge, smoking almost constantly, 3 to 50 rocks per day.
During these cycles, crack users rarely eat or sleep. This tendency to
32
binge on crack for days at a time, neglecting food, sleep, and basic
hygiene, severely compromises physical health. Consequently, crack
users appear emaciated most of the time. They lose interest in their
physical appearance. Many have scabs on their faces, arms, and
legs-the result of burns and picking at the skin to remove bugs and other
insects believed to be crawling under the skin. Crack users tend to have
burned facial hair from carelessly lighting their smoking paraphernalia,
they have mouth ulcerations and burned lips and tongues from the hot
stems of their pipes, and they seem to cough constantly. In addition,
many have self-reported and have been observed to have untreated STDs
(Inciardi et al. 1993; McCoy and Miles 1992; Ratner 1993).
Going further, although street prostitutes who barter sex for money to
purchase drugs often insist that their customers use condoms, this is not
usually the case with crack house sex. In fact, condoms are rarely seen in
crack houses. Given the health status of crack users (including a high
likelihood of compromised immune systems), the incidence of STDs
(many of which go untreated), and general lack of condom use, many of
the conditions that have contributed to the heterosexual transmission of
HIV in Africa exist in crack houses in Miami, New York, Philadelphia,
and other urban areas across the United States. In addition, given the
frequency of sex and the large number of partners associated with crack
house sex, the potential for coming into contact with HIV is of an even
greater magnitude. And it is within this context that the heterosexual
transmission of HIV to non-IDUs is most likely.
Furthermore, there are a number of situations that may make the crack
house unique in the heterosexual spread of HIV. In a study of crack
house sexual activities in Miami, for example, a number of men reported
that they could climax only through extremely vigorous masturbation or
prolonged vaginal intercourse. Many of the female partners in these
exchanges reported that the lengthy intercourse resulted in both vaginal
and penile bleeding (Inciardi et al. 1993). In situations as these, the
potential for female-to-male transmission of HIV increases. During
vaginal intercourse, the friction of the penis against the clitoris, labia
minora, and vaginal vestibule, opening, and canal causes stimulation that
can generate copious amounts of vaginal secretions. And as noted above,
HIV has been isolated from vaginal/cervical secretions. Moreover, since
women who exchange sex for crack in crack houses do so with many
different men during the course of a day or night, potentially
HIV-infected semen from a previous customer can still be present in the
vagina. It was reported by one crack house customer that he had ruptured
33
the skin on his penis while having intercourse with a crack house
prostitute while she was menstruating. This informant also indicated that
vaginal sex during menses was not a rare event. As such, genital
secretions as well as semen and blood come into direct contact with the
traumatized skin of a client’s penis during crack house sex.
In many crack houses, it is not uncommon for women to engage in
repeated oral, vaginal, and anal sexual activities, often with no time lapse
between successive customers. Since condom use is rare, not only are the
women exposed to the semen of all of their male partners, but successive
male partners also are exposed to the semen of the women’s previous
partners (Inciardi et al. 1992). As such, heterosexual transmission of HIV
can be from male to female, female to male, and male to male. It would
appear, moreover, that this phenomenon is not unique to crack house sex.
In an Orange County, California, study of Hispanic undocumented
migrant workers and heroin-addicted prostitutes (Magana 1991), a
parallel situation was found. Large numbers of men engaged in vaginal
intercourse with the same woman in rapid succession-a sexual behavior
referred to by the participants as becoming “milk brothers.”
Finally, there is the matter of oral sex. Few studies have associated orogenital sex with HIV transmission (Fischl et al. 1987; Puro et al. 1991),
and the majority of reports have examined this transmission route among
homosexual men (Keet et al. 1992; Lifson et al. 1990; Rozenbaum et al.
1988). Only one fully documented case of female-to-male transmission
through oral sex has appeared in the literature (Spitzer and Weiner 1989).
In crack houses, oral sex (both fellatio and cunnilingus) is common.
Given such risk factors and cofactors as STD infections, genital ulcers,
lesions on the lips and tongue, and abrasions on the penis and in the
vagina among those who exchange sex for crack, the potential for
female-to-male transmission of HIV through oral sex is not
inconsiderable.
CONCLUSIONS
As is apparent from this review, any examination of HIV risks among
heterosexual male non-IDUs must be done primarily through indirect
evidence. At present, the number of confirmed cases of female-to-male
AIDS cases in the United States is comparatively small, and the majority
of studies of drug-using groups find that most subjects have multiple risk
34
factors. Not surprisingly, this review raises more questions than it
answers. For example:
1. How widespread is the female-to-male heterosexual transmission of
HIV in general, and particularly among non-IDUs?
2. How widespread are sex-for-crack exchanges? Are the crack-related
sexual phenomena reported in the studies covered in this review
(Inciardi et al. 1993; Ratner 1993) unique to the cities targeted, or are
they common to the wider crack scene? Are they transitory
phenomena, characteristic only of a single phase of the crack
epidemic?
3. Where do these sex-for-crack exchanges take place? Are they
primarily in crack houses, or do they occur in other places as well?
What are the situational variables that foster occurrence in one locale
as opposed to others?
4. What is the frequency of sex in crack houses?
5 . To what extent are there untreated STDs among crack users? What
are the health risks associated with crack addiction that have
implications for HIV acquisition and transmission? To what extent
do crack users perceive risk for HIV?
6. What is the status of the crack epidemic? Has it increased, declined,
or plateaued? What are the regional differences in the status of the
epidemic? What are the implications of the status of the epidemic for
the heterosexual transmission of HIV?
7. For those deeply enmeshed in the crack life, what types of HIV/AIDS
prevention/intervention strategies are most appropriate? What are the
best mechanisms for luring crack addicts into HIV prevention
programs and into drug abuse treatment? In addition to drug
treatment and general health care, what other kinds of services does
this population need?
8 . Has the crack epidemic had any positive influences? That is, are
there any potential sources of positive behavior change that can be
utilized in either drug abuse or AIDS prevention efforts?
35
REFERENCES
Ball, J.C., and Chambers, C.D. The Epidemiology of Opiate Addiction in
the United States. Springfield, IL: Charles C. Thomas, 1970.
Barnett, T., and Blaikie, P. AIDS in Africa: Its Present and Future
Impact. New York: Guilford Press, 1992.
Beck, E.J.; Donegan, C.; Cohen, C.S.; Kenny, C.; Moss, V.; Underhill,
G.S.; Terry, P.; Jeffries, D.J.; Pinching, A.J.; Miller, D.L.;
Cunningham, D.G.; and Harris, J.R.W. Risk factors for HIV-1
infection in a British population: Lessons from a London sexually
transmitted disease clinic. AIDS 3:533-538, 1989.
Bowser, B. Crack and AIDS: An ethnographic impression. J Natl Med
Assoc 81:538-540, 1989.
Burke, D.S., and Redfield, R.R. Transmission of human
immunodeficiency virus (HIV). New Engl J Med 318:1202-1203,
1988.
Calabrese, L.H., and Gopalakrishna, K.V. Transmission of HTLV-III
infection from man to woman to man. N Engl J Med 314:987, 1986.
Cameron, D.W.; Plummer, F.A.; and Simonsen, J.N. “Female to Male
Heterosexual Transmission of HIV Infection in Nairobi.” Paper
presented at the 3rd International Conference on AIDS, Washington,
DC, June 1-5, 1987.
Carmen, A., and Moody, H. Working Women: The Subterranean World
of Street Prostitution. New York: Harper & Row, 1985.
Carswell, J.W.; Lloyd, G.; and Howells, J. Prevalence of HIV-1 in East
African lorry drivers. AIDS 3:759-761, 1989.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Atlanta, GA: Centers for Disease Control and Prevention,
February 1993.
Chiasson, M.A.; Stonebumer, R.L.; Hildebrandt, D.S.; Ewing, W.E.;
Telzak, E.E.; and Jaffee, H.W. Heterosexual transmission of HIV-1
associated with the use of smokable freebase cocaine (crack). AIDS
5:1121-1126, 1991.
Clumeck, N.; Robert-Guroff, M.; Van de Perre, P.; Jennings, A.;
Sibomana, J.; Demol, P.; Cran, S.; and Gallo, R.C.
Seroepidemiological studies of HTLV-III antibody prevalence among
selected groups of heterosexual Africans. JAMA 254:2599, 1985.
Cohen, J.B. “Sexual Transmission Associated with Commercial Sex.”
Paper presented at the National Institute on Drug Abuse AIDS
Research Planning Meeting, Rockville, MD, January 23-24, 1992.
Cohen, J.B.; Alexander, P.; and Wofsy, C. Prostitutes and AIDS: Public
policy issues. AIDS Pub Pol J 3:16-22, 1988.
36
Des Jarlais, D.C.; Abdul-Quader, A.; Minkoff, H.; Hoegsberg, B.;
Landesman, S.; and Tross, S. Crack use and multiple AIDS risk
behaviors. J Acquir Immune Defic Syndr 4:446-447, 1991.
European Study Group on Heterosexual Transmission of HIV.
Comparison of female to male and male to female transmission of
HIV in 563 stable couples. Brit Med J 304:809-813, 1992.
Evans, H. Harlots, Whores, and Hookers: A History of Prostitution. New
York: Dorset Books, 1979.
Fischl, M.A. Prevention of transmission of AIDS during sexual
intercourse. In: DeVita, V.T.; Hellman, S.; and Rosenberg, S.A., eds.
AIDS: Etiology, Diagnosis, Treatment, and Prevention. Philadelphia:
J.B. Lippincott, 1988. pp. 369-374.
Fischl, M.A.; Dickenson, G.M.; Scott, G.B.; Klimas, N.; Fletcher, M.A.;
and Parks, W. Evaluation of heterosexual partners, children, and
household contacts of adults with AIDS. JAMA 257:640-644, 1987.
Friedland, G.H., and Klein, R.S. Transmission of human
immunodeficiency virus. N Engl J Med 317:1125-1135, 1987.
Fullilove, M.T., and Fullilove, R.E. Intersecting epidemics: Black teen
crack use and sexually transmitted disease. J Am Med Worn Assoc
44:146-153, 1989.
Fullilove, R.E.; Fullilove, M.T.; Bowser, B.B.; and Gross, S.A. Risk of
sexually transmitted disease among black adolescent crack users in
Oakland and San Francisco, California. JAMA 263:851-855, 1990.
Goldstein, P.J. Prostitution and Drugs. Lexington, MA: Lexington
Books, 1979.
Grinspoon, L., and Bakalar, J.B. Cocaine: A Drug and Its Social
Evolution. New York: Basic Books, 1976.
Gross, J. A new, purified form of cocaine causes alarm as abuse
increases. New York Times, November 29, 1985.
Haverkos, H.W., and Battjes, R.J. Female-to-male transmission of HIV.
JAMA 268:1855, 1992.
Heyl, B.S. The Madam as Entrepreneur: Career Management in House
Prostitution. New Brunswick, NJ: Transaction Books, 1979.
Inciardi, J.A. Trading sex for crack among juvenile drug users: A
research note. Contemp Drug Prob 16:689-700, 1989.
Inciardi, J.A.; Chitwood, D.D.; and McCoy, C.B. Special risks for the
acquisition and transmission of HIV infection during sex in crack
houses. J Acquir Immune Defic Syndr 5:951-952, 1992.
Inciardi, J.A.; Lockwood, D.; and Pottieger, A.E. Women and CrackCocaine. New York: Macmillan, 1993.
Johnson, A.M., and Laga, M. Heterosexual transmission of HIV.
Supplement. AIDS 2:S49-S56, 1988.
37
Johnson, A.M.; Petherick, A.; Davidson, S.J.; Brettle, R.; Hooker, M.;
Howard, L.; McLean, K.A.; Osborne, L.E.M.; Robertson, R.; Sonnex,
C.; Tchamouroff, S.; Shergold, C.; and Adler, M.W. Transmission of
HIV to heterosexual partners of infected men and women. AIDS
3:367-372, 1989.
Kasalow, R.A., and Francis, D.P., eds. The Epidemiology of AIDS:
Expression, Occurrence, and Control of Human Immunodefciency
Virus Type 1 Infection. New York: Oxford University Press, 1989.
Keet, I.P.M.; van Lent, N.A.; Sandfort, T.G.M.; Coutinho, R.A.; and van
Griensven, G.J.P. Orogenital sex and the transmission of HIV among
homosexual men. AIDS 6:223-226, 1992.
Kerr, P. Crack and resurgence of syphilis spreading AIDS among the
poor. New York Times, August 20, 1989.
Knopf, A. Syphilis and crack linked in Connecticut. Substance Abuse
Report August 1:1-2, 1989a .
Knopf, A. Syphilis and gonorrhea on the rise among inner-city drug
addicts. Substance Abuse Report June 1:1-2, 1989b .
Kreiss, J.K.; Coombs, R.; Plummer, F.; Holmes, K.K.; Nikora, B.;
Cameron, W.; Ngugi, E.; Ndinya-Achola, J.O.; and Corey, L. Isolation
of human immunodeficiency virus from genital ulcers in Nairobi
prostitutes. J Infect Dis 160:380-384, 1989.
Kreiss, J.K.; Koech, D.; and Plummer, F.A. AIDS virus infection in
Nairobi prostitutes: Spread of the epidemic in East Africa. N Engl J
Med 314:414, 1986.
L’Age-Stehr, J.; Schwarz, A.; and Offermann, G. HTLV-III infection in
kidney transplant recipients. Lancet 2:1361-1362, 1985.
Lamar, J.V. Crack. Time, June 2, 1986.
Lawlor, J. USA battles souped-up cocaine. USA Today, June 16, 1986.
Lefrere, J.J.; Vittecoq, D.; North, M.L.; Smilovici, W.; Aubertin, P.;
Gueguen, M.; Nicod, A.; Lambert, T.; and Janot, C. Risk of female to
male transmission of HIV from women infected by transfusion. AIDS
2:137-138, 1988.
Lifson, A.R.; O’Malley, P.M.; Hessol, N.A.; Buchbinder, S.P.; Cannon,
L.; and Rutherford, G.W. HIV seroconversion in two homosexual men
after receptive oral intercourse with ejaculation: Implications for
counseling concerning safe sexual practices. Am J Public Health
80:1509-1511, 1990.
Ma, P., and Armstrong, D., eds. AIDS and Infections of Homosexual
Men. Boston: Butterworths, 1989.
Magana, J.R. Sex, drugs, and HIV: An ethnographic investigation. Soc
Sci Med 33:5-9, 1991.
38
McCoy, H.V., and Miles, C. A gender comparison of health status among
users of crack cocaine. J Psychoactive Drugs 24:389-397, 1992.
Miller, E.M. Street Women. Philadelphia: Temple University Press, 1986.
O’Farrell, N. Transmission of HIV: Genital ulceration, sexual behavior,
and circumcision. Lancet ii Nov.:1157, 1989.
Osmond, D. Heterosexual transmission of HIV. In: Cohen, P.T.; Sande,
M.A.; and Volberding, P.A., eds. The AIDS Knowledge Base.
Waltham, MA: Massachusetts Medical Society, 1990. pp. 1.2.4:1-9
Panos Institute. AIDS and the Third World. London: Panos Publications,
1988.
Piot, P.; Kreiss, J.K.; Ndinya-Achola, J.O.; Ngugi, E.N.; Simonsen, J.N.;
Cameron, D.W.; Tealman, H.; and Plummer, F.A. Editorial review:
Heterosexual transmission of HIV. AIDS 1:199-206, 1987.
Plummer, F.A.; Simonsen, J.N.; Cameron, D.W.; Ndinya-Achola, J.O.;
Kreiss, J.K.; Gakinya, M.N.; Waiyaki, P.; Cheang, M.; Piot, P.;
Ronald, A.R.; and Ngugi, E.N. Cofactors in male-female sexual
transmission of human immunodeficiency virus type 1. J Infect Dis
163:233-239, 1991.
Puro, V.; Narciso, N.; Girardi, E.; Antonelli, L.; Zaccarelli, M.; and
Visco, G. Male-to-female transmission of human immunodeficiency
virus infection by oro-genital sex. Eur J Clin Microbial Infect Dis
10:47, 1991.
Ratner, M.S., ed. Crack Pipe as Pimp: An Ethnographic Investigation of
Sex-for-Crack Exchanges. New York: Lexington Books, 1993.
Rosenbaum, M. Women on Heroin. New Brunswick, NJ: Rutgers
University Press, 1981.
Rozenbaum, W.; Gharakhanian, S.; Cardon, B.; Duval, E.; and Coulaud,
J.P. HIV transmission by oral sex. Lancet i June:1395, 1988.
Shannon, G.W.; Pyle, G.F.; and Bashshur, R.L: The Geography of AIDS:
Origins and Course of the Epidemic. New York: Guilford Press, 1991.
Simonsen, J.N.; Cameron, D.W.; Gakinya, M.N.; Ndinya-Achola, J.O.;
D’Costa, L.J.; Karasira, P.; Cheang, M.; Ronald, A.R.; Piot, P.; and
Plummer, F.A. Human immunodeficiency virus infection among men
with sexually transmitted diseases. N Engl J Med 319:274-278, 1988.
Singh, Y.N.; Singh, K.; Joshi, R.; Rustagi, G.K.; and Malaviya, A.N.
HIV infection among long-distance truck drivers in Delhi, India.
J Acquir Immune Defic Syndr, 6:323, 1993.
Spitzer, P.G., and Weiner, N.J. Transmission of HIV infection from a
woman to a man by oral sex. N Engl J Med 320:251, 1989.
39
Vogt, M.W.; Craven, D.E.; Crawford, D.; Witt, D.J.; Byington, R.;
Schooley, R.T.; and Hirsch, M.S. Isolation of HTLV-III/LAV from
cervical secretions of women at risk for AIDS. Lancet 1:525-527,
1986.
Vogt, M.W.; Witt, D.J.; Craven, D.E.; Byington, R.; Crawford, D.;
Hutchinson, M.S.; Schooley, R.T.; and Hirsch, M.S. Isolation patterns
of the human immunodeficiency virus from cervical secretions during
the menstrual cycle of women at risk for the Acquired
Immunodeficiency Syndrome. Ann Intern Med 106:380-382, 1987.
Weiss, R.D., and Mirin, S.M. Cocaine. Washington, DC: American
Psychiatric Press, 1987.
Winick, C., and Kinsie, P.M. The Lively Commerce: Prostitution in the
United States. Chicago: Quadrangle Books, 1971.
Wofsy, C.B.; Cohen, J.B.; Hauer, L.B.; Padian, N.; Michaelis, B.; Evans,
J.; and Levy, J.A. Isolation of AIDS-associated retrovirus from genital
secretions of women with antibodies to the virus. Lancet 1:527-529,
1986.
AUTHOR
James A. Inciardi, Ph.D.
Professor and Director
Center for Drug and Alcohol Studies
University of Delaware
77 East Main Street
Newark, DE 19716-2582
40
Context of HIV Risk Behavior
Among Female Injecting Drug
Users and Female Sexual
Partners of Injecting Drug Users
Diana Hartel
In the first decade of research on human immunodeficiency virus (HIV)
infection among women, most research was devoted to quantifying
biological risks for heterosexual and perinatal infection (Miller et al.
1990). Of necessity, most studies were designed to quantify individual
exposure risk and to develop models to predict the course of the epidemic
through the population (Brookmeyer and Gail 1994). In heterosexual
transmission studies, scientific discourse centered on the degree of risk
for each type of sex act (e.g., anal versus vaginal intercourse) and
individual conditions that might facilitate transmission (e.g., presence of
sexually transmitted infections such as syphilis). Except for prostitutes
who use needles, there was little research on female injecting drug users
(IDUs) per se; exposure to HIV, rather than gender roles related to
injecting behavior, was emphasized (Cohen et al. 1988). Female sex
partners of male IDUs only recently have been subject to study. HIV
prevention programs for women, whether IDUs or partners of IDUs, have
brought increased interest in gender-specific factors that may be barriers
to risk reduction (Brown and Weissman 1994).
In this section, the unique aspects of HIV risk behavior among women
are reviewed. Female IDUs (Allen, this volume) have been considered
separately from sexual partners of IDUs (O’Leary, this volume), although
in reality these groups are overlapping. The vast majority of female IDUs
are sexual partners of IDUs, although this is not the case for male IDUs
(Brown and Weissman 1994; Cohen et al. 1988; Turner et al. 1989). As
with male IDUs, the severity of chemical dependency and access to
needles are central determinants of needle use and needle sharing among
female IDUs (Grund et al. 1992). Women with severe drug dependency
problems are likely to be at greater risk of sexual acquisition of HIV than
men (Ickovics and Rodin 1992). IDU women are more likely than their
male counterparts to engage in high-risk sex with multiple partners for
money or drugs, share needles, and have unprotected sex with an IDU
partner. In addition, there appears to be greater efficiency in HIV
41
transmission from males to females than the reverse in the United States
and Europe (European Study Group 1992; Vermund et al. 1990). Apart
from female IDUs, other sexual partners of IDUs comprise a large
heterogenous population in which cultural, social, racial, and other factors
are major determinants of sexual risk behavior and safer sex practices.
Since there has been very little recent research on female IDUs, a brief
review of the literature on women who inject drugs that predates the HIV
epidemic is included in this introduction. The injection method has been
employed nearly universally by opiate addicts since the 1940s (Brecher
1972), with cocaine a more recent but important drug of injection since
the 1970s (Kozel and Adams 1986). Regardless of the type of drug
injected, the estimate of the ratio of male to female IDUs (Brown and
Weissman 1994; Cohen et al. 1988; Turner et al. 1989) and “hard” drug
users is 3 to 1 (Cottler et al. 1990). This ratio persists in current statistics
derived from needle exchange programs (University of California 1993).
The lower ratio of female to male IDUs has been attributed to a number
of factors: greater stigmatization and thus secrecy of women who use
illicit drugs (Barnard 1993; Kane 1991; Rosenbaum 1988); faster
removal of women from the active injecting population through treatment
or abstinence as compared to men (Hser et al. 1987; Longshore et al.
1993; Rosenbaum 1981); and reduced opportunity to initiate and sustain
hard drug use among women with resultant irregular or polydrug use
patterns (Rosenbaum 1981; Worth 1991).
There is some indication of a trend toward convergence for males and
females in population-based (i.e., not samples of IDUs alone) drug use
studies (Clayton et al. 1986). Population-based surveys show few
differences by gender among adolescents (Kaestner et al. 1986). A
community-based life study of drug use and health care behavior that has
followed Harlem adolescents for many years also shows only a small
excess of injection drug use by males compared to females since 1978
(11 percent versus 9 percent) (Brunswick et al. 1986).
Whether due to biased sampling, location-specific patterns, or an age
cohort phenomenon, gender-based differences remain among the most
severely affected drug users (Johnson 1986). The 3 to 1 gender ratio in
IDU samples most likely is the result of social factors that have a
relatively greater impact on women than men at each stage in the course
of becoming addicted. This differential dropout likely results in an IDU
population with more severely affected women compared to men.
42
The older drug injection literature probably remains relevant for women
IDUs with chronic chemical dependency problems.
In most of the literature on chemically dependent women, initiation and
continued access to drugs occur mainly through male sexual partners
(Barnard 1993; Kane 1991; Rosenbaum 1981; Worth 1991), which
partially explain the disproportionate percentage of women IDUs who
have drug-using sexual partners compared to male IDUs. This is
reflected in the HIV risk literature, which shows that women primarily
share injection equipment with sexual partners or with close-knit groups
that include the sexual partner (Barnard 1993; Kane 1991). These studies
indicate that overall rates of needle-sharing may be greater for women
compared to men. Injection of drugs with strangers in shooting
gallery-type settings or with groups of casual acquaintances with frequent
shifts in membership are predominantly male behaviors.
In samples of individuals in drug treatment, women have a faster
transition from drug use initiation to addiction (Anglin et al. 1987; Hser
et al. 1987) and greater severity of addiction as compared to their male
counterparts (Anglin et al. 1987; Hser et al. 1987; Rosenbaum 1981).
The greater ease of obtaining steady drug money through female
prostitution, compared to the more risky and erratic economic activities
of IDU males, may be a central factor in these gender differences (Anglin
et al. 1987; Hser et al. 1987; Rosenbaum 1981).
However, even if the numbers of IDUs were to become equal for males
and females, reasons for initiation of drug use, access to needles, and
needle use settings are likely to retain strong gender differences as long as
gender-based social and economic differences remain among drug users.
Illicit drug distribution economies are male-dominated hierarchies,
resulting in the restriction of female roles in drug procurement (Naffine
1987). Chemically dependent women tend to have drug-using male
sexual partners to overcome gender barriers to drug access.
When women decide to seek treatment, however, a drug-using sexual
partner may obstruct access to treatment (Mondanaro 1989). For IDU
women, drug use by male sexual partners may be the central problem in
reduction of high-risk needle use or high-risk sex associated with both
injection and noninjection drug use [e.g., crack). The chief implication is
that HIV-risk-reduction programs for women who use drugs must
consider the continued central role of drug-using male sexual partners.
As indicated in Allen’s sampling of the recent HIV literature
43
(this volume), there is little attention paid to the gender-specific
contextual issues of injection drug use as outlined above.
In her review, O’Leary (this volume) points out that sexual risk reduction
programs for women require great sensitivity to problems of male
domination of sexual roles. Prevention programs instruct women to
require male partners to use condoms. Sexual partners may be longstanding relationships, casual partners, or potential long-term partners
who may or may not be willing to comply for different reasons,
depending on the nature of the partnership or the individual psychology.
Although the majority of males do not respond with hostility to a condom
request, some do.
Violent assaults on women by male sexual partners, documented for over
a century (Gordon 1988), are estimated to occur for as many as 1 in 4
women (Herman 1992). Yet this problem, which exists in all social
classes, seldom has been considered formally in risk reduction programs.
Women who use illicit psychoactive substances are especially likely to
have a childhood history of physical and sexual abuse as well as violent
adult sexual partnerships (Boyd 1993). Prevention programs should
counsel both men and women to recognize abuse potential and to develop
strategies to reduce abusive behavior as part of safer sex awareness and
practice.
Other points in the O’Leary review bear reiteration. The crisis-driven
lifestyle of many urban poor people often results in placing a low priority
on safer sex practices (Kalickman et al. 1992). While under studied,
stabilization of lifestyle is likely an important factor in risk reduction,
Most important is the recommendation that community-based
intervention programs can be used to fine-tune HIV prevention messages
to their own communities. The advantage of establishing communitybased prevention programs is that new strategies may be tested directly
using the images, language, and styles of a community as defined by the
community itself.
What is missing in the review on sex risk also is missing in the HIV
literature in general: focused research on sexual risk behavior and its
reduction in different social strata. One suspects there are many
commonalities among women in different social strata, but there are no
comparative studies. A potential effect of more broad-based research on
women’s strategies in practicing safer sex might be to dilute the excessive
44
attention paid to impoverished women, which sometimes inadvertently
results in their stigmatization as part of the problem in HIV infection
prevention.
As experience with prevention programs for women grows during the
second decade of the HIV epidemic, a new emphasis on gender-specific
behavioral issues has been emerging. The conditions that underlie risk
behavior, particularly those not amenable to simple informational
campaigns, are most important. Increasingly, the literature points to the
importance of female subordination in social, economic, and personal
spheres as factors in the maintenance of women’s drug injection risk
behavior, sexual risk behavior related to drug use, and sexual behavior in
general. Early HIV prevention programs did not anticipate sexism as a
barrier to risk reduction, but practical experience increasingly has brought
it to researchers’ attention.
REFERENCES
Anglin, M.; Hser, Y.; and McGlothlin, W. Sex differences in addict
careers: 2. Becoming addicted. Am J Drug Alcohol Abuse 13:59-71,
1987.
Barnard, M. Needle sharing in context: Patterns of sharing among men
and women injectors and HIV risks. Addict 88:805-812, 1993.
Boyd, C. The antecedents of women’s crack cocaine abuse: Family
substance abuse, sexual abuse, depression and illicit drug use. J Subst
Abuse Treat 10:433-438, 1993.
Brecher, E. Licit & Illicit Drugs. Boston: Little Brown & Co., 1972.
Brookmeyer, R., and Gail, M. AIDS Epidemiology: A Quantitative
Approach. New York: Oxford University Press, 1994.
Brown, V., and Weissman, G. Women and men injecting drug users: An
updated look at gender differences and risk factors. In: Brown, B., and
Beschner, G., eds. At Risk for AIDS: Injection Drug Users and their
Partners. Westport, CT: Greenwood Press, 1994.
Brunswick, A.; Aidala, A.; Dobkin, J.; Howard, J.; Titus, S., and
Banaszak-Holl, J. HIV-1 seroprevalence and risk behaviors in an
urban African-American community cohort. Am J Public Health
83:1390-1394, 1993.
45
Clayton, R.; Voss, H.; Robbins, C.; and Skinner, W. Gender differences
in drug use: An epidemiological perspective. In: Ray, B., and Braude,
M., eds. Women & Drugs: A New Era for Research. National Institute
on Drug Abuse Research Monograph No. 65. DHHS Pub. No.
(ADM)90-1447. Washington, DC: Supt. of Docs., U.S. Govt. Print.
Off., 1986.
Cohen, J.; Alexander, P.; and Wofsey, C. Prostitutes and AIDS. AIDS
Pub Policy 3:16-22, 1988.
Cottler, L.; Helzer, J.; and Tipp, J. Lifetime patterns of substance use
among general population subjects engaging in high risk sexual
behaviors: Implications for HIV risk. Am J Drug Alcohol Abuse
16:207-222, 1990.
European Study Group on Heterosexual Transmission. Comparison of
male to female and female to male transmission of HIV in 563 stable
couples. BMJ 304:809-813, 1992.
Gordon, L. Heroes of Their Own Lives. The Politics & History of Family
Violence. New York: Penguin, 1988.
Grund, J.; Synn-Stem, L.; Kaplan, C.; Adriaans, N.; and Drucker, E.
Drug use contexts and HIV consequences: The effect of drug policy
on patterns of everyday drug use in Rotterdam and the Bronx. Br J
Addiction 87:381-392, 1992.
Herman, J. Trauma & Recovery. New York: Basicbooks, 1992.
Hser, Y.; Anglin, M.; and Booth, M. Sex differences in addict careers: 3.
Addiction. Am J Drug Alcohol Abuse 13:231-251, 1987.
Ickovics, J., and Rodin, J. Women and AIDS in the United States.
Epidemiology, natural history and mediating mechanisms. Health
Psychol 11:1-16, 1992.
Johnson, E. Women’s health: Issues in mental health, alcoholism and
substance abuse. Public Health Rep 1:42-48, 1986.
Kaestner, E.; Frank, B.; Marel, R.; and Schmeidler, J. Substance use
among females in New York State: Catching up with the males. Adv
Alcohol Subst Abuse 5:29-49, 1986.
Kalickman, S.; Hunter, T.; and Kelly, J. Perceptions of AIDS
susceptibility among minority and non-minority women at risk for
HIV infection. J Consult Clin Psychol 60:725-732, 1992.
Kane, S. HIV, heroin and heterosexual relations. Soc Sci Med 32:10371050, 1991.
Kozel, N., and Adams, E. Epidemiology of drug abuse. Science 234:970976, 1986.
Longshore, D.; Hsieh, S.; and Anglin, D. Ethnic and gender differences
in drug users’ perceived need for treatment. Intl J Addict 28:539-558,
1993.
46
Miller, H.; Turner, C.; and Moses, L., eds. AIDS: The Second Decade.
Washington, DC: National Academy Press, 1990.
Mondanaro, J. Treatment of Women with Chemical Dependency
Problems. Lexington, MA: Lexington Press, 1989.
Naffine, N. Female Crime. The Construction of Women in Criminology.
Boston: Allen & Unwin, 1987.
Rosenbaum, M. Sex roles among deviants: The woman addict. Intl J
Addict 16:859-877, 1981.
Rosenbaum, M. Women on Heroin. New Brunswick, NJ: Rutgers
University Press, 1988.
Turner, C.; Miller, H.; and Moses, L., eds. AIDS: Sexual Behavior and
Intravenous Drug Use. Washington, DC: National Academy Press,
1989.
University of California (Berkeley and San Francisco). The Public Health
Impact of Needle Exchange Programs in the United States and
Abroad. Vol. 1. Atlanta: Centers for Disease Control and Prevention,
1993.
Vermund, S.; Shoen, A.; Galbraith, M.; Ebner, S.; and Fisher, R.
Transmission of the human immunodeficiency virus. In: Koff, W.;
Wong-Staal, F.; and Kennedy, R., eds. Annu Rev AIDS Res 1:81-135,
1990.
Worth, D. American women and polydrug abuse. In: Roth, P., ed.
Alcohol and Drugs are Women’s Issues. Vol. 1. Metuchen, NJ:
Scarecrow Press, 1991.
AUTHOR
Diana Hartel, Ph.D.
AIDS Research
Montefiore Hospital
111 East 210th Street
Bronx. NY 10467
47
Female Drug Abusers and the
Context of Their HIV
Transmission Risk Behaviors
Karen Allen
INTRODUCTION
The following dialog is part of a conversation that took place during an
initial interview of a 29-year-old, noninjection drug using (non-IDU),
African-American female at an outpatient drug treatment program.
Counselor: “So you know you are HIV positive?”
Client:
“Yes, I just found out and I know it has something to do with
the fact that I had sex a lot.”
Counselor: “How do you feel about being HIV positive?”
Client:
“Well, I take it a day at a time. I guess I deal with it okay,
but my main concern is this cocaine. I have got to stop using
drugs! That’s why I tried to commit suicide.”
There was little or no affect in the client’s face when discussing the
human immunodeficiency virus (HIV) diagnosis. Emotion was shown
only when discussing drug use. A similar reaction was shown in the
following dialog between a physician and a 41-year-old IDU AfricanAmerican female. The interview took place on a medical unit in a
hospital, Because the patient’s energy primarily was given to routines of
her life that were devastating, the impact of receiving a diagnosis of being
HIV positive was not perceived as being devastating.
Physician:
“I just wanted to let you know that you are HIV
positive.”
Nurse:
“You are not responding. Does this diagnosis upset you?”
48
Patient:
“Why should it? It’s just one more negative problem to add
to the others. Things are already bad, this is not going to
make it much worse!”
These dialogs put faces behind reported statistics. As of December 1992,
among women ages 15 to 24, 912 deaths from acquired
immunodeficiency syndrome (AIDS) had been reported to the Centers for
Disease Control and Prevention (CDC). Among those 25 to 34 years of
age, 7,346 deaths from AIDS had been reported; for women in age
groups 35 to 44, 45 to 54, and those 55 years or older, reported deaths
from AIDS were 5,924, 1,728, and 1,556, respectively. These numbers
represent increases in AIDS death rates across all age categories for
women (Centers for Disease Control and Prevention 1993a ).
CDC (1993b) has described a number of characteristics associated with
females with the highest incidence of AIDS. These characteristics
included being of African-American or Hispanic heritage, current or past
injection drug use, being a non-IDU, and engaging in high-risk sexual
activity with IDUs or other HIV-infected persons. In addition, women
with AIDS tend to be poor and of disadvantaged socioeconomic status
(Ickovics and Rodin 1992).
Injecting drugs and having sex without condoms are the two primary
behaviors that researchers have focused on and described as being high
risk in the transmission of HIV among female drug abusers. Much of
what is known about injecting drug use and unsafe sex as related to HIV
transmission among female drug users is descriptive epidemiology.
When designing and analyzing studies, many researchers report
information that describes how many women were exposed through
specified behaviors, their levels of knowledge about the disease, or
whether certain interventions changed the number of high-risk behaviors.
However, there is a lack of information regarding the underlying factors
or contextual issues that contribute to drug-using women’s exposing
themselves to HIV. Information regarding the emotional, psychological,
familial, and social-environmental contexts in which HIV transmission
behaviors take place would yield data that would give insight into the
motivations for continued behavior. As a result, the data would provide
direction for developing effective interventions for decreasing risk
behaviors.
Some researchers have access to subjects to ask for this type of data, but
their research questions do not focus on contextual issues. Some may
49
have the data but interpret it superficially or not at all. The purpose of
this chapter is to report on representative current research to determine
how well studies have addressed the underlying issues or context in
which women, especially African-American and Hispanic women,
continue to participate in high-risk behaviors.
THEORETICAL PERSPECTIVE
A number of theoretical perspectives may be applied to understanding the
context of HIV risk in a given population. While certainly not the only
applicable perspective, one useful framework that may be applied is the
Bowen Family Systems Theory (Bowen 1978). This theory is based on
the concept of differentiation of self. The characteristics of women most
frequently diagnosed with AIDS, as profiled by CDC, are consistent with
descriptions of persons with low to moderate levels of differentiation of
self in the Bowen Family Systems Theory.
Differentiation of self refers to the degree to which a person has a “solid
self” or solidly held principles by which they live. Papero (1990)
believes that to the degree that one can thoughtfully guide personal
behavior in accordance with well-defined principles in spite of intense
anxiety, that person displays a level of differentiation. In addition,
Papero (1990) states that the basic level of differentiation of self is
manifested in the degree to which an individual manages across life to
keep thinking and emotional systems separate, to retain choice between
behavior governed by thinking and behavior governed by emotional
reactivity, and to set a life course based on carefully thought out
principles and goals.
The basic level of differentiation of self for any person is believed to
develop and become fixed early in life and is usually similar to that of the
person’s parents. It is a product of the generations of the persons who
raised the individual (the nuclear family) in that each nuclear family
embodies the emotional processes and patterns of the generations that
have preceded it. A lifestyle of behavior based on thinking or behavior
based on emotional reactivity that a child learns from the nuclear family
or caregiver is replicated in future relationships (Bowen 1978; Papero
1990).
Bowen (1978) described those with low levels of self-differentiation as
living in a world where feelings and subjectivity are dominant over
50
objective reasoning. Persons with low levels of self-differentiation are
further described as dependent on the feelings of those around them,
expending so much of their energy in maintaining the relationship system
about them (loving, being loved, reacting against failure to get love, and
achieving comfort from anxiety) that there is no life energy left for other
things. Another characteristic is that major life decisions are made on the
basis of what “feels right” or simply on being comfortable. Persons with
low levels of self-differentiation not only do not use the “differentiated I”
in relationships with others, they are incapable of doing so. Further, they
inherit a major portion of the world’s serious health, financial, and social
problems.
Bowen described persons with moderate levels of differentiation of self
as being overtly emotionally dependent on others and in lifelong pursuit
of the “ideal close relationship.” If they fail to achieve it (a likely
probability), they may withdraw, become depressed, or pursue closeness
in another relationship. Those with moderate levels of self-differentiation
are recipients of a high percentage of human problems, including the full
range of physical illness, emotional illness, and social dysfunctions
marked by many levels of impulsive and irresponsible behavior. Finally,
they tend to be involved in the use and abuse of substances to relieve the
anxiety of the moment.
Bowen then described the level of self-differentiation that he believed
contributed to good health and should be promoted among clients and
their families. Those characterized as having healthy and good levels of
self-differentiation have an intellectual system that is sufficiently
developed to begin making a few decisions without the emotional system,
and they have developed a reasonable level of solid self on most of the
essential issues in life. They are able to successfully meet new situations
and, while aware of the importance of relationships in life, determine
their life course from within themselves rather than from what others
think. Such persons are able to cope successfully with a broad range of
human situations and are remarkably free from the full range of human
problems.
Bowen’s system is subject to the criticisms of other motivational or
psychodynamic theories (e.g., difficulties in operationalizing constructs,
observed outcomes not uniquely predicted by theory), but it does provide
a context for understanding behavior and a method of integrating
observed behavior with potential meaning of that behavior. It is
51
important to examine current studies to see if this or similar theoretical
bases are being used or could be used to guide and interpret extant
research.
REVIEW OF CURRENT RESEARCH
One example is the research currently being conducted at the University
of Maryland School of Nursing. This study is aimed at reducing the
high-risk behaviors of needle sharing and sex without condoms among
African-American female IDUs. It is a 3-year experimental design study
targeted for 200 subjects in a methadone maintenance program. The
study is called the Peer Counseling and Leadership Training (PCLT)
program, and it focuses on control and self-esteem as critical issues in
development of efficacious selves, sex and sexuality as related to AIDS,
and self-reported lifestyles that have implications for drug use and sexual
behaviors. The goal of this intervention is empowerment through
instillation of a sense of inner control and motivation based on increased
self-esteem and competency involving self, relationships with others, and
involvement with the community. Particular attention is paid to
sensitivity to and management of relationships with others and to
attitudes involving trust, competence, confidence, positive thinking,
personal direction, and self-identity (R.M. Harris, personal
communication, March 18, 1993).
This research does appear to be addressing some of the contextual issues
of HIV transmission among female drug abusers. The activities in the
intervention are consistent with the Bowen Family Systems Theory and
could be seen as helping women develop a solid self as well as appealing
to their intellect for increasing their ability to determine their behavior
based on thinking rather than emotional reactivity. It could be argued
that the effect of the intervention is to change a major factor in behavior,
low level of differentiation of self. However, because the researchers are
not viewing the subjects’ self as having developed as a result of
embodying emotional processes and patterns of previous generations, the
expectation is that intended changes resulting from the intervention will
happen soon. Bowen (1978) states that an increase in the level of
differentiation of self is gradual and takes a while to accomplish.
According to Harris, one initial finding from this study is that the partners
of female drug users lend the women’s equipment to other IDUs. Hence,
some female drug users are not necessarily sharing their needles and
52
equipment; instead, they are being shared for them. Having a low level
of differentiation of self would make the women unable to say “I want
this to stop.”
Another finding of Harris and colleagues’ study is that even though the
intervention is not targeting drug use, drug use is repeatedly mentioned
by the female IDUs. Although these women are on a methadone
program, they are using drugs such as alcohol, marijuana, cocaine, and
heroin. The investigators have learned that addressing HIV transmission
among this population without addressing the issue of drug abuse is not
approaching the problem from a comprehensive perspective.
Frischer and colleagues (1993) reported on their study conducted in
Scotland that aimed at theoretically modeling the behavior and attributes
of intravenous drug users (IVDUs) in order to examine relationships
between different variables as they affect HIV risk practices. In
face-to-face interviews, the investigators administered a questionnaire
that captured 10 variables: treatment, drug use, needle sharing, harm
reduction, prison, prostitution, income, travel, sex, and AIDS awareness.
This questionnaire was designed by an international World Health
Organization (WHO) working party and used in 11 cities worldwide.
The investigators used the questionnaire to collect data from 503
drug-using subjects and submit it to the linear structural relations program
(LISREL). Data were collected in an attempt to establish relationships
predictive of HIV risk, HIV harm reduction, or protection from risk.
Although the authors reported no specifics regarding the percentage of
female subjects, data were analyzed in order to determine if gender
contributed to the predictive power of equations addressing HIV risk.
In terms of HIV risk for female IVDUs, it was found that higher levels of
drug use were associated with higher levels of sexual activity and with
lower levels of precautions taken in relation to HIV. Injecting drug use
was related to sharing of injecting equipment; those who injected more
also shared more. Female subjects, particularly younger females, those
reporting higher levels of drug use, and those reporting higher levels of
sexual activity, reported higher levels of sharing injection equipment than
males.
Injection drug use was found to be a predictor of prostitution. High
levels of drug use were associated with prostitution and illegal behavior
for obtaining drugs, Another predictor of sharing injection equipment
53
found in this study was low socioeconomic status; those at the lower end
of the socioeconomic scale and those needing housing engaged in more
sharing.
Drug use was the best predictor of harm reduction in that those who
reported a low level of drug use engaged in high levels of harm reduction
and vice versa. In other words, a low incidence of drug use was a
protective factor against HIV transmission risk for the women in this
study.
AIDS awareness was related to harm reduction. Female drug users living
with their sexual partners and having stable housing and living
arrangements reported high levels of AIDS awareness and were less
likely to engage in risky sexual behaviors. Another protective factor
discovered was that of treatment for drug abuse in that it engendered safer
practices both with regard to sexual activity and drugs.
The lack of reporting the exact number of women subjects in the study
leaves an opening for challenging gender-specific results. Also, the
reporting could have been strengthened by providing more details on the
reliability and validity of the data collection instrument. However, the
study demonstrated the utility of applying statistical models and analyses
through LISREL to identify variables contributing to risk of HIV
transmission among female drug users. An important finding in this
study is the impact of drug abuse treatment on high-risk behaviors.
Researchers found that treatment engendered safer practices with regard
to sexual activity and drugs. More treatment was reported by older
female IVDUs and those cohabiting with sexual partners.
This study by Frischer and colleagues (1993) revealed behaviors and
attributes that contribute to high-risk behaviors for HIV transmission and
provided some information about the context in which such behaviors
occur.
Kang and De Leon (1993) reported on a study of correlates of drug
injection among 152 methadone outpatients; 39 percent (59) of the
subjects were female. Data were gathered on sociodemographic
background, drug use and treatment history, needle use behavior, sexual
behavior, medical status, criminal history, and psychiatric/psychological
status. Correlation coefficients were used to examine relationships
between variables, and discriminant analysis was used to explore
predictors of classification of high- or low-risk behaviors. High levels of
54
symptomatic distress, depression, and anxiety among the 59 female
IVDUs were reported.
In an analysis of data from the Australian National AIDS and Injecting
Drug User Study, Darke and colleagues (1992) examined the context of
benzodiazepine use among IVDUs and its impact on HIV transmission
risk behaviors. The study was conducted in Australia, but the research
questions investigated have implications for female IDUs in the United
States. Two questions are particularly relevant: “What is the prevalence
of benzodiazepine use in this population?” and “Does use of this drug
cause this population to engage in more risky injecting and sexual
behavior?”
Out of the 1,245 IVDUs, 331 were females. The interview schedule used
had been pilot tested and contained sections on demographics; drug
treatment and incarceration history; drug use including type of drug used,
frequency, and method of use; patterns of needle use and reuse; social
context of injecting drug use; sexual history; knowledge of and attitudes
towards HIV; and behavior change in response to HIV.
Results showed that female IVDUs were more likely than males to report
benzodiazepine use. In turn, benzodiazepine users reported younger
mean ages at first injecting use and at first regular injection drug use than
did the rest of the sample. The benzodiazepine users also reported
significantly more injections during the last month of use, more injections
of heroin, and greater frequency of injections with borrowed equipment.
The benzodiazepine users were more likely to have engaged in
prostitution.
Other information that became clear with this study is that many risk
behaviors took place in the context of multiple drug use. It seems likely
that unless use and abuse of mood-altering drugs among these females are
addressed comprehensively in conjunction with contributing contextual
issues, HIV transmission risk behaviors will stay the same or increase.
Forney and colleagues (1992) conducted interviews with 60 female drug
users: 25 in rural Georgia and 35 in the inner city of Miami. The women
were approached through street outreach workers and contacts with local
public health and drug treatment centers. The purpose of the study was to
determine the sociodemographics of women who exchange sex for crack
cocaine; the drug use and sexual histories of these women; knowledge,
beliefs, and attitudes about AIDS and HIV transmission among women
55
who trade sex for crack; and similarities and differences in sexual
practices of rural and urban female crack users.
Of the 35 subjects from Miami, 24 were African American, 7 were white,
3 were Hispanic, and 1 was Asian American. Of the 25 subjects from
Georgia, 21 were African American, and 4 were white. The primary
source of income was prostitution for 63 percent of the Miami sample
and for 68 percent of the Georgia sample. Except for alcohol and
marijuana use, crack cocaine was used more regularly than any other
drug, and respondents in both Miami and Georgia received their crack
through barter systems. In these systems, crack is purchased in a variety
of ways, including money from drug sales, bonuses for selling drugs, in
exchange for stolen goods, in exchange for other drugs, and as pay for
sex. Occasionally, the drugs were simply given to the women. For both
samples, the most common means of obtaining crack was in exchange for
sex (65 percent in Miami, 72 percent in Georgia).
Forney and colleagues reported that almost all of the respondents
expressed concern about AIDS and seemed to understand HIV
transmission routes; however, few had changed behaviors to reduce their
risk for HIV infection except initiating or increasing condom use. Most
continued to engage in high-risk practices, particularly unsafe sex with
numerous partners, in an effort to support their chronic crack habits.
This study provides useful details and speaks to the powerful influence of
physiological addiction in maintaining risky behaviors. Future studies
should attempt to address other factors in the lives of female drug users to
clarify what life events and circumstances necessitate their continuing to
risk their lives to obtain drugs. What contextual issues present in their
lives drive them to high-risk sexual practices to achieve and continue
their high?
In focus groups with 134 African-American and Hispanic IVDUs and
HIV-positive women, Kline and colleagues (1992) explored attitudes and
behaviors that surround sexual decisionmaking in minority communities.
Approximately 85 percent of the Hispanic participants were Puerto Rican,
and 9 percent were Cuban. Forty-five percent of the entire group of
Hispanic women were born outside of the United States. Moderators of
the focus groups were provided with discussion guides containing a set of
35 to 40 core questions that all groups were asked and 10 to 15 additional
questions specific to the category of risk.
56
The investigators found that these women did place a high value on
relationships with their men, but they were not submissive, dependent, or
self-sacrificing women as is often portrayed. The investigators found that
condom use was inconsistent in the women’s relationships. Based on
differential perceptions and judgments related to physical discomfort
when using the condoms, HIV status, length and intimacy of the
relationship, perception of the partner’s risk, and personal sense of
responsibility toward the partner, the women determined whether or not
condoms would be used. Although the women believed condom use was
important with casual partners, a majority of the women agreed that use
of condoms with primary partners was not important or desirable. The
reason for this belief was based on showing trust to partners and special
concerns that using condoms would interfere with the romance of the
moment. It seems that most of the issues affecting this major decision
were related to the partner’s feelings or to what was going on in the
relationship. The emphasis on the relationship seems, characteristic of
low to moderate levels of differentiation of self as described by Bowen
(1978). The women who were HIV positive felt it was important to use
condoms with someone they cared about, but they did not feel it
necessary to extend that loyalty and concern to casual partners.
These female subjects reported a number of strategies they used with
good results to insist partners use condoms, which shows that they had
some resources for negotiating safer sex. However, this negotiation
required much effort on their part.
In view of differences in drug usage between Puerto Ricans and other
Hispanic groups (e.g., Puerto Ricans have higher rates of illicit drug use
than Mexican Americans), results from this study should not be
generalized to other groups. Also, these women were from drug
treatment programs, which likely influenced their attitudes and responses.
For a group of women not in treatment, the responses could be different,
especially for that subset of crack-abusing women who work in crack
houses and go from partner to partner for the sake of getting the drug,
securing housing, and so forth.
Kline and colleagues’ data allowed them to examine IVDUs’ condom use
and sexual behaviors. They utilized a family and cultural perspective to
understand factors affecting women’s decisions surrounding condom use.
In addition, they gathered information regarding economics and family
responsibilities that influenced behavior, despite contrary stereotypes of
these women regarding submissive behavior. It is important to put these
57
responsibilities in perspective. Such assumption of responsibility is not
unique for these women. Economic responsibility for women is not new
and not solely attributable to current high levels of drug use and high
levels of unemployment among men. These women rose to the occasion
and responded much like the women in previous generations, pitching in
and helping economically-patterns they, particularly African-American
women, were taught as children.
Corby and colleagues (1991) conducted interviews with 137 female sex
partners of male IDUs. Sixty-seven percent of these women were current
users of noninjecting drugs. Forty-seven percent of the women used
marijuana, 45 percent used crack cocaine, 19 percent used cocaine, and
14 percent used tranquilizers. Fifty-eight percent of African-American
women and 52 percent of white women had smoked marijuana in the past
6 months, compared to 16 percent of Hispanic women, AfricanAmerican women (61.5 percent) more frequently reported crack cocaine
use than white women (30 percent) or Hispanic women (16 percent).
Also, African-American women (32 percent) were more likely to report
drinking 2 or more days per week than white women (4 percent) or
Hispanic women (3 percent).
These investigators reported that a larger proportion of African-American
women (42.3 percent) reported having multiple sex partners during the
past 6 months compared to whites (26 percent) or Hispanics (16 percent).
No women reported exchanging sex exclusively for drugs. All who had
traded sex for drugs had also done so for money. A larger proportion of
African-American women acknowledged having engaged in prostitution,
and all but two prostitutes (92.9 percent) reported using crack cocaine. In
addition, women who engaged in prostitution also were more likely to
report daily drinking.
Almost all of the women across all ethnic groups reported engaging in
unprotected vaginal intercourse. However, these investigators found no
relationship between condom use, ethnicity, and use of crack cocaine or
daily drinking. Among the 95 percent of women who reported no use or
inconsistent use of condoms, the most frequently stated reasons were
dislike of condoms by their partners. Once again, relationship issues and
the feelings of those around them are very important to these women.
The women, in a sense, were unable to get beyond the barrier of needing
to be loved and avoid an intense level of anxiety in the relationship. As
Bowen (1978) states, their feeling or emotional system keeps them from
making choices governed by thinking, and such processes and patterns
58
may be passed down from generation to generation. Corby and
colleagues state that these women were concerned as well as
knowledgeable about HIV transmission, and they were fully aware of
being at risk. From a contextual perspective, these investigators obtained
information that showed the effects of socioeconomic and lifestyle status
of these subjects on behavior.
Des Jarlais and colleagues (1991) examined crack use among 87 females
who had been admitted to a municipal New York City hospital during
1988 with a diagnosis of pelvic inflammatory disease (PID). Face-to-face
interviews were conducted covering the areas of AIDS risk behaviors and
HIV counseling and testing. The subjects were 86 percent African
American, 9 percent Hispanic, 3 percent white, and 1 percent other. The
investigators found that crack use was reported by 49 (56 percent) of the
sample and was strongly associated with unsafe sexual behavior. All
21 subjects who reported recently exchanging sex for money or drugs had
smoked crack, and 21 (84 percent) of the 25 females who had recently
had casual sex partners had also smoked crack. Twelve of the 87 subjects
were HIV positive, and 10 of these were among the 49 who were crack
users.
This study shows a strong connection between female crack cocaine users
and increased risk for HIV transmission. The investigators have raised
awareness to the fact that the very addictive nature of a particular drug is
associated with the increase in HIV among females.
Lewis and Watters (1991) examined the relationship of ethnicity and
gender to high-risk sexual behavior among 457 heterosexual IVDUs,
37 percent of whom were female. The subjects were interviewed on the
street and in clinic settings in San Francisco as part of a larger ongoing
study of risk factors for HIV transmission in IVDUs. Respondents were
given a 45-minute structured interview.
Results showed that 14 percent of African-American women and
30 percent of white women had 10 or more partners over the past year.
Among the subjects who were sexually active during the study, white
women had a median of three sexual partners, and African-American
women had a median of two partners. Fifty percent of the women
reported participating in sex-for-money exchanges. More AfricanAmerican women reported prostitution than white women, and, of the
83 women who exchanged sex for money, 54 used condoms.
59
Robles and colleagues (1990) examined social relations and roles of 160
Puerto Rican sexual partners of IVDUs in Puerto Rico as they related to
prevention of HIV transmission, particularly use of condoms. They used
the AIDS Initial Assessment instrument developed by the National
Institute on Drug Abuse along with participant observation notes and
transcribed conversations. Although the women were not IDUs, they
engaged in substantial use of noninjected substances. Alcohol was the
most commonly used substance of abuse and marijuana the second most
common. Involvement with noninjection cocaine also was high.
The reasons most cited by the women for not using condoms were that
their husbands did not like it (52.6 percent), and they themselves did not
like the condoms (35.5 percent). Also, 50.3 percent reported other
reasons related to their partners. The article indicated that these women
were cognizant of their risk of getting infected with AIDS, as 84 percent
reported that they had at least some chance of developing AIDS.
These investigators obtained additional contextual information. The
ethnographic data showed that these women did not belong to or
participate in organizational activities. However, they were actively
involved in primary group relationships with family, friends, neighbors,
and peers-particularly family. In applying Bowen’s (1978) ideas, these
women in Puerto Rico would be described as expending much of their
life energy for maintaining relationship systems around them, and there
appears to be little energy left for involvement in other things. There is
no sense of solid self as described by Bowen, to the extent that energy
can be reserved and used for making decisions and changing behaviors
that impact in a major way on their health.
While all of these studies reporting associations between demographics
and specific risk behaviors or between specific drugs and risk behaviors
provide findings that are factual and informative, contextual issues are
relatively neglected. An exception to this is investigation of the impact of
multiple drug use on sexual and injecting risk behaviors. If more
investigators had probed the circumstances or situations that led these
women to use crack as opposed to another type of drug, the impact of
community and family relations on their drug use, or other situationspecific variables, a fuller sense of the context of the high-risk practices
of women would have been developed. Of particular concern is that
sometimes findings specific to females, and perhaps even females within
specified ethnic groups, could have been reported but were not.
Examples of not reporting analyses specific to females and of neglecting
60
contextual issues despite an apparent availability of some relevant data
are not unique. The author wishes to emphasize that results specific to
female subjects are needed and important.
CONCLUSION
This sample of research literature on high-risk behaviors of female drug
abusers shows much diversity in attention to contextual issues. There
often is a lack of asking research questions, collecting data, and reporting
findings from a perspective that includes enough information to form a
picture of the contextual situation in which these women exist, use drugs,
and succumb to HIV through high-risk behavior. Studies should attempt
to not only present associations, but also to explain why there is
continued high-risk behavior despite the danger. Information on the
emotional, psychological, familial, and social-environmental contexts of
high-risk behaviors among female drug users is sorely needed.
Explanatory systems, such as level of differentiation of self, could help
make sense of such statements as “I would like to get treatment for my
drug use, but I know that being drug free is like a pie in the sky dream for
me.” Bowen’s system also can be used to explain a number of current
observations on both individual and community levels.
From a community perspective, because Bowen’s theory posits that level
of self-differentiation is passed down among generations and that
children will be similar to parents in level of self-differentiation, one
would not expect problems to be limited to individuals within a family.
In fact, AIDS is affecting entire families and generations in AfricanAmerican neighborhoods. An example is a 51-year-old IVDU woman
who is HIV positive; her 36-year-old son just recently died of AIDS; and
her sister recently died of AIDS. Another example of this phenomenon is
a woman who injects heroin and is HIV positive; her daughter is cocaine
addicted and is HIV positive from high-risk sexual behavior; and her
daughter’s child is HIV positive. All of this in one family! Families in
some African-American communities are accepting the disease and
resulting death as a routine part of life, and children are growing up in
environments where the norm is for many adults to have AIDS. The
children are learning dysfunctional patterns (using drugs to deal with
issues, make life comfortable, and relieve the anxiety of relationships and
participation in high-risk behaviors) that may be passed on to other
generations.
61
The lack of rational thought separated from emotions described by
Bowen may be reflected in the following observation. When some
patients find out they have AIDS, they experience it as something
someone did to them. As a result, they are not motivated to protect others
through condom use or not sharing needles.
A low level of self-differentiation keeps people from advocating for
themselves. When diagnosed with being HIV positive, some patients in
hospitals hear about it from medical students or interns whom they do not
identify as their physician. Sometimes, the diagnosis is given when the
patient is being discharged, and no thorough information is given.
Another phenomenon is the attitudes of medical and nursing staff toward
the patients or clients. Staff may feel as if patients “do this to
themselves” and may make comments like “Why don’t you go to another
hospital or medical facility?” This response from staff contributes to poor
self-esteem and perpetuates a sense of hopelessness. Patients may go
back to using drugs to make themselves more comfortable, rather than
feeling the anger, hurt, and disillusionment. Persons with high levels of
self-differentiation would be able to negotiate the medical care system to
avoid or at least confront such insensitivities.
Finally, a consistent theme in research literature on HIV risk in women is
that treatment for drug abuse can make an impact on high-risk behaviors
for HIV transmission among female drug users. Therefore, drug
treatment for women is a crucial intervention to provide the context and
skills whereby women can lower their risk of HIV infection.
REFERENCES
Bowen, M. Family Therapy in Clinical Practice. New York: Jason
Aronson, 1978.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report: Year End Edition. Atlanta: Centers for Disease Control and
Prevention, 1993a.
Centers for Disease Control and Prevention. U.S. AIDS cases reported
through March 1993. HIV/AIDS Surveillance Report 5(1):3-19.
Atlanta: Centers for Disease Control and Prevention, 1993b .
Corby, N.; Wolitski, R.; Thornton-Johnson, S.; and Tanner, N. AIDS
knowledge, perception of risk, and behaviors among female sex
partners of injection drug users. AIDS Educ Prev 3(4):353-366, 1991.
62
Darke, S.; Hall, W.; Ross, M.; and Wodak, A. Benzodiazepine use and
HIV risk taking behavior among injecting drug users. Drug Alcohol
Depend 31:31-36, 1992.
Des Jarlais, D.; Abdul-Quader, A.; Minkoff, H.; and Tross, S. Crack use
and multiple AIDS risk behaviors. J Acquir Immune Defic Syndr
4(4):446-447, 1991.
Forney, M.; Inciardi, J.; and Lockwood, D. Exchanging sex for crackcocaine: A comparison of women from rural and urban communities.
J Community Health 17(2):73-85, 1992.
Frischer, M.; Haw, S.; Bloor, M.; Goldberg, D.; Green, S.; McKeganey,
N.; and Covell, R. Modeling the behavior and attributes of injecting
drug users: A new approach to identifying HIV risk practices. Int
J Addict 28(2):129-152, 1993.
Ickovics, J., and Rodin, J. Women and AIDS in the United States:
Epidemiology, natural history, and mediating mechanisms. Health
Psychol 11(1):1-16, 1992.
Kang, S., and De Leon, G. Correlates of drug injection behaviors among
methadone outpatients. Am J Drug Alcohol Abuse 19(1):107-118,
1993.
Kline, A.; Kline, E.; and Oken, E. Minority women and sexual choice in
the age of AIDS. Soc Sci Med 34(4):447-457, 1992.
Lewis, D., and Watters, J. Sexual risk behavior among heterosexual
intravenous drug users: Ethnic and gender variations. AIDS 5:77-83,
1991.
Papero, D. Bowen Family Systems Theory. Boston: Allyn and Bacon,
1990.
Robles, R.; Colon, H.; Gonzalez, A.; and Mateos, R. Social relations and
empowerment of sexual partners of IV drug users. P R Health Sci J
9(1):99-104, 1990.
AUTHOR
Karen Allen, Ph.D., R.N., CARN
University of Maryland School of Nursing
Department of Psychiatric, Community Health, and Adult Primary Care
Community Addictions Nursing Program
734 Parsons Hall
622 West Lombard Street
Baltimore, MD 21201-1579
63
Factors Associated With Sexual
Risk of AIDS in Women
Ann O’Leary
INTRODUCTION
Women are increasingly represented among U.S. acquired
immunodeficiency syndrome (AIDS) cases. As of December 1992,
almost 28,000 cases of AIDS in women had been reported to the Centers
for Disease Control and Prevention (CDC) (Centers for Disease Control
and Prevention 1993b). However, nearly half of these (44 percent) were
reported during 1991 and 1992 alone, reflecting a dramatic increase in the
rate of female infection (Centers for Disease Control and Prevention
1993a). The number of cases of AIDS in women has been projected to
become between 55,000 and 75,000 by the end of 1994 (Centers for
Disease Control and Prevention, 1992). As the number of women with
human immunodeficiency virus (HIV) infection and AIDS has grown,
numerous considerations specific to women have arisen. These concern
differences in disease course and treatment, issues in the prevention of
infection, and effects of stigmatization (Ickovics and Rodin 1992;
O’Leary et al. 1993b). This chapter describes the extant research
concerning the structural and personal factors related to sexual risk for
HIV infection in women, with special reference to female partners of
drug users. Following is a brief description of prevention programs that
have been evaluated, a discussion of female-centered prevention
strategies, and recommendations for future research and intervention.
WOMEN AND AIDS: WHO IS AT RISK?
Within the United States, women are a rapidly growing segment of the
population infected with HIV, the cause of AIDS. Their representation
among persons with AIDS, expressed as a percentage of the total AIDS
cases reported, increased from about 3 percent in 1981, when the first
cases were reported to the CDC, to 12 percent in 1990 (Ellerbrock et al.
1991). AIDS is now the leading cause of death among African-American
women of reproductive age (ages 15 to 44) in New York and New Jersey
(Chu et al. 1990). Heterosexual transmission, which appears to be more
efficient when the infected partner is male (Padian et al. 1991), also has
64
become a prominent source of female infection. In the United States, the
proportion of women who have contracted AIDS via heterosexual
activity has increased from 11 percent in 1984 to 34 percent in 1990, and
this proportion is increasing over time (Ellerbock et al. 1991).
Worldwide, it is estimated that 71 percent of HIV transmission occurred
through heterosexual contact (Ehrhardt 1992). Some fear that the United
States may ultimately become a “Pattern II” country with 50 percent of
those affected being women, as is the case now in some African countries
(Haverkos and Edelman 1988).
At the present time, AIDS in the United States affects predominantly the
inner-city poor, the majority of whom are people of color. AfricanAmerican and Hispanic women are disproportionately affected. While
whites account for only 27 percent of AIDS cases in women, about 50
percent of women with AIDS are African American, and about 20
percent are Hispanic (Chu et al. 1990). About 73 percent of mothers with
HIV-infected children receive public assistance (Shayne and Kaplan
1991). Commercial sex workers also are greatly at risk for HIV infection
(Centers for Disease Control and Prevention 1987; Cohen et al. 1988).
Most women with AIDS who have been infected heterosexually
(although possibly no longer most of those with early-stage HIV
infection) were infected by injecting drug-using (IDU) partners (Fordyce
et al. 1991). This is not surprising given the high seroprevalence rates of
IDUs (Centers for Disease Control and Prevention 1989), their low rates
of condom use (Lewis and Watters 1991; Weissman et al. 1991), and the
fact that sexual partners may not know of their partners’ drug use (Kane
1991).
In the absence of a cure or preventive vaccination against HIV infection,
behavior change remains the primary available means for combating this
epidemic. The most widely accepted method for preventing sexual
transmission of HIV in the United States, apart from abstinence, is the use
of latex condoms.
The following is a review of correlates of sexual risk that have been
identified among disenfranchised women. It is not restricted to research
on female partners of IDUs, although papers focusing specifically on this
population are included. A variety of factors associated with sexual risk
in women have been identified. These include factors related to poverty,
partner/relationship issues, cultural factors, beliefs and attitudes regarding
condoms, and personal characteristics. Each of these will be discussed in
turn.
65
Factors Associated With Poverty
A number of aspects of poverty affect the ability of women to protect
themselves. These include enhanced risk for drug addiction, high
prevalence of sexually transmitted diseases (STDs), and competing life
concerns. Poverty also contributes to economic dependence of some
women on their partners (Airhihenbuwa et al. 1992); relationship factors
contributing to risk are the topic of the next section.
Poverty and the pain that accompanies it contribute to the use of
psychotropic drugs, including heroin, crack and other forms of cocaine,
and alcohol. Injection of drugs created the initial vector for HIV
infection in the inner city. Further, use of psychotropic substances
militates against people’s ability to practice safer sexual behaviors (Leigh
and Stall 1993). One reason for this relationship may be the clouding of
judgment that occurs when people are under the influence of such
substances; another is that addicted women often exchange sex for the
drug that they need (Corby et al. 1991). Female partners of IDUs report
high levels of use of illegal drugs and alcohol (Corby et al. 1991;
Weissman et al. 1991) and concomitantly high levels of sexual risk
(Brown and Weissman, in press). However, it should be noted that one
study found that exchanging sex for drugs was not associated with higher
sexual risk for women (Brown and Weissman, in press).
Another aspect of inner-city life that contributes to AIDS risk for its
female inhabitants is the high prevalence of other STDs that are endemic
to these areas. Diseases such as syphilis, gonorrhea, chlamydia, pelvic
inflammatory disease, human papillomavirus, and chancroid are severe
epidemics in the inner city (Aral and Holmes 1991; Hatcher et al. 1990;
Hayes 1987; Hofferth and Hayes 1987; Rothenberg 1991). Populations
with higher rates of STDs are at increased risk for HIV infection. The
risks of STDs and HIV are similar because both are caused by
unprotected sexual activity. In addition, the presence of an STD and
associated lesions may facilitate transmission of HIV.
It has been posited that the burdened, crisis-driven lifestyle frequently
associated with poverty may contribute to AIDS risk by displacing HIV
infection among the list of concerns on peoples’ minds (Amaro 1988;
Guarnaccia et al. 1989; Marin and Marin 1991; Nyamathi and Vasquez
1989). Indeed, results from a recent study support this notion in that the
female respondents listed AIDS below several other life problems in
seriousness (Kalichman et al. 1992). In a study of IDUs, residential
66
instability was found to predict sexual risk for HIV transmission (Brown
and Weissman, in press).
Relationship Factors
Perhaps the most critical feature of many behavior changes that might
prevent the sexual transmission of HIV is that they require the
cooperation of another person, namely, the woman’s sex partner.
Unfortunately, some of the women most at risk are dependent upon their
male partners for economic security (Kane 1991). Some women also fear
that they will lose desired partners if they insist on condom use, since this
may be interpreted as an indication that the woman has been unfaithful to
her partner, believes that her partner has been unfaithful to her, or has
discovered that she is infected with HIV or other pathogenic agents (De
Bruyn 1992; Fullilove et al. 1990; Kenen and Armstrong 1992). Women
also may desire to become pregnant, militating against the use of
condoms or any other contraceptive (De Bruyn 1992; Kline et al. 1992).
Of particular concern is the plight of the woman experiencing domestic
violence or coerced sex, for whom the recommendation of condom use
may be impossible and even dangerous to execute. A recent study of the
sex partners of IDUs revealed alarmingly high rates of domestic violence
(Weissman 1991). While focus group members generally do not reveal
abuse in their own lives, they do comment that other women experience
it, as illustrated by one respondent’s statement, “It probably happens... in
90 percent of marriages and relationships where women live alone”
(Fullilove et al. 1990). Further, while the female condom (discussed
below) will give women some control over the safety of their encounters,
many of the problems that women experience in their attempts to be safe
will persist, since the presence of the female condom is quite obvious to
her partner.
Some investigators have suggested the potential utility of alternative
negotiation techniques, which involve giving reasons for condom use that
do not have to do with AIDS per se (Magana and Magana 1992; O’Leary
1991) and that therefore obviate many potential partner objections. The
woman might claim, for example, that her health care providers have
recommended condoms for contraception because she is allergic to semen
or to quell persistent yeast infections. Recommending to women that
they lie to their partners is a controversial strategy that carries obvious
risks (e.g., in abusive situations where they are most likely to be tried),
and may be seen as less healthy or desirable than other solutions such as
67
encouraging such women to leave their partners. However, in cases when
other options are not realistic or might take considerable time, careful
exploration of such possibilities would seem justified for some women.
Cultural Factors
As noted above, women of African and Hispanic descent are
disproportionately represented among women with AIDS. Thus, cultural
factors that distinguish these cultures are seen by many as being highly
relevant to AIDS prevention efforts in the communities most at risk
(Airhihenbuwa et al. 1992). While such considerations may be useful in
designing interventions aimed at these groups as a whole (e.g., in mass
media or marketing applications), it should not be assumed that that they
apply to every individual member of an ethnic group.
African-American Women. Only very limited data are available on the
prevalence and correlates of most sexual behaviors other than
contraceptive use among African-American women (Reinisch et al. 1988;
Turner et al. 1989). While African and African-American cultures are
generally regarded as very “sex-positive” and comfortable with
discussions of sexuality, it has been observed that poor African-American
women may exhibit less varied sexual repertories than white women
(Wyatt et al. 1988a, 1988b). They may equate intimacy specifically with
vaginal sex or penetration, an emphasis that may be related to an
orientation to procreation rather than erotic sex, possibly related in turn to
conservative religious values (Houston-Hamilton 1988). Caution also has
been urged in providing AIDS education that may be at odds with
established values. For example, the suggestion that women should
explore new forms of sexual behavior (e.g., nonpenetrative sexual
techniques) as a means of prevention of HIV transmission may be
received ambivalently by women subject to victim-blaming because of
stereotypes about their sexual behavior (Hine and Wittenstein 1989).
Another aspect of African-American culture relevant to AIDS prevention
efforts in that community are issues of trust in health care professionals
and mainstream health services (Airhihenbuwa et al. 1992; Thomas and
Quinn 1991). Distrust also may contribute to delay in seeking services.
It has been argued that African-American women’s resistance to admit
sexual risk is driven by fear of racial and sexual backlash (Hine and
Wittenstein 1989). There is fear that in the course of addressing the
AIDS problem, the African-American community will open itself to
additional discrimination by furthering the stereotypical association of
68
African Americans, disease, and immoral behavior. The ethnicity and
sensitivity of the deliverer of the message that AIDS is an issue of
importance to the African-American community is likely to be a critical
factor in its acceptance.
Yet another feature of some elements of African-American culture that
may impact response to AIDS messages is adhesion to folk and spiritual
beliefs regarding disease etiology and treatment (Landrine and Klonoff
1994). To the extent that an individual holds these beliefs, one might
expect low receptiveness to biomedical-based messages regarding
transmission as well as treatment. A study of low-income African
Americans living in the Los Angeles area revealed prevalent beliefs in
spiritual forces, witchcraft, and evil influences as causative agents for
HIV and AIDS (Flaskerud and Rush 1989).
Hispanic Women. Like African-American culture, Hispanic culture is
distinguished in numerous ways of relevance for AIDS. While it should
be noted that there are numerous Hispanic cultures corresponding to
different geographical regions and countries, a number of characteristics
are shared by most or all (Marin and Marin 1991). The literature has
characterized Hispanic women as having conservative religious beliefs
and gender roles in which they are willing to sacrifice themselves for
their children and are passive and subordinate to men (Canino 1982;
Rivera 1985). These findings must be interpreted with caution because
they are not derived from controlled studies and often do not account for
demographic and socioeconomic variables. One strong cultural value
that has been established is “familismo,” or an emphasis on family as the
primary social unit and source of support, with less emphasis on
individualistic achievement than in the dominant culture (Marin and
Marin 1991). Children and fertility are highly valued; motherhood brings
high status within the family structure; and desire for fertility has been
shown to be a barrier to safer sex recommendations for Hispanics (Marin,
in press; Marin and Marin 1991). However, it also has been suggested
that this family orientation can be used to enhance motivation for
behavior change to reduce the risk of AIDS, for example, by appealing to
the health of unborn children (Marin, in press).
Many Hispanic women are at risk for HIV infection due to the behavior
of their husbands (Magana and Magana 1992; Marin, in press). Injection
drug use is prevalent among Hispanics, particularly those living in the
Northeast United States and Puerto Rico (Magana and Magana 1992).
Hispanic cultural norms promote virginity and monogamous marriage for
69
Hispanic women but extramarital affairs and frequent sex with other men
for men (Magana and Magana 1992). For example, it has been estimated
that approximately 30 percent of Hispanic men have engaged in sex with
other men (Carrier 1985). An interesting but unanswered question
concerns the degree to which women know about their husbands’
behavior or admit it to themselves; however, it is clear that discussions on
this topic are very difficult for many women, making it difficult for them
to protect themselves.
It is widely believed that Hispanic women have a strong religious
orientation and that Roman Catholic doctrine has tremendous influence
upon their life decisions (Marin and Mat-in 1991). It often is assumed
that they will follow papal decrees regarding birth control and condom
use. However, contemporary studies indicate that Hispanic women use
contraception, and being Catholic is not necessarily associated with
traditional reproductive behavior (Marin and Marin 1991). In addition,
folk and spiritualistic beliefs regarding health are held by many
Hispanics. A recent study of HIV-infected Hispanics living in New
Jersey identified prevalent beliefs in folk and spiritualistic causes and
treatments for AIDS (Suarez et al. 1993).
Thus, data exist both to support the existence of cultural factors in HIV
transmission as well as to dispel myths about cultural stereotypes. It is
very important for anyone working in the area of AIDS prevention both
to attend to cultural factors that may moderate intervention effectiveness,
but also to question preexisting assumptions about the cultural
characteristics of the group with which they are working.
Male condoms have been a primary focus of most AIDS education and
behavior change efforts in the United States. Thus, cultural factors
influencing attitudes toward condoms may be especially pertinent for
AIDS prevention. A number of investigators have studied beliefs and
attitudes of at-risk women toward male condoms; the next section is
devoted to a review of this research.
Condom Beliefs and Attitudes
A number of investigators working within the frameworks of the theory
of reasoned action (Fishbein and Middlestadt 1989; Fisher and Fisher
1992; Jemmott and Jemmott 1991) and social cognitive theory (Bandura
1992; O’Leary 1992) have identified beliefs and attitudes related to safer
sex in women. Perceptions of risk have been associated with greater
70
safety in some studies (Corby and Wolitski 1992; Lo Conte et al. 1993);
it is important that questions be asked in terms of potential risk given
unsafe behavior, or subjects will base responses on current behavior,
obscuring the relationship. Self-efficacy beliefs, a central component of
social cognitive theory (Bandura 1986), are another factor that has been
examined in several studies. Self-efficacy refers to a person’s belief that
an individual can successfully execute skilled behavior necessary to
achieve a desired outcome (in the present case, safer sex). Self-efficacy
operates by affecting people’s behavior choices, the amount of effort they
will exert in performing the behavior, and the degree of persistence they
will display in the face of difficulty. Self-efficacy beliefs regarding one’s
safer sex ability are relevant in three domains (O’Leary 1992): technical
use of condoms and nonpenetrative sexual techniques; negotiation with
partners; and self-control as related to use of alcohol, other drugs, and
sexual arousal. Negotiation self-efficacy has been associated with higher
rates of condom use by commercial sex workers (Corby and Wolitski
1992), other disadvantaged women (Lo Conte et al. 1993; O’Leary et al.
1992), and adolescents (Catania et al. 1989; Jemmott et al. 1992). One
recent study found that self-efficacy perceptions were more closely linked
to behavioral safety for women than for men (Lo Conte et al. 1993).
Self-efficacy enhancement takes place through systematic skill-building,
through modeling (demonstrating the behavior), and practice of
increasingly difficult skills (Bandura 1986). A recent study of AfricanAmerican female adolescents demonstrated that increases in self-efficacy
were associated with increased behavioral intention to practice safer sex
(Jemmott and Jemmott 1992).
Expected outcomes of condom use also are important influences on
behavior. One belief is that condoms interfere with sexual pleasure
(Jemmott and Jemmott 1991). Another belief is that safer practices will
win the social approval of peers (Jemmott et al. 1992; Valdiserri et al.
1989). These beliefs, attitudes, and skills are important elements to
address in prevention interventions, and they lend themselves readily to
influence. Skill and self-efficacy–building techniques are specified
particularly well in social cognitive theory (Bandura 1986).
Personal Characteristics
Relatively little research has been conducted to identify personal
characteristics and resources that may condition women’s ability to
respond to AIDS messages. However, one recent study found that
disenfranchised women with higher self-esteem and a greater sense of
71
coherence—the sense that life is controllable and meaningful—were
practicing more protective behavior than their low–self-esteem and lowcoherence counterparts (Nyamathi 1991). The implications of these
findings for AIDS intervention are unclear, as changes in these global
personality traits may be difficult to effect, at least in brief intervention.
EVALUATED AIDS PREVENTION PROGRAMS FOR WOMEN
A small number of published studies have evaluated prevention programs
for women. The theoretical approaches that have been applied in the area
of AIDS prevention include social cognitive theory and the theory of
reasoned action. Such programs typically combine risk education with
individualized skill-building for condom use and negotiation. One study
with a pretest-posttest design demonstrated increases in condom use
intentions among African-American adolescent women following a single
session intervention (Jemmott and Jemmott 1992); another with a similar
design increased intended behavior change among low-income adult
women attending Women, Infants, and Children (WIC) clinics (Flaskerud
and Nyamathi 1990).
Two studies have used randomized experimental designs to demonstrate
intervention effects. One, by Schilling and colleagues (1991),
randomized women on methadone maintenance to receive a five-session
skill-building intervention or to a control condition. Significant increases
in condom use and in perceptions of control over AIDS risk were
obtained, although the number of sexual partners did not change.
Further, condom use remained significantly more frequent among
intervention participants at a 15-month followup (El-Bassel and Schilling
1992). Another study in which a similar intervention was delivered to
inner-city women attending a primary care clinic also demonstrated
substantial reductions in risk behavior at a 3-month followup (Kelly et al.,
in press). These studies illustrate that inner-city women can reduce their
behavioral risk for HIV infection if properly trained in negotiation and
condom use skills.
AIDS PREVENTION TECHNOLOGY
The female condom, which is undergoing safety and efficacy evaluation
in the United States, may confer superior protection both because it is
made of polyurethane, a material tougher than latex, and possibly also
72
because it covers more tissue area (Nowak 1993). As mentioned above,
the presence of the female condom is obvious to the male; thus, issues of
trust remain potentially problematic.
An issue that has become the subject of much debate but woefully little
research is the recommendation that women use spermicide alone when
the protection afforded by a condom is not a possible option (Cates et al.
1992; Rosenberg and Gollub 1992; Stein 1992). These would be
situations in which women perceive themselves as helpless to prevent
unprotected encounters, as in cases of rape and domestic violence. (Of
course, these events would have to be predictable for the woman to be
able to insert the spermicide). Indeed, many service providers in the
community are recommending this measure, which obviously does not
depend upon male cooperation. Arguments in favor of this
recommendation include the in vitro virucidal effects of nonoxynol-9
(North 1990), and the speculation that the lubrication afforded by
spermicidal gels may prevent tearing of the skin during episodes of
coerced sex. However, because frequent use of nonoxynol-9 has been
shown to disrupt the outer lining of the cervix and vagina in some women
(Niruthisard et al. 1991), it is possible that transmission may be
facilitated in some cases by this strategy. Indeed, in an experimental
study in which commercial sex workers in Nairobi were randomized to
receive either a nonoxynol-9–containing sponge or placebo cream, no
effect on HIV seroconversion was observed (Kreiss et al. 1992). This
study has been criticized as flawed in design and nongeneralizable in
results (Rosenberg and Gollub 1992); however, it suggests the need for
conservativism in recommending this prevention strategy. Furthermore,
some intervention specialists express concern about giving complex or
mixed messages, particularly given the extreme paucity of data regarding
the effectiveness of this strategy.
GENERAL ISSUES IN PREVENTION OF AIDS IN AT-RISK
WOMEN
Prevention efforts may not reach the women most in need, as many innercity women are isolated and not attached to community agencies.
Individualized outreach by trusted others is the most effective way to
recruit disenfranchised women, particularly Hispanic women. When
possible, integration of AIDS prevention intervention with ongoing
trusted services already part of women’s lives is likely to enhance
maintenance of change. Emotional support for the client on the part of
73
the change agent may be a crucial factor in promoting change (National
Institute of Mental Health 1992). Interventions must take realities of
women’s lives into consideration, for example, by providing
transportation and childcare. Interventions for couples, or at least ones
that involve the partner to some degree, must be developed. There is a
need for innovative prevention programs and programs that capitalize on
cultural strengths. Intervention will be more effective if provided within
community sources such as community-based organizations,
neighborhood groups, or tenant groups. Formal or informal community
leaders should be involved in prevention efforts. Alternative education
strategies such as “telenovelas,” street theater, and Spanish-language
radio and television programs also should be considered.
The cohesion and mutual caring that exist within the families and
communities of African-American and Hispanic individuals can be used
to advantage in AIDS prevention. Mays and Cochran (1988) have
suggested that prevention approaches for ethnic minority individuals
should focus on the individual as a responsible member of a familial or
social network. Recent work by Kelly and colleagues (1991, 1992) in
gay communities has demonstrated the promise of diffusion of innovation
models for AIDS prevention; this approach seems well-suited to the
socially cohesive Hispanic community as well.
There is a crucial need for education and services related to violence and
domestic abuse as well as creation of alternatives to remaining in abusive
relationships. AIDS prevention efforts must address issues of abuse if
they are to be effective and maintain the safety of the women they are
trying to reach.
There is also a need to document behavioral changes other than condom
use that women may be using in the belief, correct or incorrect, that they
may decrease risk. Behaviors involving partner selection, use of
withdrawal, antibody testing, and sexual history-taking (O’Leary et al.
1992) may be tried in response to AIDS prevention messages; these
behaviors need to be both documented and studied for effectiveness. A
great need also exists for research on woman-controlled technologies for
AIDS prevention.
74
ACKNOWLEDGMENTS
This research was supported by the National Institute on Drug Abuse.
Portions of the research reported here were facilitated by grants
1-RO1-MH45238 and 5-U01-MH48013 from the National Institute of
Mental Health.
REFERENCES
Airhihenbuwa, C.O.; DiClemente, R.J.; Wingood, G.M.; and Lowe, A.
HIV/AIDS education and prevention among African-Americans: A
focus on culture. AIDS Educ Prev 4:267-276, 1992.
Amaro, H. Hispanic women and AIDS. Psychol of Women Q, 12:423443, 1988.
Aral, S.O., and Holmes, K.K. Sexually transmitted diseases in the AIDS
era. Sci Am 264:62-69, 1991.
Bandura, A. Social Foundations of Thought and Action: A Social
Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall, 1986.
Bandura, A. A social cognitive approach to the exercise of control over
AIDS infection. In: DiClemente, R., ed. Adolescents and AIDS: A
Generation in Jeopardy. Beverly Hills, CA: Sage, 1992.
Brown, V.B., and Weissman, G. Women and men injection drug users:
An updated look at gender differences and risk factors. In: Brown, B.,
and Beschner, G.M., eds. Handbook on Risk for AIDS: Injection Drug
Users and Their Partners. Westport, CT: Greenwood Press, 1993.
pp. 173-194.
Canino, G. The Hispanic woman: Sociocultural influences on diagnosis
and treatment. In: Becerra, R.; Karno, M; and Escobar, J., eds. Mental
Health of Hispanic Americans. New York: Grune and Stratton, 1982.
Carrier, J.M. Mexican male bisexuality. In: Klein, F., and Wolf, T., eds.
Bisexualities: Theory and Research. New York: Haworth, 1985.
pp. 359-375.
Catania, J.A.; Dolcini, M.M.; Coates, T.J.; Kegeles, S.M.; Greenblatt,
R.M.; Puckett, S.; Corman, M.; and Miller, J. Predictors of condom
use and multiple partnered sex among sexually-active adolescent
women: Implications for AIDS-related health interventions. J Sex Res
26:514-524, 1989.
Cates, W.; Stewart, F.H.; and Trussel, J. Commentary: The quest for
women’s prophylactic methods-Hopes versus science. Am. J Pub
Health 82:1479-1482, 1992.
75
Centers for Disease Control and Prevention. Antibody to human
immunodeficiency virus in female prostitutes. MMWR Morb Mortal
Wkly Rep 36:157-161, 1987.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Atlanta: Centers for Disease Control and Prevention, 1989.
Centers for Disease Control and Prevention. Projections of the number of
persons diagnosed with AIDS and the number of immunosuppressed
HIV-infected persons: United States, 1992-1994. MMWR Morb
Mortal Wkly Rep 41:1-11, 1992.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Atlanta: Centers for Disease Control and Prevention, 1993a .
Centers for Disease Control and Prevention. Update: Acquired Immune
Deficiency Syndrome: United States, 1992. MMWR Morb Mortal
Wkly Rep 42:547-557, 1993b.
Chu, S.Y.; Buehler, J.W.; and Berkelman, R.L. Impact of the human
immunodeficiency virus epidemic on mortality in women of
reproductive age, United States. JAMA 264:225-229, 1990.
Cohen, J.B.; Alexander, P.; and Wofsy, C. Prostitutes and AIDS: Public
policy issues. AIDS Pub Policy J 3:16-22, 1988.
Corby, N.H., and Wolitski, R.J. “Relationship Between Street Sex
Workers’ Attitudes and Condom Use by Type of Partner.” Paper
presented at the Eighth International Conference on AIDS/Third STD
World Congress, Amsterdam, July 16-21, 1992.
Corby, N.H.; Wolitski, R.J.; Thornton-Johnson, S.; and Tanner, W.M.
AIDS knowledge, perception of risk, and behaviors among female sex
partners of injection drug users. AIDS Educ Prev 3:353-366, 1991.
De Bruyn, M. Women and AIDS in developing countries. Soc Sci Med
34:249-262, 1992.
Ehrhardt, A.A. Trends in sexual behavior and the HIV pandemic. Am J
Pub Health 82:1459-1461, 1992.
El-Bassel, N., and Schilling, R.F. Fifteen-month followup of women
methadone patients taught skills to reduce heterosexual HIV
transmission. Pub Health Rep 107:500-504, 1992.
Ellerbrock, T.V.; Bush, T.J.; Chamberland, M.E.; and Oxtoby, M.J.
Epidemiology of women with AIDS in the United States, 1981-1990.
JAMA 265:2971-2975, 1991.
Fishbein, M., and Middlestadt, S. Using the theory of reasoned action as
a framework for understanding and changing AIDS-related behaviors.
In: Mays, V.; Albee, G.; and Schenider, S., eds. Primary Prevention of
AIDS: Psychological Approaches. Newbury Park, CA: Sage, 1989.
pp. 93-110.
76
Fisher, J., and Fisher, W. Changing AIDS risk behavior. Psycho1 Bull
111:455-474, 1992.
Flaskerud, J.H., and Nyamathi, A.M. Effects of an AIDS education
program on the knowledge, attitudes, and practices of low income
Black and Latina women. J Comm Health 15:343-355, 1990.
Flaskerud, J.H., and Rush, C.E. AIDS and traditional health beliefs and
practices of Black women. Nursing Res 38:210-215, 1989.
Fordyce, E.J.; Blum, A.; and Stoneburner, R.L. A method for estimating
HIV transmission rates among female sex partners of male intravenous
drug users. Am J Epidemiol 133:590-598, 1991.
Fullilove, M.; Fullilove, R.E.; Haynes, K.; and Gross, S. Black women
and AIDS prevention: A view towards understanding the gender rules.
J Sex Res 27:47-64, 1990.
Guarnaccia, P.J.; Angel, R.; and Worobey, J.L. The factor structure of
CES-D in the Hispanic Health and Nutrition Examination Survey: The
influence of ethnicity, gender, and language. Soc Sci 29:85-94, 1989.
Hatcher, R.A.; Stewart, F.H.; Trussell, J.; Kowal, D.; Guest, F.; Stewart,
G.K.; and Cates, W. Contraceptive Technology, 1990-1992. New
York: Irvington Publishers, Inc., 1990.
Haverkos, H.W., and Edelman, R. The epidemiology of acquired
immunodeficiency syndrome among heterosexuals. JAMA 260:19221929, 1988.
Hayes, C. Risking the Future. Washington, DC: National Academy Press,
1987.
Hine, D., and Wittenstein, K. Female slave resistance: The economics of
sex. In: Steady, F., ed. The Black Woman Cross-Culturally.
Cambridge: Schenkman Publishing Company, 1981. pp. 189-344.
Hofferth, S., and Hayes, C. Risking the Future. Vol. 2. Washington, DC:
National Academy Press, 1987.
Houston-Hamilton, A. “Implications of Sexual and Contraceptive
Practices/Attitudes for Prevention of Heterosexual Transmission in
Blacks.” Paper presented at the National Institute on Drug Abuse
Technical Review, Washington, DC, January 1988.
Ickovics, J. R., and Rodin, J. Women and AIDS in the United States:
Epidemiology, natural history, and mediating mechanisms. Health
Psychol 11:1-16, 1992.
Jemmott, L.S., and Jemmott, J.B. Applying the theory of reasoned action
to AIDS risk behavior: Condom use among Black women. Nursing
Res 40:228-234, 1991.
Jemmott, L.S., and Jemmott, J.B. Increasing condom-use intentions
among sexually active Black adolescent women. Nursing Res 41:273278, 1992.
77
Jemmott, J.B.; Jemmott, L.S.; Spears, H.; Hewitt, N.; and Cruz Collins,
M. Self-efficacy, hedonistic expectancies, and condom-use intentions
among sexually active Black inner-city women: A social cognitive
theory approach to AIDS risk behavior. J Adolesc Health 13:512-519,
1992.
Kalichman, S.C.; Hunter, T.L.; and Kelly, J.A. Perceptions of AIDS
susceptibility among minority and nonminority women at risk for HIV
infection. J Consult Clin Psychol 60:725-732, 1992.
Kane, S. HIV, heroin, and heterosexual relations. Soc Sci Med 32:10371050, 1991.
Kelly, J.A.; Murphy, D.A.; Washington, C.D.; Wilson, T.S.; Koob, J.J.;
Davis, D.R.; Ledezma, G.; and Davantes, B. HIV/AIDS prevention for
women seen in primary health care clinics: Effects of intervention to
change high-risk sexual behavior patterns. Am J Pub Health, in press.
Kelly, J.A.; St. Lawrence, J.S.; Diaz, Y.E.; Stevenson, L.Y.; Hauth, A.C.;
Brasfield, T.L.; Kalichman, SC.; Smith, J.E.; and Andrew, M.E. HIV
risk behavior reduction following intervention with key opinion
leaders of population: An experimental analysis. Am J Pub Health
81:168-171, 1991.
Kelly, J.A.; St. Lawrence, J.S.; Stevenson, L.Y.; Hauth, A.C.; Kalichman,
S.C.; Diaz, Y.E.; Brasfield, T.L.; Koob, J.J.; and Morgan, M.G.
Community AIDS/HIV risk reduction: The effects of endorsements by
popular people in three cities. Am J Pub Health 82:1483-1489, 1992.
Kline, A.; Kline, E.; and Oken, E. Minority women and sexual choice in
the age of AIDS. Soc Sci Medicine 34:447-457, 1992.
Kreiss, J.; Ngugi, E.; Holmes. K.; Ndinya-Achola, J.; Waiyaki, P.;
Roberts, P.L.; Ruminjo, I.; Sajabi, R.; Kimata, J.; Fleming, T.R.;
Anzala, A.; Holton, D.; and Plummer, F. Efficacy of nonoxynol-9
contraceptive sponge use in preventing heterosexual acquisition of
HIV in Nairobi prostitutes. JAMA 268:477-482, 1992.
Landrine, H., and Klonoff, E.A. Cultural diversity in causal attributions
for illness: The role of the supernatural. J Behav Med 17:181-193,
1994.
Leigh, B.C., and Stall, R. Substance use and risky sexual behavior for
exposure to HIV: Issues in methodology, interpretation, and
prevention. Am Psychol 48:1035-1045, 1993.
Lewis, D.K., and Watters, J.K. Sexual risk behavior among heterosexual
intravenous drug users: Ethnic and gender variations. AIDS 5:577-583,
1991.
Lo Conte, J.S.; O’Leary, A.; and Labouvie, E. Applying social-cognitive
theory to the safety of sexual behavior of inner city AfricanAmericans. Supplement. Ann Behav Med 15:S102, 1993.
78
Magana, J.R., and Magana, H.A. Mexican-Latino children. In: Stuber,
M.L., ed. Children and AIDS. Washington, DC: American Psychiatric
Press, 1992. pp. 33-43.
Marin, B.V., and Gomez, C.A. Latinos, HIV disease and culture:
Strategies for AIDS prevention. In: Cohen, P.T.; Sande, M.A.; and
Volberding, P.A., eds. The AIDS Knowledge Base: 1993 Update.
Boston: Little Brown and Co., in press.
Marin, G., and Marin, B.V. Research with Hispanic Populations.
Newbury Park, CA: Sage, 1991.
Mays, V.M., and Cochran, S.D. Issues in the perception of AIDS risk and
risk reduction activities by Black and Hispanic/Latina women. Am
Psychol 43:949-957, 1988.
National Institute of Mental Health. Report of the working group on
women. In: Putting Knowledge to Work: HIV Prevention Research
and Community Programs. Washington, DC: Supt. of Docs., U.S.
Govt. Print. Off., 1992.
Niruthisard, S.; Roddy, R.E.; and Chutivongse, S. The effects of frequent
nonoxynol-9 use on the vaginal and cervical mucosa. Sex Transm Dis
4:176-179, 1991.
North, B.B. Effectiveness of vaginal contraceptives in prevention of
STDs. In: Alexander, N.J.; Gabelnick, H.L.; and Spieler, J.M., eds.
Heterosexual Transmission of AIDS. New York: Wiley-Liss, 1990.
pp. 273-290.
Nowak, R. Research reveals condom conundrums. J NIH Res 5:32-33,
1993.
Nyamathi, A.M. Relationship of resources to emotional distress, somatic
complaints, and high-risk behaviors in drug recovery and homeless
minority women. Res Nurs Health 14:269-277, 1991.
Nyamathi, A., and Vasquez, R. Impact of poverty, homelessness, and
drugs on Hispanic women at risk for HIV infection. Hisp J Behav Sci
11:299-314, 1989.
O’Leary, A. Self-efficacy and health: Behavioral and stress-physiological
mediation. Cog Ther Res 16:229-245, 1992.
O’Leary, A. “Social Learning Theory and Its Application to AIDS
Prevention in Developing Countries.” Paper presented at the Family
Health International AIDSTECH Task Force Meeting, Durham, NC,
April 3-5, 1991.
O’Leary, A.; Goodhart, F.; Jemmott, L.S.; and Botcher-Lattimore, D.
Predictors of safer sexual behavior on the college campus: A social
cognitive theory analysis. J Am Coll Health 40:254-263, 1992.
79
O’Leary, A.; Celentano, D.; Maibach, E.; Weisman, C.; McCormack, M.;
Raffaelli, M.; Jemmott, J.B., III; and LaVange, A. “Social Cognitive
Predictors of HIV Preventive Behavior Among Uban STD Clinic
Patients.” Paper presented at the Ninth International Conference on
AIDS, Berlin, June 6-11, 1993a.
O’Leary, A.; Jemmott, L.S.; Suarez-Al-Adam, M.; AlRoy, C.; and
Fernandez, M. I. Women and AIDS. In: Matteo, S., ed. American
Women in the 1990’s: Today’s Critical Issues. Boston: Northeastern
University Press, 1993b. pp. 173-192.
Padian, N.S.; Shiboski, S.C.; and Jewell, N.P. Female-to-male
transmission of human immunodeficiency virus. JAMA 266:16641667, 1991.
Reinisch, J.M.; Sanders, S.A.; and Ziemba-Davis, M. The study of sexual
behavior in relation to the transmission of human immunodeficiency
virus. Am Psychol 43:921-927, 1988.
Rivera, C. Hispanics and the social service system. In: Cafferty, P.S.J.,
and McCready, W.C., eds. Hispanics in the United States. New
Brunswick, NJ: Transaction Books, 1985.
Rosenberg, M.J., and Gollub, E.L. Commentary: Methods women can
use that may prevent sexually transmitted disease, including HIV. Am
J Public Health 82:1473-1478, 1992.
Rothenberg, R.B. Those other STDs. Am J Public Health 81:1250-1251,
1991.
Schilling, R.F.; El-Bassel, N.; Schinke, S.P.; Gordon, K.; and Nichols, S.
Building skills of recovering women drug users to reduce heterosexual
AIDS transmission. Pub Health Rep 106:297-304, 1991.
Shayne, V.T., and Kaplan, B.J. Double victims: Poor women and AIDS.
Women Health 17:21-37, 1991.
Stein, Z. The double bind in science policy and the protection of women
from HIV infection. [Editorial] Am J Public Health 82:1471-1472,
1992.
Suarez, M.; Raffaelli, M.; O’Leary, A.; and Lo Conte, J. “Use of
Alternative Healing Practices Among HIV-infected Hispanics.” Paper
presented at the Ninth International Conference on AIDS, Berlin,
June 6-11, 1993.
Thomas, S.B., and Quinn, S.C. The Tuskeegee syphilis study, 19321972: Implications for HIV education and AIDS risk education
programs in the Black community. Am J Public Health 81:1498-1505,
1991.
Turner, C.; Miller, H.; and Moses, L. AIDS: Sexual Behavior and
Intravenous Drug Use. Washington, DC: National Academy Press,
1989.
80
Valdiserri, R.O.; Arena, V.C.; Proctor, D.; and Bonati, F.A. The
relationship between women’s attitudes about condoms and their use:
Implications for condom promotion programs. Pub Health Briefs 79:
499-501, 1989.
Weissman, G. “Working with Women at Risk: Experience From a
National Prevention Research Program.” Plenary address presented at
the Annual Training Institute, Comprehensive HIV/AIDS Mental
Health Education Program, Detroit, MI, December 6, 1991.
Weissman, G.; Brown, V.; and The National AIDS Research Consortium.
Drug use and sexual behaviors among sex partners of injecting drugusers. MMWR Morb Mortal Wkly Rep 40:855-860, 1991.
Wyatt, G.E.; Peters, S.; and Guthrie, D. Kinsey revisited, I: Comparisons
of sexual socialization and sexual behavior of White women over 33
years. Arch Sex Behav 17:201-239, 1988a.
Wyatt, G.E.; Peters, S.; and Guthrie, D. Kinsey revisited, II: Comparisons
of sexual socialization and sexual behavior of Black women over 33
years. Arch Sex Behav 17:289-332, 1988b .
AUTHOR
Ann O’Leary, Ph.D.
Associate Professor
Department of Psychology
Rutgers University, Busch Campus
New Brunswick, NJ 08903
81
Drug Use and HIV Risk Among
Gay and Bisexual Men: An
Overview
Robert J. Battjes
The acquired immunodeficiency syndrome (AIDS) epidemic was first
identified among gay and bisexual men, and men who have sex with men
still comprise the majority of persons with AIDS in the United States.
Since the inception of the epidemic, 60 percent of AIDS cases reported to
the Centers for Disease Control and Prevention (CDC) have occurred
among men who reported having sex with men, with 54 percent of AIDS
cases occurring in men who reported sex with men but no injecting drug
use, and 6 percent of cases occurring in men reporting dual risk behaviors
of sex with men and injecting drug use (Centers for Disease Control and
Prevention 1993). The National Institute on Drug Abuse (NIDA) is the
lead Federal agency for research on drug abuse aspects of AIDS. Thus,
NIDA is concerned with AIDS associated with injecting drug use and
also with the impact of noninjecting drug use on sexual risk behaviors
and disease progression.
This section of the monograph focuses on the relationship between drug
use and HIV risk among gay and bisexual men.’ In planning this section
of the technical review, Battjes, Sloboda, and Grace recognized the
importance of focusing on two distinct subgroups of gay and bisexual
men: noninjecting drug users (non-IDUs) whose drug use may contribute
to their sexual HIV risk, and gay and bisexual men who also were IDUs.
Ostrow was asked to focus on the first group, non-IDUs. Unfortunately,
with regard to the second group (gay and bisexual IDUs), little research
has been conducted except for that focused on injecting drug use among
male prostitutes. Therefore, Waldorf was asked to focus on drug use and
HIV risk among male prostitutes. It was decided to briefly highlight the
HIV risk of other gay and bisexual IDUs in this section introduction.
In addition to the two chapters in this section, the chapter by RotheramBorus and colleagues in the section on HIV risk among adolescents
focuses on high-risk gay and bisexual adolescent males who chronically
abuse drugs, often including injecting drug use, and engage in sex to
obtain drugs and as a means of survival.
82
GAY AND BISEXUAL INJECTING DRUG USERS
NIDA’s AIDS research 5-year planning process (see Battjes et al., this
volume) identified the dual risk group of gay and bisexual men who
inject drugs as a major research gap. Relatively little research has
focused specifically on this population, although they comprise 6 percent
of all AIDS cases, 10 percent of AIDS cases occurring in men who report
sex with men, and 21 percent of AIDS cases occurring in IDUs (Centers
for Disease Control and Prevention 1993). AIDS case data indicate that
persons in this dual risk group are at substantially increased risk for AIDS
compared with persons reporting either risk behavior alone.
While men who have sex with men comprise 21 percent of AIDS cases in
IDUs, it appears that relatively few of this dual risk group are reached
through the primary HIV prevention initiatives that target drug abusers
(i.e., drug abuse treatment and AIDS outreach programs). For example,
in NIDA’s HIV Seroprevalence Survey, which studied IDUs admitted to
methadone treatment in nine U.S. cities between 1987 and 1991, only
0.8 percent of approximately 6,400 male IDUs reported having
exclusively homosexual sexual contacts within the previous year, while
another 0.9 percent reported both male and female sexual partners
(Battjes, unpublished data). In the NIDA-supported National AIDS
Demonstration Research (NADR) program, which evaluated street
outreach to IDUs at 61 sites between 1987 and 1992, only 2.2 percent of
approximately 35,000 males reported having only male sexual partners
within the prior 6 months, while 2.9 percent reported having both male
and female sexual partners (Needle, personal communication, March
1993). In the NIDA-supported HIV Outreach Cooperative Agreement
research program, which is evaluating HIV outreach and counseling to
IDUs and crack cocaine users in 22 cities, only 2 percent of
approximately 4,100 males enrolled between January 1992 and April
1993 reported having either exclusively male or both male and female
sexual partners (Needle, personal communication, March 1993).
Increased risk of HIV infection in this dual risk group compared with
non-IDU gay and bisexual men has been reported by Stall and Ostrow
(1989) in their analysis of gay and bisexual men participating in the San
Francisco Men’s Health Study. At study entry, 88.2 percent (30 of 34) of
gay and bisexual men who reported injecting drug use were HIV
seropositive, compared with 46.2 percent (189 of 409) of non-IDU gay
and bisexual men. Gay and bisexual male IDUs in NIDA’s HIV
Seroprevalence Survey were more likely to be infected than other IDUs,
83
with 34.6 percent of gay men and 28.6 percent of bisexual men HIV
seropositive compared with 13.2 percent of men reporting no male sexual
partners (Battjes, unpublished data). In the NADR study, gay IDUs were
similarly more likely to be HIV seropositive (34.6 percent), while
bisexual male IDUs were slightly less likely to be HIV seropositive
(16.4 percent) compared with males reporting no male sexual partners
(18.7 percent). Assessing gay and bisexual men together in the
Cooperative Agreement program, 39.3 percent of these men were HIV
seropositive compared with 10.7 percent of men reporting no male sexual
partners (Needle, unpublished data).
Stall and Ostrow (1989) also found that gay and bisexual male IDUs were
more likely to report high-risk sexual activity at study entry
(67.7 percent) than non-IDU gay and bisexual men (45.1 percent).
Comparing the sexual risk of gay and bisexual IDUs with male IDUs
reporting no male sexual partners in the HIV Seroprevalence Survey,
28.9 percent of gay and 46.4 percent of bisexual males reported receiving
payment for sex, compared with 7.0 percent of other men. Yet gay IDUs
also were more likely to use condoms, with 20.7 percent reporting always
using condoms, compared with 11.8 percent of bisexual male IDUs and
8.0 percent of other male IDUs (Battjes, unpublished data). Similarly in
the NADR study, 35.8 percent of gay IDUs and 64.3 percent of bisexual
male IDUs reported receiving payment for sex, compared with
6.8 percent of other male IDUs. Also, 18.8 percent of gay IDUs reported
always using condoms, compared with 8.5 percent of bisexual male IDUs
and 8.3 percent of other male IDUs (Needle, unpublished data).
This dual risk group of gay and bisexual IDUs is important not only in its
own right, but also because persons reporting dual risk may serve as a
bridge for HIV transmission between heterosexual and lesbian IDUs and
non-IDU gay and bisexual men. For example, in a study of IDUs
entering methadone treatment in four low HIV prevalence areas, Battjes
and colleagues (1989) found that the only behavioral difference between
HIV seropositive and seronegative IDUs was self-reported sharing of
needles or syringes with gay or bisexual men. Of seropositives,
25.0 percent reported such needle sharing, compared with 3.1 percent of
seronegatives (adjusted odds ratio = 5.44). In Chicago, Lampinen and
colleagues (1991) found that almost all of the earliest IDU-associated
AIDS cases (diagnosed between 1982 and 1986) were among men who
also reported sex with other men, suggesting that they may have acquired
their infection sexually rather than parenterally. They further found that
the geographic distribution of AIDS cases within Chicago suggested two
84
geographically and demographically distinct points of entry of HIV
among IDUs: one in a northside neighborhood with a large concentration
of gay men, and the other among heterosexual Puerto Ricans several
miles away. Finding such demographic distinctions within populations of
IDUs even in a single city reinforces the need for further evaluation of
contextual variables contributing to HIV risk.
GAY AND BISEXUAL NONINJECTING DRUG USERS
In the following chapter, Ostrow focuses on the relationship of illicit drug
use with sexual HIV risk behaviors and HIV infection among
noninjecting gay and bisexual men, reviewing the published scientific
literature and highlighting research findings from the Chicago site of the
Multicenter AIDS Cohort Study, funded by the National Institute of
Allergy and Infectious Diseases, and its companion behavioral/mental
health study funded by the National Institute of Mental Health. As
reported by Ostrow, nonmedical drug use is clearly associated with highrisk sexual behaviors and also with increased likelihood of HIV infection
in gay and bisexual men. However, the mechanisms that may account for
these associations are not clear; thus, it is uncertain whether high-risk
sexual behavior is causally related to drug use or if both behaviors are
manifestations of some common underlying factor. While causation is
not established, Ostrow notes that researchers currently know enough
about their association to permit targeting HIV prevention interventions.
DRUG USE AMONG MALE SEX WORKERS
In the second chapter of this section, Waldorf focuses on drug use and
HIV risk among male sex workers, reviewing the published scientific
literature and highlighting findings from two studies conducted in San
Francisco from 1987 to 1988 and from 1989 to 1991. As reported by
Waldorf, relatively few studies have assessed drug use among male sex
workers, and the extent of injecting drug use reported has varied
considerably. A significant contribution of Waldorf and his colleagues
was the differentiation of two types of male prostitutes, “hustlers” and
“call men,” with nine different subtypes, and their use of quota sampling
to permit study of the various subtypes. Waldorf reports that injectable
drugs, including methamphetamine, cocaine hydrochloride, and heroin,
were readily available in the areas of San Francisco where male sex
workers were concentrated, and high rates of injecting drug use were
85
found among both hustlers and call men in both studies. Among these
IDU male sex workers, Waldorf reports high levels of needle sharing,
including sharing in communal settings such as shooting galleries and sex
clubs. Waldorf concludes that unsafe sex and needle sharing place many
male sex workers at risk for HIV infection.
NOTE
1. While not all men who have sex with men identify themselves as gay
or bisexual, for purposes of this introduction, “gay” is used to refer to
men who report having had only male sexual partners. “Bisexual”
refers to men who report having had both male and female sexual
partners. It is recognized that such behavioral definitions do not
necessarily equate with individuals’ self-identified sexual orientation.
REFERENCES
Battjes, R.J.; Pickens, R.W.; and Amsel, Z. Introduction of HIV infection
among intravenous drug abusers in low prevalence areas. J Acquir
Immune Defic Syndr 2:533-539, 1989.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. Vol. 5, No. 3, 1993.
Lampinen, T.M.; Joo, E.; Seweryn, S.; Wiebel, W.; and Chapman, B.T.
“Race/Ethnicity Not Associated With HIV-1 Seropositivity Among
Chicago Intravenous Drug Users (IVDUs): Confounding and
Epidemic Duration.” Paper presented at the Seventh International
Conference on AIDS, Florence, Italy, June 16-21, 1991.
Stall, R., and Ostrow, D.G. Intravenous drug use, the combination of
drugs and sexual activity and HIV infection among gay and bisexual
men: The San Francisco Men’s Health Study. J Drug Issues 19:57-73,
1989.
86
AUTHOR
Robert J. Battjes, D.S.W.
Deputy Director
Division of Clinical Research
National Institute on Drug Abuse
Parklawn Building, Room 10A-38
5600 Fishers Lane
Rockville, MD 20857
87
Substance Use and
HIV-Transmitting Behaviors
Among Gay and Bisexual Men
David G. Ostrow
INTRODUCTION
This chapter reviews what is known about the relationships between
nonmedical psychoactive drug (NMPD) use and human
immunodeficiency virus type 1 (HIV-1)1 infection, HIV-related disease,
and acquired immunodeficiency syndrome (AIDS) in men who have sex
with men. The review is limited to the published literature, which is,
however, quite large in several areas of specific relevance to HIV
infection and AIDS.
By far the largest area of interest concerns intravenous (IV) drug use
because of the obvious relationship to a major and growing route of
transmission of HIV. While this chapter does not specifically address the
issues of IV drug use and HIV infection, several recently published
reviews (Des Jarlais et al. 1991; Schoenbaum et al. 1989; and Ross et al.
1992) provide comprehensive and up-to-date information regarding the
behavioral and mental health aspects of this important subject. Rather,
this chapter focuses on nonparenterally used NMPDs and their
relationship to HIV transmission behaviors, specifically high-risk sexual
behavior. It also discusses the secondary community impact of these
associations and their intervention implications. Readers interested in a
more detailed discussion of published studies concerning the potential
effects of NMPD use on the natural history of HIV infection are referred
to the recently published monograph, Cofactors in HIV-1 Infection and
AIDS (Watson 1990).
Many of the published reports and all of the tables and figures in this
chapter come from the Chicago Multicenter AIDS Cohort Study (MACS)
cohort, one of the four National Institute of Allergy and Infectious
Diseases (NIAID)-funded collaborative natural history of HIV study sites,
and its companion National Institute of Mental Health (NIMH)-funded
behavioral/mental health add-on study, the Coping & Change Study
(CCS).2 Over 1,000 gay and bisexual men volunteered for these studies
88
when they were begun in 1984, and approximately 600 men continue to
participate semiannually in this, the tenth year of the study. NMPD use
patterns have been analyzed for all men attending at each semiannual
assessment and also for the smaller number of men (approximately 350)
who have participated every year since the start of the study. Since there
were no significant differences between the serial cross-sectional and
panel patterns for commonly used substances, the former are presented
here.
PATTERNS OF NMPD USE IN A COHORT OF GAY AND
BISEXUAL MEN
Table 1 illustrates the long-term patterns (1984-1993) of use of those
continuously monitored NMPDs and cigarette smoking in the Chicago
MACS/CCS cohort, while table 2 shows the alcohol consumption
patterns for that same period. While frequency of use in the past
6 months was assessed for nine classes of drugs at each semiannual visit,
prevalence patterns suggest three general categories of NMPD use by
members of the Chicago MACS/CCS cohort seen at each semiannual
evaluation.
Frequently Used NMPDs
The prevalence rates for use of marijuana and volatile nitrates (poppers)
are extremely similar, and are the highest of any of the classes of
recreational substances throughout the study period (table 1). There was
a gradual decline during the first 3 years of the study, from approximately
70 percent of participants using either substance in 1984 to less than 50
percent reporting use from 1986 onward, but little reported change in
popularity since visit 6 or 7. At the start of the study, over 60 percent of
men reported using both marijuana and poppers; this proportion declined
to approximately 35 percent by visit 12. This drop in use of both
marijuana and poppers corresponded with a drop in the average number
of NMPDs used, from two to one (figure 1). From visits 12 to 16 there
was a further drop in popper use (to 20 percent), but a more recent
rebound (visits 17 to 19) to approximately one-third is evident in the
latest data. Cocaine was the third most popular NMPD used throughout
the study, decreasing from a prevalence of one-third of the cohort at the
study’s beginning to one-sixth from visits 6 to 7 onward. Cocaine use
was almost exclusively through nasal insufflation: Crack cocaine use was
only specifically assessed beginning with visit 12 and was reported by
89
TABLE 1. Prevalence of substance use among Chicago (MACS) men, by visit [1] (in percents)
Visit Number/Year
1
Substance:
Group A:
2
1984/85
3
4
1985/86
5
6
1986/87
7
8
1987/88
9
10
1988/89
11
12
1989/90
13
14
15
16
1991/92
1990/91
17
18
19
1992/1993
Cigarettes
Marijuana/Hashish
Cocaine
Crack [2]
39.4 35.7
70.2 64.6
33.9 28.8
34.0 32.2
62.6 57.4
25.3 22.2
32.9 30.1
50.7 49.9
20.9 16.6
31.0 29.5
45.7 44.1
15.2 15.3
30.2 31.3
43.7 41.6
15.3 14.4
30.9 31.7
41.2 39.9
12.9 12.2
30.4 28.4
34.0 32.2
10.3
8.3
28.6 27.5
29.6 31.8
9.4
6.5
27.8 28.7 26.8
35.0 31.8 33.2
7.5
5.5
5.5
Poppers
Group B:
70.6 69.0
59.3 52.9
49.0 44.5
46.2 44.1
43.3 42.5
1.5
42.4 43.4
0.9
0.6
38.7 30.2
0.6
21.1
1.7
21.6
1.2
27.9
1.2
24.7
1.6
27.1
MDA [3]
15.5 14.6
10.4
5.6
3.9
2.4
1.7
2.1
2.5
2.0
1.0
1.0
0.9
1.1
Hallucinogens [3]
Downers [3]
18.2 12.9
19.4 12.3
9.6
7.9
5.5
7.9
5.2
5.2
4.3
2.7
2.2
1.5
7.2
4.9
3.3
3.1
5.3
2.9
1.6
10.4
2.5
3.7
4.1
2.5
1.5
17.6 12.9
3.0
4.1
3.8
4.2
0.8
Uppers [3]
Other Substances:
4.1
3.3
3.3
3.0
4.4
31
3.7
1.5
2.3
1.3
2.3
3.8
1.9
1.5
2.8
1.2
2.2
2.2
0.4
2.1
0.4
3.0
1.8
0.2
1.9
0.0
1.1
0.3
1.8
0.3
1.3
2.2
1.8
0.9
1.7
6.3
6.9
6.1
5.7
824
736
704
524
494
491
506
Ethyl Chloride [4]
Heroin/Methadone [5]
Other Drugs
Total N (by visit):
5.1
1.8
4.3
1.5
1.4
2.0
1005 925
852
4.3
2.2
0.2
0.5
600 531
563
604
595 592
556 543
[1] Cases include only those Chicago MACS respondents who subsequently entered the Coping and Change Study.
The proportions represent those respondents reporting use of each substance during the prior six-month period.
[2] Crack cocaine was included as a separate category, beginning at Visit 12.
[3] These items were dropped from the MACS questionnaire after Visit 15.
[4] This item was dropped from the MACS questionnaire after Visit 7.
[5] This item was dropped from the MACS questionnaire after Visit 11.
522
FIGURE 1. Total number of drug classes* used in prior 6 months, visits 1 to 19 (with 95 percent confidence
intervals)
less than 2 percent of participants at that time. The legal but highly
psychoactive substance, nicotine, was reported consumed through
cigarette smoking by about one-third of the participants throughout the
study (table 1).
Infrequently Used Substances
All of the other NMPDs inquired about were used by less than 20 percent
of the participants at the start of the study and declined by at least
three-quarters to less than 5 percent by visit 6 (table 1). Reported use of
downers (mostly methaqualone and barbiturates) and the hallucinogen
1-(3,4-methylenedioxyphenyl)-2-aminopropane (MDA) have continued
to decline to less than 2 percent of participants in recent visits. Smoked
cocaine, or crack, was only assessed beginning with the 12th visit, and
use was less than 2 percent prevalent at that time.
The differentially greater decreases in the proportion of men using the
less popular NMPDs probably reflects their use by a subgroup of men
using three or more drugs and at significantly higher risk of prevalent
HIV infection, a group that has differentially dropped out of the study
over time. Men who regularly and consistently participate in the Chicago
MACS/CCS studies are significantly less likely to be HIV infected
(25-30 percent versus 50-60 percent HIV antibody prevalence rates
among the two groups) and use fewer NMPDs than the men who have
dropped out or been inconsistent in their study participation. In addition,
men developing AIDS in the course of their study participation have
either dropped out or been excluded from further behavioral assessment.
Given the significantly higher HIV infection rates among polysubstanceusing men of the MACS (Chmiel et al. 1987; and Easterbrook et al. 1993;
Ostrow et al. 1987), the dramatic decline in popularity of uppers,
downers, and hallucinogenic drugs seen in the study cohort may reflect
the rapid increase in AIDS and related deaths among members of the fast
track subpopulation of Chicago’s gay community between 1984 and
1987. Both of these general patterns suggest that results based on
long-term and consistent participants in the Chicago MACS/CCS may
underestimate the actual rate of current NMPD use among the original
cohort members and lessen the likelihood of detecting the relationships
between NMPD use and sexual behavior patterns.
92
Alcohol Use Patterns
Table 2 shows that while alcohol use is common in the cohort, heavy
alcohol consumption (defined as more than 60 drinks/month or an
average of 2 or more drinks per day) was reported by approximately
28 percent of the men at baseline, dropped rapidly during the first year to
approximately half that level, and has continued at around 10 to
15 percent thereafter. Most of the drop in heavy alcohol use has been
reflected in a corresponding rise in the proportion of abstainers, which
has risen slowly from less than 5 percent at baseline to approximately
10 to 15 percent during recent visits. No differences in the longitudinal
patterns of alcohol use between prevalent seropositive and seronegative
men were observed, other than that seropositive men began the study
with higher rates of alcohol use, which rapidly fell to the same levels as
reported by seronegative men by visit 3 or 4.
Other Study Findings
There are at least three other studies that lend credence to the
generalizability of the patterns of NMPD use observed in the Chicago
MACS/CCS cohort of sexually active gay/bisexual men between 1984
and 1990. The most similar study in terms of time and site is the 1985
Social Issues Survey conducted by McKiman and Peterson (1989a)
among Chicago’s gay and lesbian community. In that study,
approximately 21,000 anonymous self-report questionnaires were
distributed as inserts in a gay community newsletter, at large social
events, and through a broad variety of social, religious, political, or
professional organizations serving the community. Thirty-four hundred
(16 percent) were returned, including 2,652 from men with demographic
and socioeconomic characteristics similar to the Chicago MACS/CCS
cohort at baseline. Using the same definition of heavy drinking,
McKirnan and Peterson found that among their male respondents,
13 percent were abstainers, 70 percent were moderate users, and
17 percent qualified as heavy users. Prevalence rates for the use of
marijuana (56 percent) and cocaine (23 percent) in the past year were
lower than the baseline rates (70 percent and 34 percent, respectively) and
closer to the visit 4 rates of the Chicago MACS/CCS. However, their
reported 21 percent popper use rate among men was considerably lower
than the rate reported at any time among the Chicago MACS/CCS
participants.
93
TABLE 2. Quantity of alcohol use among Chicage (MACS) men, by visit [I] (in percents)
Visit Number/Year
1
2
3
4
5
6
7
8
Alcohol Use
Category
1984/85
1985/86
1986/87
1987/88
None
4.7 6.4
6.8 7.5
7.5 8.9
9.3 8.7
12
13 14
15 16
1988/89
1989/90
1990/91
1991/92
9.6 9.5
11.3 9.7
9
10
11
17
18
19
1992/1993
11.8 12.9 16.2 14.9 14.3 16.5 19.0
Light (<13 drinks/mo.)
27.6 29.8 33.3 32.9 36.2 37.1 36.4 36.5 40.1 38.4 39.3 38.2 38.5 39.2 36.5 38.9 37.7 40.6 38.9
Moderate (<60 drinks/mo.)
39.7 44.8 42.5 42.1 40.6 40.1 39.9 39.4 35.5 39.7 37.1 39.0 38.5 37.0 36.5 34.4 36.7 34.5 31.3
Heavy (60 + drinks/mo.)
28.0 19.0 17.4 17.5 15.7 13.9 14.4 15.4 14.7 12.5 12.3 13.2 11.1 10.9 10.8 11.8 11.4 8.4 10.7
Total N (by visit):
989 924 852 824 734 704 599 531 563 602 604 600 566 543 520 524 491 490 504
[1] Cases include only those Chicago MACS respondents who subsequently entered the Coping and Change Study. The proportions represent the
respondents’ reported average drinking during the prior 6-month period.
Using the same interview questions as the Chicago MACS, but
employing a random household survey of unmarried males living in the
San Francisco neighborhoods with the highest rates of reported AIDS
cases, Stall and Wiley (1988) reported on the NMPD use patterns of
1,034 men aged 25 to 54, of whom 286 said they were exclusively
heterosexual and 748 were gay or bisexual. Using a somewhat different
definition of heavy drinking (5 or more drinks on 2 or more nights per
week), the San Francisco Men’s Health Study (SFMHS) sample had rates
of 19 percent and 11 percent heavy drinkers among the gay/bisexual and
heterosexual subsamples, respectively, compared to rates of 28 percent
and 19 percent at the first two baseline assessments of the gay/bisexual
men in the Chicago MACS/CCS. Except for the oldest group
(45 to 54 years) in the SFMHS sample, gay/bisexual men did not differ
from heterosexual men in terms of alcohol use patterns.
In terms of NMPD use prevalence, however, the SFMHS data for
gay/bisexual men is similar to what was found in the Chicago
MACS/CCS at baseline and significantly higher than in the San Francisco
heterosexual male subgroup. Stall and Wiley (1988) reported 25 percent
and 18 percent prevalence use rates for downers and hallucinogens in
their gay/bisexual sample, which compares to the 19 percent and
18 percent rates in the Chicago cohort and 9 percent and 12 percent rates
for downers and hallucinogen use by the San Francisco heterosexual male
sample. The SFMHS gay/bisexual subgroup did report somewhat higher
baseline rates of reported use of cocaine (52 percent), opiates (4 percent),
and uppers (28 percent) than did the Chicago sample. Conversely,
popper use was reported by 58 percent of SFMHS respondents and MDA
use by 9 percent, compared to initial rates of use of 71 percent and
15 percent, respectively, by the Chicago MACS/CCS sample.
The third relevant sample is the diverse community-based sample of
746 New York City (NYC) gay/bisexual men recruited by Martin and
Dean in early 1985 (Martin et al. 1989). In three yearly interviews of
their cohort, the investigators determined both prevalence of NMPD use
and alcohol abuse/dependence disorder. They reported a 12 percent rate
of alcohol abuse or dependence in 1986, which dropped to 9 percent in
1987. In terms of NMPD use, Martin and Dean’s NYC cohort appears
similar to the Chicago MACS/CCS cohort during the period 1984-1987,
with the exception of inhaled nitrites, which were reportedly used by
45 percent of the NYC cohort in 1984/1985 and approximately
25 percent of the men in 1986/1987.
95
A fourth study of gay/bisexual men, the Boston Partners Study, has
recently published rates of NMPD use over the previous 5 years reported
at entry between 1985-1988. For the most commonly used drugs
reported by seronegative (N=275) and seropositive (N=206) men,
respectively, prevalence rates were 79 percent and 88 percent for
marijuana, 64 percent and 84 percent for poppers, and 56 percent and
73 percent for cocaine (Seage et al. 1992).
Differences among these five studies may reflect differences in sampling
methods and retrospective recall periods, or may indicate actual
differences in NMPD use preferences among gay/bisexual men in
different United States cities. These studies indicate the relative
popularity of specific NMPDs among gay/bisexual men in four cities and
recruited through differing mechanisms during the mid-1980s, a time of
particularly high rates of drug use in these communities. In fact,
approximately half of the men in both the SFMHS and Chicago
MACS/CCS cohorts reported the use of three or more categories of
NMPDs during the last 6 months of 1984, at the simultaneous start of the
two studies.
PATTERNS OF SEXUAL RISK BEHAVIOR AS RELATED TO
NMPD USE
Since the original observations that gay/bisexual men participating in
AIDS epidemiology cohort studies who used alcohol and NMPDs with
sexual partners were more likely to engage in high-risk sexual activities,
most notably unprotected anal intercourse (UAI), and were less likely to
reduce their sexual risk than abstaining men (Stall et al. 1986; Ostrow et
al. 1987), there has been a plethora of confirmatory studies. Leigh and
Stall (1993) have reviewed the published literature in this area; this
section summarizes some of the conclusions of that comprehensive and
scholarly review. Leigh and Stall’s group studies of the relationship
between NMPD use and risky sexual behavior into three general
categories.
1.
Global association studies examine NMPD use and high-risk sexual
behaviors, but these behaviors are not linked temporally. Thus, these
studies are unable to determine whether or not the high-risk sex
occurred in the context of NMPD use.
96
2. Situational association studies examine the frequency of high-risk
sexual behaviors and also examine the extent to which sexual
activities occur in conjunction with NMPD use. These studies
establish the occurrence of high-risk sex and of sex while using drugs
or alcohol, but do not determine if high-risk sex is more likely to
occur when drugs or alcohol are used.
3. Event analysis studies examine sexual behaviors and NMPD use in
specific sexual encounters. While these studies can assess the
temporal association of high-risk sexual behaviors and drug or
alcohol use, they share the limitations in terms of causal inference of
the other two categories of studies. Yet, event analysis studies are
useful in distinguishing circumstantial within-person associations.
None of the seven event analysis studies involving gay/bisexual men
cited by Leigh and Stall demonstrated a significant difference in the
likelihood of risky sex when NMPDs were used, and only one study
found an association between drinking and UAI (McCusker et al. 1990).
The most recently published event analysis studies of gay/bisexual men,
from the multisite Project Sigma of Great Britain (Weatherbum et al.
1993) and the Talking Sex Project of Toronto (Myers et al. 1992), also
failed to detect any differences in the incidence of unprotected receptive
or insertive anal intercourse between events involving the use or nonuse
of alcohol or NMPDs, respectively. The general failure of event analysis
studies to confirm a link between NMPD use in specific sexual
encounters and risky sexual behaviors does argue for possible underlying
interpersonal personality or character factors, rather than circumstantial
factors, as underlying the NMPD-risky sex associations documented
below.
MECHANISMS AND THEORIES OF ASSOCIATIONS BETWEEN
NMPD USE AND HIGH-RISK SEXUAL BEHAVIOR AMONG
GAY/BISEXUAL MEN
Leigh and Stall conclude that “it is clear that there is a positive
relationship between substance use and high risk sex; what is less clear is
the level at which this link exists.... findings from these studies are
consistent with a number of explanations-causal, correlational, and
confounding” (p. 1038). To the extent that any specific underlying
mechanisms have important HIV prevention implications, they need to be
considered as priorities for future research efforts. It is important to keep
97
in mind, however, that most of the prior studies that reported a positive
association between UAI and NMPD use can be at least partially
explained simply on the basis of a relationship between NMPD use and
increased frequency of sex or number of sexual partners, regardless of
mechanism. However, the author and colleagues believe that any
associations observed between specific drug categories and specific
high-risk activities can be cited in support of multiple causal mechanisms,
from the social, physiological, cognitive, and clinical domains of
behavioral science.
Summary of Findings from Three Ongoing Prospective
Studies
Three ongoing cohort studies of gay and bisexual men provide the most
detailed evidence of the nature and mechanisms of the association
between NMPD use and HIV-transmitting sexual behavior: the Chicago
MACS/CCS cohort, whose NMPD use history is detailed above; the
Boston Partners Study, also described above (Seage et al 1992); and the
Toronto Sexual Contact Study, which recently published an analysis
spanning the 5-year period of 1984/1985 to 1989/1990 (Calzavara et al.
1993). These three studies are reviewed according to a hierarchy of
evidence that combines the three category methodological typology of
Leigh and Stall with whether or not the outcome examined is sexual risk
behavior or rates of HIV infection.
Cross-Sectional Global Associations of NMPD Use and High-Risk
Sexual Behavior. Table 3 summarizes the significant associations
across 6 years of global NMPD use and unprotected receptive anal sex
(RAS) among members of the Chicago MACS/CCS cohort (Ostrow et al.
1993). Among the 10 categories of NMPDs examined at all
12 semiannual assessments, only popper use showed a consistent and
strong cross-sectional association with unprotected RAS. While the other
commonly used NMPDs had inconsistent associations with high-risk sex
in this cohort, the use of both poppers and either cocaine or marijuana
was associated with the highest rates of unprotected RAS throughout the
study period.
Association of Global NMPD Use and HIV Infection Rates. Both
the Chicago MACS/CCS (Chmiel et al. 1987) and Boston Partners Study
(Seage et al. 1992) demonstrate baseline prevalence rates of HIV
infection that were significantly higher among NMPD users when
compared with nonusers. Similar odds ratios for cross-sectional HIV
98
TABLE 3. Prevalence of use of substances by receptive anal sex risk: Summary of twoway table analysis,
waves 1-12 [1]
Receptive Anal Sex Risk [2] at Wave:
Substance: [3]
1
2
3
4
5
6
7
8
9
825
830
763
742
657
650
650
597
608
10
11
12
543
538
Marijuana/Hashish
Poppers
Cocaine
Crack Cocaine [4]
MDA
Hallucinogens
Downers
Ethyl Chloride [5]
Heroin [6]
Uppers
Other Drugs
Cigarettes
Alcohol Volume
Total N Per Wave:
603
[l] Based on 2x2 contingency tables analyzing the use-prevalence for each substance by the incidence of risky sexual practices at each wave. Estimates of statistical
significance were determined by Chi-Square tests. The findings from the analyses are summarized here according to the symbols:
“++”:
“+”:
“x”:
positive association and p-level < .01
positive association and p-level < .05, but > .01
no significant association, p-level > .05
[2] A “risk-index.” combining information on both number of partners for receptive anal sex (RAS) and condom usage, was constricted from the data at each wave of
the Coping and Change Study. The measures for this analysis compared two risk-levels: “safe”-respondents who, during the prior one-month period, had been
celibate, had refrained from RAS, or had had only one partner who always wore a condom: and “unsafe”-monogamous individuals, who had not used condoms, or
individuals reporting multiple RAS partners, regardless of condom usage.
[3] Substance-use data are from the Chicago MACS questionnaires for the first 12 visits, spanning the years 1984-90. The variables denote the percentage of men who
used each substance during the preceding 6-month period. The dichotomous alcohol measure contrasts abstemious or moderate with heavy drinking (an average of
60 or more drinks per month) over the previous six months.
[4] Not included as a separate item in the MACS interview until Visit 12.
[5] Dropped from the MACS interview after Visit 7.
[6] Dropped from the MACS interview after Visit 11.
seropositivity and reported NMPD use were reported in both cohorts,
with users of all categories except barbiturates (downers) showing odds
ratios with 95 percent confidence intervals greater than 1 for prevalent
HIV infection. These odds ratios ranged from 1.6 (CI=1.3-2.0) for any
NMPD use in the MACS cohort to 2.0 (CI=l.2-3.2) for marijuana use
and 2.9 (CI=1.9-4.5) for popper use among the Boston Partners Study
(Seage et al. 1992).
Association of NMPD Use with Sexual Partners and Prevalence of
HIV Infection. All three of these studies have demonstrated an increased
likelihood of HIV infection among those men reporting the use of certain
NMPDs or combinations. Since the likelihood of becoming HIV infected
is highly dependent on the numbers of partners with whom unprotected
intercourse is engaged in, these analyses have controlled for either the
numbers of sexual partners or specific practices. When this is done, it is
consistently observed that the use of poppers and/or insufflated cocaine
are associated with increased prevalence of HIV infection beyond that
expected on the basis of numbers of partners or potential anal exposures.
The investigators in the Boston Partners Study reported a particularly
strong effect on risk of HIV infection among men who reported always
using poppers whenever engaging in RAS. These men, who comprised
approximately 10 percent of the study group, were almost 34 times more
likely to be HIV infected than men with no history of unprotected RAS or
popper use, and three times more likely to be HIV infected than men who
only sometimes combined unprotected RAS and poppers. These and
other findings (summarized below) have led to the suggestion by Seage
and colleagues that the use of certain NMPDs, most notably poppers,
during UAI may actually increase the likelihood of HIV infection for any
given exposure (Seage et al. 1992). This possibility and possible
biological mechanisms for such an interaction are discussed below.
Association of Prior Change in NMPD Use and Changes in HighRisk Sexual Behavior and Incidence of HIV Infection. Among men
in the Chicago MACS/CCS cohort, those who reported the initiation of
popper use were significantly more likely to lapse to unprotected RAS
during the same 6-month period (Ostrow et al. 1993). Of particular
interest in terms of potential mechanisms was the finding that the
association was specific for lapsing and was not evident in the reverse
direction (e.g., men giving up popper use did not exhibit greater
propensity to move from unprotected to protected anal intercourse), and
was only observed for receptive (not insertive) anal sex with
nonmonogamous partners. While not accounting for all or most of the
100
reported lapses to unsafe RAS among men in the Chicago MACS/CCS
cohort, it does indicate that at least a portion of relapses to unsafe sex
occurs in conjunction with the return to use of a specific NMPD
historically associated with RAS among gay men (Mayer 1983). These
lapses, in turn, may be associated with a significant proportion of new
HIV infections given the high likelihood that the insertive partner is
already infected, and the possible facilitation of infection when poppers
are used during RAS.
Association of NMPD Use with Sexual Partners and Changes in
High-Risk Sex and Incidence of HIV Infection, Including WithinSubject Event Analysis Studies. None of the three prospective cohort
studies summarized here employed a within-subject comparison of events
that involved or did not involve specific NMPD use. However, the
Toronto Sexual Contact Study recently published data showing that only
NMPD use with sexual partners, in contrast to NMPD use outside sexual
encounters, was associated with increased scores on a sexual risk index
that summarized all reported sexual encounters (Calzavara et al. 1993).
However, less than 10 percent of reported substance use in this cohort
was not associated with sexual encounters. In multivariate analyses, only
the use of poppers, marijuana, or alcohol during sexual encounters were
significantly associated with higher sexual risk scores. When their study
participants were asked whether drugs or alcohol had an effect on their
sexual behavior, 45 percent said “yes,” and the majority of those
respondents reported that they were more likely to have more partners,
more likely to have casual partners, and less likely to use condoms. In
the Chicago MACS/CCS cohort (Ostrow et al. 1993), incident
seroconversion was associated with popper and cocaine use, although
specific use of those NMPDs with sexual partners was not separately
examined.
Intervention Data Which Simultaneously Assesses Changes in
NMPD Use and HIV Risk-Taking. While studies that simultaneously
assessed the effects of behavioral interventions on both substance use and
HIV risk-taking behaviors might provide the most persuasive evidence
for a direct causal link between NMPD use and sexual behavior change
among men at risk of HIV infection, such data are essentially nonexistent.
However, the Chicago MACS/CCS series of studies examined behavioral
and psychosocial correlates of changes in popper use (Ostrow et al.
1991). Men remitting from popper use were at significantly lower anal
sexual risk levels, more likely to have participated in individual
psychotherapy, and more likely to report higher levels of social
101
interaction and education than men continuing to use poppers. In
contrast, men lapsing to popper use were at higher anal sex risk levels,
perceived themselves to be at higher risk of AIDS, and had experienced
more adverse major life events (stressors) than men who consistently
abstained from popper use. While not a controlled intervention study,
these results indicate the potential for significant differences in sexual
behavior risk in conjunction with stopping or starting the use of poppers.
DISCUSSION
The studies reviewed above indicate that NMPD use is variably
associated with high-risk sex among gay/bisexual men. Important
variables that may affect the degree of consistency of these associations
include the extent of an individual’s NMPD use; sexual context of
NMPD use; intentions regarding changes in sexual behavior; the type of
outcome measurements obtained; geographic variables; and legal
constraints.
The extent of NMPD use, in terms of number, duration, and frequency of
the drugs used, and whether use is habitual or occasional, are important
considerations.
It should be ascertained whether the NMPD use is nonspecific or specific
for sexual contexts. For example, some NMPDs (such as poppers and
ecstasy) are sold primarily in settings where high-risk sexual activities
can take place with multiple partners (such as pornographic bookstores,
bars with backrooms, and gay bathouses), and are therefore more likely to
be associated with unprotected or multiple partnered sexual encounters if
studied only in terms of their global associations.
Measurement should be made regarding the stage of the individual’s
sexual behavior change effort—whether it is precontemplation, initiation
of change, consolidation of change, or maintenance/relapse—that is under
investigation (Peterson et al. 1992b). Temporary lapses to suppressed
high-risk sexual activities that were previously preferred may be
particularly susceptible to the disinhibiting effects of alcohol and
NMPDs.
Another variable to be considered is the nature of the outcome
measurements, i.e., whether they are self-reported sexual behaviors or
actual HIV infections, and whether they are cross-sectional, retrospective,
102
or prospective measures. While prospective studies that include objective
outcome measures, such as incidence of HIV infection rates, are
necessary to probe the causal mechanisms of any cross-sectional
associations between specific NMPD use patterns and sexual risk-taking,
the “riskiest” respondents are the most likely dropouts from prospective
studies due to either noncompliance or illness/death. Even the most
powerful study designs may significantly underestimate the associations
between NMPD use and high-risk sex and, conversely, to overestimate
any intervention effects unless differential attrition is adequately
controlled for in the analysis.
These limitations become even more serious when the study populations
move from the well-established, mostly white and well-educated, gay
identified male cohorts to more diverse and perhaps more vulnerable
populations of men who have sex with other men. Such populations
include men at extremely high risk of HIV infection, such as closeted
bisexual men, ethnic minority men (Easterbrook et al. 1993), gay youth,
homeless men, and men who engage in homosexual sex in exchange for
money, drugs, or food. Given the high levels of gay community
awareness concerning the dangers of high-risk sexual activities and the
widespread community warnings about the possibility of increased risk of
unprotected sex if intoxicated, it would not be surprising if men who do
not self-identify as gay or bisexual were more likely to engage in highrisk sex and combine NMPD use with sex (Peterson and McKinnon
1990). Also, men who engage in sex with anonymous partners or who
meet partners at settings where multiple casual sexual encounters can take
place are less likely to self-identify as gay or bisexual and more likely to
engage in high-risk sex (Peterson et al. 1992a).
Very few studies examine the impact of geographic factors, and
methodological differences make it difficult to compare findings from
various communities. Two studies performed by Gold and associates in
Sydney and Melbourne (Australia) found no association between
drinking and high-risk sex among gay men in Melbourne, but did find an
association among Sydney gay/bisexual men who were HIV-negative
(Gold et al. 1992). The investigators related these differences to the
concentration of gay/bisexual men and gay bars within central Sydney,
which made both NMPDs and casual sexual partners more available than
in Melbourne. Sydney is also a larger and less conservative city than
Melbourne, making it more attractive for gay/bisexual men to visit or
migrate there with the intention of engaging in sexual activities with
multiple partners.
103
Geographic differences found in the associations between high-risk sex
and NMPD use can be at least partially explained as social in origin;
differences in community norms regarding sexual and NMPD use
practices, legal prohibitions or restrictions on the sale or use of alcoholic
beverages or NMPDs in settings where sexual encounters may take place,
and local variations in the types of drugs available may all be reflected in
the resulting behavioral patterns of gay/bisexual men. For example,
amphetamine has never been as popular in the Midwestern United States
as in San Francisco, San Diego, or parts of New York City, and
amphetamine use has been traditional among specific sectors of the
gay/bisexual communities in those cities. To the extent that amphetamine
use has been popularized as a sex drug, one can expect it to be cited in
association with high-risk sex among subpopulations of gay/bisexual men
not yet committed to safer sexual practices. Similarly recent increases in
the popularity of the designer drugs ecstasy and ketamine (Special K)
among all-night partygoers might herald their association with high-risk
sexual activities.
In terms of legal constraints, it will be interesting to see if the nationwide
prohibition on sales of volatile nitrites in the United States, which went
into effect in mid-1992, has an effect on either the use of poppers or their
association with unprotected RAS among gay/bisexual men previously
reporting their use in sexual encounters. The recent proliferation of mailorder popper advertisements in gay publications and their continued
availability at pornographic bookstores and movie theaters in some cities
would suggest not.
Despite these reasons for variability in the observed associations between
NMPD use and high-risk sex among gay/bisexual men, there appear to be
several important trends in the existing studies reviewed above that
should be emphasized. One trend is that as the studies move up the
methodological hierarchy in terms of methods that are less likely to reveal
coincidental associations or be subject to retrospective recall biases, the
focus is increasingly on the “Big 3” drug categories of volatile nitrites,
marijuana, and cocaine. The studies reviewed here all took care to
control for the numbers of sexual partners or other potential confounders
that one might expect to be involved with commonly used NMPDs.
However, the lack of any positive findings from intrapersonal
comparisons of sexual events that did or did not involve NMPD use does
not permit rejection of mechanisms that involve common underlying
factors for both types of behaviors. In fact, even if one or more of the
direct causal mechanisms were underlying the observed associations
104
between specific NMPD use and high-risk sexual behaviors, any number
of intrapersonal and interpersonal factors could be contributing to the
observed behavioral relationships.
MECHANISMS
The author and other researchers (Cooper 1989; Leigh 1990; Stall et al.
1986) have speculated upon the types of underlying causal mechanisms
that might account for the frequently observed associations between
NMPD use and high-risk sexual activity among adolescents and gay or
bisexual men. While space constraints do not permit exhaustive
consideration of the evidence supporting or refuting them all, an
overview of the most frequently discussed potential causal mechanisms is
included in the following paragraphs.
Behavioral Disinhibition
Specific psychopharmacological activities of NMPDs and alcohol could
potentially cause the release of strongly suppressed behaviors. It is
perhaps important that alcohol, which has strong disinhibiting effects on a
variety of suppressed behaviors (particularly under stressful or anxious
conditions), has been linked to failure to use condoms by
adolescent/young adult heterosexuals (Cooper 1989) and gay/bisexual
men lapsing or relapsing to unprotected RAS after a period of change to
safer sexual practices (Kelly et al. 1991a). However, it is likely that
learned or innate expectancies about NMPD use effects are at least as
responsible for the observed associations as are specific
psychopharmacological effects (McKirnan and Peterson 1989b), again
highlighting the difficulties of distinguishing direct from indirect causal
pathways. It seems appropriate to begin to think of the combined
interaction of underlying personality characteristics, social/environmental
circumstances, and pharmacological effects that may ultimately underlie
the observed associations.
Intoxication or Cognitive Effects Models
Certainly recreational drugs and alcohol can affect information
processing and psychomotor task performance, and their use is associated
with accidents. If intoxicated enough, individuals can forget to take
precautions or improperly utilize condoms. The next question is whether
gay/bisexual men who combine NMPD use with high-risk sex decide first
105
that they are intending or desirous of the sexual or NMPD use activity. Is
it coincidental that most popular meeting spots for casual sexual
partnerships are places where alcohol and/or poppers are sold, or is this
the result of marketing that takes advantage of the synergistic interaction
of sex and drugs? The latter suggests the possibility that NMPDs differ
in their relative abilities to diminish unpleasant affects associated with
sexual intercourse (mainly anxiety and cognitive suppression of preferred
but feared activities) and maximize pleasurable feelings, such as euphoria
and sensual pleasure. The relative balance of such properties can be
conceptualized as the aphrodisiac index of each NMPD, but the
possibility should be considered that such pharmacological effects are
subject to differences in dosage as well as individual, social, and cultural
variations.
Social Influences and Social Setting Models
Use of NMPDs is frequently associated with special types of behaviors or
celebrations-a time out from normal social restraints and an inducement
to unusual or otherwise prohibited behaviors. To the extent that modern
societies have institutionalized times such as holidays and vacations, and
commercialized settings such as pick-up bars or crack houses where
NMPD use and sexual encounters are mutually available, people may
learn to associate NMPD use with high-risk sexual activities. Certain
personality types and heavy participants in either or both activities can be
expected to be most susceptible to social influence or setting effects on
subsequent behavioral patterns.
Intrapersonal Factors
Most studies concerned with possible causal relationships have made
mention of either the high-risk personality, characterized by impulsive
and sensation-seeking behavior, or the addictive personality. Either of
these personalities could predispose to both NMPD use and multiple
casual sex partners or high-risk sexual behaviors. Researchers’ inability
to disentangle any direct effects of NMPD use upon sexual behaviors
from cross-sectional or even prospective studies may ultimately be a
relatively unimportant consideration. Such personality characteristics are
probably represented more frequently among gay/bisexual men who
combine NMPD use with high-risk sexual behaviors than among men
who are abstinent from either type of activity, at least to the extent that
men with impulsive, sensation-seeking, or addictive personalities are
prevalent among gay/bisexual men.
106
The likelihood of intervention at the level of such basic personality
characteristics or behavior patterns seems diminishingly small. Still, the
examination of the use of specific NMPDs and their potential interaction
with sexual behavior and personality patterns may shed significant light
on the underlying bases of those associations, through both direct and
indirect paths.
Biological Interactions
In terms of specific drug use and HIV infection, the findings of Seage and
colleagues (1992) in terms of prevalent HIV infection, and the Chicago
MACS/CCS findings (Ostrow et al. 1993) concerning incidence of
infection both suggest potential biological interaction(s) between the
likelihood of infection and popper use during unprotected RAS. This
finding has been hypothesized as due to vasodilation within the rectal
mucosa, which increases the likelihood of rectal bleeding and the ability
of infected semen to enter the bloodstream of the receptive partner (Seage
et at. 1992). Equally plausible would be an acute immunosuppressive
effect of NMPD use during intercourse that would decrease immune
surveillance, thereby increasing the likelihood that HIV-infected cells
entering the bloodstream would infect the receptive partner. In fact, a
recent study of amyl nitrite exposure in HIV-seronegative men
demonstrated acute suppression of circulating natural killer cell activity
(Dax et al. 1991).
Researchers may never disentangle whether the observed associations
between NMPD use and sexual risk-taking behaviors or HIV infection
are the result of direct causal links, reflect common underlying
personality or coping styles, or indicate more severe psychopathology.
Obviously, a variety of mechanisms may be responsible for the
associations between sexual and NMPD use behaviors. Perhaps a single
causal mechanism may never be found. It may not matter in the long run
if researchers can find effective prevention interventions that work in
reducing the likelihood of unprotected sex occurring in the context of
NMPD use.
Prevention Implications
A crucial consideration at this time might be the issue of whether enough
is known about these associations to launch controlled intervention trials,
or whether further descriptive and causal studies need to be performed
before such interventions are tested. The author believes that, while
107
enough is known in general about these associations and their real risk of
propagating new waves of HIV infection among subpopulations of
gay/bisexual men, descriptive ethnographic studies among target
populations—in advance of or coincidental with the start of intervention
activities—are needed to increase the likelihood of beneficial outcomes.
This applies especially to those subpopulations of men who have sex with
men and who are particularly at risk of engaging in HIV-transmitting
sexual behaviors: antisocial risk takers; gay youth (especially street
youth and hustlers); high school dropouts; minority men who are not gay
self-identified; and men of lower socioeconomic background. The
consistent finding that significant proportions of HIV-transmitting sexual
behaviors among some specific populations are associated with NMPD
use argues for vigorous attempts at targeted intervention programs with
those populations.
Given the well-documented efficacy of community-based interventions
that use either cognitive-behavioral techniques within small groups
(Kelly et al. 1989) or community leadership norm-changing and
dissemination (Kelly et al. 1991b, 1992), it seems appropriate to examine
whether a combination of these approaches might be useful. While small
group interventions are effective in helping individuals to recognize the
behavioral patterns, including NMPD use, that place them at risk for
lapse to unprotected sexual behavior, the skills for negotiating safer sex
learned in such group interventions are frequently insufficient to prevent
lapse when alcohol or NMPD use is involved. On the other hand,
changing community norms about NMPD use through opinion leader
interventions and social marketing techniques may help at-risk
individuals avoid using substances or alter their use patterns to avoid
them in sexual encounters.
The author recommends development of community-based interventions
that combine knowledge about recreational drug use and sexual behavior
interventions into effective programs to target this specific aspect of HIV
prevention. Such interventions would explicitly acknowledge the links
between recreational substance use and high-risk sexual behavior, while
accepting the fact that there are no simple answers or explanations for
those links.
Alternative intervention approaches may be needed for specific at-risk
populations. For example, minority men who are not self-identified as
gay or bisexual are unlikely to be reached through gay community-based
media or social organizations. However, through preliminary
108
ethnographic research and focus groups, it is possible to identify those
information sources and community leaders whom such men respect and
find credible.
CONCLUSION
By targeting multiple, culturally appropriate approaches that seek to
support abstinence from both recreational substance use and unprotected
sexual intercourse, researchers may be able to prevent future epidemics of
sexually transmitted HIV and AIDS among specific at-risk populations.
By simultaneously evaluating both the process and outcome of such
interventions, researchers can learn how to mount more effective
interventions and also improve the understanding of the mechanisms
linking recreational substance use and HIV-transmitting sexual behaviors.
NOTES
1. Throughout this chapter, the abbreviation HIV is used to indicate
HIV-1 (or the human immunodeficiency virus, type 1), the
predominant form of HIV in all but Western Africa.
2. The Multicenter AIDS Cohort Study (MACS) in Chicago is
performed at the Howard Brown Health Center and Northwestern
University Medical School under the direction of John Phair, M.D.,
and Joan Chmiel, Ph.D., and is supported by NIAID Contract #N01Al-32535. The Coping and Change Study (CCS) of Men at-Risk of
AIDS is performed at the Chicago MACS study sites and the
University of Michigan, under the direction of John Maassab, Ph.D.,
David Ostrow, M.D., Ph.D., and Jill Joseph, Ph.D., M.D., and
supported by NIMH grant #R01 MH39346. I am extremely grateful
to the MACS/CCS participants and the staff and investigators of both
studies for making these observations possible. In particular,
Wayne DiFranceisco, M.S., prepared the table and figures used in
this chapter. By answering detailed questionnaires about sexual and
drug use behaviors every 6 months for almost 10 years, the study
participants have generously permitted us to access the most intimate
details of their lives. I hope that the resulting observations and
insights will help them and others in their battle to overcome and
recover from HIV infection.
109
REFERENCES
Calzavara, L.M.; Coates, R.A.; Raboud, J.M.; Farewell, V.T.; Read, S.E.;
Fanning, M.M.; and MacFadden, D.L. Ongoing high-risk sexual
behaviors in relation to recreational drug use in sexual encounters:
Analysis of 5 years of data from the Toronto Sexual Contact Study.
Ann Epidemiol 3:272-280, 1993.
Chmiel, J.S.; Detels, R.; Kaslow, R.A.; Van Raden, M.; Kingsley, L.A.;
Brookmeyer, R.; and the Multicenter AIDS Cohort Study. Factors
associated with prevalent human immunodeficiency virus (HIV)
infection in the Multicenter AIDS Cohort Study. Am J Epid 126:568577, 1987.
Cooper, M.L. “Adolescent High-Risk Sexual Activities and Alcohol
Use.” Paper presented at the Symposium on Recreational Drug Use
and High-Risk Sexual Behaviours at the 4th ASAM Conference on
Aids and Substance Abuse, Miami, FL, February 1989.
Dax, E.M.; Adler, H.; Nagel, J.E.; Lange, W.R.; and Jaffe, J.H. Amy1
nitrite alters human in vitro immune function. Immunopharmacol
Immunotoxicol 13:577-587, 1991.
Des Jarlais, D.C.; Friedman, S.R.; and Woods, J.C. Intravenous drug use
and AIDS. In: Ostrow, D.G., ed. Behavioral Aspects of AIDS. New
York: Plenum Publishing, 1991, pp. 139-155.
Easterbrook, P.J.; Chmiel, J.S.; Saah, A.J.; Hoover, D.R.; Kaslow, R.A.;
Rinaldo, R.A.; and Detels, R. Racial and ethnic differences in HIV-1
seroprevalence among homosexual and bisexual men. Am J Epid
138:415-429, 1993.
Gold, R.S.; Skinner, M.J.; Grant, P.J.; and Plummer, D.C. Situational
factors and thought processes associated with unprotected intercourse
in gay men. Psychol Health 5:259-278, 1991.
Kelly, J.A.; St. Lawrence, J.S.; and Brasfield, T.L. Predictors of
vulnerability to AIDS risk behavior relapse. J Consult Clin Psychol
59:163-166, 1991a.
Kelly, J.A.; St. Lawrence, J.S.; Diaz, Y.E.; Stevenson, L.Y.; Hauth, A.C.;
Brasfield, T.L.; Kalichman, S.C.; Smith, J.E.; and Andrews, M.E. HIV
risk behavior reduction following intervention with key opinion
leaders of population: An experimental analysis. Am J Public Health
81:168-171, 1991b.
Kelly, J.A.; St. Lawrence, J.S.; Hood, H.V.; and Brasfield, T.L.
Behavioral intervention to reduce AIDS risk activities. J Consult Clin
Psychol 57:60-67, 1989.
110
Kelly, J.A.; Sikkema, K.J.; Winett, R.A.; Solomon, L.J.; Roffman, R.E.;
Kalichman, S.C.; and Stevenson, L.Y. Outcomes of a 16-city field trial
of a community-level HIV risk reduction intervention. Proc 8th Int
Conf AIDS, Amsterdam, 1992.
Leigh, B. The relationship of substance use during sex to high-risk sexual
behavior. J Sex Res 27:199-213, 1990.
Leigh, B.C., and Stall, R. Substance use and risky sexual behavior for
exposure to HIV: Issues in methodology, interpretation, and
prevention. Am Psychologist 48:1035-1045, 1993.
Martin, J.L.; Dean, L.; Garcia, M.; and Hall, W. The impact of AIDS on a
gay community: Changes in sexual behavior, substance use, and
mental health. Am J Community Psychol 17:269-293, 1989.
Mayer, K.H. Medical consequences of the inhalation of volatile nitrites.
In: Ostrow, D.G.; Sandholzer, T.A.; and Felman, Y.M., eds. Sexually
Transmitted Diseases in Homosexual Men. New York: Plenum Press,
1983, pp. 237-242.
McCusker, J.; Westenhouse, J.; Stoddard, A.M.; Zapka, J.G.; Zorn,
M.W.; and Mayer, K.H. Use of drugs and alcohol by homosexually
active men in relation to sexual practices. J Acq Immune Defic Synd
3:729-736, 1990.
McKirnan, D.J., and Peterson, P.L. Alcohol and drug use among
homosexual men and women: Epidemiology and population
characteristics. Addict Behav 14:545-553, 1989a .
McKirnan, D.J., and Peterson, P.L. Psychosocial and cultural factors in
alcohol and drug abuse: An analysis of a homosexual community.
Addict Behav 14:555-563, 1989b .
Myers, T.; Rowe, C.J.; Tudiver, F.G.; Kurtz, R.G.; Jackson, E.A.;
Orr, K.W.; and Bullock, S.L. HIV, substance use, and related behavior
of gay and bisexual men: An examination of the Talking Sex Project
cohort. Br J Addict 87:207-214, 1992.
Ostrow, D.G. Barriers to the recognition of links between drug and
alcohol abuse and AIDS. In: Petrakis, P.L., ed. Acquired Immune
Deficiency Syndrome and Chemical Dependency. U.S. Public Health
Service Monograph. DHHS Pub. No. (ADM)87-1513. Washington,
DC: Supt. of Docs., U.S. Govt. Print. Off., 1987.
Ostrow, D.G.; Van Raden, M.; Kingsley, L.; Fox, R.; Dudley, J.; and
Kaslow, R.A. Drug use and sexual behavioral change in a cohort of
homosexual men. Proc 3rd Int Conf AIDS, Washington, DC, 1987.
Ostrow, D.G.; Beltran, E.; and Wesch, J. “Changes in Sexual and Drug
Use Behaviors, 1984-1989.” Proc Annu Meet Am Psychiatr Assoc
135:234, 1990.
111
Ostrow, D.G.; Beltran, E.; Chmiel, J.; Wesch, J.; and Joseph, J. Predictors
of volatile nitrite use among the Chicago MACS cohort of
homosexual men, Proc 7th Int Conf AIDS, Florence, 1991.
Ostrow, D.G.; Beltran, E.D.; Joseph, J.G.; DiFranceisco, W.; Wesch, J.;
and Chmiel, J.S. Recreational drugs and sexual behavior in the
Chicago MACS/CCS cohort of homosexually active men. J Substance
Abuse 5:311-325, 1993.
Peterson, P.L., and McKirnan, D.J. Gay identity, substance abuse and
AIDS risk. CME Syllabus and Proc 143rd Annu Meet Am Psychiatr
Assoc 143:186, 1990.
Peterson, J.L.; Coates, T.J.; Catania, J.A.; Middleton, L.; Hilliard, B.; and
Hearst, N. High-risk sexual behavior and condom use among gay and
bisexual African-American men. Am J Pub Health 82:1490-1494,
1992a.
Peterson, P.L.; Ostrow, D.G.; and McKirnan, D.J. Behavioral
interventions for primary prevention of HIV infection among
homosexual/bisexual men. J Primary Prev 12:19-34, 1992b .
Ray, O. Drugs, Society and Human Behavior. 2d ed. St. Louis: C.V.
Mosby, 1978.
Ross, M.W.; Wodak, A.; and Gold, J. Sexual behavior in injecting drug
users. In: Schonnesson, L.N., ed. Sexual Transmission of HIV
Infection: Risk Reduction, Trauma, and Adaptation. New York:
Harrington Park Press, 1992, pp. 89-104.
Schoenbaum, E.E.; Hartel, D.; Selwyn, P.A.; Klein, R.S.; Davenny, K.;
Rogers, M.; Feiner, C.; and Friedland, G. Risk factors for human
immunodeficiency virus infection in intravenous drug users. N Engl J
Med 321:874-879, 1989.
Seage, G.R.; Mayer, K.H.; Horsburgh, CR.; Holmberg, S.D.; Moon,
M.W.; and Lamb, G.A. The relation between nitrite inhalants,
unprotected receptive anal intercourse, and the risk of human
immunodeficiency virus infection. Am J Epidem 135:1-11, 1992.
Stall, R.; McKusick, L.; Wiley, J.; Coates, T.J.; and Ostrow, D.G.
Alcohol and drug use during sexual activity and compliance with safe
sex guidelines for AIDS: The AIDS Behavioral Research Project.
Health Educ Q 13:359-371, 1986.
Stall, R., and Wiley, J. A comparison of alcohol and drug use patterns of
homosexual and heterosexual men: The San Francisco Men’s Health
Study. Drug Alcohol Depend 22:63-73, 1988.
Watson, R.D., ed. Cofactors in HIV-1 Infection and AIDS. Boca Raton,
FL: CRC Press, 1990.
112
Weatherbum, P.; Davies, P.M.; Hickson, F.C.I.; Hunt, A.J.; McManus,
T.J.; and Coxon, A.P.M. No connection between alcohol use and
unsafe sex among gay and bisexual men. AIDS 7:115-119, 1993.
AUTHOR
David G. Ostrow, M.D., Ph.D.
Center for AIDS Intervention Research
Department of Psychiatry and Mental Health Sciences
Medical College of Wisconsin
8701 Watertown Plank Road
Milwaukee, WI 53226
113
Drug Use and HIV Risk Among
Male Sex Workers: Results of
Two Samples in San Francisco
Dan Waldorf
INTRODUCTION
Shortly after it was discovered that the human immunodeficiency virus
(HIV) could be transmitted by sexual activities, there was concern that
prostitutes might be active HIV transmitters and might also be vectors of
transmission to low-risk groups such as heterosexual men, women who
were not prostitutes, and children. The first concerns were about female
prostitutes, particularly after it was learned that prostitutes who worked
along arterial highways in several African countries were transmitting
HIV to men who traveled such routes. Somewhat later, male prostitutes
also were considered as a potential high-risk group and a possible vector
to bisexual family men, their wives, and children. This chapter
summarizes the research that describes the contexts of HIV risk among
prostitutes, with a particular emphasis on males.
HIV SEROPREVALENCE TESTING AMONG PROSTITUTES
In response to the concern about female prostitution and the spread of
HIV in the United States, the Centers for Disease Control and Prevention
(CDC) undertook eight different studies of 1,396 female sex workers in
the United States during 1986. A summary of these findings was
reported in Scotland during 1989 and was subsequently published in
1990 (Darrow et al. 1990). The studies found that only 4.8 percent of the
non-injecting drug using (non-IDU) female prostitutes were HIV positive.
Among IDU female prostitutes, the percentage who were HIV positive
was considerably greater at 19.9 percent. Among the 8 sites for the
studies, HIV rates varied considerably for IDU female prostitutes, from 0
percent to 58.3 percent. The highest percentages of HIV-positive IDU
female prostitutes were in New Jersey (42.9 percent in southern New
Jersey and 58.3 percent in northern New Jersey) and Miami (26.6
percent). In Atlanta, the percentage was only 1.5 percent.
114
HIV seroprevelance studies among female sex workers outside the United
States reveal rates of 0 to 88 percent. The highest rates have been in
Nagoma, Rwanda, where 88 percent of 33 female prostitutes (IDU status
unknown) were found to be HIV positive. In a study in Zurich only
1 percent of 103 prostitutes who were not IDUs were HIV positive, while
78 percent of the 18 IDU prostitutes were positive. Most of the studies
that find high rates of HIV infection are those conducted in Africa
(Darrow et al. 1990).
In general, there have been far fewer HIV seroprevalence studies of male
sex workers for unknown reasons. Perhaps it is because most people do
not realize how widespread the practice is and how criminal justice
agencies respond to male prostitution; male prostitutes are not arrested for
solicitation at the rate that females are arrested. A review of the literature
reveals only five HIV seroprevalence studies of male prostitutes-one in
New York City, another in New Orleans, a third in northeast Italy, and
two (the most recent) in Atlanta. Chaisson and associates were the first to
offer findings about male prostitutes. During 1986 and 1987, tests were
conducted on 84 male prostitutes appearing at a New York City sexually
transmitted disease (STD) clinic; more than half (52) reported exchanging
sex for money with females only and 32 with males only. Among those
who reported receiving money for sex from men, 17 of 32 (53 percent)
were HIV positive, and 1 of 8 (12 percent) who reported intravenous (IV)
drug use were positive (Chaisson et al. 1988).
Morse and associates conducted the study of New Orleans male street
prostitutes in 1988 and 1989 (Morse et al. 1991). They found that
17.5 percent of 211 prostitutes were HIV positive, but there were no
differences in rates of seropositivity between African-American and white
prostitutes. Among IDUs (20.2 percent of 109), there was a slightly
higher rate than among non-IDUs (14.7 percent of 102), but the
differences were not statistically significant.
The third study took place in northeast Italy during 1987 and 1988 and
included only 27 male prostitutes; 1 of 4 (25 percent) IDUs were positive,
and only 1 of 7 (14 percent) homosexuals and 1 of 16 (6 percent)
transvestites were positive (Tirelli et al. 1987, 1988).
The last two studies were funded by CDC and undertaken in Atlanta.
The first, which was of 235 male street prostitutes (hustlers) working in
1988 and 1989, found that 29 percent were HIV positive. IDUs were
HIV positive at greater rates than non-IDUs: 35 percent of 120 IDU male
115
prostitutes were HIV positive, as compared with 23.5 percent of the
115 non-IDUs. African Americans had higher rates than whites
(33.9 percent versus 28.6 percent); self-identified homosexuals
(43.9 percent) and bisexuals (35.3 percent) had greater seroprevalance
rates than heterosexuals (17.9 percent). The longer respondents had
engaged in prostitution, the greater the prevalence of HIV (Elifson et al.
1993a).
The second study in Atlanta took place during 1990 and 1991 and tested
53 transvestite prostitutes (Elifson et al. 1993b). This study found that
more than 2 out of 3 (68 percent) were positive for HIV, and more than
7 out of 10 had seromarkers for syphilis (79 percent) and hepatitis B
(76 percent). Among the African-American prostitutes in this study,
seromarkers for syphilis were significantly associated with HIV infection.
INJECTION DRUG USE OF MALE SEX WORKERS
In general, knowledge about the IDU behavior of male sex workers is
sparse and lacking in detail. There have been only five studies that have
reported drug use of male sex workers in the United States with any detail
(Elifson et al. 1993a, b; Inciardi et al. 1991; Morse et al. 1991; Pleak et
al. 1990). There have, however, been two other studies that report
samples from Amsterdam (Coutinho 1988) and Edinburgh (Plant 1990;
Thomas 1990).
In the U.S. studies, there are mixed findings of low and high percentages
of IDU. Pleak found low levels among 52 street workers interviewed in
New York City (3 of 52, or 5.8 percent), while both Morse and Elifson
found relatively high levels. Among the street workers studied by Morse,
more than half (51.7 percent) were IDUs. Among Elifson’s samples,
51 percent of a sample of 235 male street sex workers were IDUs, but
only 6 percent of 53 transvestite street workers were IDUs.
Inciardi, in a large study of seriously delinquent youth in Miami, found
that 8 of 20 (40 percent) male prostitutes were presently IDUs and
another 6 (30 percent) had been IDUs in the past; if one combines the two
groups, then 70 percent were IDUs.
Among the European studies, again there are both low and high
percentages. An Amsterdam study found only low levels; 1 of 37
(or 2.7 percent) of male sex workers working in brothels were IDUs. The
116
Edinburgh study found greater percentages—13 of 110
(11.8 percent)—and reported that female sex workers (28 of 101, or
27.7 percent) were more likely to be IDUs than males.
TWO STUDIES CONDUCTED IN SAN FRANCISCO
During the period from 1987 through 1991, two different studies of male
sex workers were undertaken in San Francisco. The first study, from
1987 to 1988, interviewed 360 male sex workers. The study was funded
by the National Institute on Drug Abuse (NIDA) and focused on injection
drug use and syringe-sharing practices. The second study was undertaken
between 1989 and 1991 and interviewed 552 male sex workers. This
study was funded by the National Institute of Child Health and Human
Development. The study focused primarily on condom use but did ask
questions about drug use as well. Neither study sample was random, as
there was not enough information available to use such sampling
methods. Both studies did, however, attempt to include substantial
numbers of all the major types and subtypes of sex workers. It should be
noted that these are the largest samples ever done of male sex workers.
The sampling methodology for the two studies was similar. The
existence of different types of male sex workers was established from
ethnographic studies conducted earlier by one of the project staff, and
interview quotas were established for each type of sex worker. The first
study identified two general types of sex workers and seven subtypes.’
The two general types were hustlers and call men. The first type,
hustlers, solicited clients face-to-face, most often in public places-on the
streets, in bars, and at sex magazine and paraphernalia shops. The second
type solicited clients most often over the telephone; this group is known
as call men.
Among hustlers or workers who solicit clients face-to-face in public
places, there are three subtypes: gay-identified youth, trade hustlers, and
drag queen hustlers.
Gay-identified youth are young, gay-identified males who present
themselves as being innocent and naive in the ways of the world but who
participate in a wide spectrum of sexual activities. A substantial number
of this group are juvenile runaways who generally present themselves as
being adults when confronted by authorities.
117
Trade hustlers are usually heterosexual (straight) or bisexual males who
typically express a rather ostentatious and aggressive maleness, trade sex
for money, do not admit homosexual inclinations, and often profess no
enjoyment of sex with men. They usually offer passive oral sex and
seldom engage in anal intercourse.
Drag queen hustlers are transvestites and transsexuals who assume an
exaggerated female identity, usually specialize in oral sex, and are
localized in particular streets and bars in the Tenderloin, a sex trade zone.
The second type of prostitutes was call men who solicited clients over the
telephone. Very often this type of sex worker is middle class, with more
education and more stable living arrangements than prostitutes who
solicit in public places. There were four subtypes of call men: call book
men, models and escorts, erotic masseurs, and stars of the erotic industry.
Call book men are predominantly gay-identified and bisexual men who
generally work from a call book of regular customers and who provide a
wide range of sexual services. New clients usually are generated through
referrals from other clients.
Models and escorts are men who generally obtain customers through
advertisements placed in mainstream and special interest publications.
They often entertain clients socially as well as sexually and participate in
a broad range of sexual activities. They tend to develop networks of
regular customers and also may operate simultaneously from a call book.
Erotic masseurs are men who advertise for new clients while serving
regular customers and provide some elements of the legitimate massage
business with an erotic twist. Very often they are certified by licensed
massage schools and are guided by well-developed philosophical
rationales for their work. Their primary sexual services are masturbation
and oral sex, but they may participate in anal intercourse with selected
clients. They charge relatively low prices compared to other call men and
may be the least expensive in the industry. Some masseurs also operate
out of athletic clubs.
Stars of the erotic industry are individuals who are well known from
stage, screen, and magazines. They are an elite group among male sex
workers, are very small in number, and charge the highest prices in the
industry. The most common of this type of prostitute in the San
Francisco Bay area are erotic dancers who strip in local nightclubs and
118
theaters for both male and female audiences and often provide sexual
services to both male and female clients.
It should be noted that the first study, from 1987 to 1988, did not locate
male sex worker agencies operating in San Francisco, although there
were several female agencies operating. Male sex worker agencies had
operated in the past in San Francisco, but police had conducted
crackdowns against such agencies just prior to the first study, and not one
was operating at that time.
The two additional subtypes identified during the second study were drag
queen call men and agency-affiliated call men. During 1989,
transvestite/transsexual sex workers began to place advertisements in
specialty newspapers and magazines offering their services; this was a
new development. Both sex worker agencies began operating in 1990
and advertised regularly up to the end of data collection in 1991. The
second study gained unusual access to these agencies and interviewed all
the workers at both agencies. Therefore, the second study conducted
interviews of nine different subtypes of male sex workers rather than the
seven subtypes interviewed during the previous study (table 1).
Drag queen call men are transvestite/transsexual men who generally
provide the same kinds of services as drag queen hustlers but use
advertisements and telephones to solicit clients.
Agency-affiliated call men are gay-identified men who operate
similarly to models and escorts but get their clients through an agency
rather than by advertisements.
Concurrent with the acquired immunodeficiency syndrome (AIDS)
epidemic, there has been a paradoxical increase in the number of
advertisements for call men appearing in San Francisco’s gay and erotic
newspapers. There may be many more call men than there were before
the AIDS epidemic. Explicit erotic massage advertisements are also a
fairly recent phenomena in San Francisco newspapers and have increased
dramatically since the AIDS epidemic began. This increase in
advertisements for sex workers seems to indicate that sex in San
Francisco may have become much more commercialized with the AIDS
epidemic.
119
TABLE 1. Number of Completed Interviews According to Type of Sex
Workers for the Two Samples
First Sample
(N = 360)
Second Sample
(N = 552)
Hustlers
Gay-Identified Youth
Trade Hustlers
Drag Queen Hustlers
Subtotal
60
60
60
180
76
75
75
226
Call Men
Call Book Men
Model/Escorts
Erotic Masseurs
Drag Queen Call Men
Agency Affiliated
Erotic Stars
Other
Subtotal
60
60
59
1
180
75
75
75
69
25
3
2
324
Type of
Sex Worker
Location Techniques
Both studies utilized a variety of location techniques: observation of
street hustling activities, personal contacts of a gay ethnographer, chain
referrals (Biernacki and Waldorf 1981; Watters and Biernacki 1989), and
telephone solicitation of telephone numbers listed for call men,
model/escorts, and erotic masseurs. The study also employed former sex
workers, gay and lesbian interviewers who had good contacts in sex work
and gay communities. Transvestite/transsexuals were interviewed by
female interviewers, a tactic that facilitated the development of rapport.
All respondents were screened with a questionnaire that asked questions
about the length of time they had worked as sex workers and the number
of clients they had served during the previous 6 months. In general, the
typology of various types of sex workers served as a theoretical overview
and guide to locate sex workers for the study.
In general, the studies were represented in the gay community by a gayidentified ethnographer who worked on the project for the full duration of
120
the first study and most of the second. This was a conscious strategy to
offset possible hostile reactions in the gay community to nongay staff.
From the onset, the study endeavored to develop good relationships in the
gay community to facilitate interviewing.
During the second survey, each respondent was asked if he had been
interviewed during the previous study conducted 2 years earlier. Only
1 in 12 (8.0 percent) reported that they had been interviewed previously;
this seems to indicate that the population of male sex workers may be
large or transitory with a high turnover. The high turnover of the
population was further revealed in data about other cities where workers
had sold sex. Only a third (34 percent) reported that they had worked as
sex workers only in San Francisco, while two-thirds (66 percent)
mentioned at least two cities. More specifically, 8 percent mentioned
6 different cities, 14 percent 5 cities, 25 percent 4 cities, and 40 percent
3 cities. Hustlers had worked in more cities than call men. The most
frequently mentioned other cities were Los Angeles and New York City.
SELF-REPORTS ABOUT AIDS AND ARC DIAGNOSES AND
HIV TESTING
The first study was conducted during the period when HIV testing was
very controversial in San Francisco. Many people in the gay community
were suspicious of and hostile to agencies and projects that proposed
widespread HIV testing. Testing was considered during that study, but
the idea was abandoned because the climate around testing was so
volatile at the time. The study did gather data about self-reports of
medical diagnoses of AIDS and AIDS-related complex (ARC) and results
from HIV testing from both samples and found very consistent findings.
During the first study, self-reports revealed that 12.7 percent were either
HIV positive or had been diagnosed as having AIDS or ARC. One in
20 (5.5 percent) reported that they had been diagnosed as having AIDS or
ARC, and another 7.2 percent reported that they were HIV positive. Call
men reported being HIV positive and diagnosed with ARC or AIDS
slightly more often than hustlers, but the differences were not statistically
significant.2
The second study did not consider or plan HIV testing, but interviewers
asked about diagnoses and the results of the last HIV test. Reports were
similar to the first sample; 12.1 percent reported being either HIV
positive or diagnosed, and again call men reported being positive or
121
diagnosed slightly more often than hustlers. In general, self-reports may
under-estimate the actual rates.
DRUG USE SCENES IN SAN FRANCISCO
The Tenderloin is an area in San Francisco where a large number of male
prostitutes work and live. Unlike other cities in the United States, there
appears to be considerable diversity of drug use in San Francisco. For
example, phencyclidine (PCP) and methamphetamines, which are only
occasionally available in many cities and appear to be used by only small
percentages of people in most cities, are readily available in San
Francisco and are widely used in selected drug scenes. In the Mission
District, a Hispanic community, PCP is readily available and is used quite
extensively by Hispanic youth, but it is seldom used by African
Americans or whites. Methamphetamines are readily available in the
Tenderloin and Castro Districts of the city, which are areas where many
male sex workers live and work, and are used regularly by gay-identified
males and sex workers who live in those communities. These drugs are
in addition to staple drugs such as marijuana, cocaine hydrochloride, and
heroin, which are also readily available. Despite the ready availability of
methamphetamines, there is very little “ice” (a smokeable
methamphetamine) available in the city, so use of that drug is low.
Cocaine hydrochloride has been widely available since the mid-1970s
and continues to be readily available up to the present day. Rock cocaine,
or crack, was relatively slow arriving in San Francisco; it did not become
widespread until 1988 and 1989, unlike in other cities such as Miami,
New York, and Los Angeles. It was not readily available during the first
study but was during the second study.
The use of amyl nitrates (or poppers), widely used in the San Francisco
gay community during the 1970s and the early 1980s, declined sharply
during the early days of the AIDS epidemic when it was believed that
AIDS was a direct result of amyl nitrate use. Amyl nitrates never
reemerged as a widely used drug.
Injection Drug Use
Drug use was generally high for both samples; virtually everyone
interviewed for both studies reported some illicit drug use. Respondents
in the first study reported very high incidence of injection drug use; more
122
than two-thirds (67.8 percent) of the hustlers and nearly two-fifths
(38.9 percent) of call men reported having ever injected (table 2). Levels
of educational achievement were associated with injection drug use; men
who reported low levels of education more often reported injection drug
use than those with high levels. In the second sample, the percentage of
those reporting ever injecting drugs was even greater; nearly
three-quarters (74.8 percent) of the hustlers and more than half
(56.1 percent) of the call men had injected illicit drugs (table 3). In both
instances, the most frequently reported injected drugs were
methamphetamines, cocaine, and heroin.
Syringe Sharing and Problems Designing Questions To
Explore the Topic
In general, there were considerable problems formulating and asking
questions about syringe sharing. The surveys were limited by the amount
of time required to gather good quality data, the caution and reluctance
that many drug injectors feel about admitting unsafe injection practices,
and the variable ways that IDUs define sharing. These limitations all
became evident during the first study.
The initial wording for all syringe questions (32 questions in all) in the
first study was devised to include a simple screening question that, if
answered ‘yes,’ was followed by a series of questions that explored who
respondents shared with, occasions of sharing, and situations of sharing.
The screening question was worded as follows: “Have you ever shared
needles or syringes with any person? By needle sharing, we mean
sharing with anyone.” The codes for this answer were simple
enough-shared or not shared. And if a person said that he did not share,
then interviewers skipped the series of other sharing questions.
Before actual interviewing took place for the first study, the questions
were tested with 20 respondents. Some of the other sharing questions
were revised, but no problems were observed with the screening question.
After the first 90 interviews were completed, the results were reviewed
and investigators came to the conclusion that respondents were
underreporting syringe sharing. Underreporting was suspected because
of the sizeable proportion of injectors who never reported sharing.
Roughly 30 percent of those who reported drug injection also reported
that they had never shared, although some of these responses were from
men who reported injecting more than 1,000 times. In San Francisco,
123
TABLE 2. Injection Drug Use by Type of Sex Worker—First Sample
(N=360) 1987-1988
Injection
Drug Use
Ever Injected
Drugs
Drugs Injected*
Methamphetamines
Cocaine
Heroin
Ever Shared
Syringes”
Ever Shared
Syringes at
Shooting
Gallery’*
Ever Shared
Syringes at
Bath House or
Sex Club**
Type of Sex Worker
Hustlers
(N = 180)
Call Men
(N = 180)
Total
(N=360)
122 67.8%
70 38.9%
192 53.3%
113 92.6%
72 59.0%
52 42.6%
62 88.6%
45 64.3%
26 37.1%
175 91.1%
117 60.9%
78 40.6%
88 72.1%
51 72.9%
139 79.4%
30 24.6%
12 23.5%
42 30.2%
9 10.2%
12 23.5%
21
15.1%
* Of those sex workers who reported ever injecting drugs.
** Of those sex workers who reported injecting drugs and ever
sharing syringes.
syringes were not readily available; California has had a prescription law
for syringes since 1929 and a paraphernalia law since 1958, and illicit
syringes were selling for $3 to $5 on the street.3 Furthermore, the
investigator’s long experience with IDUs over several studies
(1966-1986) indicates that it is the exceptional person who does not share
syringes, especially when they first use the drug. Beginning users
seldom, if ever, have their own syringes; they almost always share with
persons who introduce them to the drug and the injection.
124
TABLE 3. Injection Drug Use by Type of Sex Worker--Latest Sample
(N = 552) 1989-1991
Injection
Drug Use
Ever Injected
Drugs
Number of
Times Injected*
1-10
11-250
250-1,000
> 1,000
Type of Sex Worker
Hustlers
(N = 226)
Call Men
(N = 326)
Total
(N = 552)
169 74.8%
183 56.1%
352 63.8%
27.9%
15.3%
14.2%
42.6%
74 21.0%
61 17.3%
47 13.4%
170 48.3%
157 92.9%
110 65.1%
110 65.1%
159 86.9%
94 51.4%
93 50.8%
316 89.8%
204 58.0%
203 57.7%
62 36.7%
5 1 30.2%
56 33.1%
118 64.4%
47 25.6%
18 9.8%
180 51.1%
98 27.8%
74 21.0%
Shared Syringe
The First Time
Injected*
97 57.4%
101 55.2%
198 56.3%
Ever Used
a Used Syringe
Without
Cleaning It
97 57.4%
96 52.5%
193 54.8%
Interviewed in
Previous
Study
1 7 7.5%
27
Drug Injected.
Methamphetamine
Cocaine
Heroin
Number of
Times Injected
Last Week
Zero
1-5
6 or more
23
33
21
92
13.6%
19.5%
12.4%
54.4%
51
28
26
78
8.3%
44
* Of those sex workers who reported ever injecting drugs.
** Of the total number of sex workers.
125
8.0%
By the 100th interview, the screening question was revised to open up the
answers. The revised questions were: “Tell me about the first time you
ever injected any illicit drugs. How did you get the needle or syringe to
inject the drug?” “For how long did you ever use other people’s syringes
or works?” “How many people did you share a syringe with?” “Did you
ever share needles with anyone else after the first time?” At the
conclusion of the remaining 260 interviews, the same kinds of disparities
were found as in the first 100: roughly 30 percent who reported injection
stated that they had never shared. In other words, the revised wording
did not elicit any better responses than the simple screening question had.
The investigator decided to simply take people at their word and used two
codes for the question: shared or did not share. However, the investigator
was suspicious of these reports and believes that a sizeable number of the
respondents underreported sharing.
Despite these problems, the study found that there was considerable
sharing and a good deal of it took place in communal settings such as
shooting galleries, sex clubs, and bath houses. More than 7 out of
10 (79.4 percent) of 192 injectors from the first sample reported sharing
syringes. Syringe sharing occurred among a variety of relationships and
social situations: with friends, roommates, and intimates and in
communal settings such as shooting galleries, sex clubs, and bath houses.
More than 3 out of 10 (30.2 percent) of those who had shared syringes
reported sharing in shooting galleries, and a sixth (15.1 percent) reported
sharing in bath houses and sex clubs. (Refer to Waldorf et al. 1990 and
Waldorf and Murphy 1990 for more detailed findings.)
For the second survey, which focused primarily on condom use but also
asked about drug injection, the syringe-sharing questions were redesigned
and consolidated into two questions. Here is how the questions were
worded: “Tell me about the first time you ever injected any drug. Did
you have your own syringe, or did you use someone else’s?” “Have you
ever been forced by circumstances to use a needle that someone else had
used previously without cleaning it with bleach?”
This revision produced the following results for the second survey:
nearly two-thirds of the sample reported injecting drugs (63.8 percent),
and only 56.3 percent of injectors reported that they shared syringes the
first time they injected. Roughly half (54.8 percent) reported that they
had used another person’s syringe without cleaning it with bleach.
126
The investigator believes that the questions in both studies were not as
good or extensive as they should have been to overcome the reluctance of
respondents to report unsafe injection practices and gather good reliable
data. Perhaps a future study should ask a number of questions about the
setting and occasion of first injection and subsequent injections to gather
more details about first injection and other occasions when sharing might
take place. Hopefully, it will reveal better results.
Condom Use
In addition to HIV risks involved with drug injection, male sex workers
also are at risk for unsafe sexual practices. Self-reports about condom use
were generally high in the second survey; more than 9 out of 10 reported
the use of condoms in the previous year, and nearly three-quarters in the
previous week they had sex. There were significant differences in the
rates of condom use for different types of sexual behaviors, different
partners, and different types of sex workers. Hustlers, or sex workers
who solicit clients in public places, reported less frequent condom use
during anal intercourse than call men. Oral sex with condoms varied by
levels of education, but unexpectedly; men who reported attending some
college (usually call men) also reported that they used condoms during
oral sex less frequently than men with lower education levels.4 The
variation by different sexual partners was as expected; condoms were
used more frequently with customers than intimate or other nonpaying
partners (Waldorf and Lauderback 1992). It also was found that condoms
slip off and break 2.39 percent of the time and that hustlers report more
condom failures than call men (Waldorf and Lauderback 1993). These
findings suggest that many male prostitutes, and hustlers in particular,
regularly engage in unsafe sexual practices.
AIDS Prevention Efforts Among Sex Workers
San Francisco has generally done a good job of providing AIDS
prevention information, condoms, and bleach to sex workers in the city.
During the second study, sex workers reported they had received free
condoms from 15 different social agencies, as well as information about
HIV risks. In many bars in the Tenderloin, condoms are provided in
large bowls as a regular service. During January 1992, Prevention Point,
the large and successful needle exchange program, began to provide
syringe exchanges near Polk Street with a particular focus on sex workers
in the area. Another site in the Tenderloin generally did not reach male
sex workers.
127
SUMMARY AND CONCLUSIONS
In general, large numbers of male sex workers in San Francisco
participate in considerable injection drug use, most particularly
methamphetamines and cocaine. It is expected that much of this has to
do with the particular drug scenes of Polk Street, the Tenderloin, and the
Castro District, neighborhoods that have very active drug scenes and
where many sex workers work and live. Male sex workers also report
considerable syringe sharing, but it is felt that many workers
underreported sharing. Both practices, along with unsafe sexual
activities, place many sex workers at risk of HIV. Other studies of male
street sex workers in Atlanta and New Orleans also have found
substantial percentages of IDUs among street workers in those cities, but
this was not the case in New York City.
With the exception of the San Francisco study, most studies of male sex
workers have focused on street workers and have neglected call men, or
men who solicit clients over the telephone. To date there has not been
any study that has tested call men for HIV, and none of the other studies
included them in their samples. Any future work should consider call
men as well as street workers.
Studies that undertake to explore syringe sharing should take care in how
questions are worded and attempt to find ways to overcome sex workers’
reluctance to report unsafe injection practices. Simple screening
questions will not be adequate. The author recommends that future
research use indepth qualitative questions to explore the phenomena.
NOTES
1. In general, most male sex workers will take clients wherever they find
them-in the street, in bars, at social gatherings, via
advertisements-but most have some general modus operandi that
allows them to be categorized by the principal method they use to
locate clients.
2. Whenever differences are mentioned they will be statistically
significant (at the .01 level). When differences are not significant,
they will be designated as not statistically significant.
128
3. Prevention Point, the large needle exchange program that began its
exchanges in the Tenderloin, was not operating at the time, so there
was no sudden availability of syringes.
4. Shortly before and during the second survey there was an ongoing
debate in the gay community, particularly in the gay press, about the
safety of oral sex. Many believed that HIV could not be transmitted
via oral sex, and this may have had some impact on call men’s
attitudes and behaviors about condom use and oral sex.
REFERENCES
Biernacki, P., and Waldorf, D. Snowball sampling: Problems and
techniques of chain referral sampling. Sociol Meth Res 10
(2):141-163, 1981.
Chaisson, M.A.; Lifson, A.R.; Stoneburner, R.L.; Ewing, W.;
Hildebrandt, D.; and Jaffe, H.W. HIV-1 seroprevalence in male and
female prostitutes in New York City. Abstract. 6th International
Conference on AIDS, Stockholm, Sweden, June 1988.
Coutinho, R.A.; van Andel, R.L.M.; and Rysdyk, T.J. Role of male
prostitutes in spread of sexually transmitted diseases and human
immunodeficiency virus. [Letter] Genitourin Med 64:207-208, 1988.
Darrow, W., and the Centers for Disease Control Collaborative Group for
the Study of HIV-1 in Selected Women. Prostitution, intravenous drug
use, and HIV-1 in the United States. In: Plant, M., ed. AIDS, Drugs
and Prostitution. London: Tavistock/Routledge, 1990.
Elifson, K.W.; Boles, J.; and Sweat, M. Risk factors associated with HIV
infection among male prostitutes. Am J Pub Health 83(1):79-83,
1993a .
Elifson, K.W.; Boles, J.; Posey, E.; Sweat, M.; Darrow, W.; and
Elsea, W. Male transvestite prostitutes and HIV risk. Am J Pub Health
83(2), 1993b .
Inciardi, J.A.; Pottieger, A.E.; Forney, M.A.; Chitwood, D.D.; and
McBride, D.C. Prostitution, IV drug use, and sex-for-crack exchanges
among serious delinquents: Risks for HIV infection. Criminology
29(2):221-235, 1991.
Morse, E.B.; Simon, P.M.; Osofsky, H.J.; Balson, P.M.; and
Gaumer, H.R. The male street prostitute: A vector for transmission of
HIV infection into the heterosexual world. Soc Sci Med
32(5):535-539, 1991.
129
Plant, M. Sex work, alcohol, drugs and AIDS. In: Plant, M., ed. AIDS,
Drugs and Prostitution. London: Tavistock/Routledge, 1990.
Pleak, R., and Meyer-Bahlberg, H. Sexual behavior and AIDS knowledge
of young male prostitutes in Manhattan. J Sex Res 27:557-587, 1990.
Thomas, R.M. AIDS risks, alcohol, drugs, and the sex industry: a
Scottish study. Plant, M., ed. In: AIDS, Drugs and Prostitution
London: Tavistock/Routledge, 1990.
Tirelli, U.; Vaccher, E.; Diodatao, S.; Bosio, P.; De Paoli, P.; and
Crotti, D. HIV infection among female and male prostitutes. Abstract.
Third International Conference on AIDS, June 1-5, 1987. Washington,
DC: U.S. Department of Health and Human Services and the World
Health Organization, 1987.
Tirelli, U.; Vaccher, E.; Bullian, P.; Saracchini, S.; Errante, D.; Zagonel,
V.; and Serraino, D. HIV-1 seroprevalence in male prostitutes in
Northeastern Italy. [Letter] J Acquir Immune Defic Syndr 1:414-415,
1988.
Waldorf, D., and Murphy, S. IV drug use and syringe sharing practices of
call men and hustlers. In: Plant, M., ed. AIDS, Drugs and Prostitution.
London: Tavistock/Routledge, 1990.
Waldorf, D.; Murphy, S.; Lauderback, D.; Reinarman, C.; and
Marotta, T. Needle sharing among male prostitutes: preliminary
findings of the Prospero Project. J Drug Issues 20:309-334, 1990.
Waldorf, D., and Lauderback, D. The condom use of male sex workers in
San Francisco. AIDS Pub Pol J Summer: 108-119, 1992.
Waldorf, D., and Lauderback, D. Condom failure among male sex
workers in San Francisco. AIDS Pub Pol J Summer 8(2):79-90, 1993.
Watters, L., and Biernacki, P. Targeted sampling: options for the study of
hidden populations. Soc Probl 6:1989.
ACKNOWLEDGMENTS
The data for this paper were generated with support by grants from the
National Institute on Drug Abuse (#R01-DA04535) and the National
Institute of Child Health and Human Development (#5 R01 HD26282).
130
AUTHOR
Dan Waldorf, M.A.
Institute for Scientific Analysis
2719 Encinal Avenue
Alameda, CA 94501
131
HIV Risk in Drug-Using
Adolescents
Vincent L. Smeriglio
Studies of adolescents and young adults have examined the prevalence of
acquired immunodeficiency syndrome (AIDS) and human
immunodeficiency virus (HIV) infection, knowledge and attitudes about
HIV and AIDS, alcohol and other drug use, sexual behavior, and sexually
transmitted diseases (STDs). Findings document substantial risks for
HIV and AIDS and provide insights into similarities and differences in
risk across subgroups of adolescents. The current body of knowledge on
adolescents and their risk behaviors is valuable in the development and
targeting of interventions (e.g., Bowler et al. 1992; DiClemente 1992).
Design and targeting of HIV prevention strategies would be enhanced by
further elucidation of factors that may contribute to or protect from risk.
Potentially critical factors include the context of risk behaviors,
developmental variation in risk behaviors, and effects of competing
risks-all of which are likely to be diverse across adolescent subgroups
and changeable over the course of adolescent development. Contextual
variables include the individual’s cognitive and emotional status, the
social grouping in which behavior occurs, economic conditions, and
prevalence of substance abuse and HIV infection in the environment.
Reviews of data provided by Boyer and Ellen (this volume) and by others
(e.g., Bowler et al. 1992) indicate that although the cumulative number of
AIDS diagnoses in the 13- to 19-year-old age group is low, when the
long intervals between HIV infection and AIDS are taken into account
and HIV seroprevalence data are considered, infection risks during the
adolescent years are striking. Through June 1993, a cumulative total of
1,301 AIDS cases among adolescents 13 to 19 years of age had been
reported in the United States (Centers for Disease Control and Prevention
1993). For the same period, the cumulative number of AIDS cases was
11,840 in young adults 20 to 24 years of age (the majority of whom are
assumed to have been infected during adolescence), and 47,777 in the
25- to 29-year-old age group (a substantial portion of whom are likely to
have been infected during adolescence). The 60,918 AIDS cases for the
13- to 29-year-old age group constitute almost 20 percent of all reported
AIDS cases in the United States. HIV seroprevalence data indicate wide
variation in infection rates within the adolescent population, including,
132
for example, 3.7 per 1,000 in a sample of 13- to 19-year-olds receiving
ambulatory health services in a Washington, D.C., hospital (D’Angelo et
al. 1991), and 53 per 1,000 in a high-risk group of 15- to 20-year-olds
who were clients at a facility for runaway and homeless youth in New
York City (Stricof et al. 1991). Gender, racial, ethnic, and geographic
comparisons in infection rates have been provided in numerous reports
(e.g., Bowler et al. 1992).
Boyer and Ellen (this volume) review and discuss adolescent risk-taking
behaviors and adolescent developmental processes as they may relate to
HIV infection risk. They consider specific behavioral risk factors
(e.g., age at onset of sexual activity and use of alcohol and other drugs)
and possible antecedents of some of these behaviors (e.g., personality,
school, and family factors). Rotheram-Borus and colleagues (this
volume), following a discussion of developmental changes and
challenges in adolescence, focus on high-risk adolescents. By examining
case studies developed during ethnographic research, these authors
provide indepth analyses of the context of HIV risk behaviors, routes into
a specific drug-using subculture, and life patterns while using illicit drugs
(i.e., methamphetamine). Rotheram-Borus and colleagues then make
suggestions for intervention strategies to help adolescents cease or
modulate drug-related activities, reduce HIV risk behaviors, and continue
the identity search that is interrupted by contact with the
methamphetamine subculture.
REFERENCES
Bowler, S.; Sheon, A.R.; D’Angelo, L.J.; and Vermund, S.H. HIV and
AIDS among adolescents in the United States: Increasing risk in the
1990’s. J Adolesc 15:345-371, 1992.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance.
Second Quarter Edition. Atlanta: Centers for Disease Control and
Prevention, 1993.
D’Angelo, L.J.; Getson, P.R.; Luban, N.L.C.; and Gayle, H.D. Human
immunodeficiency virus infection in urban adolescents: Can we
predict who is at risk? Pediatrics 88:982-986, 1991.
DiClemente, R.J., ed. Adolescents and AIDS: A generation in jeopardy.
Newbury Park, CA: Sage, 1992.
Stricof, R.L.; Kennedy, J.T.; Nattell, T.C.; Weisfuse, I.B.; and Novick,
L.F. HIV seroprevalence in a facility for runaway and homeless
adolescents. Supplement. Am J Public Health 81:50-53, 1991.
133
AUTHOR
Vincent L. Smeriglio, Ph.D.
Health Scientist Administrator
Clinical Medicine Branch
Division of Clinical Research
National Institute on Drug Abuse
Parklawn Building, Room 11A-33
5600 Fishers Lane
Rockville, MD 20857
134
HIV Risk in Adolescents: The
Role of Sexual Activity and
Substance Use Behaviors
Cherrie B. Boyer and Jonathan M. Ellen
INTRODUCTION
In order to understand adolescents’ risk for acquiring the human
immunodeficiency virus (HIV), it is first imperative to have knowledge
of the sociodemographic characteristics of the adolescent population as
well as the significance of developmental and psychosocial changes
during adolescence that influence their behavior. Moreover, it is
necessary to understand differences in the prevalence and patterns of risk
behaviors in adolescents as well as the context in which these behaviors
occur, since exposure to HIV is not equal among all adolescents.
Therefore, this chapter includes an overview of the prevalence of
negative health outcomes of adolescent risk-taking behaviors, with a
particular focus on acquired immunodeficiency syndrome (AIDS), HIV,
sexually transmitted diseases (STDs), and pregnancy; a description of the
primary risk factors that are associated with acquisition and transmission
of STDs/HIV; and a brief overview of the antecedents and correlates of
sexual and alcohol/drug-using behaviors, with an emphasis on
sociodemographic and psychosocial factors.
Sociodemographic Factors: Who Are the Youth at Risk?
There are approximately 28 million youth and adolescents in the United
States between the ages of 10 and 17. Most individuals between these
ages are white (81 percent); 15 percent are African American; and
4 percent are of other ethnic and racial backgrounds. Of all adolescents,
10 percent are categorized as being of Spanish/Latin ancestry. However,
in the West and Southwest regions of the country (primarily California
and Texas), the majority of the schoolage children are Hispanic or of
other nonwhite origin. The nonwhite population is increasing, due
largely to higher rates of birth and immigration. By the year 2000,
20 percent of adolescents in the United States will be African American
and 18 percent will be Hispanic; by 2010, Hispanic adolescents will
135
comprise 23 percent of all adolescents, while African-American teens will
make up 21 percent of this population (Dryfoos 1990).1,2
One in four of the Nation’s youth and adolescents currently live in
single-parent-headed households. However, among African-American
adolescents, more than half live in female-headed households. Although
70 percent of women with teens between the ages of 14 and 17 are in the
workforce, living in a female-headed household increases the likelihood
that an adolescent lives in poverty. While 11 percent of all families live
in poverty, 46 percent of female-headed households with schoolage
children live in poverty (Dryfoos 1990).
Race and geographic place of residence are strongly associated with
poverty status. African-American and Hispanic youths are more likely to
be impoverished than white youth; 45 percent of African-American and
41 percent of Hispanic youths live in poverty, compared to 13 percent of
white youths who also live in poverty (Dryfoos 1990). Almost half of all
individuals under the age of 18 live in suburban areas, 30 percent live in
inner cities, and 23 percent live in rural areas; however, AfricanAmerican teens are more likely to live in inner cities. While 35 percent
of all white inner-city teens live in poverty, 57 percent of all AfricanAmerican inner-city teens live in poverty (Dryfoos 1990). These factors
are significant in that poverty is strongly associated with poor health
status. In general, adolescence is a time of excellent physical health, but
rates of morbidity and mortality in this age group have increased
dramatically in recent decades (Irwin 1990; Rosen et al. 1990), due
largely to many of the behaviors that are prevalent in adolescents.
Adolescent Development and Risk-Taking Behavior
Adolescence is the period between childhood and adulthood that is
marked by rapid biological, emotional, cognitive, and social change. It is
also a period of tremendous exploration and experimentation, when many
teenagers perceive themselves to be invulnerable to accidents and disease
and often engage in a wide range of risk-taking behaviors. The term
“risk-taking“ describes the patterns of behaviors initiated during
adolescence that are responsible for many negative health outcomes that
occur during this time period. Specifically, adolescent risk-taking is
defined as follows: “young people with limited or no experience engage
in behaviors with anticipation of benefit and without understanding the
immediate or long-term consequences of their actions” (Irwin 1990).
136
In addition, risk-taking behaviors are deemed to be volitional in nature
and account for the majority of morbidity during adolescence.
Although risk-taking behaviors such as initiation of sexual intercourse
and experimentation with alcohol and drugs are thought to be a part of
“normal” adolescent development, often these behaviors have grave
negative health consequences that last well into adulthood. Adolescents
who engage in such risk-taking behaviors are at increased risk for a host
of problems, including acquisition and transmission of HIV and other
STDs.
OUTCOME OF ADOLESCENT RISK-TAKING BEHAVIOR:
SCOPE OF THE PROBLEM
AIDS and HIV Infection
As of June 1993, the Centers for Disease Control and Prevention (CDC)
had reported 315,390 cases of AIDS in the United States. While these
numbers are of epidemic proportions, adolescents ages 13 to 19 represent
1,301 or less than 1 percent of the total cases. Adolescent males represent
913 cases (less than 1 percent of all the male cases), and adolescent
females comprise 318 cases (reflecting 1 percent of all the female cases)
(CDC 1993a). The number of AIDS cases may not reflect the actual rate
of HIV infection among adolescents, since the incubation period is long
and varied (Gayle et al. 1989). Therefore, it is probable that many of the
59,617 young adults ages 20 to 29 diagnosed with AIDS (19 percent of
reported cases) acquired HIV during their teen years.
A profile of the currently reported adolescent AIDS cases in the United
States by exposure category reveals that most adolescents with AIDS
were infected as a result of high-risk sexual behavior or injecting drug use
(IDU): 299 (23 percent) are males who reported sex with other males;
143 (11 percent) reported IDU behavior; 42 (3 percent) are males who
reported sex with males and IDU behavior; and 236 (18 percent) reported
heterosexual contact with a person who was HIV-infected or at high risk.
Of the individuals in the last group, 208 (88 percent) are females.
National AIDS surveillance data indicate that minority adolescents are
overrepresented among persons with AIDS relative to their proportion in
the population: whites represent 530 (41 percent) of the adolescent cases;
African Americans total 515 (40 percent) of the cases; Hispanics total
137
232 (18 percent) of the cases; and Asians, Pacific Islanders, Alaskans,
and Native Americans together comprise 23 (2 percent) of all adolescent
cases (CDC 1993a).
Although AIDS surveillance data reveal patterns of HIV transmission, the
extent of asymptomatic HIV infection remains largely unknown. To
date, seroprevalence data regarding adolescents have been reported from
military recruits, Job Corps entrants, STD clinic patients, and from street
and other disenfranchised youth. The rates have ranged from 22 per
1,000 among STD clinic patients (Quinn et al. 1988) to 3.6 per 1,000
among Job Corps students (St. Louis et al. 1991). While these data offer
some insights into the prevalence of HIV infection among certain groups
of adolescents, they are not generalizable to all adolescents.
However, a more accurate indicator of trends of HIV infection may be
found in rates of STDs in adolescents, since the behaviors associated with
the acquisition and transmission of STDs also are associated with HIV
transmission. In addition, other STDs may serve as cofactors for HIV
acquisition. It is suggested that genital tract ulceration, which is
associated with STDs such as syphilis, chancroid, and herpes, can
increase the likelihood of HIV transmission (Greenblatt et al. 1987;
Holmes 1988).
STDs
STDs are the most pervasive and destructive infectious diseases
confronting adolescents in the United States. Of the estimated 20 million
cases reported annually, two-thirds are among individuals under the age
of 25 (CDC 1991a), with one-fourth of the cases occurring in adolescents
ages 15 to 19. The high rate of STDs among sexually experienced
adolescents is accompanied by a parallel increase in their sequelae of
pelvic pain, tubal infertility, ectopic pregnancy, genital and cervical
carcinoma, neonatal transmission of infections causing perinatal death or
illness, and HIV leading to AIDS (Cates 1990; CDC 1991a; Shafer and
Sweet 1990).
Given the large number of cases and the stability of reporting, trends in
gonorrhea and syphilis (the only two reported diseases) provide the most
accurate assessment of STDs in adolescents (Cates 1990). Overall, the
CDC reported a decline of gonorrhea cases between 1975 and 1989
(Cates 1990), yet rates have increased among teenage males 15 to 19
years of age in the past 2 years and have remained unchanged among
138
15-to 19-year-old teenage females during the same time period
(CDC 1991a). Racial and gender differences among adolescents also are
apparent in trends of gonorrhea; while rates among white males and
females have steadily declined over the last decade, rates among AfricanAmerican males and females have increased during this time period
(Cates 1990). Currently, 15- to 19-year-old females have the highest
rates of gonorrhea among all females, and 15 to 19-year old males have
the second highest rates among all males (CDC 1991a). Moreover,
recent data indicate that in 1990, rates of primary and secondary syphilis
increased among adolescent males and females ages 15 to 19
(CDC 1991a) .
Chlamydia trachomatis, which causes nongonococcal urethritis in males
and mucopurulent cervicitis in females, is estimated to be the most
prevalent bacterial STD in the United States (Schydlower and Shafer
1990); it is most commonly associated with pelvic inflammatory disease
(PID) in sexually experienced adolescent females (Coupey and Klerman
1992; Schydlower and Shafer 1990). In adolescent males, the prevalence
rates of asymptomatic chlamydia range from 8 percent to 35 percent in
various populations. Although complications of chlamydial infections in
males are unusual, a major risk of transmission to female sexual partners
is of tremendous concern. It is estimated that 30 to 60 percent of
adolescent females have asymptomatic chlamydial infections, and 10 to
30 percent of cervical chlamydial infections infect the fallopian tubes
(Schydlower and Shafer 1990). Chlamydia is frequently asymptomatic
and goes undiagnosed and untreated unless routine clinical screening
occurs.
The human papillomavirus (HPV) causes genital warts and is associated
with cervical cancer many years after the first infection. Prevalence rates
of up to 38 percent have been documented in adolescent females,
depending on the population and the method used to detect the infection
(Moscicki 1990; Rosenfeld et al. 1989).
In essence, the prevalence data reported from various studies indicate that
inner-city, racial or ethnic minority adolescents have higher rates of STDs
when compared to their white counterparts (Boyer 1990; Cates 1990). In
addition, socioeconomic factors are associated with STDs in adolescents.
This finding comes from data that is reported from urban juvenile
detention centers where poor, inner-city minority teens are
overrepresented (Boyer 1990; Cates 1990).
139
Although these data provide an overview of trends of STDs in
adolescents, they are subject to biases due to differences in the reporting
by public and private-sector clinics. While public health clinics are more
likely to report STDs of poor and minority patients who often utilize
these clinics, reports from private physicians who tend to treat more
affluent patients are affected by the lack of diagnostic validation of their
findings (Cates 1990). Thus, no firm conclusions can be drawn from
STD surveillance data. Nonetheless, specific research on determinants of
STDs in various groups of teens is necessary to control future acquisition
of STDs, especially HIV infection in adolescents.
Pregnancy
The risk of STDs/HIV being vertically transmitted from an infected
mother to her infant is reflected in recent data on teenage pregnancies.
The pregnancy rates among teenagers have remained high over the last
decade and are reflective of the increasingly high level of sexual activity
among adolescents (Rosen et al. 1990). Almost one in four teens
(23 percent) who engage in sexual activity experience pregnancy,
resulting in 1 million pregnancies each year (Dryfoos 1990). Of these
teens, approximately 32 percent are under the age of 15 (Rosen et al.
1990). Eighty-four percent of the pregnancies among adolescents are
unintended, of which 45 percent result in live births, 42 percent end in
abortions, and 10 percent result in either miscarriages or stillbirths
(Coupey and Klerman 1992). It is estimated that among teens who
experience pregnancy, 6 percent are white and 13 percent are African
American between the ages of 15 and 17 years. Fifteen percent are white
and 26 percent are African Americans between the ages of 18 and
19 years. Data regarding other racial/ethnic groups were not reported.
BEHAVIORAL RISK FACTORS ASSOCIATED WITH STDs/HIV
INFECTION
To further place adolescents’ exposure to HIV and other STDs in context,
it is imperative to understand the prevalence and patterns of the primary
risk behaviors that are associated with disease transmission, including a
young age at sexual debut, multiple sexual partners, anal intercourse,
inadequate or no use of barrier-method contraceptives, and use of alcohol
and drugs (including both injectable and noninjectable substances).
These behavioral risk factors are interrelated and are prevalent among
adolescents.
140
Sexual Risk Behavior
More adolescents are commencing sexual intercourse at younger ages
than ever before; therefore, they are placing themselves at risk for STDs
and unintended pregnancies at younger ages than ever before. The
incidence of sexual behavior increased significantly from 1971 to the
mid-1980s in all age cohorts between the ages of 15 and 19. In 1988,
26 percent of white females, 24 percent of African-American females,
24 percent of white males, and 69 percent of African-American males
initiated sexual intercourse by age 15 (Irwin and Shafer 1992).
Epidemiologic data of high school students grades 9 to 12 indicate that
54 percent reported experiencing sexual intercourse at some point, of
which 39 percent reported sexual experience during the 3 months prior to
being surveyed. Differences in the prevalence of sexual experience by
gender and race also were found. Sixty-one percent of males and 48
percent of females reported sexual contact, as did 52 percent of white
students, 72 percent of African-American students, and 53 percent of the
Hispanic students. The percentages of students engaging in sexual
intercourse increased significantly with each higher grade (CDC 1992).
In a sample of urban middle school students ages 10 to 14, 21 percent
reported engaging in sexual intercourse at least once. Males also were
more likely than females to have engaged in sexual intercourse
(35 percent versus 8 percent, respectively) (Irwin and Shafer 1992).
In addition to engaging in sexual intercourse at early ages, some
adolescents have nonmonogamous sexual relationships or have multiple
sexual partners within a short period of time in a pattern of serial
monogamy. Having multiple sexual partners, in addition to inadequate
use of barrier-method contraceptives, increases the risk of STDs, HIV
transmission, and unintended pregnancies. Based on national samples of
sexually experienced adolescent and young adult females ages 15 to 24,
75 percent had two or more lifetime sexual partners, and 45 percent
reported four or more lifetime sexual partners (Forrest and Singh 1990).
Among males 15 to 19 years of age, over 50 percent reported more than
one lifetime sexual partner, and 32 percent reported six or more lifetime
sexual partners (Forrest and Singh 1990; Sonenstein et al. 1989). Recent
data from an urban public STD clinic reveal that among adolescents and
young adults ages 12 to 20, the mean number of lifetime sexual partners
among the males was 33 (45 percent had 7 to 25 sexual partners), and the
mean number for females was 9 (35 percent had 4 to 6 sexual partners).
141
Most of these were unprotected sexual encounters; 48 percent of males
and 64 percent of females never or rarely used condoms (Boyer et al.
1993).
In addition, some teenagers engage in other sexual risk behaviors that
also increase their risk of negative health outcomes, especially HIV
transmission. For example, some adolescent males engage in anal
intercourse without the use of barrier methods to protect against disease
transmission (Remafedi 1990). Clinical surveys of adolescent females
attending family planning clinics found that 12 to 26 percent engaged in
anal intercourse (either as a form of birth control or as a means of sexual
experimentation), which also may increase their risk for certain STDs
(Catania et al. 1990; Moscicki et al. 1988).
Contraceptive Use Behavior
The concurrent epidemics of STDs and unintended pregnancies among
adolescents suggest that most teens, like many adults, do not use
contraceptives effectively. The role oral contraceptives play in the
acquisition of STDs and their sequelae of PID is unclear. It appears that
while the use of oral contraceptives provides some protection against
gonococcus-related PID, it may increase the risk of chlamydia-related
PID (Shafer and Sweet 1990).
Use of the latex condom is effective in reducing the risk of STDs,
including HIV. However, the extent to which teens use them is unclear;
differences in samples and in questions asked of adolescents make it
difficult to compare results of the studies.
National epidemiologic data from high school students reveal that, among
students who engaged in sexual activity within the 3 months prior to the
survey, 78 percent of females and 79 percent of males reported use of
some form of contraception (birth control pills, condoms, withdrawal, or
another method) during their last sexual encounter. However, when
queried specifically about the use of condoms, the number decreased;
only 40 percent of the females and 49 percent of the males reported use
of this form of protection (CDC 1992). In addition, ethnic/racial
differences in the use of contraceptives have been found. A large survey
of adolescent and adult women ages 15 to 44 indicates that, compared to
their white counterparts, African-American women are significantly more
likely to use oral contraceptives and are less likely to use condoms or
142
diaphragms, suggesting that these women are at increased risk of
acquiring a disease (Mosher and Pratt 1990).
Other patterns in teens’ use of contraceptives also have been noted. As
sexual experience increases, the use of oral contraceptives also increases,
and the use of condoms and other barrier methods decreases (Morrison
1985). Thus, many teens initially attend family planning clinics with the
intent of changing from the use of condoms to the use of oral
contraceptives. Moreover, with the recent push for use of hormonal
implants among adolescents, research is needed to determine the role of
nonbarrier forms of contraception in influencing adolescents’ risk for
acquiring sexually transmitted infections. It is also evident that
adolescents need to be educated about the important differences in the use
of contraceptives for birth control and for disease prevention.
Antecedents of Sexual Risk Behaviors
Adolescents’ decisions to initiate and consistently use barrier-method
contraceptives are as complex as their reasons for initiating sexual
intercourse or for engaging in behaviors that place them at risk for STDs
and HIV transmission. Many reviews exist that attempt to explain this
phenomenon (Boyer 1990; Brooks-Gunn and Furstenberg 1989; Irwin
and Shafer 1992; Kegeles et al. 1988). This research addresses the need
to understand the role of sociodemographic, cultural, biological,
developmental, psychosocial, and academic factors. For example,
Brooks-Gunn and Furstenberg (1989) underscore the importance of
understanding the psychosocial and emotional needs that sexual
intercourse fills for some adolescents. In contrast, Talmadge (1985)
describes the influence of contextual factors such as attitudes, beliefs, and
social interactions.
Alcohol and Drug Use Behavior3
The high prevalence of alcohol and drug use in adolescents poses a
significant threat to their health and well-being. Use of these substances
is associated with motor vehicle accidents, homicides, and suicides
among adolescents and is responsible for major medical, psychological,
and social morbidity in teens (CDC 1989; Dryfoos 1990).
Although adolescents’ use of illicit substances has declined in recent
years, their overall level of use continues to be high, while their use of
alcohol has remained constantly high (Bachman et al. 1991). A recent
143
national epidemiologic survey of high school students grades 9 to 12
indicates that 31 percent of the students used marijuana at least once in
their lifetime, and 14 percent used this substance within 30 days prior to
the survey. Cocaine was tried at least once by 7 percent of the students
and recently (within 30 days) by 2 percent of the population. In addition,
these data clearly demonstrate that alcohol is the substance of choice for
high school students: 88 percent of the students consumed alcohol at
some point in their lifetime, and 59 percent consumed alcohol at least
once within the preceding 30 days, with 37 percent consuming five or
more drinks on one occasion (CDC 1991b). A national study of high
school students from both public and private schools found a prevalence
of 2.7 percent for students who reported a history of illicit drug injection.
Of these students, 0.8 percent reported sharing a needle at some point in
time. Similarly, within the previous year, 1.7 percent of the students
reported IDU behavior and 0.5 percent shared a needle (Holtzman et al.
1991). A more recent national study by CDC reported IDU behaviors
among 2 percent of its student population (CDC 1993b). Grade, age, and
gender differences in overall rates of illicit substance and alcohol use
have been noted; use of all substances increased with each higher grade
level (seniors demonstrated highest rates of use for all substances) and
was more prevalent among males than females for both lifetime and
recent use (CDC 1991b).
A national longitudinal study (combined data from 1985 to 1989) of licit
and illicit use of substances among high school seniors revealed
significant ethnic/racial differences (Bachman et al. 1991). Annual
prevalence rates for marijuana were found to be highest among white
males and females, followed by Native American males and females and
Mexican-American males. Although rates among other Hispanic and
African-American males were somewhat lower, rates were lowest among
Hispanic females, African-American females, and Asian-American males
and females. Regarding use of cocaine, prevalence rates were highest
among Native American males and females and Hispanic males;
somewhat lower rates were found for Hispanic females and white males
and females, and the lowest rates were among Asian Americans and
African Americans. However, among Hispanic males and AfricanAmerican males and females, the prevalence was almost twice as high for
males than females.
Alcohol consumption among white and Native American males and
females was relatively higher than among Asian-American and AfricanAmerican students; about half of the males and one-third of the females
144
consumed alcohol in the previous month. Moreover, significant
ethnic/racial differences also were found; almost half of the white,
Mexican-American, and Native American males reported consuming five
or more drinks in one setting in the 2 weeks preceding the survey.
Significantly fewer Hispanic, Asian-American, and African-American
males reported such abuse of alcohol. The prevalence of alcohol abuse
among females generally paralleled that of males, albeit at lower levels.
As striking as these data are, they are not reflective of teens who drop out
or are chronically absent from school, who are in juvenile detention
centers, and who are homeless; these teens are thought to be at highest
risk. Therefore, these data clearly underrepresent the actual prevalence of
alcohol and drug use among all adolescents.
Antecedents of Substance Use and Abuse Behavior
Many studies have examined risks for substance abuse and initial patterns
of use. Numerous factors have been identified as significant markers of
substance use (Dryfoos 1990; Hawkins et al. 1992); however, there is
little empirical evidence that describes the relative significance and
interactions between the risk factors (Hawkins et al. 1992). Moreover,
there is no consensus about how to categorize or measure substance abuse
risk factors. For example, Dryfoos (1990) identified 21 risk factors
across 4 domains: (1) demographic (e.g., age, gender, race/ethnicity);
(2) personal (e.g., school performance, peer influences, psychological
factors); (3) family (e.g., poverty, parental influences, culture); and
(4) community (e.g., quality of school and neighborhood).
In contrast, Hawkins and his colleagues (1992) identified 17 factors
across two broad domains: (1) contextual factors, including societal and
cultural influences related to legal and normative expectations for
substance use behaviors as well as economic factors; and (2) individual
and interpersonal factors that take into account the individual within the
context of social, behavioral, and biological influences on adolescent
decisionmaking and subsequent substance use behavior.
Although there is no consensus about risk domains, there is agreement
regarding the significance of several key risk factors (Dryfoos 1990;
145
Hawkins et al. 1992; Newcomb et al. 1987). The risk factors are as
follows:
Early initiation of drug use predicts later abuse. Most research
concurs that there is a linear relationship to early onset and severity of
abuse.
School problems such as academic failure in school and a low degree
of commitment to education have been identified as significant
predictors of substance abuse.
Personality characteristics such as rebelliousness, nonconformity
(alienation), and a strong sense of independence are positively
associated with substance abuse.
Lack of family support and guidance, resulting in family conflict,
noninvolvement, and detachment between adolescents and their
parents have been related most significantly with initiation of
substance use. Conversely, adolescents’ reports of trust, warmth, and
involvement with parents have been associated with nonuse or less
frequent use of cigarettes, alcohol, and marijuana.
Of particular significance are findings related to the role of peers in
determining adolescents’ use of substances. Peer influences on
adolescents’ substance use have been empirically demonstrated by
adolescents’ association with peers who also use substances,
conforming to peer pressure and perceptions that peers are engaging
in substance use behaviors (perceived social norms), as well as
relying on the opinions of peers rather than on parents or other adults.
While knowledge about risk factors for substance abuse in adolescents
does not offer a total solution for preventing substance use behavior, it
does point to key elements to target when designing intervention
strategies for prevention. Moreover, prevention strategies are most likely
to be effective if they are based on an understanding of the antecedents to
substance use. Thus, it is unlikely that a focus on one risk factor will lead
to prevention, nor is it likely that one prevention strategy will be
effective.
146
Co-Occurrence of Risk Behaviors and Negative Health
Outcomes
Although alcohol consumption and use of noninjecting drugs are not
directly linked to acquisition of STDs/HIV, youths who use one or more
of these substances also tend to engage in sexual risk behaviors, including
earlier initiation of sexual intercourse and inconsistent use of barriermethod contraceptives, when compared to youth who abstain from using
these substances (Bagnall et al. 1990; Ensminger 1990; Keller et al. 1991;
Mott and Haurin 1988; Rosenbaum and Kandel 1990). Moreover, a
recent study found a significant association between abuse of alcohol and
drugs and failure to use condoms (Hingson et al. 1990). Specifically, the
authors reported that adolescents who consumed five or more drinks daily
or used marijuana within the month prior to the survey were significantly
less likely to use condoms. In addition, of the teens who reported sexual
activity after drinking alcohol, 16 percent indicated that they were less
likely to use condoms, and 25 percent reported using condoms less often
after using drugs.
While the co-occurrence of adolescent substance use behavior and sexual
activity is well established, the role of this relationship to transmission of
STD/HIV infection has recently received attention. Use of crack cocaine
has been linked to increased rates of syphilis and gonorrhea in some
urban adolescent populations; this increase is postulated to be due to
drug-related sexual disinhibition and the exchange of sex for drugs
(Fullilove et al. 1990; Goldsmith 1988; Schwarcz et al. 1992; Zweig et al.
1991).
Summary of Adolescents’ Risk for STDs/HIV
From the data reported thus far, one can conclude that:
risk-taking behavior among adolescents is prevalent and interrelated;
adolescents commence sexual intercourse at earlier ages than ever
before, and many of them do not use effective measures to protect
themselves from diseases;
STD and pregnancy rates among adolescents are high;
use of substances, especially alcohol, is common in teens and thought
to be associated with increased risk of STDs/HIV;
147
higher rates of HIV/AIDS in adolescents than those currently
reported are probable;
males who engage in sexual intercourse with other males, as well as
African-American and Hispanic youth, are over-represented among
teens with AIDS relative to their proportion in the population;
given the high rates of pregnancy in teens, higher rates of perinatal
HIV transmission are likely to occur in teen mothers; and
peers and family play an important role in determining adolescents’
decisions to engage in risk-taking behaviors. Together, these factors
must be taken into account when planning strategies to prevent
further spread of STDs/HIV infection in adolescents.
CONCLUSIONS
Currently, AIDS in adolescents is rare; however, many more teens are
thought to be HIV-infected given the long latency period between
infection and manifestation of disease. Teenagers are at increased risk for
HIV transmission because many of them engage in sexual and substance
use behaviors; they are initiating sexual intercourse at younger ages than
ever before; many of them have multiple sexual partners; and some use
substances that may directly or indirectly place them at risk of exposure
to STDs/HIV infection. In addition, many teens do not consistently use
condoms when engaging in sexual intercourse.
To prevent further spread of sexually transmitted infections, prevention
programs that emphasize cognitive and behavioral skills-building are
necessary (Bandura 1992; Boyer and Kegeles 1991). Such programs
should be theory based and must take into account the risk behaviors and
their associated antecedents. To be most effective, these programs should
include a variety of educational and communication strategies that may
differ depending on the target population and the site where they are to be
implemented. Finally, although current research has begun to elucidate
the mechanisms by which adolescents make health decisions and
subsequently engage in health-promoting or health-damaging behaviors,
more research is critically needed.
148
NOTES
1.
The data reported from Dryfoos (1990) was compiled primarily from
the US. Bureau of the Census, Statistical Abstract of the United
States, Washington, DC: U.S. Government Printing Office, 1988.
2. Adolescents, teens, teenagers, and youth(s) are used interchangeably
in this document. However, when referring to 10- to 12-year-olds,
the term “youths” is used to denote this younger age group.
3. Alcohol and drug use is interchanged with substance use or substance
use behavior.
ACKNOWLEDGMENTS
Dr. Boyer was funded, in part, by the National Institutes of Health,
National Institute of Allergy and Infectious Diseases, Sexually
Transmitted Branch (grant number U01A131499-01), and Dr. Ellen was
funded by the Maternal and Child Health Bureau (grant number
MCJ00978).
REFERENCES
Bachman, J.G.; Wallace, J.M., Jr.; O’Malley, P.M.; Johnston, L.D.;
Kurth, C.L.; and Neighbors, H.W. Racial/ethnic differences in
smoking, drinking, and illicit drug use among American high school
seniors, 1976-89. Am J Public Health 81:372-377, 1991.
Bagnall, G.; Plant, M.; and Warwick, W. Alcohol, drugs, and AIDSrelated risks: Results from a prospective study. AIDS Care 2:309-317,
1990.
Bandura, A. A social cognitive approach to the exercise of control over
AIDS infection. In: DiClemente, R.J., ed. Adolescents and AIDS: A
Generation in Jeopardy. Newbury Park, CA: Sage, 1992. pp. 89-116.
Boyer, C.; Moscicki, A.B.; Barrett, D.; and Flynn, P. “Evaluation of
AIDS Intervention in High Risk Youth.” Paper presented at the 10th
Annual AIDS Investigators Meeting of the University of California,
Universitywide AIDS Research Program and Task Force on AIDS.
San Francisco, March 9, 1993.
149
Boyer, C.B. Psychosocial, behavioral, and educational factors in
preventing sexually transmitted diseases. In: Schydlower, M., and
Shafer, M.A., eds. Adolescent Medicine: State of the Art Reviews:
AIDS and Other Sexually Transmitted Diseases. Vol. I(3).
Philadelphia: Hanley & Belfus, Inc., 1990. pp. 597-613.
Boyer, C.B., and Kegeles, S.M. AIDS risk and prevention among
adolescents. Soc Sci Med 33(1):11-23, 1991.
Brooks-Gunn, J., and Furstenberg, F.F. Adolescent sexual behavior. Am
Psychol 44:249, 1989.
Catania, J.A.; Coates, T.J.; Greenblatt, R.M.; Puckett, S.; Carman, M.;
and Miller, J. Predictors of condom use and multiple partnered sex
among sexually active adolescent women: Implications for AIDS
related health interventions. J Sex Res 26:514-524, 1990.
Cates, W., Jr. The epidemiology and control of sexually transmitted
diseases in adolescents. In: Schydlower, M., and Shafer, M.A., eds.
Adolescent Medicine: State of the Art Reviews: AIDS and Other
Sexually Transmitted Diseases. Vol. I(3). Philadelphia: Hanley &
Belfus, Inc., 1990. pp. 409-427.
Centers for Disease Control. Results from the National Adolescent
Student Health Survey. MMWR 38:147-150, 1989.
Centers for Disease Control and Prevention. Centerfor Prevention
Services, Division of STD Prevention, Annual Report. Atlanta: Centers
for Disease Control and Prevention, 1991a .
Centers for Disease Control and Prevention. Alcohol and other drug use
among high school students-United States, 1990. MMWR
40(45):776-777 and 783-784, 1991b .
Centers for Disease Control and Prevention. Health objectives for the
nation: Current tobacco, alcohol, marijuana, and cocaine use among
high school students-United States, 1990. MMWR 40(38):659-663,
1991c .
Centers for Disease Control and Prevention. Health objectives for the
nation: Sexual behavior among high school students-United States,
1990. MMWR 40(51/52):885-888, 1992.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report. July 5(2):10-11, 1993a .
Centers for Disease Control and Prevention. Selected behaviors that
increase risk for HIV infection, other sexually transmitted diseases,
and unintended pregnancy among high school students-United
States, 1991. JAMA 269(3):329-330, 1993b .
Coupey, S.M., and Klerman, L.V. Adolescent Medicine: State of the Art
Reviews: Adolescent Sexuality-Preventing Unhealthy Consequences.
Philadelphia: Hanley & Belfus, Inc., 1992.
150
Dryfoos, J.G. Adolescents at Risk: Prevalence and Prevention. New
York: Oxford University Press, 1990.
Ensminger, M.E. Sexual activity and problem behaviors among black,
urban adolescents. Child Dev 61:2032-2046, 1990.
Forrest, J.D., and Singh, S. The sexual and reproductive behavior of
American women, 1982-1988. Fam Plann Perspect 22:206-214,
1990.
Fullilove, R.E.; Fullilove, M.T.; Bowser, B.P.; and Gross, S.A. Risk of
sexually transmitted disease among black adolescent crack users in
Oakland and San Francisco, CA. JAMA 263:851-855, 1990.
Gayle, H.; Manoff, S.; and Rogers, M. “Epidemiology of AIDS in
adolescents USA.” Paper presented at the Fifth International AIDS
Conference, Montreal, June 1989.
Goldsmith, M.F. Sex tied to drugs = STD spread. JAMA 260(14):2009,
1988.
Greenblatt, R.M.; Lukehart, S.L.; Plummer, F.A.; Quinn, T.C.;
Critchlow, C.W.; and D’Costa, L.J. “Genital Ulceration as a Risk
Factor for Human Immunodeficiency Virus Infection in Kenya.” Paper
presented at the Third International Conference on AIDS, Washington,
DC, June 1987.
Hawkins, J.D.; Catalano, R.F.; and Miller, J.Y. Risk and protective
factors for alcohol and other drug problems in adolescence and early
adulthood: Implications for substance abuse prevention. Psychol Bull
112(1):64-105, 1992.
Hingson, R.W.; Strunin, L.; Berlin, B.M.; and Heeren, T. Beliefs about
AIDS, use of alcohol and drugs, and unprotected sex among
Massachusetts adolescents. Am J Public Health 80:295-299, 1990.
Holmes, K.K. “Heterosexual Transmission of HIV: Current Evidence and
Future Prospects.” Paper presented at the Fourth International
Conference on AIDS, Stockholm, Sweden, June 1988.
Holtzman, D.; Anderson, J.E.; Kann, L.; Arday, S.L.; Truman, B.I.; and
Kolbe, L.J. HIV instruction, HIV knowledge, and drug injection
among high school students in the United States. Am J Public Health
81(12):1596-1601, 1991.
Irwin, C.E., Jr. The theoretical concept of at-risk adolescents. In:
Strasburger, V.C., and Greydanus, D.E., eds. Adolescent Medicine:
State of the Art Reviews: The At-Risk Adolescent. Vol. I(1).
Philadelphia: Hanley & Belfus, Inc., 1990. pp. 1-14.
151
Irwin, C.E., Jr., and Shafer, M.A. Adolescent sexuality: Negative
outcomes of a normative behavior. In: Rogers, D.E., and Ginzberg, E.,
eds. Adolescents at Risk: Medical and Social Perspectives.
Proceedings of the Cornell University Medical College Seventh
Conference on Health Policy. Boulder, CO: Westview Press, 1992. pp.
35-79.
Kegeles, S.M.; Adler, N.E.; and Irwin, C.E., Jr. Sexually active
adolescents and condoms: Changes over one year in knowledge,
attitudes, and use. Am J Public Health 78:460, 1988.
Keller, S.E.; Bartlett, J.A.; Schleifer, S.J.; Johnson, R.L.; Pinner, E.; and
Delaney, B. HIV-relevant sexual behavior among a healthy inner-city
heterosexual adolescent population in an endemic area of HIV.
J Adolesc Health 12:44-48, 1991.
Morrison, D.M. Adolescent contraceptive behavior: A review. Psychol
Bull 98:538, 1985.
Moscicki, A.B. Genital human papillomavirus infections. In: Schydlower,
M., and Shafer, M.A., eds. Adolescent Medicine: State of the Art
Reviews: AIDS and Other Sexually Transmitted Diseases. Vol. I(3).
Philadelphia: Hanley & Belfus, Inc., 1990. pp. 451-470.
Moscicki, B.; Millstein, S.G.; Broering, J.; and Irwin, C.E., Jr.
“Adolescent Beliefs and Behaviors Concerning Sexually Transmitted
Disease and Acquired Immunodeficiency Syndrome.” Paper presented
at the Annual Conference of the Society for Adolescent Medicine,
New York, March 1988.
Mosher, W.D., and Pratt, W.F. Contraceptive Use in the United States,
1973-88. National Center for Health Statistics, Series 81, No. 1.
Washington, DC: Supt. of Docs., U.S. Govt. Print. Off., 1990.
Mott, F.L., and Haurin, R.J. Linkages between sexual activity and alcohol
and drug use among American adolescents. Fam Plann Perspect
20(3):128-136, 1988.
Newcomb, M.D.; Maddahian, E.; Skager, R.; and Bentler, P.M.
Substance abuse and psychosocial risk factors among teenagers:
Associations with sex, age, ethnicity, and type of school. Am J Drug
Alcohol Abuse 13(4):413-433, 1987.
Planalp, S., and Honeycutt, J. Events that increase uncertainty in personal
relationships. Hum Commun Res 11:593-604, 1985.
Quinn, T.C.; Glaser, D.; Cannon, R.O.; Matuszak, D.L.; Dunning, R.W.;
Kline, R.L.; Cambell, C.H.; Israel, E.; Fauci, A.S.; and Hook, E. W.
Human immunodeficiency virus infection among patients attending
clinics for sexually transmitted diseases. N Engl J Med 318:197-203,
1988.
152
Remafedi, G. Sexually transmitted diseases in homosexual youth. In:
Schydlower, M., and Shafer, M.A., eds. Adolescent Medicine: State of
the Art Review: AIDS and Other Sexually Transmitted Diseases. Vol.
I(3). Philadelphia: Hanley & Belfus, Inc., 1990. pp. 565-582.
Rosen, D.S.; Xiangdong, M.; and Blum, R.W. Adolescent health: Current
trends and critical issues. In: Strasburger, V.C., and Greydanus, D.E.,
eds. Adolescent Medicine: State of the Art Reviews: The At-Risk
Adolescent. Vol. I(1). Philadelphia: Hanley & Belfus, Inc., 1990.
pp. 15-32.
Rosenbaum, E., and Kandel, D.B. Early onset of adolescent sexual
behavior and drug involvement. J Marriage Pam 52:783-798, 1990.
Rosenfeld, W.D.; Vermund, S.H.; Wentz, S.J.; and Burke, R.D. High
prevalence rate of human papillomavirus infection and association
with abnormal Papanicolaou smears in sexually active adolescents. Am
J Dis Child 143:1443-1447, 1989.
Schwartz, S.K.; Bolan, G.A.; Fullilove, M.; McCright, J.; Fullilove, R.;
Kohn, R.; and Rolfs, R.T. Crack cocaine and the exchange of sex for
money or drugs: Risk factors for gonorrhea among black adolescents
in San Francisco. Sex Transm Dis 19(1):7-13, 1992.
Schydlower, M., and Shafer, M.A. Chlamydia trachomatis infections in
adolescents. In: Schydlower, M., and Shafer, M.A., eds. Adolescent
Medicine: State of the Art Review, AIDS and Other Sexually
Transmitted Diseases. Vol. I(3). Philadelphia: Hanley & Belfus, Inc.,
1990. pp. 615-628.
Shafer, M.A., and Sweet, R. Pelvic inflammatory disease in adolescent
females. In: Schydlower, M., and Shafer, M.A., eds. Adolescent
Medicine: State of the Art Reviews: AIDS and Other Sexually
Transmitted Diseases. Vol. I(3). Philadelphia: Hanley & Belfus, Inc.,
1990. pp. 545-563.
Sonenstein, F.L.; Pleck, J.H.; and Ku, L.L. Sexual activity, condom use
and AIDS awareness among adolescent males. Fam Plann Perspect
21:152-158, 1989.
St. Louis, M.E.; Conway, G.A.; Hayman, C.R.; Miller, C.; Petersen, L.R.;
and Dondero, T.J. Human immunodeficiency virus infection in
disadvantaged adolescents: Findings from the U.S. Job Corps. JAMA
266:2387-2391, 1991.
Talmadge, W.C. Premarital sexuality. In: L’Abate, L., ed. The Handbook
of Family Psychology and Therapy. Homewood, IL: The Dorsey
Press, 1985.
153
Zweig Greenberg, MS.; Singh, T.; Htoo, M.; and Schultz, S. The
association between congenital syphilis and cocaine/crack use in New
York City: A case-control study. Am J Public Health 81(10):13161318, 199l.
AUTHORS
Cherrie B. Boyer, Ph.D.
Health Psychologist
and
Assistant Adjunct Professor
Jonathan M. Ellen, M.D.
Senior Fellow
Department of Pediatrics
Division of Adolescent Medicine
University of California, San Francisco
400 Parnassus Avenue, Room AC-01
San Francisco, CA 94143-0374
154
Going Nowhere Fast:
Methamphetamine Use and HIV
Infection
Mary Jane Rotheram-Borus, G. Cajetan Luna,
Toby Marotta, and Hilarie Kelly
INTRODUCTION
Adolescents who are unconventional or atypical often are drawn and
pushed into environments and subcultures where illicit drug use is
endemic. Their sexual behavior is influenced by the settings, norms, and
values of other members of these substance-using subcultures. To
decrease the level of human immunodeficiency virus (HIV) risk for these
youths, it is necessary to understand the contexts of adolescents’ sexual
behavior in order to develop successful intervention programs. In
addition to focusing on the age of initiation, types of acts, and rates of
sexual activities by specific members of marginalized subcultures, one
must understand the behavioral practices and patterns that typify
subcultural groups and consider the accompanying developmental and
life challenges (Commission on Behavioral and Social Sciences
Education 1993). This chapter describes the contexts and activities of
gay and bisexual male youths that may place them at high risk for HIV
infection. To understand and appreciate the characteristics of these
subcultures, descriptive and ethnographic research strategies are required.
This chapter presents selected life episodes and case studies of three gay
or bisexual youth living with HIV or acquired immunodeficiency
syndrome (AIDS). Their lives were characterized by substance use and
methamphetamine abuse. This chapter examines how their
developmental course was effected as the youths came to terms with their
sexual orientations while functioning within methamphetamine-addicted
subcultures. The chapter discusses common paths into methamphetamine
use or “getting high”; describes the behavioral patterns while addicted,
including “tweaking, freaking, and over-amping”; and explores individual
attempts to adapt, function, curtail, or discontinue methamphetamine use
and self-destructive practices that pervaded their environments. After
examining the context of their HIV-risk behaviors, the authors suggest
155
the types of intervention strategies that appear to help youths modulate or
cease methamphetamine use and thereby to reduce future HIV-risk
activity.
BACKGROUND
Sexual activity typically begins during adolescence. About 80 percent of
adolescents experience intercourse by the age of 18. Adolescent
intercourse often is unprotected, placing youths at risk for acquiring HIV
(Rotheram-Borus et al., in press). In developing interventions to reduce
youths’ risk for HIV infection, adolescent sexuality cannot be separated
from their broader developmental task of exploring and committing to
their sexual role and personal identity (Erikson 1968; Miller and
Rotheram-Borus in press; Rotheram-Borus 1989). The context of
adolescent sexual behaviors and substance use (with whom, what acts,
where, at what age, in what place, why, and under what circumstances)
critically influences the risk of contracting HIV and the type of
intervention program needed. It also impacts the adolescent’s long-term
adjustment.
Knowledge of the contexts of adolescent sexual behavior and substance
use is very limited (Commission on Behavioral and Social Sciences
Education 1993). Epidemiological data from schools and national
household surveys indicate that the age ‘and sequencing of sexual
activities (e.g., oral, anal, or vaginal sex) vary by gender, ethnicity, and
socioeconomic status (Boyer, this volume). In the last 7 years,
researchers have identified subgroups of youths at relatively higher risk
of HIV infection than their peers: gay and bisexual youths, runaway and
homeless youths, delinquents, those who are sexually abused, teenage
mothers, and injecting drug users (IDUs) and their partners (RotheramBorus et al., 1994). Among these subgroups of youths at high risk for
HIV infection, sexuality is linked to other problems in their lives (Ibid).
These youths are not only at risk for unprotected sexual activity, but are
also more likely to abuse alcohol and drugs, experience academic and
behavioral problems at school, have contact with the criminal justice
system, and feel depressed or suicidal (Dryfoos 1990; Jessor and Jessor
1977).
Certain youths are at risk for HIV infection, are involved in early sexual
activity, and have multiple problem behaviors in part because they are
atypical from their peers in some way. For example, youths who
156
self-identify as homosexual may feel different from the majority of their
peers in their sexual orientation. Often youths who recognize they are
homosexual have no adult role models for developing sexual identity,
feel isolated from their peers, and lack opportunities to establish
relationships that are not highly sexualized with same-sex partners
(Martin and Hetrick 1988). Fearing discovery as homosexual, many of
these youths leave home or lead multiple lives, masking their
homosexuality (Ibid). Parents who learn of a youth’s homosexual
orientation often reject and then eject these youths from home, leaving
them with few housing options (Hunter and Schaecher 1990). Without a
stable living situation or social support network, some of these youths
become marginalized. Marginalization often is associated with
involvement in substance-abusing subcultures.
While some unconventional adolescents seek situations that may involve
them in multiple problem behaviors, others involuntarily find themselves
in social settings or life situations where high-risk HIV activities are
common. One-third of the girls identified as HIV seropositive at one
hospital in New York City had previously had fewer than three sexual
partners at the time of their diagnosis and had never injected drugs
(Futterman et al. 1993). Yet these girls live in neighborhoods that have
high HIV seroprevalence rates, which is related to the endemic substance
abuse in these same neighborhoods. Girls living in rural environments
with low seroprevalence rates may engage in the same behaviors as those
in urban environments, yet have very little risk of HIV infection.
Similarly, homeless youths are at increased risk for HIV infection due to
their living situation. Some of these youths become homeless because of
unconventional behaviors. However, many are homeless because their
families are homeless (20 percent), their families have rejected them, or
the family environment is victimizing. For example, 42 percent of
runaway girls were found to be victims of sexual abuse (Rotheram-Borus
et al. 1992). Many youths at high risk for HIV infection are at risk due to
their living situations-situations they were not responsible for creating.
METHODOLOGY
The life episodes and case studies that follow were obtained from
October 1992 to April 1993 in ethnographic interviews and observations
as part of a larger project on secondary prevention of HIV and AIDS
conducted in San Francisco and Los Angeles and supported by the
National Institute on Drug Abuse (NIDA). These cases were selected
157
from a study group of 52 youths who were HIV seropositive and had
been interviewed repeatedly. Focus was on youths already
seroconverted. The ranges of lifestyles, scenes, and subcultures that
characterize their lives were identified. Youths were recruited from
community-based agencies, support groups, and snowball sampling from
street contacts. Some had participated in other studies before this project
began.
The strength of the qualitative data presented here rests on the
observational and networking skills of the researchers. Therefore, the
personal and professional backgrounds and expertise of the ethnographers
in working with the population and in conducting ethnographic work
were central assets of the current study.’ This expertise allowed for rapid
identification of the ranges and characteristics of the social roles and
niches occupied by youth living with HIV and AIDS.
Involvement in homosexual activity was a pathway for each of these
youths’ becoming substance users and their subsequent activities.
Because the youths were gay-identified, they often were stigmatized and
marginalized from families or peers. The degree of marginalization
varied depending on the degree of financial disenfranchisement. Some
lived in underclass sex and drug trade zones in the inner city and were
totally disconnected. Other middle-class youths were connected to the
support services of community-based agencies.
The cases selected for this chapter represent themes emerging from the
total study group. In general, youths varied in how and when they
identified themselves as gay or bisexual, in their socioeconomic
backgrounds and resources, and in the circumstances of their initiation
and continuation of substance use. Nevertheless, similarities in
behavioral patterns and themes illustrate how social settings, subcultural
groupings, and activities affect HIV risk-seeking and risk-taking
behavior. There were also parallels in their reasons for involvement in
methamphetamine-using subcultures. However, the youths’ methods and
abilities for ending methamphetamine use and disassociating from these
social worlds differed, and those differences suggest factors that must
influence prevention programs.
158
ADOLESCENT PATHS, TASKS, AND RESOURCES
Methamphetamine (known as meth, speed, crystal, and crank) has gained
popularity as a hard drug in many western cities (Currie 1993).
Thousands of illicit methamphetamine laboratories are estimated to be in
the United States, mainly in the West and Southwest; as evidence, there
were 775 reported seizures of methamphetamine labs in 1987 alone
(Currie 1993). In San Francisco and Los Angeles, methamphetamine use
defines a lifestyle and subculture characterized by activities that place
disenfranchised youths at higher risk for HIV and dominate their lives
(Marotta 1989, 1991, 1992). Many of these marginalized young people
already have become infected with HIV (Bermudez and Shalwitz 1992).
Youths who become enmeshed in this subculture are not short-term
runaways who are rescued easily and reintegrated into foster care or
families. These are long-term disenfranchised youths. Physically, these
youths may sometimes appear threatening, offensive, obnoxious, or
generally disagreeable. These features are consequences of lives
permeated with chronic substance use and the effects of permanent or
intermittent street life. Three case studies illustrate the social and
behavioral contexts that characterize HIV-risk activity for three young
men enmeshed in methamphetamine-using subcultures and social worlds.
Mark
Mark was 13 when he first arrived in San Francisco from the suburbs. He
was ejected from his home for being gay. As a result, he has been
responsible for his own survival for the last 6 years. Similar to other
disenfranchised youth, he soon learned that he could earn easy money
through full-time sexual activity for money on the streets and in gay bars.
He liked the attention paid to him; his previous family and peer
relationships were isolating. He remembers that he was “high” the first
time that he had sex with another male, an adult, when he was 13 years
old. Marijuana made the whole “sexual thing” easier. He later
experimented with crack cocaine and heroin and had a long stretch of
alcohol abuse that “clouded over” his life the past 5 years. In fact, he has
a hard time remembering his specific activities during this period because
his long- and short-term memory have been affected. He does remember
clearly that he began using speed in pill form at age 16 and has injected
speed almost continuously since he was 17 years old. His first injectable
speed experience was with an older man who “kept” him for a short
period of time. He remembers the older man “shooting him up” in the
159
forearm during a sexual episode. At that time he believed he was in love.
He stated that he has not had similar beliefs since that time.
Jim
Jim is an outgoing 21-year-old male who identifies himself as bisexual.
He reported a troubled childhood in Tucson. His mother was a heavy
drinker, a “bar-woman” who ignored him and her other children. He was
frequently cared for by his stepfather’s father, who regularly got him
drunk on alcohol when he was a child. His older brother taught him how
to smoke marijuana and chew tobacco before he was 10 years old. For as
long as he can remember, he used illicit substances. He “got high” every
day, except during the long stretches of time he spent in reform school or
jail.
Throughout his early adolescence he was in and out of reform school,
usually taking “raps” for his older brother, until one particularly
threatening crime and the prospect of jail caused him to run away.
Previous problems with the criminal justice system led him to the
Southeast, where he found that his relatives and friends smoked locally
grown marijuana.
Subsequently he hitchhiked to Portland, OR where he stole a motor,
scooter and traded it for some speed. After getting into trouble with the
local police when he was “wired,” Jim decided to go to Los Angeles and
then San Francisco.
“From the way I heard it, this was the place to be if, it was real liberal.
They called it the ‘Gay Bay,’ down here. I figured I’ll go check out the
Gay Bay! (laughs) I never left! This is where I fit in! I think I fit in.”
Rob
Rob discovered that he was homosexual at 15. He was able to develop a
gay lifestyle in the Northwest United States without losing the love and
support of his family. His upbringing was middle class. In his early
adolescence, Rob encountered a variety of drugs being used among his
friends:
“Acid basically, acid, cocaine, marijuana. I never injected anything, I
would never put a needle in my arm. I never have. I’ve tried crack,
160
I’ve freebased, I’ve done acid, I’ve done mescaline, mushrooms,
cocaine, crank.”
Rob’s first homosexual experiences were with men in their 20s who also
had “good jobs” (a disk jockey, a grocery store bagger). His first
boyfriend was a man he met in a “cruising” area in downtown Denver.
He turned out to be a hustler who was hooked on methamphetamine. His
mother moved to California with Rob and his sister about 5 years ago.
She began to use speed when forced to work graveyard shifts. One day,
when Rob “was going through one of her boxes,” he discovered her
“cross-tops.” With the move to California, methamphetamine became the
central feature of his life. He arrived in the Bay Area when he was 20.
Mark, Jim, and Rob came from different geographic areas, and their
families had varying personal and economic resources at their disposal.
However, their lives took a similar path after they began their drug-using
activities. Three central themes permeate their stories: “getting high” or
the initiation into substance use and the perceived functional value of this
activity; “tweaking and freaking” or the behavioral patterns during
substance-using episodes or careers, including unprotected sexual
activity; and “over-amping” or an extended state of anxiety and
discomfort (“hitting the wall”) when substance use activities are
perceived as clearly self-destructive or negatively affect the course of
their lives. Usually at this point they realize drug use must stop.
Examples of each of these themes provide insight into the motivations for
methamphetamine use and the rhythm and life patterns once involved in
the drug-using subculture, and suggest intervention strategies for ceasing
or moderating speed use.
GETTING HIGH: PATTERNS OF INITIATION AND
METHAMPHETAMINE USE
Getting high serves many functions and means many things to
disenfranchised inner-city youths, as demonstrated by the actions and
beliefs of Mark, Jim, and Rob. For example, methamphetamine use
results in a strong pleasurable feeling, so strong that it can become the
meaning of one’s existence, as Jim reported after his first high.
“He hit me (i.e., injected drugs). I started coughing and the rush hit
me! The room started shaking and right from then, I knew what my
mission was! (laughs) To keep that feeling up!...It made me feel like
161
I was nice looking, like I can do anything, really. Made me feel like I
was it, like I was really good.”
Mark’s experience of intense physiological pleasure was similarly
overwhelming:
“It made my hair feel like it was crawling up the back of my neck.
Like I was buzzing, wired... I just snorted a couple of lines, so, it
lasted probably 12 hours. I was high for 12 hours.”
Similarly, Rob reported:
“I used speed with everything, wherever. To stay awake. I used it
with him for sex... I used it just to feel good. There was always a
reason to use it.”
These euphoric feelings often are accompanied by strong feelings of
sexual potency when first using methamphetamine. Jim summarized the
aphrodisiac qualities.
“It made me feel real sexy; it made your sex drive real strong, too.”
Rob reported the same experience.
“I mean it’s sort of euphoric in a sexual way.”
To youths who feel depressed, uprooted, disconnected, or unsure of their
future, the rush of a drug high initially provides meaning and a purpose to
their lives. The need for meaning and structure that drugs seem to
provide is even stronger among gay youths, whose lives are typified by
social isolation. According to Rob,
“There wasn’t anyone my age, there still isn’t hardly...I mean there’s
more as I get older, but at the time, 17 years old, walking around in
this gay neighborhood, there wasn’t very many 17-year-olds that
weren’t screaming queens, you know, that were just normal people,
walking around.”
162
Given this isolation, methamphetamine use often provides a way for gay
youths to connect with others. As Mark stated:
“I had been alone for so long that I was desperate, and I had found
someone and I wasn’t alone anymore. I felt like I had a real
somebody in the world. I had a place in the world and I didn’t want
to lose that.”
This isolation may enhance the importance of romantic partnerships, and
lead to shared needle use and unprotected sexual activity as a way of
demonstrating love and commitment to partners. For example, Jim had
unprotected sex with an HIV seropositive partner to show that the partner
was “somebody in the world.” Jim “loved” the partner and wanted to
share his fate.
“Finally, I knew that’s how you caught it but I didn’t care at the time.
Me and him have a pact that when we go, we’re just gonna go
together.”
Rob expressed his fears that if he did not engage in unprotected sex, he
may lose his partners.
“I thought if I didn’t make them use it (the condom) maybe they’d
want to be with me, or stay, and I just didn’t care, I didn’t think it
would happen...I mean, I don’t know if I could say, give a reason for
every time why, there’s too many times to count.”
Thus, drug use and sexual risk often occurred while searching for love
and affection.
In addition to enabling youth to connect with others, one effect of
methamphetamine use is to disinhibit behaviors otherwise perceived as
taboo. Adolescents who are uncomfortable exploring homosexual or
bisexual feelings may place responsibility for their homosexual acts on
their drug-induced state. Jim described his first homosexual encounter
while under the influence of methamphetamine.
“I was staring at my brother and I didn’t know why (laughs). My
brother did bad things too, he used to play with me and stuff. When
he was high. Even when he wasn’t high, when we was kids, he’d
touch me and stuff....that’s the first time I really thought about it. In
a homosexual way.”
163
It is unclear whether the disinhibition reported by youths is solely the
physiological consequence of the speed use or whether it provides a
rationale that allows the youth to engage in taboo behavior and maintain
self-esteem. Other youths may be less honest with themselves regarding
sexual desires that may only be acted upon while under the influence of
methamphetamine. Rob, for example, minimized the differences in his
sexual behaviors in a “stoned” versus an “unstoned” state.
“I mean maybe they made me a little less inhibited, yeah, but I
wouldn’t blame anything on these drugs... Because I was doing these
things anyway, or was wanting to do these things anyway.”
The association between drug use and sexual orientation existed in the
recollections of each youth but in different contexts. For example, Rob
identified as gay at 15. Social isolation and masking his orientation were
primary precipitants of Rob’s drug use. He did well in high school,
where he was “out.”
“People usually didn’t mess with me. I very rarely had any trouble,
and when I did, I just ignored it, so I figured the only way it’s gonna
hurt me is if I let it hurt me. Of course, it wasn’t the best feeling, to
be excluded for that, but I had my own friends.”
Rob was a successful and productive student. However, at age 16, his
partner was using methamphetamine, and the relationship ended abruptly
and dramatically.
“I was tired of the Midwest, I was upset. Ben (my partner) was doing
things I didn’t like...he would go out and not come home, and even
hustling or whatever he was doing....I just needed to get out.”
Rob left home and moved to San Francisco. Because there were no
openly gay students in his new suburban high school, Rob went “back
into the closet.” He explained:
“Well, I knew I was gay. My mom knew I was gay, but she and I
never really talked about it, and when I went to high school, I just let
everyone think I was straight. I mean I didn’t act gay, so... I went out
with girls, and I went with the guys, and whatever... I don’t
remember having any sex during that time. It was about 6 months...
my brother who is gay, one of my brothers from my father... came
down a couple of weeks after I moved out here and he took me up to
164
the city. It was the first time I’d ever been to this one gay area, and I
said ‘Wow. I don’t believe this.’ There was guys holding hands
walking down the street together, it’s amazing... But I didn’t go back
up there for a while just because I was busy doing things with straight
people, with people my own age, which was also something very
strange, cause I never hung out with people my own age in the
Midwest.”
It was during this period of masking his sexual orientation that Rob
started using drugs.
Most of the youth in the larger study group experienced conflict over
“coming out” and choosing a gay or bisexual lifestyle. This conflict was
interconnected with methamphetamine use for Mark, Jim, and Rob.
Many youths who identify as gay or bisexual do not adopt substanceabusing patterns. However, negative societal reactions to gay and
bisexual youths appear to lead some youths out of traditional mainstream
cultural practices in order to find acceptance, personal worth, and value.
Some social worlds and subcultures are more accepting of diversity and
welcome gay youths. Once accepted into alternative subcultures, some
youth become vulnerable to the negative or destructive activities or
practices characteristic of these adopted social worlds.
While many youths experienced an increased sense of meaning,
self-worth, pleasure, and disinhibition associated with initial
methamphetamine use, eventually there were negative consequences of
chronic use. For each benefit, a contrasting cost arose-costs far higher
than the benefits and almost unavoidable. As Jim summarized:
“So it just kept going down, down, down, and so he moved, and kept
doing more drugs, and I did drugs with him. We stayed up all night
and we’d fight, and we’d get back together again and have great sex.”
Most youths experienced a downward spiral and were addicted to the
drug and to the lifestyle that surround usage. Activities and relationships
of physical and emotional ecstasy and disinhibition became
dysfunctional. For example, while methamphetamine was initially an
aphrodisiac, it soon became necessary for sexual arousal. Episodic or
165
continual sexual dysfunction was reported in each interview with young
homosexually active methamphetamine users. As Rob stated:
“You know, it’s hard to get an erection when you’re on crystal or a
lot of crank.”
Jim’s account also points to how methamphetamine use potentially may
lead to high-risk situations.
“You can’t get an erection when you’re on speed, it’s concentration.
He (meaning, his lover) can never get an erection! (laughs) Hardly
ever. When he’s high, if we play games, it’s easier for him to get it
up ... (laughs). Like play sex games like handcuffs.”
Mark also became disinterested in sex and only engaged in sexual
activities for money or as a means of getting personal validation from
others.
“I have to be on drugs to have sex with anybody. Period, ‘cause I do
not enjoy having sex. I do not go out and look for sex. I don’t even
masturbate. I don’t, like it’s part of my manic depressiveness.
Depressed people do not like to have sex... That doesn’t appeal to
me. I won’t (have sex) unless I’m on speed.”
Methamphetamine use is characterized by increased sexual activity not
only because of myths suggesting increased potency but also in circular
fashion due to an interdependency with sex work as a means of support
for the “meth habit.” Two of the three youths described in this paper
traded sex for drugs or money. As Jim’s comments illustrate,
involvement in the drug-using subculture may place youths in situations
where sex is bartered.
“We bought this bag and we went back to my hotel...and I split it
with him. He did half and I did the other half, and it instantly
sobered me up, I remember. I didn’t know what was going on and
stuff like that, and I thought I was beautiful and all this stuff, and I
came out to try to make some more money, and I did, and that was
how I got into that whole trap of money, making money for sex and
constantly doing speed.”
The methamphetamine subculture has clear normative roles for sex work.
For example, Mark distinguishes between “johns” and “tricks.” “Tricks”
166
are short-term sexual partners from whom Mark receives money, whereas
“johns” can be “worked” over long periods of time, as is the case with
“sugar daddies.” The youths’ descriptions of bartering sex demonstrate
how initial attributions of self-worth and potency associated with chronic
drug use eventually are replaced by alienation, as demonstrated by
Mark’ s account.
“He is in love with me too, another sorry guy who fell in love with
me (laughs), I kept saying, well I want to buy some dope with you...
so she (sic) says, here’s the money go buy the coke...She likes to get
people real full (high) and give them dope and have sex.”
Methamphetamine use as a means of connecting and experiencing
intimacy with others eventually evolves into situations where drug use
increases, and the relationships dissolve in conflict. Rob’s description of
his problems with Alex demonstrates this pattern.
“Alex was addicted and at first he told me he was gonna quit. And
then he didn’t, and then, when he and I started to fall apart, he started
to do methamphetamine a lot more, and then I started to do it with
him, cause it was one way we could be together.”
Youths get high and use methamphetamine for various reasons that often
influence their present circumstances. Rob succinctly summarized the
reasons for his former partner’s speed use.
“He said he was doing it to lose weight, but that wasn’t the truth.
The truth was he was trying to forget that his father abandoned him,
his uncle molested him, or his mother didn’t love him; his brothers
were messed up, and he was gay.”
Youths who become part of the methamphetamine-using subculture find
that their sexual satisfaction is enmeshed with their survival needs and
feelings of self- worth. To them, getting high is characterized by a
euphoric physiological rush, and is used to increase personal feelings of
self-worth and sexual potency, to connect with others, and to explore
one’s sexual orientation without guilt. Getting high also makes it easier
to participate in undesirable but necessary sex work in order to provide
for one’s immediate survival needs. However, in time, the long-term
consequences of chronic methamphetamine use become more apparent:
sexual dysfunction, alienation from others, and marginalization into
destructive interdependent subcultural social networks.
167
TWEAKING, FREAKING, AND OVER-AMPING: LIFE ON
METHAMPHETAMINE
In street vernacular, shooting speed and having sex are linked together
and referred to as “tweaking and freaking.” Tweaking and freaking
represent a pattern and rhythm to the lives of some young people. For
example, Jim reported that once he became a regular tweaker, he
exhibited mood swings and self-destructive phases as effects of his drug
use. Every time he injected speed, he would try to stay as high as he
could for as long as he could for as many days as possible. His goal was
to get as high as he could (get full) without over-amping (too many drugs
for too long). Over-amping would make him too nervous and too
irritable to be comfortable. To avoid the manic-like state that
accompanied crashing, he stayed high by shooting up “just the right
amount” of “just the right stuff’ at “just the right time” intervals. The
goal was to stay awake until he was so exhausted that he could put
himself to sleep with a little alcohol.
This cycle of tweaking, freaking, and over-amping methamphetamine
leads to acts that further marginalize youths. Theft and petty crime are
characteristic activities of the methamphetamine-using subculture. For
example, Jim was hustling full-time when he joined in the petty crime
that is a feature of the local speed scene.
“If I was tweaked enough, I’d get a trick and he’d invite me to stay.
I’d spend the night and tweak around, drawing and scribbling. If
they were dumb enough to go to sleep when I was tweaking (laughs)
in their house, they’d get robbed most of the time....Because that way
I wouldn’t have to go back on the street and hustle....I’d go to my
dealers. With a new flight jacket, get 1/2 gram for it. Jewelry, I had
a gold class ring that I took, it had a sapphire in it, I got 1/16th for it!”
In addition to the economic relationship between drugs and sex, drugs
may have a psychologically numbing effect in sexual activity. Pleasure
and self-worth are disassociated from behaviors. Mark explained how
drugs allowed him to participate in otherwise unpleasurable sexual
activity.
“Just like the big fat guy that picked me up one time, gave me $300.
But she (sic) wasn’t the most attractive thing in the world... but all
she wanted to do was lay on top of (you), and do nothing, but lay on
168
top of (you). If I wouldn’t have been high on something I never
would have went with her (laughs).”
Survival often is dependent on drug connections. Other “players” and
their gatherings provide the context for unprotected sexual activity to
occur. For example, Mark identified Daniel as the person he would go to
for help. Daniel is his 34-year-old “former lover” who is also one of the
main methamphetamine suppliers in his neighborhood. Daniel provides
Mark with a free $20 bag of speed daily, and does not make any sexual or
financial demands in return. According to Mark, “He loves me.” Mark
reported that he is not now actively involved full-time in sexual activity
for money, although he does sometimes run into former clients and “falls
into old patterns.” Occasionally he will walk down the main male sex
worker street (“my daily stroll”) while high on methamphetamine, and if
a car stops, he does not say no to advances.
HIV-risk behavior can escalate within the context of sex-for-money
exchanges. Jim’s first experience with anal receptive intercourse
occurred with a trick. He had not anticipated being the receptive partner
when sexual activity began.
“The first time I did do it and it was for money, we had already got to
his house and we agreed that he’d pay me $50 to have oral sex.
Right in the middle of it, he told me he’d give me an extra $100 if I’d
let him do that (anal sex). I said okay and he put on a rubber and did
it....It hurt! I felt that feeling, for months, ugh! I couldn’t stand it, I
hated it! It made me feel like a woman! I felt real cheap.”
Similar routine, negotiated, or forced sexual situations or behaviors are
dictated by immediate survival needs. Relationships often are entered
into only for support, protection, or survival. Mark, for example,
emphasized that he is not attracted to men his own age because he feels
they are “generally stupid and without direction.” While Mark likes the
financial stability usually characteristic of older men (age 40 to 50), he
describes these men as generally dysfunctional, saying they “fall in love
too easily.” While discounting those his own age, Mark’s choice of
partners usually is based on the potential benefits he can obtain. Money
to purchase speed is a frequent benefit of his cross-generational
relationships.
“...My track record, I mean it’s not exactly, you know, making the
best time, the best mileage, I’ve left a lot of accidents (laughs) on my
169
highway of love. It’s like, I’d be the first person to tell you, excuse
me, but you really don’t want to be in love with me (laughs).
Because I’ll work you, because if I know that somebody is going to
do anything for me, I will work them. Not really bad, but I will work
them, just like Daniel. Daniel came to pick me up in jail and gave me
money and sex, that’s why she’s (sic) the first person that I called, I
knew he’d pick me up and I knew he’d have sex with me.”
For Mark, chronic methamphetamine use has also had negative social
side-effects, including periods of self-imposed isolation and paranoia.
“I mean I was so tweaked out from the weekend it wasn’t even funny,
I was in my room, I had my little friend, my little mouse, see, to this
day I don’t even know if I actually saw that little mouse or not, but I
thought I saw this little mouse, and for a whole weekend I was in my
room and I’d pick up my coat, and I was trying to leave, and I go like
this (gestures) making sure there was no mouse in my coat...I woke
up Sunday morning and the first thing I did was, like throw up
stomach acid, and I said gee, girl, (sic), when was the last time you ate
(laughs)?”
Following long periods of sleep, the postcrash depression is less
pronounced and the exhaustion gone. The task then is to arrange for a
new supply of speed, the sooner to begin another “high”—another “run.”
MODERATING OR DISCONTINUING METHAMPHETAMINE
USE
Given the interconnected nature of sexual activity and speed use, it is
difficult to prevent HIV sexual-risk activity if methamphetamine use
continues. Some youth, especially those already living with HIV and
AIDS, are fatalistic, experience hopelessness, and do not want to change
their drug use practices. For example, Mark described his unwillingness
to change his life situation and patterns in the following way:
“... as far as I can see, I’m just going to die anyway, so why bother
getting my life together when I’ll probably die before I do it (laughs).
Tomorrow I could be in the hospital with some tubes coming out of
me.”
170
Hopelessness is the rationale for Mark to continue using drugs. While
Rob’s explanation is different, similarly he was not motivated to protect
himself from HIV infection. Rob initially perceived himself as
invulnerable.
“Of course there were general things out and around (about AIDS),
you know, in the city or whatever, but not in the suburbs. I don’t
remember anything...It wasn’t a priority with me, and it never
became a priority with me...I don’t know why. I guess it never got
into my head that I could be infected....I had problems with
self-esteem. I didn’t care, for a long time. I didn’t care if I got
infected, it was like, what does it matter.”
Even if these adolescents were motivated to change their drug use and
sexual behaviors, most do not know what to do and how to change. In
fact, many say there were few prevention messages. For example, Jim’s
knowledge about HIV transmission upon arrival in San Francisco was
incomplete.
“I thought the only way to get AIDS was through anal sex. I thought
that as long as I didn’t do that, then I was cool!... I couldn’t figure
out why them outreach workers were passing bleach out on the
streets (laughs). Then I learned about bleach your needles, clean ‘em.
I figured my needle works just fine, I clean them with water...Within
a week or two, them outreach workers said don’t share needles,
you’ll get AIDS! I had already done a bunch of them things.”
Information is understood and acted upon slowly and over time in
informal settings and communities. Eventually, group norms and
subcultural practices have to be changed for consistent self-protective
behavior to occur. Rob, for example, would only use condoms if a
partner took the initiative, supplied them, and insisted that he put one on.
First and foremost, he wanted to please.
“There may have been very few times that I did (use a condom). I
mean, this is the Northwest, it (condom use) didn’t happen in the
Northwest, you know, it was like no one knew anyone that had ever
had HIV there...There wasn’t even really any prevention messages. I
went through sex education in high school and I never heard diddly
about it....it seemed to have been a far-off thing to me, you know...it
was not there.”
171
Unprotected sexual activity is more complicated and difficult to
understand when one partner is an AIDS educator and is aware of
infection precautions. One of Jim’s partners was employed on a study of
male prostitution, HIV, and intravenous (IV) drug use. Jim received his
AIDS education face-to-face from his first “boyfriend.”
“Yeah, he told me about sharing needles. We’d usually buy a
10-pack, but we’d get so tweaked that we’d just grab any one! He
told me to be careful. He told me he’d been at this a long time and
he’d taken a lot of chances, that he didn’t want me to take any
chances. At that time, I didn’t care ‘cause I wanted me and him, it
sounds real stupid and it is real stupid, I figured that if he’s gonna die,
then I want to die! That’s weird, huh?”
Most methamphetamine-using youths have tried at one point or another
to quit and get their “lives in order.” Total cessation of speed use is very
difficult, as reflected in Jim’s account of his first experience with a courtordered drug treatment program.
“You had to go to groups. Wake up in the morning, go to breakfast
and go to group. After that group, you’d have a small break. After
break, go back to group (laughs). Groups all day long, talking about
your drug history, drug abuse. All it really does is make you want to
use more drugs! Talking about drugs all day, you kind of want to use
them!”
The 6-week treatment program Jim was forced to enter was intended to
discourage him from patterns of drinking and drug use he had known
since childhood. Jim did not take his treatment sessions seriously.
“I didn’t really want to be there in the first place! I told them what
they wanted to hear! I never planned on quitting!”
Jim’s primary partner, Scott, overtly discouraged and yet subtly
encouraged Jim’s substance abuse. Scott took Jim to various treatment
programs in San Francisco without lasting success. Finally, Scott
succeeded in getting Jim to give up alcohol by persuading him that booze
made Jim too contentious and violent. However, speed was tolerable so
as long as Jim refrained from over-amping or using too much of a batch
that was too strong. Jim’s own efforts to stop regular speed use ran
aground because his partner had grown so accustomed to tweaking and
172
freaking that he could no longer have sex without using
methamphetamine. Their lives contained many poignant contradictions.
As evident in the recollections of all three young men, methamphetamine
use was initiated and maintained in order to facilitate intimacy with
others. Implementation of HIV prevention strategies by using condoms
and stopping drug use was limited by fears of alienating others, conflict,
or losing a partner. Rob reported that he was depressed because he did
not have an intimate partner. He began to attend a youth rap group
sponsored by a gay-affiliated advocacy group. He also “cruised” gay
bars. Regardless of the setting, Rob’s condom use was dependent on his
partner’s desires.
“If I wanted someone, if I wanted to date someone, I wouldn’t make
them wear one, but if it was just a one-night thing, I would.”
The same fear of rejection was evident in Jim’s relationship, even when it
was known that his partner was seropositive. Jim and his partner used
condoms inconsistently. Fear of HIV was less salient than demonstrating
their love and loyalty.
Given these complicated circumstances, it might appear impossible for
youth to be able to stop using methamphetamine. However, in almost
every instance an opportunity existed when interventions could be
successful. Chronic methamphetamine use characteristically is
sequential, from getting high and running to the inevitable crash or
coming down. Youth seem to be especially receptive to interventions
during that period when they are coming down (or immediately
following) and before they begin another run.
Unfortunately, Mark rarely comes down, and his periods of sobriety are
few. He currently is living alone in a downtown Tenderloin “skid row”
hotel room provided by a program for seropositive youths. He listens to
the radio when home and while high on speed. The rest of his day is
spent borrowing, stealing, or “scamming” money for “shopping sprees” to
obtain more speed. He does not socialize with others at the hotel, most of
whom are also drug users and sex workers. His immediate primary social
world is composed of his speed supplier, his regular johns, and other
youths from the seropositive youth program.
The goals for HIV prevention with an adolescent like Mark are to
promote his self-esteem, to provide him with realistic and achievable life
173
alternatives and opportunities, and to develop within him a future
orientation and a will to live. He currently describes his life as “running
on autopilot.” He does not feel he controls his destiny but is resigned to
his fate. Mark continues his speed usage without any plans to quit. He
remains in a sort of social and developmental limbo, without plans for the
future or employment possibilities. He lives with the assistance of
service providers who have taken care of him for years. Currently, he is
awaiting general assistance and eventually social security. His high-risk
sexual activity ceased years ago, and while his IV drug use continues,
clean needles are more accessible. He plans to live in this state
indefinitely.
For Jim, intervention opportunities existed following a drug-related
incident when Scott, after a 9-day drug “run,” “raped” Jim. The police
were called, and Scott was put in jail. Jim was again destitute and denied
access to his belongings. Following this episode, Jim spoke about giving
up methamphetamine. He talked of getting a job and securing his own
living space. This period was short in duration. Unfortunately, Jim could
not act quickly enough to take advantage of his resolve to improve his
life situation. Given his lack of knowledge of available services and the
long waiting list to access drug treatment, he soon returned to familiar
people and activities and old behavioral patterns.
After a couple of weeks, Jim resumed a life of getting high and sexual
activity for money in adult bookstores. After the period had passed when
interventions could have helped, his concern about himself and his risk
for HIV infection diminished. Jim considered HIV the least of his
worries. Most of all, he struggled with daily survival, and he could not
imagine doing so without speed. As long as he was out of jail, he
remained trapped in his world of tweaking and freaking. Drugs, alcohol,
and sexual activity for money were the foundations of his street-based
lifestyle. He continued his drug-related activities. This pattern had
become the only life he knew.
Having more personal abilities and resources, Rob experienced fewer
stressful events. Rob had accepting parents and a home life, had
succeeded academically, held a job, and owned a car. Nevertheless,
when Rob decided to stop using methamphetamine, his desire for the
intimacy he associated with speed use continued. This need led Rob to
seek social support at a community-based agency. He decided to try a
rap group for gay youth.
174
“There were like 20 people or so, but most of them I didn’t, I wasn’t
interested in being associated with, they were queeny or whatever...
But it was good to have it, because I did meet some people that I
could relate to, that I’m still friends with today, and I met someone I
became involved with.”
Rob successfully used this therapeutic youth group to connect with
others. After donating to a campus blood drive in the fall of 1992, Rob
learned that he had been infected with HIV. For help adjusting to his
new status, Rob went to a support group for youth living with HIV and
AIDS. In the winter of 1993, he was working up the courage to tell his
family and his lover his HIV status. He had started using condoms
regularly, stopped using drugs, and decided never again to use speed.
The steps Rob had taken-joining a support group, seeing a therapist,
finding new employment, taking an apartment with another person living
with HIV, beginning an intimate relationship with a non-drug-user-were
examples of the dramatic life changes that are associated with drug use
cessation (Kandel and Raveis 1989; Stall and Biernacki 1986). With a
variety of social supports, Rob’s resolve to change the conditions of his
life may be reinforced and succeed. Unfortunately Mark and Jim do not
share the same resolve or have positive support networks available.
IMPLICATIONS FOR INTERVENTIONS AND CONCLUSIONS
Methamphetamine use can provide youths with an illusion of excitement
and satisfaction in their lives. Initially it is a means to an end: to get
high, to increase sexual potency, to facilitate intimacy with others.
Eventually, chronic use becomes an end in itself, as waking hours are
spent from one run to the next and all efforts are geared towards getting
more drugs. Instead of sexual potency, sexual dysfunction results;
instead of intimacy, further isolation and paranoia occur. The
developmental tasks that face all youth are distorted or retarded.
Metaphorically and in reality, methamphetamine use is equivalent to
going nowhere fast.
Youths like Mark, Jim, and Rob who participate in methamphetamineusing activities and subcultures clearly are at risk for HIV infection.
These youths have multiple problems in their lives, and the contextual
factors that influence their HIV risk are on many levels. At the most
basic level, some youths frequently are concerned with finding shelter for
175
the night and food for the morning. With such immediate and basic
concerns, youths are not thinking about dying from AIDS 10 years in the
future. Unless the youth’s life is considered holistically, the salience of
primary and secondary HIV prevention will be transient at best. A needshierarchy approach must be adopted to the design of intervention
strategies (Maslow 1970). Therefore, the authors make following
recommendations.
1. Unless immediate needs for survival (food, shelter), security (food
and shelter tomorrow), safety (some protection from violence), and
social support are met, youths will be unlikely to practice
HIV-preventive behaviors consistently. Youths whose survival needs
are not met often need to trade sex for money and drugs (AbelPeterson 1992). In such bartering, the person who needs money is in
a less powerful negotiating position (Cohen 1992). Reid and
colleagues (1993) found that gay youths who bartered sex were far
less likely to decrease their HIV-risk behaviors. Youths living with
HIV or AIDS who have received housing supplements feel more
secure and may choose to remain at home instead of going out to
barter sex. They know they will have shelter the next day. The need
to raise quick money for housing is eliminated.
2.
Alternatives must be developed and encouraged whereby youths can
meet their needs for intimacy and community in ways other than
unprotected sexual activity and shared drug use. Providing youths
with a positive social support network can reduce the need for
seeking sexual intimacy in risky ways. The challenge in developing
interventions and programs for these youths is whether programs
should address short- or long-term needs, or both.
3. There are not enough long-term placements to serve the numbers of
disenfranchised, alienated youths (Rotheram-Borus et al. in press).
Federal, State, and municipal agencies typically have devoted their
resources to emergency or temporary shelters instead of more stable,
long-term facilities. A major goal for social service agencies should
be to develop a system to relocate youths gradually from drugsaturated environments to less seductive and less toxic long-term
settings. They also must provide vocational and employment
opportunities whereby youths can satisfy their basic survival needs in
more conventional ways. The program at Larkin Street Youth Center
in San Francisco is an example of a successful approach (Kennedy
1989). The Center offers street outreach services, daily drop-in
176
counseling, meals, a medical clinic, crisis management, long-term
housing, and job development and placement opportunities.
Community-based agencies like the Larkin Street Center with active
outreach and drop-in services are a bridge for youths to begin
building new social support networks.
4. An understanding of youths’ substance use practices and patterns is
central to determining the features of interventions to reduce
HIV-risk activity. When youths are high, condom use often is
overlooked. Chronic substance use allows for the disinhibition of
unprotected sexual acts, avoidance of difficult life situations and
negative emotions, and temporary enhancement of feelings of selfworth. It eventually prohibits the development of truly intimate and
supportive relationships with others. These factors derail
adolescents’ healthy development and eventually may have
consequences in their adult life. Counselors must help each youth
identify and chronicle their behavioral patterns and activities.
5. Repeated and consistent consciousness-raising activities on drug use
and sexual activity are needed. These activities are crucial for youth
functioning in the “fast lane” who have little time for introspection.
The development of personal projects could be beneficial (Etwart
1991). For some, personal projects are a prerequisite to behavioral
change. A project could be disclosing sexual orientation or HIV
status to family, returning to school, or learning skills to reduce stress
and increase relaxation. Underlying feelings of low self-worth
(e.g., “I’m no good; I’m a user”) associated with youths’ selfdestructive acts must be confronted and addressed in personal
projects. Youths must be convinced they are entitled to better lives.
Positive reinforcements should be included in the definition of
personal projects. Setting career goals, finding employment
opportunities, and establishing specific plans for the future are basic
to motivating behavioral change. Each of these components of a
comprehensive program with case management appears necessary to
effect behavioral change and facilitate exiting methamphetamineusing social worlds.
6. The existential dimension of life should be emphasized in
interventions. This area typically is not addressed in most programs
or theories. Posing questions such as “Why am I here?” or “Where
am I going?” and the discussions that follow can help individuals
develop meaning and purpose in their lives. Discussions of this type
177
also can help youths explore and commit to personal projects and
encourage a future orientation.
7. In addition to the necessary service interventions, a research agenda
is needed to detail the risk behaviors associated with
methamphetamine use and HIV transmission. Longitudinal field
studies must be conducted, following youth in depth and over time to
ascertain strategies individuals have employed for moderating or
discontinuing their drug-use habits. Specific attention must be given
to identifying and chronicling the “turning points“ in their drug-using
patterns and sexual careers. These turning points may be the only
times when interventions and outreach programs can be truly
effective.
NOTES
1.
G. Cajetan Luna has conducted local, national, and international
research and policy development on street youth and AIDS for 12
years (Luna 1987, 1991; Luna and Rotheram-Borus 1992). He
worked out of the main San Francisco youth service agencies. He
interviewed youth in natural locations and studied their friends living
with HIV or contacts who were not clients of service agencies. All
but one of these youth were currently using methamphetamines.
None was receiving experimental AIDS treatments; all believed that
their drug use was not detrimental to their HIV status; and all lived
on a day-to-day basis. Mark is typical of the youth Luna interviewed.
Toby Marotta has been studying sexual and drug-using subcultures in
San Francisco since 1976 (Marotta et al. 1982). While doing
ethnography with youth living with HIV in San Francisco, he has
been interviewing moderate and heavy speed users for the first
behavioral study funded by the National Institute of Mental Health
entitled “Ice and Methamphetamine Use: A Three Year
Exploration.” Marotta employed longtime key informants to access
and study youth living with HIV and AIDS in the downtown area
who were not attending youth service agencies. To compare the lives
of these youth with those who receive support, he subsequently is
studying youth served through the Public Health Department. Jim
and Rob are typical of the youth he has been studying through both
informal and formal channels.
178
Hilarie Kelly is an anthropologist and ethnographer with crosscultural experience studying sexual and drug use practices in Africa
and young people in Los Angeles and Orange County. She accessed
youth through the main service agency for gay and lesbian youth in
Los Angeles and through formal and informal networks in Orange
County, CA.
REFERENCES
Abel-Peterson, T. “Street Youths and HIV.” Paper presented at the
Second International Encounter on the Health Problems of Street
Youths. Rio de Janeiro, Brazil, September 3-5, 1992.
Bermudez, R., and Shalwitz, J. Report on HIV seroprevalence and
transmission factors in young people using special programs for youth
clinics. In: Baxter, M., and Reyes, M., eds. HIV and Adolescence:
Issues in San Francisco - 1992. San Francisco: Special Programs for
Youth, 1992.
Cohen, J. “Different Types of Prostitution Show Wide Variation in HIV
and Other Sexually Transmitted Disease Risk.” Paper presented at the
Eighth International Conference on AIDS, Amsterdam, the
Netherlands, July 1992.
Commission on Behavioral and Social Sciences and Education. Losing
Generations: Adolescents in High-Risk Settings. National Research
Council. Washington, DC: National Academy Press, 1993.
Currie, E. Reckoning: Drugs, the Cities, and the American Future. New
York: Hill and Wang, 1993.
Dryfoos, J. Adolescents at Risk: Prevalence and Prevention. New York:
Oxford University Press, 1990.
Erikson, E. Identity: Youth and Crisis. New York: Norton, 1968.
Etwart, C. Social action theory for a public health psychology. Am
Psychol 46:931-946, 1991.
Futterman, D.; Hein, K.; Reuben, N.; Dell, R.; and Schaffer, N. HIV virus
infected adolescents: The first fifty patients in a New York City
program. Pediatrics 91(4):730-735, 1993.
Hunter, J., and Schaecher, R. Lesbian and gay youth. In: RotheramBorus, M.J.; Bradley, J.; and Oblolensky, J., eds. Planning to Live:
Evaluating and Treating Suicidal Teens in Community Settings.
National Resource Center for Youth Services. Tulsa, OK: University
of Oklahoma, 1990.
179
Jessor, R., and Jessor, S.L. Problem Behavior and Psychosocial
Development: A Longitudinal Study of Youth. New York: Academic
Press, 1977.
Kandel, D.B., and Raveis, V.H. Cessation of illicit drug use in young
adulthood. Arch Gen Psychiatry 46:109-116, 1989.
Kennedy, M. “HIV Prevention with Adolescents Who are HIV
Seropositive.” Paper presented at the Annual Conference on HIV
Prevention, Washington, D.C., March 1989.
Luna, G.C. HIV and homeless youth. Focus 2(10), 1987.
Luna, G.C. Street youth: Adaptation and survival in the AIDS decade.
J Adolescent Health 12(7):497-582, 1991.
Luna, G.C., and Rotheram-Borus, M.J. Street youth and the AIDS
pandemic. Supplement. AIDS Educ Prev 4:1-13, 1992.
Marotta, T.; Fisher, B.; and Pincus, M. Adolescent Male Prostitution,
Pornography, and Other Forms of Sexual Exploitation. San Francisco:
Urban and Rural Systems Associates, 1982.
Marotta, T. Midnight Cowboys. San Francisco: Self-Published
Ethnography, 1989.
Marotta, T. Snowballing Sex Workers. San Francisco: Self-Published
Ethnography, 1991.
Marotta, T. Doing Sex Work. Self-Published Ethnography, 1992.
Martin, A.D., and Hetrick, E.S. The stigmatization of the gay and lesbian
adolescent. J Homosex 15:163-183, 1988.
Maslow, A.H. Motivation and Personality. 2d ed. New York: McGraw
Hill, 1970.
Miller, R., and Rotheram-Borus, M.J. Growing up biracial in the United
States. In: Salett, E.P., and Koslow, D., eds. Race, Ethnicity, and Self
Identity in Multicultural Perspective. Washington, DC: National
Mutlicultural Institute Publications, 1994.
Reid, H.; Rotheram-Borus, M.J.; Rosario, M.; and Gwadz, M.
“Effectiveness of HIV Prevention with Homeless Youth Over Two
Years.” Paper presented at the Ninth International Conference on
AIDS, Berlin, June 1993.
Rotheram-Borus, M.J. Therapeutic issues in assertiveness training.
Psychol Q J Hum Behav 21(1):28-33, 1984.
Rotheram-Borus, M.J. Ethnic differences in adolescents’ identity status
and associated behavior problems. J Adolesc 12:361-374, 1989.
Rotheram-Borus, M.J., and Tsemberis, S.J. The role of ethnicity in social
competency training programs. In: Bond, L; Compass, B.; and Swift,
C., eds. Prevention Research in the Schools. Fontana, CA: Sage
Publications, 1989. pp. 297-3 18.
180
Rotheram-Borus, M.J.; Feldman, J.; Rosario, M.; and Dunne, E.
Preventing HIV among runaways: Victims and victimization. In:
Diclemente, R., and Peterson, J., eds. Preventing AIDS: Theories and
Methods of Behavioral Interventions. Newbury Park, CA: Sage
Publications, 1994. pp. 175-188.
Rotheram-Borus, M.J.; Koopman, C.; and Ehrhardt, A. Homeless and
runaway youth and HIV infection. Am Psychol 46:758-766, 1991.
Rotheram-Borus, M.J.; Meyer-Bahlburg, H.; Koopman, C.; Rosario, M.;
Exner, T.; Henderson, R.; Matthieu, M.; and Gruen, R. Lifetime
sexual behaviors among runaway males and females. J Sex Research
29:15-29, 1992.
Rotheram-Borus, M.J.; Rosario, M.; and Koopman, C. Minority youths at
high risk: Gay males and runaways. In: Gore, S., and Colten, M.E.,
eds. Adolescent Stress: Causes and Consequences. Hawthorne, NY:
Aldine de Gruyter, 1991.
Rotheram-Borus, M.J.; Mahler, K.; and Rosario, M. AIDS prevention
with adolescents. AIDS Educ Prev, in press.
Stall, R., and Biernacki, P. Alcohol and drug use during sexual activity
and compliance with safe sex guidelines for AIDS: The AIDS
Behavioral Research Project. Health Educ Q 13:1-13, 1986.
ACKNOWLEDGMENTS
This chapter was prepared with support from a grant from the National
Institute on Drug Abuse (1R01DA07903-01) and a Faculty Scholars
Award to the first author from the William T. Grant Foundation. We
acknowledge the collaboration of Dr. Robert B. Edgerton of UCLA and
Dr. Anselm L. Strauss of UCSF. The community-based agencies that
provided access to youths and that offer the basic support for helping
youths cope with difficult living situations were critical to initiating this
project. We especially appreciate the youths who opened their lives to us,
allowing for the preparation of this paper.
181
AUTHORS
Mary Jane Rotheram-Borus, Ph.D.
Director, Adolescent Prevention Studies Unit
G. Cajetan Luna, M.A.
Director, Hoven Project
Toby Marotta, Ph.D.
Project Ethnographer
Hilarie Kelly, Ph.D.
Project Ethnographer
The Adolescent Studies Group
University of California, Los Angeles
10920 Wilshire Boulevard, Suite 1103
Los Angeles, CA 90024
182
The Context of Risk:
Methodological Issues
Zili Sloboda
The behaviors that put most adolescents and adults at risk for human
immunodeficiency virus (HIV) infection are those that enable the
transmission of the virus through the sharing of bodily fluids from an
infected to an uninfected person. This is the epidemiologic approach to
the problem. Much of the available information that is derived from both
epidemiologic and prevention intervention studies has been based on the
numeric counts of specific behaviors that become surrogate measures of
this exposure. Although useful, these measures generally fail to take into
consideration partners’ HIV status and to eludicate the behaviors
themselves; the decisionmaking processes that underlie the choices to
perform these behaviors; the values or perceptions of norms and
expectancies that are associated with these behaviors; and other
environmental, social, and cultural determinants of the behaviors.
The previous sections of this monograph discuss the context and, within
population groups, the values and circumstances that have been found or
are thought to influence the engagement in and performance of these
behaviors. This section presents three chapters that grapple with the
issues related to measurement of these behaviors in the context of the
dyadic or group relationships involved and the social, temporal, cultural,
and environmental circumstances that define their performance.
Brunswick, in her chapter, Bringing the Context in From the Cold:
Substantive, Technical, and Statistical Issues for AIDS Research in the
Second Decade, challenges the reader and her research colleagues with a
multidimensional matrix that she terms an ecological model, representing
several levels of influence on the behaviors of individuals and
populations. Brunswick suggests that the theoretical models that underlie
the HIV prevention strategies currently being assessed depend greatly on
the Health Belief Model. The limitations of such a cognitive model are
that it principally examines behaviors from the individual viewpoint and
does not adequately include the social influences of the family, peers, and
the community on the extent to which individual behaviors are
self-determined and self-controlled.
183
Brunswick goes on to discuss the levels of circumstances that modify
HIV risk behaviors, including the macrosystem, the exosystem, the
microsystem, and the ontogenic system. In the macrosystem, the major
influences are gender, cohort, ethnoracial factors, and historic period.
The importance of these influences creates a need to assess and measure
not only how these factors shape the expression of risk behaviors, but
also how society dictates the roles and patterns of behaviors for people
within these characteristics.
The exosystem includes situational and institutional arrangements that
impact behaviors. The status of knowledge regarding HIV infection and
its prevention as well as the availability of testing, counseling, and
treatment services obviously impact risk perceptions and behaviors.
Simply knowing the prevalence of HIV infection and AIDS diagnoses
within a geographic area is an important determinant of risk behaviors in
some instances.
Roles and expectancies within interpersonal settings such as social
networks, the family, peer groups, and the neighborhood
(the microsystem) impact the performance of risk behaviors. Brunswick
specifically addresses the need to measure “perceived” behaviors as well
as “real” behaviors, using self-report, collaborative information, and
direct observations. The last system, the ontogenic, refers to
characteristics within the individual: attitudes, beliefs, and self-image.
It is apparent that there are many contexts that influence behaviors. In
Brunswick’s chapter, as in the following chapter by Koester, the
importance of determining what factors are most salient to the behaviors
under question is underscored. The need to integrate both qualitative and
quantitative measures of these factors also is emphasized in Brunswick’s
chapter. She discusses the technical issues associated with studies of risk
behaviors, specifically those related to sampling design, instrument
development, and analysis. She concludes with an emphasis on the
phenomenon of serendipity and on the impact of the data collection
process itself on individual behaviors.
The next chapter, The Context of Risk: Ethnographic Contributions to
the Study of Drug Use and HIV, is a natural extension of the framework
established by Brunswick. In this chapter, Koester discusses the many
contributions made by ethnographic research to the understanding of HIV
risk behaviors, particularly those relative to drug use. He points out that
the focus of ethnography is to understand “how people organize and
184
make sense of their world.” Koester discusses the two ethnographic
approaches to data collection, participant observation and open-ended
interviews, and how these approaches are combined over time to gain
insights and a fuller understanding of a particular group of people.
However, he observes that the traditional ethnographic approaches have
been modified within the context of HIV research to solve a narrow
research question.
Specifically, Koester discusses the contributions of ethnographic research
on the meaning of risk and on the micro and macro contexts in which risk
behaviors take place. In addition to the review of the qualitative
methodology for the collection of information on HIV risk behaviors, he
points out the importance of the phrasing and context of the questions
asked by the researcher in a variety of settings and the relationship of the
responses received to time, situation, and role of the respondent.
In conclusion, Koester relates the importance of having a
multidimensional measure of risk behaviors in order to gain sufficient
information to intervene and alter or modify these behaviors through
planned community, environmental, and individual programs.
Finally, Gibson and Young in their chapter, Assessing the Reliability and
Validity of Self-Reported Risk Behavior, examine the issues of the
meaning of data collected through quantitative approaches, specifically
structured questionnaires. This chapter reviews the authors and others’
research that examines the extent to which drug abusers report certain
behaviors reliably and the degrees to which self-report instruments obtain
valid responses. Three aspects of these issues are discussed: (1) the
truthfulness of the information on HIV risk practices provided by
intravenous drug users; (2) the extent to which the tendency to provide
socially desirable responses “contaminate” self-report(s); and (3) methods
researchers can use to reduce response bias when collecting information
on sensitive behaviors.
Gibson and Young present the analyses of their studies with intravenous
drug users, which show that questions regarding drug use-particularly
sharing of injection equipment-were threatening to the respondents, and
questions regarding oral and anal sex were sensitive areas of inquiry.
They tested different data collection techniques to determine which were
best for these two behavioral areas. They found that response bias was
reduced by utilizing a questionnaire format that used long,
185
permission-giving introductions to each set of questions and by asking
sensitive questions later in the interview.
It is clear that these chapters do not cover all the methodological issues
associated with using multidimensional approaches to collecting
information to better understand HIV risk-taking behaviors. One obvious
omission is analytic approaches that can address the needed temporal and
multilevel analyses. However, conceptualization of substantive issues
must of necessity precede a focus on analytic techniques. Until this
technical review, substantive contextual issues had not been addressed by
such a diverse group. This monograph should serve as a stimulus to
researchers focusing on HIV transmission to begin to broaden their
conceptualization of their measures and to be challenged not only to
include the many dimensions suggested in these chapters, but also to
include researcher colleagues who represent disciplines other than their
own to work with them on studying this very serious public health
problem. It is only through a clear understanding of these
decisionmaking processes that the scientific community will be able to
prevent the spread of HIV and other infectious diseases.
AUTHOR
Zili Sloboda, Sc.D.
Acting Director
Division of Epidemiology and Prevention Research
National Institute on Drug Abuse
Parklawn Building, Room 10A-38
5600 Fishers Lane
Rockville, MD 20857
186
Bringing the Context in From the
Cold: Substantive, Technical, and
Statistical Issues for AIDS
Research in the Second Decade
Ann F. Brunswick
This chapter discusses an ecological model of health that has guided the
author’s research for 25 years. Its articulation is derived from
Bronfenbrenner’s (1979) paradigm of behavioral influences. The idea of
applying a behavioral model to guide conceptualizations about an illness
follows from a cross-fertilization of social psychology and health. This
model is particularly apt for studying illnesses incurred through human
immunodeficiency virus (HIV) and acquired immunodeficiency
syndrome (AIDS), which undeniably have strong behavioral linkages.
The behavioral underpinning refers not only to understanding that
contagion comes about through specific behaviors, but also to the
behavioral responses HIV exacts at multiple social levels and the crosscurrents of social response it has evoked at different social levels.
Following from table 1, which summarizes the model discussed here,
these include: moral (macrosystem), legal and institutional (exosystem),
interpersonal (microsystem), and individual knowledge, belief, and
attitudes (ontogenic or intrapersonal system) (Brunswick 1985).
The ecological model is a reminder of what needs to be considered when
attempting to bring the context in from the cold. It bridges the spheres of
interest of different social science disciplines: political science,
anthropology, social psychology, and sociology. The model provides a
heuristic device for conceptualizing the nature of the epidemic(s) and
identifying how HIV infection is experienced differently by low-income
people of color, as compared to the experience of the mostly middle-class
gay community on whom earlier scientific findings were based.
This basic ecological paradigm serves as a guide to variables to be
included in HIV/AIDS studies and identifies some measurement pitfalls
to avoid in the context of injection drug users (IDUs). The need for a
contextual approach in HIV-related research, policies, and programs is
underscored by the fact that HIV-related disease is not randomly
distributed, either geographically or socially, nor does symptom
187
conversion and perhaps even seroconversion occur pari passu among all
who are exposed (Brunswick et al., forthcoming).
It is easier to give lip service to the social nature of the disease and to the
need for incorporating different social levels into its analysis than to do
so. The most frequently tested research model in AIDS prevention
research is the health belief model (HBM). Yet the HBM is of
questionable validity when applied to IDU and minority populations
because of its predominantly cognitive dimensions of cost-benefit and
vulnerability appraisals, behavioral and outcome efficacy, and because
the model operates primarily if not solely in the ontogenic (within the
individual) sphere (Brunswick and Banaszak-Holl, unpublished data). A
major architect of the HBM movement recently has critiqued the current
health promotion movement in words that are equally applicable to the
neglect of the social context in HIV research, policy, and programs:
Finally, I turn to what, in my estimation, is the most disturbing aspect
of the contemporary health promotion movement: its tendency to
locate the responsibility for the cause and the cure of health problems
in the individual. Three assumptions appear to underlie this
approach: personal health-related behaviors are discrete and
independently modifiable; anyone can decide to alter his/her behavior
and then go on to do so successfully; and everyone has a personal
responsibility to ‘live well’ through self-discipline and behavior
modification.... These assumptions do not fit very well with what we
know about the major determinants of health and the prevention of
illness.
I would argue, first, that health habits are acquired within social
groups (i.e., family, peers, the subculture).... The lifestyle approach
enables us to ignore the more difficult, but at least equally important,
problem of the social environment which both creates some lifestyles
and inhibits the initiation and/or maintenance of others (Becker 1993,
p. 3-4).
The three areas identified in this chapter’s title are discussed in the
remainder of the chapter: substantive, technical, and statistical issues that
might be helpful in rethinking existing research models and methods for
studying HIV. They are guided by the ecological model and concern for
situational and environmental factors that have been articulated above.
188
SUBSTANTIVE ISSUES
Following the ecological model, structural or macrosystem attributes are
considered first, including gender, cohort, period, and ethnoracial issues
that are closely tied to social stratification. Analysis of these factors is
expected to capture differences, not only in types of risk behaviors and
risk exposures, but also in the societally imposed roles and patterns of
behavior that apply to minority versus white groups and women versus
men within those groups. Macrosystem factors would take into account
not only the patterns and norms that prevail generally in society, but also
those within specified subcultures. Examples of some of these issues are:
Do women IDUs share works with their sex partners relatively more
often than do men? Does this sharing occur in Hispanic subgroups
(which ones) or among African Americans? Do men universally go first
when sharing occurs? Cohort effects also need to be considered at this
level (e.g., broad changes over time in norms governing standards for
drug or sexual behavior and for sex roles). Cohorts born at varying times
experience these changing norms differently. In addition, the
macrosystem calls into question singular historic changes that impact all
cohorts simultaneously (although not necessarily all social strata), such as
economic and industrial “revolutions” and economic recession and
depression.
Secular time also enters at the next level of analysis—the exosystem
(table 1). Policy and related factors that influence institutional allocation
of resources are subsumed here. These reflect situational variations in
time and place. Particularly in HIV/AIDS research, time cannot be
ignored as a situational variable: the time when data are collected,
analyzed, and reported must be given primary consideration because of
the rapid changes in primary transmission vectors, host population,
biotechnological and pharmaceutical developments, and even in the agent
itself (e.g., HIV-2 now is reported in West Africa and among West
African immigrants) (Onorato et al. 1993). For HIV/AIDS research, it is
of interest, for example, to note what effect an announcement of infection
on the part of a public figure such as Magic Johnson has upon public
perceptions and awareness. It is less clear whether changes at this level
are translated into changes in risk and risk avoidance behavior
(Kalichman and Hunter 1992).
Some have analyzed changes in rates of infection, latency periods, and
the nature of symptoms as the epidemic progresses over time.
Additionally, data need to be specified by place—region and specific
189
TABLE 1. Ecological paradigm of contextual influences on behavior
Moral
Legal and Institutional
Interpersonal
Intrapersonal
MACROSYSTEM
EXOSYSTEM
MICROSYSTEM
ONTOGENIC SYSTEM
(within person
Values, Norms and Social
Expectancies Governing
Behavior, Defined by:
Situational and Institutional
Arrangements Spatial and
Temporal Factors:
Interpersonal Settings, Social
Networks and Interpersonal
Relationships of All Types and
Sizes:
Biopsychosocial Attributes and
Life Histories:
General economic and political
climate
Gender
Ethnicity/race
Birth cohort
Timing of specific events:
Isolation of virus
Pharmacological
advances (AZT, etc.)
Availability of and access to
sterile injection equipment
Treatment availability
Neighborhood violence
Familial
(Spousal, parental, children)
Peer
Attitudes
Beliefs
Self-image
Health
SOURCE:
Adapted from Brunswick 1985
Educational
Neighborhood groups and
organizations
Networks within and outside own
neighborhood
locale. More discussion of geographic variations in HIV/AIDS
epidemiology appears in this chapter under the section on technical
issues. For now, attention is called to such place-related concerns as
availability of injection drugs and of free sterilized needles and proximity
to crack houses, abandoned housing, and red light districts when
considering the role of exosystem variables in HIV/AIDS risk-related
behavior.
The next level addressed in the ecological model is the microsystem.
This level subsumes the life settings in which the individual interacts with
others. Research on peer networks and on sibling sets falls into this
domain (Brunswick et al., forthcoming). The difference between
perceived and so-called real measures at this level—between what a
respondent reports about network characteristics and behavior (perceived)
and actual observations or interviews with the network (real)—is a
particularly cogent issue. Both of these are measuring real (though
distinct) domains and not substitutes for one another. Ideally, both
should be modeled.
The fourth level in table 1, that of ontogenic factors, requires little
discussion because most of the HBM research has been focused on these
factors. Allusion was made earlier to the shortcomings of the HBM
because it ignores the social settings of the individual’s life and, further, it
imbues that life with a regularity and coherence that ofttimes are not
present. As an example, when the HBM was tested with cross-sectional
data from the Longitudinal Harlem Health Study, those individuals who
reported the most concern about HIV/AIDS were the ones who engaged
in riskier behaviors. While seemingly logical, this finding turns the HBM
on its head, since the model suggests that the most concerned individuals
will undertake risk avoidance practices.
Regardless of what other particular concerns arise from the ecological
model, one closing note must be sounded of a more general nature:
always leave room for serendipity. Though guided by theory, researchers
need to be open to the unexpected correlation-and to report it. Too
often, research is conducted either without any conceptual-theoretical
orientation or with too much of it, resulting in the neglect or oversight of
unexpected patterns that might appear in the data. While research needs
to be focused, it should be done with sufficient latitude for the
unexpected to come into view.
191
TECHNICAL ISSUES
Technical issues include those factors that affect the validity
(predictiveness) and reliability (consistency/generalizability) of research
findings. Quite obviously these issues have statistical as well as
procedural implications, and in some cases their classification here is
moot. Placing what is measured and how it is measured in context
requires that new methods be improvised or old methods be adapted or
readapted, always with a mind to preserving (not merely measuring)
reliability and validity.
The need to study HIV in the social contexts within which individuals are
embedded requires a wedding or merger between ethnographic and
quantitative statistical methods. In survey research in the “good old
days,” this combined approach was pro forma during the exploratory or
pilot stage in developing the larger survey. Site observation and
unstructured interviews were conducted to determine appropriate content,
wording, and sequence of the final survey items. A beneficial byproduct
of integrating these types of procedures in the early stages of research
was the investigator’s increased involvement in, and knowledge of, data
collection. Recent trends in computerization and specialization have led
to increased separation of data analysis from knowledge of how the data
were collected and from the raw data. The author speculates that this
separation has led at times to acceptance of erroneous results and
misinterpretation of findings-severely limiting both their validity and
reliability, Cronbach’s alpha notwithstanding (Nunnally 1978).
An example of a simple technique that demonstrates the ethnographicquantitative merger is use of open-ended questions in developing
quantitative surveys. With such questions, categories are built up from
the responses. These categories can then be quantified. A case in point
from the author’s research concerns data collected beginning in 1975 in
reply to the simple question, “How have things been going for you?”
(very well/pretty well/so-so, etc.), followed by “How do you mean?”
Replies led to the development of a catastrophe scale, akin to what, in the
literature now, is usually referred to as a life events scale. Since then, the
author has used the scale in closed-end form as a measure of stress
(Brunswick et al. 1992). In the fifth round of study, now in the field, the
open-ended followup is being used again to identify new events and
circumstances wrought over time in the lives of an inner-city AfricanAmerican cohort. The categories of reply comprise a population
192
subgroup relevant measure, a topic discussed later in this chapter as
important in studying contextual effects.
Laumann and colleagues (1993) have applied a social network and
statistical strategy to investigate the undercount in HIV/AIDS prevalence
estimates. Their strategy is an application of multiplicity sampling, and it
is essentially useful as a proxy method for counting cases of a
phenomenon too rare to measure by using normal area probability
sampling procedures. This technique takes into account the respondent’s
probability of selection into the sample and the number and
characteristics of infected individuals identified in the network.
Demographic characteristics of network cases from the multiplicity
sample then are compared to those in official counts to estimate the
location(s) and rate(s) of undercounting. Such a method, combined with
what is known about geographic concentration of HIV/AIDS cases, might
provide important insights not only into changing rates and undercounts
but networks of infection diffusion as well. It is an example of how the
requirements for valid and reliable HIV/AIDS reports might lead to
improvisation of existing methods while preserving rigorous standards
for reliability and validity in measurement.
In that same article, Laumann and colleagues (1993) identified sources of
bias or error in AIDS case reporting and specifically cited those factors
they inferred as biasing against report of middle-class white infection at
the various reporting levels. They cite an estimated undercount of up to
30 percent in 1989 across the three levels of reporting that comprise the
Centers for Disease Control and Prevention’s (CDC’s) case counts
(Laumann et al. 1993).
Record bias of a different kind appeared in the author’s study and
concerned the failure to report AIDS infection on African-American
females’ death certificates in the mid-1980s. Deaths from pneumonia
were reported for three female IDUs, and a fourth death was reported
from acute renal failure with notations about narcotics use. The women
were not tested for HIV/AIDS from 1985 to 1987; consequently, none of
these deaths was ascribed to HIV infection or AIDS. Of course, failure to
report an AIDS/HIV diagnosis on a death certificate also occurred before
CDC’s modifications of the AIDS case definition (Centers for Disease
Control 1987, 1992).
The issues of question wording, question order, question context, and
question format are discussed next, and all have repercussions for data
193
reliability. It seems obvious that questions need to be clear to the
respondent, whether structured or nonstructured interview methods are
used. For example, one woman self-identified as bisexual because “it’s
only me and my husband.” Words should be selected that are familiar to
the population(s) under study. Involving representatives of the study
group during instrument preparation is one helpful approach. Pilot
testing and pretesting are essential. In HIV/AIDS studies using
questions—whether written or oral, structured or unstructured—
deliberate steps need to be taken to vary the positive-negative direction of
questions and prevent respondents’ inference of desirable responses.
Risk behaviors, especially when evaluating before-after interventions, are
quite transparent with respect to what is good and bad. The same
information needs to be sought in different ways and in contexts that are
not conducive to bias. A case in point occurred in the author’s research,
when an individual gave an affirmative reply that he or she had increased
condom use since hearing about AIDS (Brunswick, unpublished data).
Earlier in the interview, however, the same respondent reported no
current condom use. The author has little faith in direct questions about
changing specified actions because of AIDS as a valid measure of actual
behavior, although responses might be useful as measures of changing
awareness of risk behaviors.
Wording questions to elicit intentions to perform certain actions, behavior
preferences, and actual behavior practices are distinctions demanding
particular attention. Each of these questions reflects a different reality,
just as the perceived environment (e.g., what is reported by the individual
about another’s behavior) differs from objective reality. The distinctions
can be informative, but these questions cannot be used as proxies for one
another (e.g., Would you want to know if you were infected? Would you
want your partner to know? Would you tell him/her? Did you tell
him/her?).
The order or context in which questions are presented also influences
responses. Colasanto and colleagues (1992) elicited a significant
difference (9 percent) in agreement responses by varying the order of
survey questions regarding transmissibility of AIDS through blood
donation. The difference hinged on whether a question about AIDS and
blood donation preceded or followed another question about whether
AIDS could be transmitted through blood transfusion. The sample who
was asked about donation first had the higher proportion who accepted
the erroneous mode of transmission. The authors believe that asking first
194
about transfusion clarified the meaning of and the distinction with the
subsequent question on blood donation.
Concern for the social context implies a recognition of the variability of
human behavior. Some things are done with some people at some times.
To obtain complete and accurate accounts, time (e.g., last time, first
time), type of relationship, and type of activity (when, what, with whom)
need to be specified in inquiries into risk behaviors and risk exposures.
Many are guilty of framing questions to ask about what the respondent
usually does, which elicits answers that depend on the varying ability of
respondents to abstract and generalize about behavior.
The serendipity or new knowledge principle discussed earlier as a
substantive issue has technical implications as well. Observations
(whether verbal or visual, structured or unstructured) should be broad
enough to capture unanticipated responses. Some examples have been
given above.
Changing temporal and spatial (geographic) dimensions of HIV/AIDS in
the second decade exact technical considerations for data collection and
design, as well as substantive considerations. How frequently must new
data be collected to capture changing incidence and prevalence, vectors
of infection, and rates of infection? Spatial and geographic concentration
of infection and the requisite consideration of contextual factors require
integration of data from different sources which are often recorded in
different units of observation.
By way of example, in attempting a test of neighborhood influence on
HIV rates in the longitudinal African-American study, the area HIV rates
that were available were in broadly defined neighborhoods, which
included up to eight zip codes in a single neighborhood, and between
4 and 45 census tracts within one zip code. Special permission and
requests were required in order to obtain infection rates at the zip code
level.
The contextual parameters that were sought included crime statistics
(which were available at the police precinct level) and socioeconomic
characteristics (provided at the census tract level). Zip codes appeared to
be the most useful level for uniform measurement, since that was the
smallest geographic unit for reporting AIDS cases. However, these areas
do not embrace census tracts in an orderly fashion. Even though they are
smaller, census tracts are frequently split between two zip codes and
195
sometimes as many as three zip codes. Collecting geographic area
information, therefore, involves different units of analysis that have to be
disaggregated or reaggregated into uniform measures.
Another consideration is that appropriate boundaries of HIV risk areas are
yet to be determined. High-rate cities have been identified but, as the
National Research Council (NRC) (1993) report emphasized, cases are
concentrated in certain areas within those cities. These areas coincide
with concentrations of gays or impoverished minority groups. A future
study may determine whether these concentrated areas conform better to
census tract boundaries, zip code boundaries, health areas, or none of the
above. Investigating the natural geographic boundaries of infection
might yield information about the spread of disease as well as the role of
spatial factors (e.g., crack houses, sex trade) in that spread (National
Research Council 1993; Wallace 1988). The author’s study compared
proportions of infected and noninfected individuals in New York City
neighborhoods identified as high infection centers. The study found that
equivalent proportions of infected and uninfected individuals resided in
those neighborhoods.
Perhaps the preeminent technical challenge facing researchers is the need
for ethnic, gender, and cultural sensitivity and specificity in research
methods. Ethnogender response differences arise from the differential
norms governing the same behaviors that are of particular interest and
relevance in assessing HIV/AIDS risk. These differences are more
readily understood in reference to the ecological model that enlightens a
comparison of HIV/AIDS infection among gays and minorities and
between different minorities as well (Diaz et al. 1993).
Gender differences are important both within and among cultural groups.
One example from the author’s experience concerns the honesty that
African-American men and women attributed to replies regarding sexual
behavior (Brunswick, unpublished data). The differences were split not
only by gender but, as further evidence of differential gender norms, by
HIV infection status as well. When asked how honestly they thought
others in the survey had answered questions about sexual behavior,
infected men were more likely to answer “completely honestly”
(31 percent) compared to uninfected men (12 percent) and infected
women, none of whom attributed complete honesty to others. Uninfected
women (20 percent) were more likely than infected women and
uninfected men to attribute complete honesty to others in reporting sexual
experiences. The question was intended as an indirect measure of
196
reliability of self-report, but instead was a further example of perceived
reality being a phenomenon in its own right.
Assumptions about attitudes toward transmission risk behaviors, toward
the disease itself, and toward medical care for HIV infections need to be
subjected to empirical testing, especially when those assumptions may
impede accumulation of knowledge. As most researchers have learned,
often the reticence is with the questioner, not the respondent. Varying
sensitivities among ethnogender and cultural groups can only be
established empirically.
What researchers are trying to study and the people being studied are
criteria for selecting the particular data collection mode. In the case of
perceived or self-reported data, choices need to be made between written
and oral responses and between self-administered protocols and those
administered by an interviewer or other trained research worker. If oral
responses are solicited, a choice must be made between in-person and
telephone interviews. The reliability and validity of the selected mode
may vary with the ethnic and gender identity of the study population.
In a study focused specifically on interracial differences in reliability of
drug use reports, Aquilino and LoSciuto (1990) found that distortions in
alcohol and drug use reports on telephone surveys as compared to
personal interviews were greater for African-American respondents than
for whites. This finding was independent of the sample bias that resulted
from the exclusion of households without telephones (sample coverage)
and the consequent demographic bias in the obtained sample
(the telephone sample disproportionately included upscale African
Americans). Distribution on educational attainment is a good indicator of
an African-American adult sample’s reliability or representativeness. If
the findings of a community-drawn sample deviate markedly from
20 percent who fail to complete high school, 60 percent who are high
school graduates, and 20 percent who have gone beyond high school, the
author would infer that selection factors have been at work (Brunswick et
al. 1993).
Note the distinction between the three types of bias that may be incurred
according to the selected study mode: sample exclusion (e.g., exclusion
of those with no telephone or even not household-based); biased
representation of those included; and response bias—the quality of
response obtained from in-person interview settings versus those
conducted by telephone or self-administered instruments.
197
Resources and effort are required to surmount potential biases that are
linked to study mode. In the author’s longitudinal study of African
Americans, HIV risk behaviors differentially impacted on men’s and
women’s accessibility to interview. Women who reported multiple HIV
risk behaviors required more attempts to interview. The correlation was
less predictable for men; some with multiple risk factors could be
accessed readily, while others required so many attempts to contact that
interviewers did not keep accurate records of the number of attempted
contacts (Brunswick 1991).
The difficulty in predicting AIDS risk behavior among minorities and
drug users based upon their attitudes and knowledge (Brunswick and
Banaszak-Holl, unpublished data; Longshore et al. 1992) is not so much
one of mode of study as of insufficiently sensitive measurement tools.
The difficulty reflects the need to develop sensitive and contextually
relevant subpopulation measures and to improve the reliability of those
measurements. Longshore and coworkers (1992) noted that additional
difficulties arise in interpreting findings when the measuring instruments
are found to have differential reliability between ethnic or gender
population subgroups. These differences can lead to erroneous
interpretation of data and erroneous conclusions unless transformations or
other corrective steps are undertaken to remove the measurement
heterogeneity.
STATISTICAL ISSUES
Statistical issues, for the most part, have been addressed in the discussion
of technical issues, especially those concerning sampling bias and
response bias. If modifications of conventional random or probability
sampling methods occur, standard error calculations also require
modification. Thus, as the disciplinary representation of a research team
is broadened to reflect a multidimensional inquiry, the inclusion of a
sampling statistician is crucial to ensure that study methods and statistical
assumptions are congruent.
CONCLUSION
In conclusion, the context to be brought in from the cold in HIV/AIDS
research must include broadened health, psychosocial, and social
contexts. A model for doing this has been described in this chapter,
198
along with discussion of some technical and statistical difficulties in
conducting research (and by extension, intervention programs) that is
more sensitive and specific to people in their settings. Like the ecological
model, other research models need to become multidimensional and to
incorporate indicators and measures across different strata of social
complexity.
The issue of serendipity referred to above has application to contextual
effects of the study mode itself. The interview process itself may (and
does) have an effect. When both groups have had before-and-after
interviews, this serendipitous effect could work to restrict the differences
that are observed between experimental and control groups in
intervention studies and between alternative intervention strategies.
Whether or not this effect is a variation of the Hawthorne effect
(Roethlisberger and Dickson 1939) of test-retest, it does suggest the need
to modify the study design. The study could be modified by adding
another group to be interviewed only at followup (carefully matched to
the before-and-after experimental and control groups) or by statistically
matching the results for the control and experimental groups with data
obtained in the same period from equivalent populations outside of the
study. These supplementary data could be used to estimate how much of
an intervention effect being interviewed and being part of a study group
comprises in and of itself.
REFERENCES
Aquilino, W.S., and LoSciuto, L.A. Effects of interview mode on selfreported drug use. Pub Opin Quart 54:362-395, 1990.
Becker, M.H. A medical sociologist looks at health promotion. J Health
Soc Behav 34:1-6, 1993.
Bronfenbrenner, U. The Ecology of Human Development. Cambridge,
MA: Harvard University Press, 1979.
Brunswick, A.F. Impairment, disability and/or handicap: An ecological
paradigm. J Adolesc Health Care 6(2):141-151, 1985.
Brunswick, A.F. Health and substance use behavior: The Longitudinal
Harlem Health Study. J Addict Dis 11:119-137, 1991.
Brunswick, A.F.; Aidala, A.; Dobkin, J.; Howard, J.; Titus, S.P.; and
Banaszak-Holl, J. HIV-1 seroprevalence and risk behaviors in an
African American community cohort. Am J Public Health 83:13901394, 1993.
199
Brunswick, A.F.; Lewis, C.S.; and Messeri, P.A. Drug use and stress:
Testing a coping model in an urban African-American sample.
J Comm Psychol 20:148-162, 1992.
Brunswick, A.F.; Messeri, P.A.; Dobkin, J.; Flood, M.; and Yang, A.
Sibling homophily in HIV infection: Biopsychosocial linkages in an
urban African American sample. In: Needle, R.H.; Genser, S.; and
Coyle, S.L. Social Networks, Drug Abuse, and HIV Transmission.
National Institute on Drug Abuse Research Monograph. Washington,
DC: Supt. of Docs., U.S. Govt. Print. Off., forthcoming.
Buehler, J.W.; Berkelman, R.; and Devine, O. “Estimate of HIV-related
Deaths in Young Adult Men, United States, 1988.” Abstract presented
at the Fifth International Conference on AIDS, Montreal, June 1989.
Centers for Disease Control and Prevention. Revision of the CDC
surveillance case definition for acquired immunodeficiency syndrome.
MMWR 36:1-15S, 1987.
Centers for Disease Control and Prevention. 1993 revised classification
system for HIV infection and expanded surveillance case definition for
AIDS among adolescents and adults. MMWR 41:RR-17, 1993.
Colasanto, D.; Singer, E.; and Rogers, T.F. Context effects on responses
to questions about AIDS. Pub Opin Quart 56:515-518, 1992.
Diaz, T.; Buehler, J.W.; Castro, K.G.; and Ward, J.W. AIDS trends
among Hispanics in the United States. Am J Public Health 83:504509, 1993.
Kalichman, S.C., and Hunter, T.L. The disclosure of celebrity HIV
infection: Its effects on public attitudes. Am J Public Health 82:13741376, 1992.
Laumann, E.O.; Gagnon, J.H.; Michaels, S.; Michael, R.T.; and Schumm,
L.P. Monitoring AIDS and other rare population events: A network
approach. J Health Soc Behav 34:7-22, 1993.
Longshore, D.; Hsieh, S.; and Anglin, M.D. AIDS knowledge and
attitudes among injection drug users: The issue of reliability. AIDS
Educ Prev 4(1)29-40, 1992.
National Research Council. The Social Impact of AIDS in the United
States. Washington, DC: National Academy Press, 1993.
Nunnally, J. Psychometric Theory, 2d ed. New York: McGraw-Hill,
1978.
Onorato, I.M.; O’Brien, T.R.; Schable, C.A.; Spruill, C.; and Holmberg,
S.D. Sentinel surveillance for HIV-2 infection in high-risk US
populations. Am J Public Health 83:515-519, 1993.
Roethlisberger, F.J., and Dickson, W.J. Management and the Worker.
Cambridge, MA: Harvard Press, 1939.
200
Wallace, R. A synergism of plagues: “Planned shrinkage,” contagious
housing destruction and AIDS in the Bronx. Environ Res 47:1-33,
1988.
AUTHOR
Ann F. Brunswick, Ph.D.
Senior Research Scientist
Columbia University
Public Health/Sociomedical Sciences
60 Haven Avenue, Box 394
New York, NY 10032
201
The Context of Risk:
Ethnographic Contributions to the
Study of Drug Use and HIV
Stephen K. Koester
INTRODUCTION
Ethnographic research has made significant contributions to the
understanding of how and why injection drug use can facilitate human
immunodeficiency virus (HIV) transmission. Ethnographic studies have
described drug users and their behavior and informed the design of
interventions that are culturally relevant and capable of reaching these
individuals (Wiebel 1988). In this chapter, the author briefly summarizes
ethnographic methods and illustrates how this research approach is
especially useful for providing research depth. In particular, the chapter
focuses on the contribution of ethnographic research to detailing the
meanings people attribute to their behavior and to identifying and
describing the contexts in which behavior is embedded.
Understanding how other people organize and make sense of their world
is the primary objective of ethnographic research. To accomplish this
objective, an ethnographer studies a group by experiencing its members
lives firsthand—interacting with them on their terms and in their
environment. This in-depth research method is called participant
observation and is at the heart of anthropological fieldwork. The
assumption behind participant observation is that much can be learned by
listening and observing a group’s members in their natural setting.
Conducting participant observation with a group over time enables
researchers to gain an insider perspective, enabling them to comprehend
both the cognitive and material reality of others (Agar 1980).
Ethnographers augment their ongoing observations with a variety of
interviewing techniques. These techniques are designed to solicit
information about ways of life and patterns of behavior that may appear
fundamentally different than their own. Interviewing techniques extend
from completely unstructured discussions to survey instruments.
However, since ethnography is most useful for exploring new areas of
inquiry and for providing depth rather than breadth, most ethnographic
202
interviews are designed to promote discussion. The subjects are
encouraged to “tell their story,” to describe their world and explain their
behavior. As a result, ethnographic interviews employ open, contrasted
to closed, questions. Combining participant observation with
ethnographic interviews and conducting both over an extended period of
time enable the ethnographer to identify and describe patterns of behavior
as well as the factors influencing behavior.
Two recent examples of in-depth ethnographic research pertinent to this
discussion are Eli Anderson’s book “Streetwise” (1990) and Philippe
Bourgois’ forthcoming study of a neighborhood crack scene in Spanish
Harlem (Bourgois, unpublished observations). Anderson details
changing economic and social relations in a Philadelphia community, and
Bourgois describes how local residents resist the limited economic
options open to them in the formal economy by participating in the
underground crack trade, a process of resistance that culminates in a
culture of terror. Both of these researchers lived with the people they
studied and participated in their daily lives.
In many cases, current ethnographic research on HIV and drug use does
not achieve the degree of researcher involvement illustrated by these two
studies. Much drug and acquired immunodeficiency syndrome
(AIDS)-related research is “applied,” meaning that it is oriented toward
solving or contributing to the solution of a particular problem rather than
providing a comprehensive study of a particular group or culture.
Nonetheless, applied researchers share the same perspective and methods.
Applying the ethnographic perspective and its accompanying research
methods to the study of high-risk activity among drug users has resulted
in a much more intimate and holistic understanding of drug users and
their behavior. It has uncovered the meanings drug users attach to their
behavior and has led to a greater understanding of the complex set of
factors that influence their lives.
THE MEANING OF RISK
Even when social groups are part of a larger social and cultural system,
their specific experiences and circumstances may result in differences in
meaning or perspective. It is essential to identify and comprehend such
differences to understand drug users and the risks they take. For
example, needle cleaning is a term public health researchers would
assume refers to a hygienic procedure for disinfecting a needle. This
203
same meaning may be extended to drug injectors who typically perform
needle cleaning immediately after injection. Researchers then might
assume that injecting drug users (IDUs) already are conscious about
needle cleanliness and that encouraging them to take the additional step
of using bleach is a relatively straightforward and effortless behavioral
change. However, for many drug injectors this postinjection rinse is not
conducted to disinfect the syringe; rather, its purpose is to unclog the
blood and drug residue from the needle so that it can be reused.
Although this may seem like a minor point, it indicates how even
behavior that appears obvious may in fact have very different meanings
for those engaged in it. Ethnographic research is particularly well-suited
for identifying and describing these kinds of cognitive differences.
In another illustration, the author found interpretations of what constitutes
sexual behavior may be, in part, dependent upon one’s sexual orientation,
the purpose for which a sexual act is performed, and the individual with
whom the act occurs. An interviewer on Denver’s National Institute on
Drug Abuse (NIDA)-funded cooperative agreement reported
inconsistencies among some subjects regarding their answers to a risk
assessment instrument’s questions on sexual behavior. He noticed that
gay women who exchange sex for money often revealed that they regard
sex as a pleasurable, intimate act they do with their female lover, while
the oral sex and heterosexual intercourse they engage in with male
customers is perceived as work (Anderson, personal communication,
January 1993). In fact, ethnographers in several cities have reported that
among both gay and heterosexual male and female subjects, oral sex and
anal sex are frequently viewed as something other than “real” sex
(Herdt 1992). Obviously, these differences in interpretation have
important implications for research aimed at measuring these high-risk
behaviors and intervention projects aimed at reducing them.
Perhaps the most notable example of the disparity in meanings among
drug injectors and public health researchers is over the term “syringe
sharing.” Throughout the 1980s HIV transmission among drug injectors
was directly linked to syringe sharing. From a biomedical perspective,
the term “sharing” succinctly captured the act that facilitates the
transmission of HIV among IDUs. Efforts to stop injectors from using
one another’s syringes were summarized in messages emphasizing “don’t
share your syringe.” As ethnographers began studying high-risk
injection, they began reporting that injectors do not necessarily use this
all-encompassing term to describe episodes in which a single syringe is
used. Many injectors do not consider that they are sharing if they use the
204
syringe first, if they share with their sexual partners, if they jointly
purchase a syringe, if they anonymously use a previously used needle, or
if they clean a jointly used syringe with bleach.
Some behavioral scientists described syringe sharing as a ritualized social
practice, implying that the exchange of a syringe between injectors was a
conscious act of reciprocity. In this interpretation, syringe sharing was
thought to function as a means for injectors to bond and develop trust
(Des Jarlais et al. 1986; Howard and Borges 1972). Observations of
injecting situations and detailed interviews with injectors led several
ethnographers to suggest additional explanations for this behavior. They
found that, in many cases, syringe sharing is motivated by the need to get
high, and has little to do with ritual or cementing social relationships
(Carlson et al. in press; Clatts et al. in press; Fernando 1991; Kane and
Mason 1992; Koester 1992, 1994; Murphy 1987; Page et al. 1990).
Ethnographers have suggested the terms “needle transfer” and “needle
circulation” as more accurately describing the multiple ways injectors
pass syringes among themselves (Carlson 1991; Carlson et al. in press).
Usually, the IDU who gives a syringe to another injector does not want it
back. Needle pooling describes how injectors use common stocks of
used syringes available at shooting galleries as well as what happens
when an injector “stashes” a syringe in a location where other IDUs can
find it (Page et al. 1990).
Studies by Carlson and colleagues (in press) and by Koester (1994)
explained multiple reasons, besides HIV, that injectors have for not
sharing their syringes. Used syringes are more difficult and painful to use
because the point quickly becomes dull, an important consideration for
injectors who experience difficulty locating a vein or who want to
prolong the use of their remaining uncollapsed veins. Also, the most
common syringes now used by injectors are disposable insulin syringes
that become inoperable from repeated use; needles clog and the rubber
plunger begins to lose its pliancy and thus its suction after more than a
few injections.
These illustrations suggest the importance of understanding the meaning
of behavior from the point of view of those who engage in it and confirm
the necessity of identifying and understanding the “complex and broader
constraints” influencing these behaviors (Kane and Mason 1992, p. 201).
Identifying when sex becomes work or when syringe sharing becomes an
anonymous act requires an understanding of the context in which these
behaviors take place.
205
THE CONTEXT OF RISK
Context is an abstract concept used to describe the environment in which
human behavior occurs. It refers to those conditions and circumstances
that affect action and thought. As such, context is manifested in different
dimensions and on many levels of human experience. These conditions
may be personal, social, or part of the physical environment; they may
operate on a local or microlevel; or they may be macrolevel structural
forces. Contextual factors are not static. Factors influencing people’s
lives always are changing; people are constantly adjusting and adapting
as well as resisting the conditions and circumstances affecting them.
By grounding drug users’ beliefs and behavior within their everyday
reality-the conditions and circumstances in which they
live-researchers can resist the tendency to explain their views and
behaviors as evidence of some unique cultural system that can only be
comprehended through its own internal logic. Instead, researchers can
demystify and in a sense “detribalize” drug users, rather than see them as
some exotic but deviant other. They can begin to view much of users’
behavior and beliefs as pragmatic responses to their life circumstances.
CONTEXT AT THE MICROLEVEL
A variety of contextual factors has been identified by ethnographers as
influencing high-risk drug injecting. These include social and situational
factors at the immediate microlevel as well as macrolevel social,
economic, and political forces. Immediate, microlevel contextual factors
include the stage users are at in their drug use, the emotional or
physiological state of the injector at the time of injection, the nature of the
relationships between users, the physical and social setting in which
drugs are used, the kinds of drugs being injected, injectors’ economic
status, and the hustles or occupations they employ. For example, an
injector’s physiological state may affect significantly an injection
episode. It is unlikely that an injector who is in withdrawal, desperate to
get high, or intoxicated will take the precautions necessary to reduce the
possibility of HIV contamination.
The social network or group within which drugs are used and obtained
may influence members’ injection behavior. The relationship between
the members of a network may determine whether they use common
injecting equipment. Injectors who are sexual partners or close friends
206
may regard a refusal to share a syringe as a sign of mistrust. On the other
hand, familiarity might enable drug injectors to be assertive about proper
needle hygiene. An individual’s position relative to the other people with
whom they are engaging in sex or drug use may also determine whether
they engage in risky activity. For example, a woman who is dependent
upon a male injector for access to drugs may be limited in her ability to
negotiate for safer drug use or the use of condoms (Guydish et al. 1991).
Non-drug-using members of a drug user’s personal network also may
influence drug-using behavior. For example, in San Juan, Puerto Rico,
IDUs who live with their families often frequent shooting galleries, where
high-risk drug use is common, in order to keep their drug use hidden
(Ann Finlinson, personal communication, May 2, 1994). This pattern is
common among injectors in Denver as well. However, among some
Denver injectors, an opposite trend is apparent. Some IDUs are able to
use drugs in their own homes because their families support them and
tolerate their drug use.
The specific setting in which drugs are used has been shown to play an
important role in influencing drug-using behavior. Researchers in several
cities have described the relationship between type of injecting location
and the drug-using behavioral complex that occurs there (Carlson et al. in
press; Clatts et al. in press; Des Jarlais and Friedman 1990; Oulette et al.
1991; Watters 1989). The shooting galleries and “get offs” (locally
known places injectors go to use drugs) that they describe vary in size,
organization, and function, and include establishments that service over
100 injectors a day as well as apartments or hotel rooms that are used
only by members of a closed drug-injecting network. Some shooting
galleries have tightly enforced rules governing injecting behavior; others
have no rules at all. Some of these establishments’ customers include
both crack smokers and injectors, some allow sexual barter to occur, and
some facilitate completely anonymous syringe sharing.
In Denver, drug injection occurs in a variety of settings, from relatively
small organized galleries to abandoned buildings, private homes, and the
backseats of automobiles. Each of these locations presents the injector
with certain conditions that may affect drug injection hygiene. For
example, the number of participants and their familiarity will vary with
the setting. Necessary ingredients like clean water may or may not be
present. Privacy for injection may be of particularly brief duration. IDUs
who inject in public areas, behind buildings, or in automobiles must do so
quickly if they are to avoid detection. As a result, the procedure is often
207
hurried, and adequate needle cleaning is unlikely. Such immediate,
microlevel contextual elements often have a causal association with more
encompassing macrolevel social, economic, and political forces that
shape drug injectors’ lives.
CONTEXT AT THE MACROLEVEL
Several researchers have linked drug abuse, escalating violence, and the
overwhelmingly uneven distribution of HIV among inner-city minority
populations to fundamental changes in the American economy. They
contend that there is a causal link between the economic restructuring of
the American economy as characterized by corporate downsizing and
deindustrialization—the shift from manufacturing to information
processing and service—and the social pathology that seems endemic to
America’s inner cities. They argue that these structural changes have
been particularly detrimental to urban poor people and have resulted in
job loss, changing job requirements, increasing social isolation, and the
withdrawal of municipal services from inner-city neighborhoods (Jencks
1988; Massey 1990; Newman 1992; Phillips 1990; Wallace 1993; Wilson
1987). These conditions have become persistent and dominant forces in
the lives of urban poor people, and are manifested in myriad ways that
have direct public health implications.
The importance of considering these macrolevel, structural conditions in
epidemiological studies of HIV transmission is readily apparent when
considering the individuals who are most vulnerable. Increasingly, they
are impoverished people of color. As Page and colleagues suggest in the
introduction to an article about Miami drug injectors, “All of the
described hustles, theft and dating activities operate in a cultural
environment where hopelessness and poverty have been endemic for
decades” (Page et al. 1990, p. 63). For drug users, these conditions not
only contribute to their drug use; they also exacerbate the dangers
involved in using drugs.
Bourgois (1989a) anchored his study of Spanish Harlem and its
expanding underground, crack-based economy within the changing
economy of New York City and described how the shift from a
manufacturing-based economy to a service-based economy has affected
the lives of people who formerly found employment as workers in the
city’s factories. As he explains, “The underground economy beckons
seductively as the ultimate ‘equal opportunity employer.’ The rate of
208
unemployment for Harlem youth is at least twice the citywide rate of
8.1 percent, and the economic incentive to participate in the burgeoning
crack economy is overwhelming” (1989b, p. 63). He describes how those
pursuing careers in the crack economy are no longer exploitable by the
“white man.” They speak with anger of their former legal jobs and the
exploitation they endured, and “... make fun of friends and
acquaintances—many of whom come to buy drugs from them-who are
still employed in factories or in service jobs... All of them have, at one
time or another, held the jobs-delivery boys, supermarket baggers,
hospital orderlies—that are objectively recognized as among the least
desirable jobs in American society. They see the illegal, underground
economy as not only offering superior wages, but also a more dignified
work place” (1989b, p.63).
Similarly, Clatts and colleagues (in press), in their ethnographic studies of
New York’s Lower Eastside, describe how poverty and loss of lowincome housing units to conversion, urban renewal, and reductions in
federally subsidized housing contribute to homelessness and how
homelessness leads to poor health. For drug injectors, homelessness
undermines the degree of risk reduction that can be accomplished and
sustained. These injectors are without reliable places to use drugs or keep
drug-using paraphernalia.
In an ethnographic study of crack-using women in New York City,
Maher and Curtis (1992) contend that “...the position of women crack
smokers can only be understood by locating their lives, their illicit drug
use and their income generating activities within the context of a specific
set of localized socio-economic and cultural developments” (p. 221).
Specifically, they contend that these women’s crack use, criminality, and
experiences with violence only can be understood within the context of
gender relations and their opportunities in both the formal and informal
economies. Similar findings were noted by several of the ethnographers
in the NIDA-funded eight-city ethnographic study examining the
phenomenon of sex-for-crack exchanges (Ratner 1993). Several of these
studies linked the powerlessness that characterized these women’s lives,
as well as the lives of their addicted male counterparts, to comparable
socioeconomic and cultural factors. A male crack user in Denver,
commenting on why he often asked crack-addicted women to engage in
209
degrading sexual acts, poignantly made the connection between this
behavior and his position within the larger society:
Power. I think it’s this feeling that you can dominate somebody.
You know, cause here you know they beg for it. It just turns some
guys on. Like, gimme that and I’ll do this for you. Most guys if
they’ve got cocaine they think they have power cause cocaine bring
them money. It make you think you have power. I don’t know, that
you can dominate. Yeah I don’t know, I like power. I’m just dying
sort of like. But then every once in awhile I like to feel like I’m
important and all that, like I got something you know (Koester and
Schwartz 1993, p. 197).
Ethnographers also have studied the role the criminal justice system and
law enforcement play in the lives of drug users. Mason described the
drug subculture she studied in Baltimore as “being constituted and
reconstituted in opposition to law enforcement and criminal justice
representatives...” (1989, p. 8). Several studies have demonstrated how
drug users’ responses to legal constraints on syringe access and the
strategies they use to avoid interactions with the police may lead to highrisk injection episodes (Clatts et al. in press; Carlson et al. in press;
Feldman and Biernacki 1988; Fernando 1991; Koester 1989, 1994;
Mason 1989; Oulette et al. 1991; Page et al. 1990; Singer 1991; Watters
1989).
In Chicago, ethnographers led by Oulette demonstrated how differences
in the frequency of needle sharing among clients of shooting galleries
were due, in part, to the socioeconomic status and political power of the
neighborhood where they were located. These researchers found a
relationship between a community’s economic and political power and
the level of police attention it received. In poor neighborhoods, police
largely ignored shooting gallery operations. As a result, galleries
operated as businesses, with relatively little danger of being raided. They
were well-managed with strict house rules, including rules governing
needle hygiene. They also were accessible to HIV intervention efforts.
In contrast, syringe sharing was common in the more hidden, low-volume
galleries that typified neighborhoods undergoing gentrification. Greater
police presence in these neighborhoods discouraged larger, more
established galleries. Instead, private “taste” galleries were the norm.
These were users’ apartments, where a small group of injectors could
exchange a portion of their drugs for a place to inject. These small,
210
closed galleries were not as accessible to intervention efforts because the
operators were more fearful of being arrested. Their structure was less
authoritarian, house rules were not rigidly enforced, and needle hygiene
was regarded as an injector’s own business (Oulette et al. 1991).
Another example of how the legal system can affect high-risk behavior
concerns the role paraphernalia laws play in restricting IDUs’ access to
new syringes, a condition injectors frequently overcome by transferring
used syringes. Laws regulating the purchase or possession of drug
paraphernalia, including syringes, are currently in effect in 44 states and
numerous municipalities (Des Jarlais et al. 1992; Fernando 1991; Koester
1989, 1994; Pascal 1988).
In Denver, needle access is restricted by both a State law and a municipal
ordinance. Possession of a syringe with intent to inject an illicit
substance is a misdemeanor, and is usually punishable by a fine of
between $50 and $100. For an injector this is a serious matter.
Committing this offense identifies the offender as a drug user to the
police. It results in a court appearance and fine, and it occasionally leads
to incarceration. Jail time results because drug users often fail to appear
for their court date. In addition, a paraphernalia violation goes on the
injector’s record, making it more difficult to plead not guilty to future
drug-related charges. As IDUs explained it, having violations for
paraphernalia on their record would make it difficult to convince a judge
of their innocence regarding more serious drug-related offenses. As
Feldman and Biemacki (1988) contend:
The illegality of possessing hypodermic syringes...accounts for the
scarcity or unpredictable supply of hypodermic syringes, the chronic
fear of arrest, and the necessity of constructing social arrangements
that involve needle sharing (p. 35).
Denver injectors appear to agree with this assessment, For many, arrest
and incarceration are more immediate fears than HIV. Being jailed for
even a few days is not only unpleasant for an addicted IDU but often
physically and psychologically painful as well. As one injector
explained:
That is how they busted me that one time for drugs-because they
busted me with a fit (syringe). They took me down there and strip
searched me and found heroin and coke on me. They have done it
other times... they did it one time and took me downtown and strip
211
searched me and I had stashed the coke in my mouth and they didn’t
find it. They use that (the paraphernalia law) as an excuse. They
know you’ll forget the court date half the time. They know that even
if you make the court dates you might not be able to pay the fine;
then the next time they see you they know they are going to run an
ID check on you—you will have a warrant out and they—l take you
in for that. It’s just one big hassle (Koester, unpublished data).
Other contextual aspects of drug injectors’ lives combine with these legal
constraints to increase the likelihood that they will not have a syringe
when they are ready to inject. Laws criminalizing syringe possession are
particularly troublesome for street-based injectors who, because of their
impoverished lifestyle, often are visible and thus vulnerable to arrest.
Street-based injectors may have only temporary living arrangements and
no access to transportation. In many instances, they may not have a place
of their own to go and consume drugs. In addition, they must often
expose themselves to view because the specific hustles they perform
demand it.
Poor injectors frequently pool their limited cash to purchase drugs. Even
a quarter gram of cocaine at $20, the smallest quantity sold, may be
beyond their individual means, and some heroin dealers will not sell any
less than a $60 “piece.” To overcome this condition of the market,
injectors form temporary “business” arrangements with other injectors; to
do so, they go to known copping sites to meet. It is at known copping
sites where police presence is most constant; as a result, injectors seldom
carry a syringe when going to buy drugs. Injectors who support their drug
use by copping drugs for others, linking buyers with dealers, also are
forced to be somewhat visible, and they too refrain from carrying
syringes (Koester 1994).
Paraphernalia laws are not the only impediment to syringe access.
Although syringes are legal to purchase without a prescription in
Colorado, this does not necessarily ensure that they can easily be obtained
by IDUs. Drug stores’ policies, the attitudes of individual pharmacists,
the price a pharmacy charges for syringes, and the proximity of a store to
areas where injectors buy and use drugs are important determinants of
accessibility. The author found that some pharmacists refused to sell
syringes to individuals they thought were drug injectors, while others
required the potential buyer to show proof that he or she was a diabetic.
Pharmacies that readily sell to injectors often are located several blocks
away from drug copping scenes, or they charge exorbitant prices for
212
syringes. One pharmacy located close to a high-profile drug copping
scene kept syringes at the front counter but charged $1.50 a piece or
$5 for a 10-pack of U-100 insulin syringes. In contrast, pharmacies
attached to national food chain stores sold lo-packs of insulin syringes
for as little as $1.70. However, not all of these chain stores would sell to
injectors.
CONCLUSION
Attempting to understand the multiple reasons why risk reduction is not
necessarily easy to achieve or maintain requires that researchers identify
and analyze impediments to behavior change, and thus examine the larger
contextual dimensions of drug users’ lives. What microlevel and
macrolevel conditions and circumstances affect people’s ability to make
potentially life-preserving changes? By emphasizing context in their
studies, researchers can temper the tendency to explain high-risk behavior
and the attitudes that support it as owing to personal inadequacy or
maladaptive subcultural traits.
By identifying significant contextual factors and describing how they
exacerbate or inhibit risk behavior, researchers can augment current
interventions aimed at individual behavior change. Intervention
programs can be developed to support individuals’ attempts at risk
reduction by addressing the local conditions that encourage risk behavior
while at the same time promoting conditions that encourage risk
reduction. For example, in other forums, the author has advocated a
reassessment of current paraphernalia laws that make syringes difficult to
obtain or illegal to possess. Because of the HIV epidemic, these laws
may no longer serve the public interest. Likewise, pharmacists in states
where syringes are sold legally could be encouraged not to discriminate
against IDUs. Clearly, such proactive approaches must be localized.
They will only be effective if they are based on locally significant
behavioral influences and if information provided by active drug users is
considered in these approaches.
Ethnographic research has made substantial contributions to the fields of
epidemiology, public health, and HIV prevention. By giving drug users a
voice, ethnographic research has provided a means for those at risk to
inform the research and interventions. By describing how the
environment in which people live influences their behavior,
ethnographers have presented the field of public health with new
213
REFERENCES
Agar, M. The Professional Stranger: An Informal Introduction to
Ethnography. New York: Academic Press, 1980.
Anderson, E. Streetwise. Chicago: University of Chicago, 1990.
Bourgois, P. In search of Horatio Alger: Culture and ideology in the
crack economy. Contemp Drug Probl 16(4):619-651, 1989a .
Bourgois, P. “Just Another Night on Crack Street.” New York Times
Magazine, 12 November 1989b. p. 61.
Carlson, R.G. “HIV Needle Risk Behavior among IDUs in West Central
Ohio: An Ethnographic Overview.” Paper presented at the Conference
on AIDS and Anthropology in the United States. Atlanta, GA,
October 16, 1991.
Carlson, R.G.; Siegal, H.A.; and Falck, R.S. Ethnography, epidemiology
and public policy: Needle use practices and risk reduction among IV
drug users in the midwest. In: Feldman, D.A., ed. Anthropology and
Global AIDS Policy. Westport, CT: Greenwood Press, in press.
Clatts, M.; Davis, W.R.; Deren, S.; Goldsmith, D.S.; and Tortu, S. AIDS
risk behavior among drug injectors in New York City: Critical gaps in
prevention policy. In: Feldman, D., ed. Anthropology and Global
AIDS Policy. Westport, CT: Greenwood Press, in press.
Des Jarlais, D.C. and Friedman, S.R. Shooting galleries and AIDS:
Infection probabilities and ‘tough’ policies. Am J Public Health 80:
142-144, 1990.
Des Jarlais, D.C. and Friedman, S.R. AIDS and legal access to sterile
drug injection equipment. Ann Am Acad Political Soc Sci 521: 42-65,
1992.
Des Jarlais, D.C.; Friedman, S.; and Strug, D. AIDS and needle sharing
within the IV drug use subculture. In: Feldman, D., and Johnson, T.,
eds. The Social Dimension of AIDS: Method and Theory. New York:
Praeger, 1986.
Feldman, H.W., and Biemacki, P. The ethnography of needle sharing
among intravenous drug users and implications for public policies and
intervention strategies. In: Battjes, R.J., and Pickens, R.W., eds.
Needle Sharing Among Intravenous Drug Abusers: National and
International Perspectives. National Institute on Drug Abuse Research
Monograph No. 80. DHHS Pub. No. (ADM)88-1567. Washington,
D.C.: Supt. of Docs., U.S. Govt. Print. Off., 1988.
Fernando, D. Fundamental limitations of needle-exchange programs as a
strategy for HIV prevention among IVDUs in the US. AIDS Public
Pol J 6:116-120, 1991.
214
Guydish, J.R.; Golden, E.; and Hembry, K. Needle sharing, needle
cleaning and risk behavior change among injection drug users. In:
Sorenson, J.L.; Wermuth, L.A.; and Gibson, D.R., eds. Preventing
AIDS in Drug Users and Their Sexual Partners. New York: Guilford
Press, 1991.
Herdt, G. Introduction. In: Herdt, G., and Lindenbaum, S., eds. The Time
of AIDS. Newbury Park, CA: Sage, 1992.
Howard, J., and Borges, P. Needle sharing in the Haight: Some social and
psychological functions. In: Smith, D., and Gay, G., eds. It’s So Good,
Don’t Even Try It Once. Englewood Cliffs, NJ: Prentice Hall, 1972.
Jencks, C. “Deadly neighborhoods.” The New Republic 198:23-32, 1988.
Kane, S. and Mason, T. The limits of epidemiological categories and the
ethnography of risk. In: Herdt, G., and Lindenbaum, S., eds. The Time
of AIDS. Newbury Park, CA: Sage, 1992.
Koester, S.K. “When Push Comes to Shove: Poverty, Law Enforcement
and High Risk Behavior.” Paper presented at the annual meeting of the
Society for Applied Anthropology, Santa Fe, NM, April 9, 1989.
Koester, S.K. “Ethnography and High Risk Drug Use.” Paper presented at
the annual meeting of the College on the Problems of Drug
Dependence, Keystone, CO, June 17, 1992.
Koester, S. and Schwartz, J. Crack, gangs, sex and powerlessness: A view
from Denver. In: Ratner, M., ed. Crack Pipe as Pimp. New York:
Lexington Books, 1993.
Koester, S.K. Copping, running and paraphernalia laws: Contextual
variables and needle risk behavior among injection drug users in
Denver. Hum Organiz 53(3), forthcoming.
Maher, L., and Curtis, R. Women on the edge of crime: Crack cocaine
and the changing contexts of street-level sex work in New York City.
Crime Law Soc Chang 18:221-258, 1992.
Mason, T. A preliminary look at social and economic dynamics
influencing drug markets, drug use patterns, and HIV risk behaviors
among injecting drug users in two Baltimore networks. In:
Proceedings of the Community Epidemiology Work Group.
Washington, DC: National Institute on Drug Abuse, 1989.
Massey, D. American apartheid: Segregation and the making of the
underclass. AJS 96:329-357, 1990.
Murphy, S. Intravenous drug use and AIDS: Notes on the social economy
of needle sharing. Contemp Drug Probl 14:373-395, 1987.
Newman, K.S. Culture and structure in the truly disadvantaged. City Soci
6(1):3-25, 1992.
215
Ouellet, L. J.; Jimenez, A.D.; Johnson, W.A.; and Wiebel, W.W.
Shooting galleries and HIV disease: Variations in places for injecting
illicit drugs. Crime Delinq 37:64-85, 1991.
Page, J.; Chitwood, D.; Smith, P.; Kane, N.; and McBride, D. Intravenous
drug use and HIV infection in Miami. Med Anthropol Q 1:56-71,
1990.
Pascal, C.B. Intravenous drug abuse and AIDS transmission: Federal and
state laws regulating needle availability. In: Battjes, R.J., and Pickens,
R.W., eds. Needle Sharing Among Intravenous Drug Abusers:
National and International Perspectives. National Institute on Drug
Abuse Research Monograph No. 80. DHHS Pub. No. (ADM)88-1567.
Washington, DC. Supt of Docs.: U.S. Govt. Print. Off., 1988.
Phillips, K. The Politics of Rich and Poor: Wealth and the American
Electorate in the Reagan Aftermath. New York: Random House,
1990.
Ratner, M. Crack Pipe As Pimp. New York: Lexington Books, 1993.
Singer, M.; Irizarry, R.; and Schensul, J.J. Needle access as an AIDS
prevention strategy for IV drug users: A research perspective. Hum
Organiz 50:142-153, 1991.
Wallace, R. Social disintegration and spread of AIDS - II. Soc Sci Med
37(7):887-896, 1993.
Watters, J. Observations on the importance of social context in HIV
transmission among intravenous drug users. J Drug Issues 19:9-26,
1989.
Wiebel, W. Combining ethnographic and epidemiologic methods in
targeted AIDS interventions: The Chicago model. In: Battjes, R.J., and
Pickens, R.W., eds. Needle Sharing Among Intravenous Drug
Abusers: National and International Perspectives. National Institute
on Drug Abuse Research Monograph No. 80. DHHS Pub. No.
(ADM)88-1567. Washington, DC: Supt of Docs., U.S. Govt. Print.
Off., 1988.
Wilson, W.J. The Truly Disadvantaged. Chicago: University of Chicago
Press, 1987.
216
AUTHOR
Stephen K. Koester, Ph.D.
Assistant Professor
Department of Psychiatry
University of Colorado School of Medicine
1643 Boulder Street
Denver, CO 80211
217
Assessing the Reliability and
Validity of Self-Reported Risk
Behavior
David R. Gibson and Martin Young
INTRODUCTION
This chapter presents preliminary data from the AIDS Risk Measurement
Study (ARMS), a survey of intravenous drug users (IVDUs) interviewed
in the San Francisco Bay area from 1991 to 1992. While a principal
objective of the study was to examine correlates of human
immunodeficiency virus (HIV) risk behavior, it included several
methodological experiments both to assess and to improve the reliability
and validity of drug users’ self-reports. The experiments were inspired
by survey research literature (Bradbum and Sudman 1979) that examined
the truthfulness of survey respondents’ answers to questions about such
threatening or sensitive behaviors as drunken driving and bankruptcy.
This small literature documents the extent to which respondents tend to
underreport such behaviors and suggests procedures for minimizing
underreporting. Since HIV risk behavior includes many illicit and
socially stigmatized practices, it seemed worthwhile to undertake a
similar investigation.
This chapter addresses three questions. First, how truthful or accurate is
information that IVDUs provide to researchers concerning practices that
put them at potential risk of infection of HIV? Second, to what extent are
behavioral self-reports contaminated by self-presentation bias-the
tendency of drug users to present a socially desirable image of
themselves? And third, what steps can researchers take to reduce
response bias when asking sensitive or threatening questions?
The sample from which the data for this report were taken was 508
IVDUs who were interviewed at entry to heroin detoxification treatment
at four sites in the San Francisco Bay area between September 1991 and
December 1992. Since the treatment is of brief duration (maximum of
21 days), many clients soon return to active drug use. One hundred
fifty-one subjects were recruited from the street with coupons that made
them eligible for immediate free treatment at the clinics. While a
218
condition of receiving the coupons was that subjects not have been in
treatment during the previous year, clients recruited in this way did not
differ demographically or in terms of reported risk behavior from clients
who referred themselves to the clinics. Thus the results reported here
may be roughly reflective of those that would be obtained with an out-oftreatment population. This is of some importance, since out-of-treatment
opiate addicts account for as many as six in seven opiate addicts in the
United States (Watters and Biernacki 1989).
A sociodemographic profile of the sample appears in table 1. The
ethnically diverse sample consisted mainly of males between 30 and 40
years old. About 60 percent of the sample had never been in methadone
maintenance treatment, and nearly half had made fewer than five attempts
to detoxify from heroin.
ACCURACY OF SELF-REPORTS
There is a critical need for accurate estimates of the prevalence of
behaviors that put people at risk of HIV infection. Reliable estimates are
needed to identify subpopulations to be targeted for intervention, to
discover correlates or antecedents of high-risk practices, and to properly
evaluate prevention programs to reduce transmission of HIV.
Behavioral research on transmission of HIV has been hampered by
dependence on people’s self-reports about their behavior and by the
researchers’ inability to verify those reports. The validity of self-reports
is suspect, since many risk behaviors include illicit or otherwise
stigmatized practices that respondents may be reluctant to acknowledge.
Methodological studies reveal that substantial underreporting of behavior
occurs when respondents find questions embarrassing or threatening. For
example, 50 percent of drunken drivers surveyed by Bradbum and
Sudman (1979) denied having been arrested, while 30 percent of those
who had gone through bankruptcy failed to acknowledge this in
interviews.
Studies that have attempted to document opiate addicts’ self-reports of
criminal or drug behavior (e.g., Amsel et al. 1976; Bale et al. 1976;
Maddux and Desmond 1975; Maisto et al. 1982) nevertheless have found
them to be remarkably accurate. Most of these studies, however,
gathered a relatively limited range of information in settings such as
prisons and drug treatment clinics where admitting drug habits or
219
TABLE 1. Sociodemographic profile of sample (N = 508)
Percent
Age (M = 39)
19-29
30-39
40-49
50 and over
10.1
43.7
38.5
7.7
Gender
Male
Female
63.0
37.0
Living situation
Married/living with
partner
Single
59.1
40.9
Housing
House or apartment
Other
90.7
9.3
Sexual orientation
Heterosexual
Homosexual
Bisexual
84.8
1.6
13.6
Years of methadone maintenance
treatment (M = 2.0)
Zero
1-4
5 or more
58.2
24.4
17.4
Number of previous detoxes
Zero
1-4
5 or more
1.4
47.4
51.2
Age at first heroin use (M = 21)
20 and under
21-30
30 and over
60.9
31.8
7.3
220
criminal history was unlikely to be threatening. Gibson and colleagues
(1987) found heroin addicts’ answers to questions about their
psychological or social functioning to be moderately to highly correlated
with a short form of the Marlowe-Crowne social desirability scale. In
particular, the answers were highly skewed by the tendency to deny
negative information about themselves and their families.
Test-Retest Reliability
In the present study, two attempts were made to assess the reliability of
IVDUs’ self-reports. First, part of the sample was reinterviewed to
determine whether their answers would be consistent across two
interviews; second, IVDU heterosexual couples (48 women and 48 men)
were interviewed separately to determine how closely their reports about
sexual practices agreed.
Test-retest correlations and correlations of partners’ self-reports have
been examined in a number of studies (for a review, see Catania et al.
1990) and frequently have been interpreted as validity indicators.
Reliability, however, while necessary to validity, is not sufficient to
establish the accuracy of self-reports. Reporting biases are as likely to be
reflected in a retest as in an initial interview, and partners’ reports may be
spuriously consistent due to shared biases. Unfortunately, there is not a
gold standard against which to evaluate the validity of self-reports, at
least at the individual level.
The retest interviews were conducted an average of 10 days following the
initial interview. Both interviews were conducted by trained interviewers
during clients’ daily visits to the drug treatment clinics. On both
occasions, respondents were asked to report their sexual and injection
practices during the 30 days prior to their admission to the detoxification
program (for several behaviors, a 6-month reporting period was also
used, and clients were asked whether the behaviors had ever been
practiced). Respondents were not told of the principal reason for the
second interview, only that some additional information not obtained in
the first interview was needed (additional items were added to the second
interview to disguise the intent). The interval of 10 days was selected to
be long enough so respondents would not remember what they had said
at the first interview but short enough so they would be able to accurately
report about their behavior during the month prior to their entry to
treatment.
221
Table 2 shows percentage agreement in answers across the two
interviews. The figures indicate a reasonable degree of reliability,
ranging from 77 percent to 90 percent. The behavior least reliably
reported was borrowing of injection equipment (30-day indicator), with
77 percent agreement. The most accurate measures were for the rarest
behaviors such as anal intercourse, with only 12 and 2 respondents,
respectively, reporting this behavior at either interview.
Reliability of Sexual Behavior Reported by Heterosexual
Couples
Table 3 shows sexual data from 29 couples who reported mutually
monogamous sexual behavior for the 30 days prior to entering drug
treatment. The couples were identified discreetly by clinic intake workers
and were not aware that a purpose of the study was to assess the
reliability of self-reported sexual practices. Interpartner agreement about
sexual practices was very close to agreement between the test and retest.
The behavior least reliably reported was oral sex (79 percent), while there
was complete agreement about whether the couple had been sexually
abstinent.
Respondents as Informants
The lack of a gold standard for validating self-reports makes assessment
of their accuracy difficult at best. Two attempts were made to validate
self-reports of needle sharing that shed some light on the probable
accuracy of the information given.
The first was suggested by experiments conducted by Bradbum and
Sudman (1979) in which college students were interviewed about their
own and their friends’ use of alcohol and marijuana. The assumption was
made that the level of alcohol and marijuana actually used would be
similar for respondents and friends. Analysis of the data, however,
indicated overall that the students reported their friends used alcohol and
marijuana more often than they did themselves, suggesting that they may
have underreported their own use. A further suggestion that self-reported
use was less accurate was that the variances of measures of their own use
were larger than for measures of their friends’.
In the present study, in addition to asking respondents about their own
borrowing of needles and syringes, they were asked whether they had
observed others doing so during that time (past 30 days). While only
222
TABLE 2. Percent agreement in answers to questions about selected
HIV risk behaviors, T1 and T2 (N = 94)
Behaviors
Injection behaviors
Borrowed dirty syringe
Shared cotton or cooker
Sexual behaviors
2 or more sexual partners
Used condoms some
Used condoms 100 percent
Anal intercourse
*Agreement significant at p
Ever
Past 6
months
Past 30
days
87.9*
82.7*
77.6*
82.8*
84.9*
91.9*
88.9
87.7*
92.3*
97.0*
0.001.
TABLE 3. Concordance of self-reports of monogamous couples about
selected sexual behaviors (N = 29)
Percent
agreement
Sexual Behaviors
Sexually abstinent, past 30 days
Used condoms some, past 30 days
Used condoms 100 percent, past 30 days
Oral sex, past 30 days
Anal intercourse, past 6 months
*Agreement significant at p
100.0*
87.0*
91.3*
79.3*
93.1*
0.001.
36 percent of respondents reported that they had borrowed a syringe
themselves, 63 percent said that they had observed others borrowing
(p < 0.001); “borrowing” refers to using a needle that someone else had
used without first cleaning it with bleach. While it is difficult to attribute
this difference entirely to underreporting of their own risk behavior, it is
probable that respondents had far fewer opportunities to observe others’
‘behavior than their own.
223
FORENSIC TECHNIQUES TO DOCUMENT SELF-REPORTS OF
NEEDLE SHARING
The authors attempted to verify self-reported needle sharing by
performing forensic tests on syringes provided by drug users to determine
whether the tests corroborated what drug users reported they had done
with the syringes (Gibson et al. 1991).
A convenience sample was recruited in a San Francisco street setting and
asked to provide syringes for an anonymous study of injection
equipment. Fifty-nine drug users presented an average of 1.7 syringes for
inspection, with 31 presenting at least one syringe containing visible
blood. With each drug user, a staff member selected a single syringe as
the basis for a brief interview concerning its use by the donor and others,
After completing the 10-item interview, drug users provided a reference
fingerstick blood specimen and were paid $10 for their time. They were
not told of the specific purpose of the study.
Blood residue in the syringes was analyzed for genetic traits using tests
for the isoenzyme phosphoglumatase (PGM), Gamma marker (Gm) and
Kappa marker (Km) immunoglobulin, and HLA-DQ alpha-type. Of the
31 syringes, 27 contained sufficient blood to conduct one or more of the
tests. Genetic markers were obtained in 10 PGM tests, 23 Gm tests,
25 Km tests, and 23 DQ-alpha tests. The reference blood samples
provided by the donors of the 27 syringes were analyzed to determine the
genetic traits of the donors.
In only four syringes were there genetic traits of two drug users. The
analysis indicated, however, that 15 of the 27 syringes had been used by
someone other than the donor. Table 4 shows a cross-tabulation of the
laboratory results with self-report of sharing a dirty needle (i.e., a needle
that had been used by someone else). The table indicates a reasonable
degree of agreement between the laboratory results and self-reports, with
there being a correspondence in 20 of 27 cases (74 percent). However,
more than a third (6 out of 17) of drug users who denied sharing were
contradicted by the laboratory tests. Moreover, in six of the seven cases
where there was a discrepancy between the lab test and self-report, the
drug user denied that he or she had borrowed. Self-reports of needle
sharing thus appear to understate “true” levels of sharing.
224
TABLE 4. Cross-tabulation of laboratory results with self-reported
borrowing of a dirty needle (N = 27)
Laboratory test
Self-report
Did not borrow
Did borrow
Total
Did not borrow
Did borrow
Total
11
1
12
6
9
15
17
10
27
X2= 8.33, p < 0.005; phi = 0.56.
SELF-PRESENTATION BIAS IN SELF-REPORTS
Self-presentation bias is a problem ubiquitous in social research that
derives from the natural tendency of people to present themselves in a
socially desirable light. Past research on social desirability indicates this
tendency is more likely to manifest itself in situations that people find
particularly threatening or embarrassing (for a review, see DeMaio 1984).
Because the questions asked concerned sexual practices and illicit drug
use and usually are considered sensitive, the authors hypothesized that
self-presentation bias would be reflected to some degree in
underreporting of risk behavior.
Social Desirability
To determine whether this was true, the authors examined zero-order
relationships between self-reported risk behavior and a 20-item short
form of the Marlowe-Crowne social desirability scale (Strahan and
Gerbasi 1972). The short form demonstrated adequate reliability when
used with a college sample and correlated highly with the original
33-item scale.
Respondents endorsed a mean of half the items in a socially desirable
direction, which is comparable to results obtained with the college
sample.
To better examine the relationship between the scale and the self-reported
risk behaviors, the scores were collapsed into quartiles and
225
cross-tabulated with dichotomous measures of risk behavior, for example,
borrowing of syringes (none versus some).
The data in table 5 indicate that self-reported injection behavior may be
significantly underreported to staff in drug treatment settings due to selfpresentation bias, while, with one exception (anal intercourse), reported
sexual practices appeared to be unrelated with social desirability. The
26-point spread between the first and fourth quartiles in the percentage of
respondents reporting that they borrowed needles (past 30 days) suggests
that at least some respondents found this question threatening. The scores
for the lowest quartile of social desirability may approach those that
would be obtained if social desirability did not color self-reports.
Question Threat
The social desirability measure assumes a generalized disposition to
project a socially desirable image. In a related analysis, the authors
examined whether self-reports might be a function less of persons than
the perceived social sensitivity of specific sets of questions. Following
questions about injection behaviors, respondents were told: “We now
want to ask you about the questions we’ve been asking-about sharing of
outfits and the like.” Respondents were then asked four questions such
as, “How (sensitive, uncomfortable, nervous, uneasy) do you think these
questions would make most drug users?” A similar set of items followed
the questions about sexual practices: “How (sensitive, uncomfortable,
nervous, uneasy) do you think most people would find these questions?”
(about condoms and number of sex partners and so forth). The questions
were intended to be unobtrusive measures of respondents’ own feelings
toward the items. The response choices ranged from “not at all,” to
“somewhat,” to “quite a bit.”
Descriptive statistics indicated that the average respondent found both
sets of questions to be “somewhat” intimidating, with there being a fair
degree of variability in the perceived threat of the questions (standard
deviation of about half a point on a 3-point scale). The measures were
collapsed into thirds by level of threat (“not at all,” “somewhat,” “quite a
bit”) before cross-tabulating with self-reported behaviors. The analysis
(see table 6) showed that question threat was related to only one behavior,
oral sex, with there being a 20-point plus difference between the lowest
and highest third in the percentage of respondents reporting oral sex.
226
TABLE 5. Relationship of social desirability to self-reported risk-related behaviors (N = 508)
Social desirability
Percent who reported
Low
Moderate
High
Very high
P
Injection behaviors
Borrowed dirty needle, past 6 months
Borrowed dirty needle, past 30 days
Shared cooker or cotton, past 30 days
61.7
52.6
82.1
45.5
35.0
68.0
48.1
33.3
74.1
36.8
27.1
61.5
0.08
0.000
0.003
Sexual behaviors
Ever used condoms
Some condom use, past 30 days
100 percent condom use, past 30 days
Oral sex, past 30 days
Anal intercourse, ever
Anal intercourse, past 6 months
12.7
34.7
16.8
58.1
67.9
15.3
16.5
39.1
15.9
61.8
66.0
13.6
12.3
36.9
22.3
62.3
61.5
15.4
15.5
39.3
19.1
53.5
50.9
11.2
NS
NS
NS
NS
0.03
NS
NS = nonsignificant
TABLE 6. Relationship of question threat to self-reported risk-related
behaviors (N = 508)
Question threat
Percent who reported
Low Moderate High p
Injection behaviors
Borrowed dirty needle, past 6 months
Borrowed dirty needle, past 30 days
Shared cooker or cotton, past 30 days
50.5
39.8
74.6
47.8
37.7
72.6
49.0
32.7
59.6
Sexual behaviors
Ever used condoms
Some condom use, past 30 days
100 percent condom use, past 30 days
Oral sex, past 30 days
Anal intercourse, ever
Anal intercourse, past 6 months
85.3
38.1
19.1
46.8
58.1
8.1
87.1
37.3
19.2
57.8
63
14.8
82.1 NS
36.8 NS
19.1 NS
69.1 0.03
59.5 NS
13.1 NS
NS
NS
NS
NS = nonsignificant
Sexual Self-Disclosure
A final analysis examined covariance of self-reported sexual practices
with a measure of respondents’ comfort with disclosing information
about sexual practices. In a previous study, a variant of this measure
predicted refusal to answer questions in a human sexuality study (Catania
et al. 1986). The authors hypothesized that high-disclosing respondents
would acknowledge a greater degree of risk behavior. As with social
desirability, the measure was collapsed into quartiles to examine its
relationship with self-reported risk behavior. There was one relationship,
a trend, in which the lowest quartile reported less oral sex (see table 7).
Improving the Validity of Self-Reports
The findings reported in the previous section, particularly those showing
self-reports to be downwardly biased by social desirability, point to the
need for measures that minimize self-presentation bias. In Response
Effects in Surveys (1974), Sudman and Bradburn reviewed evidence that
response effects were generally greater for threatening than for
nonthreatening questions and suggested that threatening questions may
228
TABLE 7. Relationship of sexual self-disclosure to self-reported risk-related sexual behaviors (N = 508)
Sexual self-disclosure
Percent who reported
Ever used condoms
Some condom use, past 30 days
100 percent condom use, past 30 days
Oral sex, past 30 days
Anal intercourse, ever
Anal intercourse, past 6 months
NS = nonsignificant
Low
81.7
37.7
16.9
47.0
58.3
9.6
Moderate
87.1
36.2
19.2
60.3
62.1
13.8
High
86.2
31.5
14.1
61.4
60.5
13.4
Very high
89.0
40.5
24.7
61.2
65.6
15.1
p
NS
NS
NS
0.07
NS
NS
aggravate self-presentation bias. This evidence, which was culled from a
large number of studies, also suggested that task variables (e.g., the
structure and length of questions and how the interview was
administered) were important factors influencing response, with memory
variables and the characteristics of interviewers and respondents being
much less important.
In a later study, Bradburn and Sudman (1979) attempted to determine
whether more anonymous forms of data collection (self-administered
questionnaire or telephone survey) might reduce response bias relative to
that which occurs when interviews are conducted face-to-face. The
questions included highly undesirable behaviors such as drunken driving
and bankruptcy. While the more anonymous interview did not
consistently increase levels of behavior reported, variations in question
structure and length and structure did. Subjects were randomized to
closed-ended versus open-ended questions, long versus short questions,
and questions with standard versus familiar wording. Familiarly worded
questions are questions that employ language that respondents themselves
would use. The results indicated that long, open-ended, familiarly
worded questions yielded higher levels of reporting than short, closedended questions with standard wording. The pattern was evident even
when differences in formats were examined on an item-by-item basis.
Very few interactions of formats were noted.
Impact of Questionnaire Format
In the present study, the authors examined the impact of question length
and question order on the extent of underreporting and zero-reporting of
high-risk practices. Without a gold standard against which to validate
self-reports, the authors made the assumption that higher levels of
reported risk behaviors would on average be more accurate since selfreports are downwardly biased when respondents find questions to be
threatening or to invite socially desirable responses (Catania et al. 1990).
The authors hypothesized that long questions and questions asked late
rather than early in the interview would result in higher levels of
reporting. The long questions (15 words or more) were preceded by
preambles that essentially encouraged respondents to acknowledge
socially undesirable (risk) behavior. For example, the question about
needle sharing was prefaced with the statement, “Now we have some
questions about needles and syringes. There are reasons people share
used outfits, such as being dope sick, not having your own works, and not
230
having bleach available.” Similar statements preceded questions about
condom use and anal intercourse.
The hypothesis concerning questions asked late versus early in the
interview was predicated on the belief that rapport develops between an
interviewer and a respondent in the course of an interview, increasing the
comfort level of the respondent and thereby making it more likely that the
respondent will reveal unfavorable information. While the literature on
order effects reveals a mixed pattern of findings (Sudman and Bradbum
1974; Bradbum and Sudman 1979), the authors’ experience working with
IVDUs suggested this would be the case.
Table 8 shows the results of the experiment with interview formats. The
long questions elicited significantly higher levels of reported borrowing
of syringes and a higher reporting level of lifetime practice of anal
intercourse. They appeared to have little impact, however, on selfreported condom use. Order (early versus late) effects were observed for
only one variable, condom use for the previous 30 days; respondents
asked this question late in the interview were more likely to disclose
condom use. This finding is difficult to interpret, but it may have been
related to an association of condoms with loose sexual morals, making it
less likely that respondents would admit to using them early in the
interview. There were no significant interactions between the long versus
short and early versus late manipulations.
Matching Gender of Interviewers and Respondents
In a separate experiment, the authors randomized male and female
respondents to male and female interviewers to determine whether
responses to questions about sexual behavior might be influenced by the
gender of interviewer, respondent, or both. Although Sudman and
Bradbum’s (1974) review of the literature suggested the characteristics of
interviewers and respondents have little effect on responses to questions
about behavior, an apparent exception is situations in which
interviewer/respondent characteristics are related to the subject matter.
Table 9 shows the results of the randomized experiment. There were no
statistically significant main effects of interviewer gender. In two trends,
women respondents were less likely to acknowledge to a female
interviewer that they had vaginal intercourse but more likely to report
using condoms consistently. Other patterns were evident but not
231
TABLE 8. Percent of subjects reporting behaviors in response to long
versus short questions and to questions asked early versus
late in the interview (N = 508)
Percent of subjects
reporting behavior
Long versus short questions
Long
question
Borrowed dirty syringe, past 6 months
Borrowed dirty syringe, past 30 days
Used condoms, ever
Used condoms some, past 30 days
Used condoms 100 percent, past 30 days
Anal intercourse, ever
Anal intercourse, past 6 months
47.8
43.3*
86.6
38.4
21.5
66.4*
13.6
Short
question
46.9
32.8*
85.6
36.5
16.9
57.8*
14.0
Percent of subjects
reporting behavior
Early versus late questions
Early
question
Borrowed dirty syringe, past 6 months
Borrowed dirty syringe, past 30 days
Used condoms, ever
Used condoms some, past 30 days
Used condoms 100 percent, past 30 days
Anal intercourse, ever
Anal intercourse, past 6 months
48.2
39.0
84.3
31.5*
16.9
60.2
11.2
Late
question
46.5
37.0
87.8
43.1*
21.3
63.8
16.3
*Differences significant at p < 0.05.
statistically significant. Female respondents were less likely to report
condom use to male interviewers, perhaps reflecting an association of
condoms with easy virtue; male respondents were more likely to report
vaginal, oral, and anal sex to male interviewers, suggesting what a
colleague has called a braggadocio effect, men exaggerating their sexual
prowess to other males. It is possible that research with larger samples
might find these relationships to be statistically significant.
232
TABLE 9. Percent of male and female subjects reporting sexual
behaviors to male and female interviewers (N = 93)
Male
interviewer
Female
interviewer
Male respondents
Vaginal intercourse, past 30 days
Used condoms some, past 30 days
Used condoms 100 percent, past 30 days
Oral sex, past 30 days
Anal intercourse, past 6 months
82.1
26.1
21.7
55.6
17.9
75.9
40.9
27.3
55.2
10.3
Female respondents
Vaginal intercourse, past 30 days
Used condoms some, past 30 days
Used condoms 100 percent, past 30 days
Oral sex, past 30 days
Anal intercourse, past 6 months
89.5
35.3
5.9
44.4
15.8
64.7†
54.6
27.3†
64.7
5.9
Sexual Behavior
†
Differences between male and female interviewers significant at
p < 0.10.
DISCUSSION
The present study yielded some indirect but nevertheless interesting
information concerning the probable validity of self-reported risk
behavior. The pattern of findings across the experiments suggested that
the most threatening questions for respondents were the items pertaining
to sharing of injection equipment. Answers to questions about borrowing
of syringes appeared to be less reliable than questions about sexual
practices and much more strongly correlated with social desirability.
Also, there appeared to be a large discrepancy between respondents’
reports of their own injection behavior and the behavior of fellow drug
users whom they had the opportunity to observe.
Among the sexual variables, questions about oral and anal sex proved to
be the most sensitive, while respondents’ answers to questions about
condoms appeared to be both reasonably reliable and uncontaminated by
self-presentation bias. This finding is encouraging, since oral sex may be
less risky than other sexual practices, such as unprotected vaginal
233
intercourse, and since anal intercourse appears to be a relatively rare
practice in this population.
The experiments with questionnaire format suggest the problem of
response bias can to some extent be alleviated by use of long, permissiongiving introductions to questions and perhaps also by asking sensitive
questions late versus early in the questionnaire. While response bias
probably cannot be eliminated, it can be corrected in multivariate
analyses by inclusion of a measure of social desirability or other measure
of presentation bias. In experimental studies, such bias is ordinarily
corrected by randomization, although it adds to random measurement
error.
A limitation of several of the findings reported here is that they are based
on small sample sizes. These findings in particular need to be replicated
with larger samples. With larger samples, patterns not evident in the
present data may also be discovered.
REFERENCES
Amsel, Z.; Mandell, W.; Matthias, W.; Mason, C.; and Hocherman, I.
Reliability and validity of self-reported illegal activities and drug use
collected from narcotic addicts. Int J Addict 11:325-336, 1976.
Bale, R.; Van Stone, W.; Engelsing, T.; Zarcone, V.; and Kuldav, J. The
validity of self-reported heroin use. Int J Addict 11:325-336, 1976.
Bradbum, N.M., and Sudman, S. Improving Interview Method and
Questionnaire Design. San Francisco: Jossey Bass, 1979.
Catania, J.; Gibson, D.R.; Chitwood, D.D.; and Coates, T.J.
Methodological problems in AIDS behavioral research: Influences on
measurement error and participation bias in studies of sexual behavior.
Psychol Bull 108:339-362, 1990.
Catania, J.; McDermott, L.; and Pollack, L. Questionnaire response bias
and face-to-face interview samples bias in sexuality research. J Sex
Res 22:52-72, 1986.
DeMaio, J. Social desirability and survey measurement: A review. In:
Turner, CF. and Martin, E., eds., Surveying Subjective Phenomena,
Vol. 2. New York: Russell Sage Foundation, 1984.
Gibson, D.R.; Guydish, J.; Wraxall, B.G.D.; Blake, E.T.; Clark, G.; and
Case, P. Using forensic techniques to verify self-reports of needlesharing (letter). AIDS 5:1149-50, 1991.
234
Gibson, D.R.; Wermuth, L.; Sorenson, J.L.; Menicucci, L.; and
Bernal, G. Approval motive in self-reports of addicts and family
members. Int J Addict 22:895-903, 1987.
Maddux, J., and Desmond, D. Reliability and validity of information
from chronic heroin users. J Psychiatr Res 12:87-95, 1975.
Maisto, S.; Sobell, L.; and Sobel, M. Corroboration of drug users’ selfreports through the use of multiple sources. Am J Alcohol Drug Abuse
9:301-308, 1982.
Strahan, R., and Gerbasi, K.C. Short, homogenous form of the MarloweCrowne social desirability scale. J Clin Psychol 28:191-193, 1972.
Sudman, S., and Bradburn, N.M. Response Effects in Surveys. Chicago:
Aldine, 1974.
Watters, J., and Biemacki, P. Targeted sampling: Options for the study of
hidden populations. Soc Probl 36:416-430, 1989.
AUTHORS
David R. Gibson, Ph.D.
Associate Adjunct Professor
Martin Young
Research Associate
Department of Psychiatry and
Center for AIDS Prevention Studies
University of California
San Francisco, CA 94143
and
Haight Ashbury Free Clinics
San Francisco, CA 94118
235
Future Directions for Studies on
the Context of HIV Risk
William C. Grace, Robert J. Battjes, and Zili Sloboda
The purpose of research on the context of human immunodeficiency virus
(HIV) risk behaviors is to understand the myriad factors that influence
and determine whether one will engage in risk behaviors. In particular,
understanding contextual issues in injecting and noninjecting drug users
(IDUs), sex partners of drug users, and special populations such as
women, minorities, and gay or bisexual men remains an important
challenge to the acquired immunodeficiency syndrome (AIDS) research
community as means of preventing HIV transmission are sought and
developed. The chapters in this monograph demonstrate that while
important progress in understanding contextual issues has occurred, much
remains to be understood. Yet, gaps in knowledge were not
unanticipated, and comments from the panel reactors, Agar, Des Jarlais,
and Robles were specifically solicited to identify and address
understudied issues.
This chapter summarizes ideas from a combination of sources. These
sources include the reactors’ comments after each of the five panels in the
meeting (Heterosexual Males, Women, Men Who Have Sex with Men,
Adolescents, and Methodology), discussion sessions throughout the
technical review meeting, and ideas developed by this chapter’s authors
upon reflection on the monograph. Because the meeting did not attempt
to develop agreement on issues, this chapter should not be viewed as
reflecting a consensus of opinion. Themes in the discussions generally
transcended specific population issues. Therefore, this chapter is not
organized by populations as was the technical review itself, but presents
general research issues and opportunities.
MUCH REMAINS TO BE UNDERSTOOD ABOUT THE
CONTEXT OF HIV RISK BEHAVIORS
Participants agreed that research on the context of HIV risk behaviors
remains relatively underdeveloped. Agar summed up much of the
difficulty with conducting contextual research when he noted, “It is
difficult to see new things when you are busy counting.” Early research
236
in the United States was directed toward urgent issues of identifying and
enumerating “risk groups,” transmission routes, risk behaviors, and HIV
outcomes. Public health urgency, researchers’ familiarity and comfort
with quantitative methodology, and a lack of basic information on the
disease combined to create an atmosphere in which contextual issues
received relatively little attention.
While enumeration of risk behaviors is important for understanding the
extent of the epidemic, assessing the potential for spread, and targeting
prevention efforts, such efforts do not adequately clarify processes
underlying initiation and continuation of risk, recovery from risk, and
relapse to risk. Alternately, a focus on enumeration does not elucidate the
temporal and situational variability in risk behaviors. In sum, simply
counting risk behaviors does not adequately inform development of
prevention activities.
A CONTEXTUAL PERSPECTIVE MUST CONSIDER MULTIPLE
FACTORS THAT IMPINGE ON RISK
HIV risk behaviors are complex and cannot be adequately described by
focusing on enumeration of behaviors and groups at risk. A shift in
perspective is needed. This shift would move research from a primary
focus on individual factors toward a contextual perspective that fully
considers social and other environmental factors that impinge on risk
decisions. Risk behaviors are integrally tied to basic physiological,
social, and emotional functions of humans and are multiply determined.
A contextual perspective demands consideration of familial structure and
function, peer groups’ structures and processes, relationships with sex
partners, relationships with drug-using partners, and the effects of drug
use.
Beyond these individual and small group levels, contextual perspectives
require consideration of broader community factors such as economic
conditions, access to employment, needle availability, and similar factors
that impinge on and interact with interpersonal ones.
It is clear that to understand the context of HIV risk behaviors related to
the abuse of drugs, one must understand the epidemiology and natural
history of drug abuse itself. HIV risk is imposed upon and incorporated
into a larger context, and answers to questions about the temporal
variance of drug abuse, the natural history of drug abuse, norms among
237
abusers, and other variables are needed to aid in developing specific
hypotheses for study.
Understanding of the many factors involved will require sophisticated
skills in management and analysis of data collected from multiple
sources. Univariate analyses that cannot account for interactions and
shared variance among variables are not likely to provide significant
advances in information. Understanding of family interactions and
structures, of dyadic and sexual interactions, and of drug-using and
nonusing social networks must be combined with an understanding of
individual factors to clarify how risk behaviors are developed and
maintained. Further, broad social factors beyond an individual’s control,
such as economic conditions and needle or drug availability in the
community, must be evaluated for their interactive effects with other
determinants of risk. Such multifaceted studies will require sophisticated
quantitative and qualitative methods.
RESEARCH ON CONTEXT OF HIV RISK MUST BE GUIDED BY
RELEVANT THEORETICAL UNDERSTANDINGS OF
BEHAVIOR AND DISEASE
Review participants noted that research on risk context has not been
adequately guided by theory. This is especially unfortunate because
epidemiological patterns tend to recur in drug abuse, and it is possible to
use past experience and general theories to guide research. For example,
patterns of drug use seen in the crack epidemic are being repeated as the
use of methamphetamine rises. Similarly, patterns of sexual behavior
associated with crack use may be seen in methamphetamine users.
Theoretical perspectives (e.g., social learning theory) have been used to
plan intervention and prevention programs, but these perspectives have
been less utilized to develop understanding of risk. Theory-based
research on risk behaviors is difficult to develop and conduct because of
the needs to operationalize relevant constructs and to specify relationships
of constructs to behaviors. In many cases, theory must be developed in
conjunction with the collection of data on risk behaviors, so theory
development, theory validation, and understanding of context will occur
in a dynamic, reciprocal relationship. Certainly studies that can validate
theoretical predictions have merit, but in many cases specific predictions
must await data to help develop the theoretical framework. In cases when
theory validation is not possible, to be guided by theory means that
238
variables should be selected and studies designed in a manner that allows
results to be used systematically to inform and develop a theoretical
framework.
CONTEXTUAL ISSUES SHOULD BE UNDERSTOOD AND
EXAMINED WITH BOTH QUALITATIVE AND QUANTITATIVE
METHODOLOGIES
One of the benefits of a technical review is that researchers with various
backgrounds and perspectives exchange ideas, and this process highlights
the complementarity of approaches. Quantitative and qualitative research
methods are often presented as opposite ends of a continuum, but such a
presentation ignores the fact that they are both methods of systematic
inquiry and learning whose shared opposite is ignorance. From that
perspective, it becomes incumbent on researchers to use multiple research
approaches to address a problem.
Important concepts in this regard are those of “experience near” and
“experience distant,” terms used by anthropologists to indicate proximity
to the language and concepts of the observed subject. Experience-near
language, such as backloading, is close to the world view and
phenomenology of the subject, whereas experience-distant language
reflects the observer’s terminology and conceptual organization.
Qualitative researchers use these ideas to evaluate how well their
observations reflect the subject’s experiences and interpretations and not
only the researcher’s views. Ethnographers present at the technical
review stressed the importance of experience-near language and thinking
in order to clarify contextual issues and to raise hypotheses. They also
noted that data provide insights into different levels of contextual
determinants. For example, the researcher arrives at different conclusions
if a self-report is taken as objective fact rather than as the subjects’ world
view. People neither recall nor report accurately and reliably their
experiences and motivations, tending to err in reports and to respond to
environmental contingencies, some of which they are not aware. They
often deny the influence of stimulus control and psychological
motivations that can be objectively related to their behavior (Critchfield
1993; Nisbett and Wilson 1977). Therefore, qualitative and quantitative
data both must be carefully examined to determine the realms of behavior
and context to which they refer.
239
Behavior patterns can be clarified by both quantitative and qualitative
approaches. Qualitative approaches can develop an understanding of the
meaning and purpose of patterns of behavior and, in combination with
quantitative epidemiology and other quantitative approaches, provide an
understanding of the behavioral context. Qualitative methods can help
define relevant variables, while quantitative methods determine the
distribution of the variables. In quantitative methods, patterns are noted
and verified by descriptive and inferential statistics; in qualitative
methods, data are examined for concurrence of themes and consistency
with general propositions or theories.
Through developing an appreciation for the broad spectrum of research
approaches, it is possible to design studies that effectively lead to an
understanding of contextual issues. However, studies should be designed
to use complementary approaches from the beginning, rather than trying
to add a component when the study is underway. An example of this
would be to test an intervention using quantitative analyses of outcomes
while ethnographers examine individual and social responses, particularly
looking for unanticipated contextual determinants of risk maintenance.
Qualitative and quantitative methods should not simply be applied in
parallel and independent fashion within a study: they should inform each
other. Qualitative data may be analyzed through quantitative methods. A
frequent error of quantitative epidemiology, according to the discussion,
is to become so far removed from the data collection process that
meaning gets distorted in the analyses (e.g., through failure to appreciate
the affect in respondents as they selected response options). Further,
some data do not lend themselves to quantitative methods for either
collection or analysis.
THE UNITS OF STUDY AND ANALYSIS MUST BE CAREFULLY
SELECTED AND DEFINED CONSISTENT WITH THE
RESEARCH HYPOTHESES
During the HIV epidemic, there has been a general reluctance to shift
from the concept of risk groups to understanding behavioral variation and
occurrence at the levels of the individual, dyad, and small group. Studies
of patterns of how a disease is spread throughout a population do not
sufficiently address the determinants of that spread, such as motivations
for engaging in risk behavior, determinants of relapse, and predictors of
self-protective behaviors. For example, women are defined and studied
240
as a population, and discussions of women’s risk often center around
their roles in the family, the community, and peer groups. These roles
raise questions about the appropriate unit of study and analysis for
research on women.
Similar issues about unit of analysis also apply to other populations.
Should studies be designed and analyzed on the level of the individual,
the community, the family, or the sexually active couple? Such questions
have profound implications for the design, statistical analysis, and cost of
investigations, and the answers must derive from the hypotheses of the
study. It makes little sense, for example, to study empowerment in
women without examining other people in the woman’s life, as
empowerment implies an interpersonal dynamic as well as an
intrapersonal change.
A helpful way to think about units of analysis is to think in terms of
systems elements that must be understood and affected in a risk-reduction
program. For example, Dr. Robles noted that Puerto Rican IDUs were
unlike the typical IDU in the continental United States in that they tend to
live with families. The families do not generally allow injection in the
home, so a number of elements must be linked for the IDU to use drugs.
The family, the shooting gallery, and sometimes houses of prostitution
constitute linked elements where greater understanding is needed and
interventions might occur. In other settings, families may not be as
important; elements such as peer networks and separate sexual networks
may need to be addressed instead. Disparate settings provide an
opportunity for examining microstructural issues, such as how a crack
house serves as an economic institution, in relationship to broader
macrostructures (e.g., community economic distress).
Analyzing data at a community or population level is often necessary, but
it is fraught with complexities. Populations overlap and do not neatly fit
usual research definitions. For example, the overlap of men who have
sex with men and IDUs may serve as a bridge to HIV transmission if an
infected IDU has sex with a uninfected man. But are the men who are in
both groups and who constitute the points of overlap best understood as
drug users, as gay men, as non-gay-identified men who engage in
homosexual activity, or as a unique group? Is the unit of analysis the
man within the group? If one designs a community intervention for this
group, would it look more like an intervention for IDUs or for men who
have sex with men, or would it be notably unique?
241
As Des Jarlais noted, designs and analyses must account for the
instability within populations. IDUs start and cease injection every day
within a given city, so the overall rate of injection can be constant within
a city while numerous significant changes in injection behavior are
occurring at other levels of analysis. The increase in use of heroin by
sniffing or insufflation, for example, represents a significant shift in risk
behavior that would not be captured by statistics on heroin use unless
heroin use were disaggregated by route of administration. Drug abuse
itself is a chronic disorder with a varying course over time, and variations
in the disease course may be associated with risk behaviors and related
cognitions, motivations, and appraisals. Factors that account for changes
within a population must be studied to truly understand behavior.
CULTURAL DIFFERENCES ARE NOT SUFFICIENT
EXPLANATIONS OF THE CONTEXT OF RISK BEHAVIOR
Culture is a concept that must be treated with respect because it refers to
exceedingly complex issues that have significant impacts on behavior.
Cultural variables may include observable phenomena such as family
composition and types of buildings as well as elements not directly
observable such as national identification, religious beliefs, and affective
style. As an example of the complexity of culture, when one considers
the diverse countries of origin, the range of socioeconomic resources, the
linguistic differences, and the geographic dispersion represented in
groups that all legitimately may be called Hispanic, the term “Hispanic
culture” has no clear and unitary meaning. Even if its meaning were
clear, how would one tease apart the contributions of Hispanic culture
and U.S. culture?
In drug-using populations there are norms that may constitute a drug
culture that overlies and interacts with other cultural variables. This drug
culture is not homogeneous across groups. Drug norms themselves shift,
as the growing acceptance by drug sellers of marketing to young
adolescents demonstrates. It is clear that norms and values of different
subpopulations affect each other, and conceptualization of research issues
must take this into account. One example cited in the review of one norm
influencing another was how adolescents’ ideas about sexual identity are
influenced by larger societal attitudes towards sexual functioning.
Research designs that compare one cultural group with another often lead
to conceptual dichotomization of continuous traits, so that it becomes
242
misleading to talk of one group as, say, family oriented while another is
not. Such characteristics not only exist on a continuum, but their group
distributions overlap. General statements about cultural differences
between groups provide a useful framework for research and some
interventions but can only suggest what may be occurring at an individual
level. Rather than simply attempting to compare groups or cultures,
research on HIV risk context would benefit from the suggestion of
Betancourt and Lopez (1993) on cultural research in general:
Investigators specify what it is about the group that hypothetically
influences behavior and directly examine and measure that phenomenon
in a theoretical framework. Group identity in itself should be considered
inadequate for explaining group differences in behavioral and
psychological research (Zuckerman 1990).
Explicit hypothesis testing would also help obviate another pitfall of
cultural research identified in the discussion, that of confounding culturebased explanations with the investigator’s own values in areas such as
rational choice, free will, morality, and biological influence. An example
of specific hypothesis testing would be in studies of adolescent risk,
where comparing adolescents with adults is not nearly as useful as
identifying and evaluating the impact of unique aspects of adolescent peer
group structures on how they evaluate and appraise potential outcomes.
As group compositions change and develop, effects on HIV risk appraisal
should be evidenced.
RESEARCH SHOULD LOOK FOR CONTEXT OF PROTECTIVE
BEHAVIOR, NOT FOCUS SOLELY ON NEGATIVE BEHAVIOR
Determinants of risk behavior are not necessarily converse determinants
of protective behavior. Research that specifically examines protective
contextual factors and determinants in individuals or groups that either
initially avoided risk behavior or reduced risk behavior is needed to guide
prevention planning.
Specific protective effects are likely to operate only at specific
developmental stages, and advantage must be taken of the opportunity to
study these. Studies of children in high-risk families before they reach
adolescence, or even school age, could be quite useful in determining
child-rearing contextual variables related to risk behaviors. In studies
when both negative and positive influences are sought, qualitative
methods can be especially useful in helping detect protective factors and
243
strengths within a setting or individual, as these are often not specified a
priori in designing studies.
Individuals and communities have made changes in risk behavior
patterns. Better understanding of how these changes have developed is
needed, and an understanding of discrepant norms is needed. An
example given is that one norm promoted by HIV prevention programs is
“a man is not supposed to bring a disease home,” yet these programs also
promote that “a woman is not supposed to trust a man to protect her.”
What are the positive and negative ramifications for the individuals and
the relationships from such discrepant messages?
ALTHOUGH FURTHER UNDERSTANDING OF CONTEXTUAL
ISSUES IS NEEDED, IT IS ALREADY CLEAR THAT
FUNDAMENTAL CHANGES IN PEOPLE’S LIVES WILL BE
REQUIRED TO ADDRESS HIV RISK
High rates of homelessness, unemployment, and general economic
disadvantage have been found in studies of drug users and their sexual
partners at risk for HIV infection (Brown and Beschner 1993), although
the severity of these social and environmental problems differs with
geographic region. Dr. Gibson (this volume) and others (Malow et al.
1992) noted an association of depression and anxiety with HIV risk. It is
likely that social and intrapersonal factors not only affect risk behaviors,
but that they interact to determine risk.
Because of this interaction, and without denying the enormous impact of
social problems, changes in social environments alone are not likely to
reduce all risk. For example, poverty is associated with HIV risk.
However, Dr. Robles noted that in Puerto Rico, women may have
resources such as an apartment and income, yet repeatedly become
involved with one IDU boyfriend after another. Clearly, lack of
economic empowerment, which the women have at levels necessary for
self-sustainment, does not explain the self-defeating pattern.
Companionship, prestige, and other psychological factors need to be
examined, and fundamental changes in people’s cognitions, attitudes, and
feelings about themselves may be required for long-term risk reduction.
Research must take into account the interaction of social and other
environmental factors and psychological factors in the development of
risk behaviors, and fundamental changes in both spheres will be needed
to sustain health protective behaviors.
244
REFERENCES
Betancourt, H., and Lopez, S.R. The study of culture, ethnicity, and race
in American psychology. Am Psychol 48:629-637, 1993.
Brown, B.S., and Beschner, G.M., eds. Handbook on Risk of AIDS:
Injection Drug Users and Sexual Partners. Westport, Connecticut:
Greenwood, 1993.
Critchfield, T.S. Signal-detection properties of verbal self-reports. J Exp
Anal Behav 60:495-514, 1993.
Malow, R. M.; Corrigan, S.A.; Pena, J.M.; Calkins, A.M.; and Bannister,
T.M. Mood and HIV risk behavior among drug dependent veterans.
Psychol Addict Behav 6:131-134, 1992.
Nisbett, R.E., and Wilson, T.D. Telling more than we can know: Verbal
reports on mental processes. Psychol Rev 84:231-259, 1977.
Zuckerman, M. Some dubious premises in research and theory on racial
differences: Scientific, social, and ethical issues. Am Psychol 45:
1297-1303, 1990.
AUTHORS
William C. Grace, Ph.D.
Research Psychologist
Robert J. Battjes, D.S.W.
Deputy Director
Zili Sloboda, Sc.D.
Acting Director
Division of Epidemiology and Prevention Research
Division of Clinical Research
National Institute on Drug Abuse
Parklawn Building, Room 10A-38
5600 Fishers Lane
Rockville, MD 20857
245
National
Institute on
While limited supplies last, single copies of the monographs may be obtained free of
charge from the National Clearinghouse for Alcohol and Drug Information (NCADI).
Please also contact NCADI for information about availability of coming issues and
other publications of the National Institute on Drug Abuse relevant to drug abuse
research.
Additional copies may be purchased from the U.S. Government Printing Office
(GPO) and/or the National Technical Information Service (NTIS) as indicated. NTIS
prices are for paper copy; add $3.00 handling charge for each order. Microfiche
copies are also available from NTIS. Prices from either source are subject to
change.
Addresses are:
NCADI
National Clearinghouse for Alcohol and Drug Information
P.O. Box 2345
Rockville, MD 20852
(301) 468-2600
(800) 729-6686
GPO
Superintendent of Documents
U.S. Government Printing Office
P.O. Box 371954
Pittsburgh, PA 15220-7954
(202) 738-3238
FAX (202) 512-2233
NTIS
National Technical Information Service
U.S. Department of Commerce
Springfield, VA 22161
(703) 487-4650
For information on availability of NIDA Research Monographs from 1975-1993 and
those not listed, write to NIDA, Community and Professional Education Branch,
Room 10A-39, 5600 Fishers Lane, Rockville, MD 20857.
246
26 THE BEHAVIORAL ASPECTS OF SMOKING. Norman A. Krasnegor, Ph.D.,
ed. (Reprint from 1979 Surgeon General’s Report on Smoking and Health.)
NTIS PB #80-118755/AS (A09) $27.00
NCADI #M26
42 THE ANALYSIS OF CANNABINOIDS IN BIOLOGICAL FLUIDS. Richard L.
Hawks, Ph.D., ed.
NTIS PB #83-136044/AS (A07) $27.00
NCADI #M42
50 COCAINE: PHARMACOLOGY, EFFECTS, AND TREATMENT OF ABUSE.
John Grabowski, Ph.D., ed.
NTIS PB #85-150381/AS (A07) $27.00
NCADI #M50
52 TESTING DRUGS FOR PHYSICAL DEPENDENCE POTENTIAL AND ABUSE
LIABILITY. Joseph V. Brady, Ph.D., and Scott E. Lukas, Ph.D., eds.
NTIS PB #85-150373/AS (A08) $27.00
NCADI #M52
53 PHARMACOLOGICAL ADJUNCTS IN SMOKING CESSATION. John
Grabowski, Ph.D., and Sharon M. Hall, Ph.D., eds.
NTIS PB #89-123186/AS (A07) $27.00
NCADI #M53
54 MECHANISMS OF TOLERANCE AND DEPENDENCE. Charles Wm. Sharp,
Ph.D., ed.
NTIS PB #89-103279/AS (A19) $52.00
NCADI #M54
56 ETIOLOGY OF DRUG ABUSE: IMPLICATIONS FOR PREVENTION. Coryl
LaRue Jones, Ph.D., and Robert J. Battjes, D.S.W., eds.
NTIS PB #89-123160/AS (A13) $36.50
NCADI #M56
61 COCAINE USE IN AMERICA: EPIDEMIOLOGIC AND CLINICAL
PERSPECTIVES. Nicholas J. Kozel, M.S., and Edgar H. Adams, M.S., eds.
NTIS PB #89-131866/AS (A11) $36.50
NCADI #M61
247
62 NEUROSCIENCE METHODS IN DRUG ABUSE RESEARCH. Roger M.
Brown, Ph.D., and David P. Friedman, Ph.D., eds.
NTIS PB #89-130660/AS (A08) $27.00
NCADI #M62
63 PREVENTION RESEARCH: DETERRING DRUG ABUSE AMONG CHILDREN
AND ADOLESCENTS. Catherine S. Bell, M.S., and Robert J. Battjes, D.S.W., eds.
NTIS PB #89-103287/AS (A11) $36.50
NCADI #M63
64 PHENCYCLIDINE: AN UPDATE. Doris H. Clouet, Ph.D., ed.
NTIS PB #89-131858/AS (A12) $36.50
NCADI #M64
65 WOMEN AND DRUGS: A NEW ERA FOR RESEARCH. Barbara A. Ray,
Ph.D., and Monique C. Braude, Ph.D., eds.
NTIS PB #89-130637/AS (A06) $27.00
NCADI #M65
69 OPIOID PEPTIDES: MEDICINAL CHEMISTRY. Rao S. Rapaka, Ph.D.; Gene
Barnett, Ph.D.; and Richard L. Hawks, Ph.D., eds.
NTIS PB #89-158422/AS (A17) $44.50
NCADI #M69
70 OPIOID PEPTIDES: MOLECULAR PHARMACOLOGY, BIOSYNTHESIS, AND
ANALYSIS. Rao S. Rapaka, Ph.D., and Richard L. Hawks, Ph.D., eds.
NTIS PB #89-158430/AS (A18) $52.00
NCADI #M70
72 RELAPSE AND RECOVERY IN DRUG ABUSE. Frank M. Tims, Ph.D., and
Carl G. Leukefeld, D.S.W., eds.
NTIS PB #89-151963/AS (A09) $36.50
NCADI #M72
74 NEUROBIOLOGY OF BEHAVIORAL CONTROL IN DRUG ABUSE. Stephen I.
Szara, M.D., D.Sc., ed.
NTIS PB #89-151989/AS (A07) $27.00
NCADI #M74
78 THE ROLE OF NEUROPLASTICITY IN THE RESPONSE TO DRUGS. David
P. Friedman, Ph.D., and Doris H. Clouet, Ph.D., eds.
NTIS PB #88-245683/AS (A10) $36.50
NCADI #M78
248
79 STRUCTURE-ACTIVITY RELATIONSHIPS OF THE CANNABINOIDS. Rao S.
Rapaka, Ph.D., and Alexandros Makriyannis, Ph.D., eds.
NTIS PB #89-109201/AS (A10) $36.50
NCADI #M79
80 NEEDLE SHARING AMONG INTRAVENOUS DRUG ABUSERS: NATIONAL
AND INTERNATIONAL PERSPECTIVES. Robert J. Battjes, D.S.W., and Roy W.
Pickens, Ph.D., eds.
NTIS PB #88-236138/AS (A09) $36.50
NCADI #M80
82 OPIOIDS IN THE HIPPOCAMPUS. Jacqueline F. McGinty, Ph.D., and David P.
Friedman, Ph.D. eds.
NTIS PB #88-245691/AS (A06) $27.00
NCADI #M82
83 HEALTH HAZARDS OF NITRITE INHALANTS. Harry W. Haverkos, M.D., and
John A. Dougherty, Ph.D., eds.
NTIS PB #89-125496/AS (A06) $27.00
NCADI #M83
84 LEARNING FACTORS IN SUBSTANCE ABUSE. Barbara A. Ray, Ph.D., ed.
NTIS PB #89-125504/AS (A10) $36.50
NCADI #M84
85 EPIDEMIOLOGY OF INHALANT ABUSE: AN UPDATE. Raquel A. Crider,
Ph.D., and Beatrice A. Rouse, Ph.D., eds.
NTIS PB #89-123178/AS (A10) $36.50
NCADI #M85
87 OPIOID PEPTIDES: AN UPDATE. Rao S. Rapaka, Ph.D. and Bhola N.
Dhawan, M.D., eds.
NTIS PB #89-158430/AS (A11) $36.50
NCADI #M87
88 MECHANISMS OF COCAINE ABUSE AND TOXICITY. Doris H. Clouet, Ph.D.,
Khursheed Asghar, Ph.D., and Roger M. Brown, Ph.D., eds.
NTIS PB #89-125512/AS (A16) $44.50
NCADI #M88
249
89 BIOLOGICAL VULNERABILITY TO DRUG ABUSE. Roy W. Pickens, Ph.D.,
and Date S. Svikis, B.A., eds.
NTIS PB #89-125520/AS (A09) $27.00
NCADI #M89
92 TESTING FOR ABUSE LIABILITY OF DRUGS IN HUMANS. Marian W.
Fischman, Ph.D.; and Nancy K. Mello, Ph.D., eds.
NTIS PB #90-148933/AS (A17) $44.50
NCADI #M92
94 PHARMACOLOGY AND TOXICOLOGY OF AMPHETAMINE AND RELATED
DESIGNER DRUGS. Khursheed Asghar, Ph.D.; Errol De Souza, Ph.D., eds.
NTIS PB #90-148958/AS (A16) $44.50
NCADI #M94
95 PROBLEMS OF DRUG DEPENDENCE, 1989. PROCEEDINGS OF THE 51st
ANNUAL SCIENTIFIC MEETING. THE COMMITTEE ON PROBLEMS OF DRUG
DEPENDENCE, INC., Louis S. Harris, Ph.D., ed.
NTIS PB #90-237660/AS (A99) $67.00
NCADI #M95
96 DRUGS OF ABUSE: CHEMISTRY, PHARMACOLOGY, IMMUNOLOGY, AND
AIDS. Phuong Thi Kim Pham, Ph.D. and Kenner Rice, Ph.D., eds.
NTIS PB #90-237678/AS (A11) $36.50
NCADI #M96
97 NEUROBIOLOGY OF DRUG ABUSE: LEARNING AND MEMORY. Lynda
Erinoff, ed.
NTIS PB #90-237686/AS (A11) $36.50
NCADI #M97
98 THE COLLECTION AND INTERPRETATION OF DATA FROM HIDDEN
POPULATIONS. Elizabeth Y. Lambert, M.S., ed.
NTIS PB #90-237694/AS (A08) $27.00
NCADI #M98
99 RESEARCH FINDINGS ON SMOKING OF ABUSED SUBSTANCES. C. Nora
Chiang, Ph.D. and Richard L. Hawks, Ph.D., eds.
NTIS PB #91-141119 (A09) $27.00
NCADI #M99
250
100 DRUGS IN THE WORKPLACE: RESEARCH AND EVALUATION DATA. VOL
II. Steven W. Gust, Ph.D.; and J. Michael Walsh, Ph.D., eds.
GPO Stock #017-024-01458-3 $8.00
NCADI #M100
101 RESIDUAL EFFECTS OF ABUSED DRUGS ON BEHAVIOR. John W.
Spencer, Ph.D. and John J. Boren, Ph.D., eds.
NTIS PB #91-172858/AS (A09) $27.00
NCADI #M101
102 ANABOLIC STEROID ABUSE. Geraline C. Lin, Ph.D. and Lynda Erinoff,
Ph.D., eds.
NTIS PB #91-172866/AS (A11) $36.50
NCADI #M102
106 IMPROVING DRUG ABUSE TREATMENT. Roy W. Pickens, Ph.D.; Carl G.
Leukefeld, D.S.W.; and Charles R. Schuster, Ph.D., eds.
NTIS PB #92-105873 (A18) $50.00
NCADI #M106
107 DRUG ABUSE PREVENTION INTERVENTION RESEARCH:
METHODOLOGICAL ISSUES. Carl G. Leukefeld, D.S.W., and William J. Bukoski,
Ph.D., eds.
NTIS PB #92-160985 (A13) $36.50
NCADI #M107
108 CARDIOVASCULAR TOXICITY OF COCAINE: UNDERLYING MECHANISMS.
Pushpa V. Thadani, Ph.D., ed.
NTIS PB #92-106608 (A11) $36.50
NCADI #M108
109 LONGITUDINAL STUDIES OF HIV INFECTION IN INTRAVENOUS DRUG
USERS: METHODOLOGICAL ISSUES IN NATURAL HISTORY RESEARCH. Peter
Hartsock, Dr.P.H., and Sander G. Genser, M.D., M.P.H., eds.
NTIS PB #92-106616 (A08) $27.00
NCADI #M109
251
111 MOLECULAR APPROACHES TO DRUG ABUSE RESEARCH: VOLUME I.
Theresa N.H. Lee, Ph.D., ed.
NTIS PB #92-135743 (A10) $36.50
NCADI #M111
112 EMERGING TECHNOLOGIES AND NEW DIRECTIONS IN DRUG ABUSE
RESEARCH. Rao S. Rapaka, Ph.D.; Alexandros Makriyannis, Ph.D.; and Michael
J. Kuhar, Ph.D., eds.
NTIS PB #92-155449 (A15) $44.50
NCADI #M112
113 ECONOMIC COSTS, COST EFFECTIVENESS, FINANCING, AND
COMMUNITY-BASED DRUG TREATMENT. William S. Cartwright, Ph.D., and
James M. Kaple, Ph.D., eds.
NTIS PB #92-155795 (A10) $36.50
NCADI #M113
114 METHODOLOGICAL ISSUES IN CONTROLLED STUDIES ON EFFECTS OF
PRENATAL EXPOSURE TO DRUG ABUSE. M. Marlyne Kilbey, Ph.D., and
Khursheed Asghar, Ph.D., eds.
NTIS PB #92-146216 (A16) $44.50
NCADI #M114
115 METHAMPHETAMINE ABUSE: EPIDEMIOLOGIC ISSUES AND
IMPLICATIONS. Marissa A. Miller, D.V.M., M.P.H., and Nicholas J. Kozel, M.S.,
eds.
NTIS PB # 92-146224/II (A07) $27.00
NCADI #M115
116 DRUG DISCRIMINATION: APPLICATIONS TO DRUG ABUSE RESEARCH.
R.A. Glennon, Ph.D., T.U.C. Jarbe, Ph.D., and J. Frankenheim, Ph.D., eds.
NTIS PB # 94-169471 (A20) $52.00
NCADI #M116
117 METHODOLOGICAL ISSUES IN EPIDEMIOLOGY, PREVENTION, AND
TREATMENT RESEARCH ON DRUG-EXPOSED WOMEN AND THEIR
CHILDREN. M. M. Kilbey, Ph.D. and K. Asghar, Ph.D., eds.
GPO Stock #017-024-01472-9 $12.00
NTIS PB #93-102101/LL (A18) $52.00
NCADI #M117
252
118 DRUG ABUSE TREATMENT IN PRISONS AND JAILS. C.G. Leukefeld, D.S.W.
and F. M. Tims, Ph.D., eds.
GPO Stock #017-024-01473-7 $16.00
NTIS PB #93-102143/LL (A14) $44.50
NCADI #M118
120 BIOAVAILABILITY OF DRUGS TO THE BRAIN AND THE BLOOD-BRAIN
BARRIER. Jerry Frankenheim, Ph.D., and Roger M. Brown, Ph.D., eds.
GPO Stock #017-024-01481-8 $10.00
NTIS PB #92-214956/LL (A12) $36.50
NCADI #M120
121 BUPRENORPHINE: AN ALTERNATIVE TREATMENT FOR OPIOID
DEPENDENCE. Jack D. Blaine, Ph.D. ed.
GPO Stock #017-024-01482-6 $5.00
NTIS PB #93-129781/LL (A08) $27.00
NCADI #M121
123 ACUTE COCAINE INTOXICATION: CURRENT METHODS OF TREATMENT.
Heinz Sorer, Ph.D., ed.
GPO# 017-024-01501-6 $6.50
NTIS PB #94-115433/LL (A09) $27.00
NCADI #M123
124 NEUROBIOLOGICAL APPROACHES TO BRAIN-BEHAVIOR INTERACTION.
Roger M. Brown, Ph.D., and Joseph Fracella, Ph.D., eds.
GPO #017-024-01492-3 $9.00
NTIS PB #93-203834/LL (A12) $36.50
NCADI #M124
125 ACTIVATION OF IMMEDIATE EARLY GENES BY DRUGS OF ABUSE.
Reinhard Grzanna, Ph.D., and Roger M. Brown, Ph.D., eds.
GPO# 017-024-01503-2 $7.50
NTIS PB # 94-169489 (A12) $36.50
NCADI #M125
126 MOLECULAR APPROACHES TO DRUG ABUSE RESEARCH VOLUME II:
STRUCTURE, FUNCTION, AND EXPRESSION. Theresa N.H. Lee, Ph.D., eds.
NTIS PB # 94-169497 (A08) $27.00
NCADI #M126
253
127 PROGRESS AND ISSUES IN CASE MANAGEMENT. Rebecca Sager Ashery,
D.S.W., ed.
NTIS PB # 94-169505 (A18) $52.00
NCADI #M127
128 STATISTICAL ISSUES IN CLINICAL TRIALS FOR TREATMENT OF OPIATE
DEPENDENCE. Ram B. Jain, Ph.D., ed.
NTIS PB #93-203826/LL (A09) $27.00
NCADI #M128
129 INHALANT ABUSE: A VOLATILE RESEARCH AGENDA. Charles Wm. Sharp,
Ph.D., Fred Beauvais, Ph.D., and Richard Spence, Ph.D., eds.
GPO #017-024-01496-6 $12.00
NTIS PB #93-183119/LL (A15) $44.50
NCADI #M129
130 DRUG ABUSE AMONG MINORITY YOUTH: ADVANCES IN RESEARCH AND
METHODOLOGY. Mario De La Rosa, Ph.D., Juan-Luis Recio Adrados, Ph.D., eds.
GPO #017-024-01506-7 $14.00
NTIS PB # 94-169513 (A15) $44.50
NCADI #M130
131 IMPACT OF PRESCRIPTION DRUG DIVERSION CONTROL SYSTEMS ON
MEDICAL PRACTICE AND PATIENT CARE. James R. Cooper, Ph.D., Dorynne J.
Czechowicz, M.D., Stephen P. Molinari, J.D., R.Ph., and Robert C. Peterson, Ph.D.,
eds.
GPO #017-024-01505-9 $14.00
NTIS PB # 94-169521 (A15) $44.50
NCADI #M131
132 PROBLEMS OF DRUG DEPENDENCE, 1992: PROCEEDINGS OF THE 54TH
ANNUAL SCIENTIFIC MEETING OF THE COLLEGE ON PROBLEMS OF DRUG
DEPENDENCE. Louis Harris, Ph.D., ed.
GPO# 017-024-01502-4 $23.00
NTIS PB #94-115508/LL (A99)
NCADI #M132
133 SIGMA, PCP, AND NMDA RECEPTORS. Errol B. De Souza, Ph.D., Doris
Clouet, Ph.D., and Edythe D. London, Ph.D., eds.
NTIS PB # 94-169539 (A12) $36.50
NCADI #M133
254
134 MEDICATIONS DEVELOPMENT: DRUG DISCOVERY, DATABASES, AND
COMPUTER-AIDED DRUG DESIGN. Rao S. Rapaka, Ph.D and Richard L. Hawks,
Ph.D., eds.
GPO #017-024-01511-3 $11.00
NTIS PB # 94-169547 (A14) $44.50
NCADI #M134
135 COCAINE TREATMENT: RESEARCH AND CLINICAL PERSPECTIVES.
Frank M. Tims, Ph.D. and Carl G. Leukefeld, D.S.W., eds.
GPO #017-024-01520-2 $11.00
NTIS PB # 94-169554 (A13) $36.50
NCADI #M135
136 ASSESSING NEUROTOXICITY OF DRUGS OF ABUSE. Lynda Erinoff, Ph.D.,
ed.
GPO #017-024-01518-1 $11.00
NTIS PB # 94-169562 (A13) $36.50
NCADI #M136
137 BEHAVIORAL TREATMENTS FOR DRUG ABUSE AND DEPENDENCE. Lisa
Simon Onken, Ph.D., Jack D. Blaine, M.D., and John J. Boren, Ph.D., eds.
GPO #017-024-01519-9 $13.00
NTIS PB # 94-169570 (A15) $44.50
NCADI #M137
138 IMAGING TECHNIQUES IN MEDICATIONS DEVELOPMENT: CLINICAL AND
PRECLINICAL ASPECTS. Heinz Sorer, Ph.D. and Rao S. Rapaka, Ph.D., eds.
NCADI #M138
139 SCIENTIFIC METHODS FOR PREVENTION INTERVENTION RESEARCH.
Arturo Cazares, M.D., M.P.H. and Lula A. Beatty, Ph.D., eds.
NCADI #M139
140 PROBLEMS OF DRUG DEPENDENCE, 1993: PROCEEDINGS OF THE 55TH
ANNUAL SCIENTIFIC MEETING, THE COLLEGE ON PROBLEMS OF DRUG
DEPENDENCE. VOLUME I: PLENARY SESSION SYMPOSIA AND ANNUAL
REPORTS. Louis S. Harris, Ph.D., ed.
NCADI #M140
141 PROBLEMS OF DRUG DEPENDENCE, 1993: PROCEEDINGS OF THE 55TH
ANNUAL SCIENTIFIC MEETING, THE COLLEGE ON PROBLEMS OF DRUG
DEPENDENCE. VOLUME II: ABSTRACTS. Louis S. Harris, Ph.D., ed.
NCADI #M141
255
U.S. GOVERNMENT PRINTING OFFICE: 1994
300-973/00002
NIH Publication No. 94-3750
Printed 1994