CBT for High Anxiety Sensitivity: Alcohol Outcomes

Addictive Behaviors 46 (2015) 19–24 Contents lists available at ScienceDirect Addictive Behaviors Short Communication CBT for high anxiety sensitivity: Alcohol outcomes Janine V. Olthuis a,⁎, Margo C. Watt b,a, Sean P. Mackinnon c,a, Sherry H. Stewart c,a a b c Department of Psychology and Neuroscience, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada Department of Psychology, Saint Francis Xavier University, Antigonish, Nova Scotia B2G 2W5, Canada Department of Psychiatry, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada H I G H L I G H T S • • • • • Cognitive behavioral therapy (vs. a waiting list) reduced anxiety sensitivity. Treatment also specifically reduced drinking to cope with anxiety. Physical and intrapersonal alcohol-related problems were also reduced. Change in anxiety sensitivity mediated reductions in drinking to cope with anxiety. Reductions in drinking to cope with anxiety mediated changes in alcohol problems. a r t i c l e i n f o Available online 26 February 2015 Keywords: Anxiety sensitivity Alcohol use CBT Drinking motives Problem drinking a b s t r a c t Introduction: High anxiety sensitivity (AS) has been associated with greater alcohol consumption and alcoholrelated problems as well as greater sensitivity to the anxiety-reducing effects of alcohol and greater risky negative reinforcement motives for drinking. The present study reported on the alcohol-related outcomes of a telephone-delivered cognitive behavioral treatment (CBT) designed to reduce high AS. Methods: Eighty individuals with high AS (M age = 36 years; 79% women; 76% Caucasian) seeking treatment for their AS-related concerns participated in the study and were randomly assigned to an eight week telephone CBT program or a waiting list control. Participants completed measures of drinking motives and problem drinking at pre- and post-treatment. Results: Multilevel modeling showed that the treatment was successful in reducing AS. The treatment also resulted in specific reductions in drinking to cope with anxiety motives as well as physical alcohol-related problems. Mediated moderation analyses showed treatment-related changes in AS mediated changes in drinking to cope with anxiety motives. Changes in drinking to cope with anxiety motives mediated changes in physical alcoholrelated problems. Conclusions: Results of the present study suggest that an AS-targeted intervention may have implications for reducing risky alcohol use cognitions and behaviors. Further research is needed in a sample of problem drinkers. © 2015 Elsevier Ltd. All rights reserved. 1. Introduction Anxiety sensitivity (AS) is an enduring fear of arousal-related body sensations (e.g., rapid heart rate) arising from the tendency to interpret these sensations catastrophically (Reiss, 1991; Reiss & McNally, 1985). Research has linked high AS to problematic alcohol use (Norton, 2001; Stewart, Samoluk, & MacDonald, 1999). Individuals high (vs. low) in AS report more frequently drinking to excess (Stewart, Peterson, & Pihl, 1995), endorse more alcohol problems (Watt, Stewart, Birch, & ⁎ Corresponding author at: Department of Psychology and Neuroscience, Dalhousie University, 1355 Oxford Street, PO Box 15000, Halifax, Nova Scotia B3H 4R2, Canada. Tel.: +1 902 494 3793; fax: +1 902 494 6585. E-mail address: janine.olthuis@dal.ca (J.V. Olthuis). http://dx.doi.org/10.1016/j.addbeh.2015.02.018 0306-4603/© 2015 Elsevier Ltd. All rights reserved. Bernier, 2006), and develop alcohol disorders at higher rates (Schmidt, Buckner, & Keough, 2007). Motivational theories of alcohol use propose that specific personality characteristics (e.g., AS) are associated with differential activation of brain motivation systems and susceptibility to certain drug reinforcement properties (Conrod, Pihl, Stewart, & Dongier, 2000). High AS individuals may thus be motivated to use alcohol to reduce, control, and/or avoid their fear of aversive arousal sensations and the sensations themselves (McNally, 1996; Stewart et al., 1999). Notably, high (vs. low) AS individuals are more sensitive to alcohol's anxiolytic effects (MacDonald, Baker, Stewart, & Skinner, 2000; Zack, Poulos, Aramakis, Khamba, & MacLeod, 2007). In addition, AS is related to negative reinforcement motives associated with problem drinking (Martens et al., 2008), including coping and conformity motives (DeMartini & Carey, 2011; Stewart, Zvolensky, & Eifert, 2001). Recent evidence shows the AS-alcohol 20 J.V. Olthuis et al. / Addictive Behaviors 46 (2015) 19–24 problems link to be mediated through anxiety symptoms and in turn coping motives (Allan, Albanese, Norr, Zvolensky, & Schmidt, 2015). Research has also shown that high AS can interfere with substance use treatment, by increasing risk of dropout (Lejuez et al., 2008) and/ or relapse (Zvolensky et al., 2007). Without addressing high AS, individuals' fear of somatic sensations – a contributor to their substance use – may persist, serving as a diathesis for relapse to cope with this fear (Otto, Safren, & Pollack, 2004) or with aversive withdrawal sensations (Johnson, Stewart, Steeves, & Zvolensky, 2012). Recently, researchers have incorporated AS reduction into substance use interventions (Conrod, Stewart, Comeau, & Maclean, 2006; Tull, Schulzinger, Schmidt, Zvolensky, & Lejuez, 2007). This study investigated the alcohol use outcomes of an AS reduction intervention. Comprehensive details of the RCT testing this intervention are described elsewhere (Olthuis, Watt, Mackinnon, & Stewart, 2014). A brief form of this intervention (Watt et al., 2006) showed decreases in problem drinking and conformity-motivated drinking in high AS university women. We hypothesized that our AS-focused intervention would reduce negative reinforcement drinking motives and problem drinking. 2. Material and methods 2.1. Participants We used newspaper and flyer advertisements to recruit participants from the community who were high in AS and were interested in seeking treatment for AS-related concerns. Participants were not recruited on the basis of drinking-related behaviors. To be eligible, individuals had to: be ≥18 years, have access to a telephone, and score ≥23 on the Anxiety Sensitivity Index—3 (ASI-3; Taylor et al., 2007), which is one standard deviation above the population mean (M = 12.8, SD = 10.6; Reiss, Peterson, Taylor, Schmidt, & Weems, 2008). Exclusion criteria were: contraindications to exercise, current psychotherapy, new pharmacotherapy in the last three months, or current psychosis or suicidal ideation. Overall, 182 individuals expressed interest, 109 were eligible, and 80 consented (M age = 36.3, SD = 11.3; 79% women; 76% Caucasian; 38% concurrent pharmacotherapy) and were randomized (n = 40 to the intervention and n = 40 to the control). Structured Clinical Interviews for DSM-IV-TR (SCID; First, Spitzer, Gibbon, & Williams, 1997) conducted at pre-treatment revealed that 67.5% of participants qualified for a primary diagnosis of an Anxiety Disorder, 10% a primary Mood Disorder, 5% a primary Adjustment Disorder, 1.3% a primary diagnosis of Hypochondriasis, and 16.3% did not qualify for a current DSM-IV Axis I diagnosis. Thirty-four participants had at least one comorbid diagnosis. Only 3.8% of participants qualified for a current Alcohol Use Disorder (AUD) while 1.3% qualified for an AUD in partial remission and 21.3% qualified for an AUD in full remission. Participant flow details are presented in Fig. 1 (the full PRISMA diagram is published in Olthuis et al., 2014). 2.2. Procedure Participants were randomized using an online number generator. Measures were completed at baseline, eight and twelve weeks. 2.2.1. Intervention (CBT) Participants received eight weeks of telephone-delivered cognitive behavior therapy (CBT) for high AS based on a brief empiricallyvalidated intervention (Watt & Stewart, 2008; Watt et al., 2006). A therapist provided weekly 50-minute telephone sessions encompassing psychoeducation, cognitive restructuring, interoceptive exposure via physical exercise, and prevention of relapse of high AS. Participants continued interoceptive exposure four weeks after telephone therapy concluded. All aspects of the CBT program were aimed at reducing high AS. Alcohol use was only addressed briefly, in terms of: (1) psychoeducation around the link between high AS and alcohol use problems, and (2) a brief discussion about the use of alcohol as a maladaptive coping strategy for AS-related concerns. 2.2.2. Waiting list control (WLC) Participants received no treatment. After the waiting list period had concluded, participants in the WLC were offered the CBT intervention, if interested. 2.3. Materials 2.3.1. Anxiety Sensitivity Index — 3 (ASI-3; Taylor et al., 2007; derived from Peterson & Reiss, 1992) The ASI-3 is a self-report measure of AS. Participants indicate the extent to which they agree or disagree with 18 items (0 = very little to 4 = very much). The ASI-3 has good internal reliability and criterion validity (Taylor et al., 2007). 2.3.2. Modified Drinking Motives Questionnaire — Revised (MDMQ-R; Grant, Stewart, O'Connor, Blackwell, & Conrod, 2007) The 28-item self-report MDMQ-R measures social, enhancement, coping-with-anxiety, coping-with-depression, and conformity drinking motives. Participants indicate how often they drink for each reason (1 = never/almost never to 5 = almost always/always). The MDMQ-R has good test–retest reliability and concurrent and predictive validity (Grant et al., 2007). 2.3.3. Short Inventory of Problems — Recent (SIP-R; Miller, Tonigan, & Longabaugh, 1995) The SIP-R was used to assess problem drinking. Participants indicated how often they experienced each of 15 alcohol problems in the past month1 across five subscales: physical, interpersonal, intrapersonal, impulse control, and social responsibility problems. SIP-R subscale scores have modest internal consistency and test–retest reliability (Feinn, Tennen, & Kranzler, 2003; Miller et al., 1995). 2.4. Data analytic plan Multilevel modeling with HLM 7.0 software (Scientific Software International, Inc., Lincolnwood, IL) was used to accommodate unequal time points between assessments (Gueorguieva & Krystal, 2004), use a maximum likelihood approach for missing data (Graham, 2009), and account for repeated measures (Garson, 2013). A two-level model was specified with repeated measures (level 1) nested within people (level 2). We estimated separate models for each outcome variable using restricted maximum likelihood estimation. Time was a predictor at level 1 and was coded as 0 (pre-treatment), 2 (8 weeks), and 3 (12 weeks) to capture the unequal time between assessments. We tested a linear growth curve with random slopes and random intercepts and quadratic growth curves using fixed slopes and random intercepts with orthogonal contrast coding.2 Treatment group was a predictor at level 2 (WLC = 0; CBT = 1). We also tested a time ∗ group interaction by including a cross-level effect between time at level 1 and group at level 2. The equations for analyses are provided elsewhere (Olthuis et al., 2014). We probed significant cross-level interactions using simple slopes (Preacher, Curran, & Bauer, 2006). We used the formula for Cohen's d adapted for use in growth-curve models as a measure of effect size of the pre-post change (Feingold, 2009). To test mediated moderation, we calculated significance of indirect effects using a Monte Carlo method with 20,000 resamples (Preacher & Selig, 2012). We used percent 1 Typical time reference is 3 months. At least four measurement occasions would be required for random slopes in a quadratic growth curve (Mroczek & Griffin, 2007). 2 J.V. Olthuis et al. / Addictive Behaviors 46 (2015) 19–24 21 Fig. 1. Abbreviated PRISMA diagram of participant flow through the trial. Adapted from Olthuis et al. (2014). mediation (PM; ratio of the indirect effect to the direct effect, calculated by ab / [ab + |c′|]) as an effect size for the indirect effect. 3. Results ANOVAs and chi-squares showed that the groups did not differ on sex, age, or medication use (Olthuis et al., 2014). We log10 transformed skewed scores (N ± 1.00). Mean ASI-3 scores at pre-treatment (Table 1) were as high as the levels found in panic disorder (M = 32.6) and social phobia (M = 31.4; Reiss et al., 2008). Mean coping-with-anxiety motives were somewhat elevated and mean enhancement motives somewhat reduced relative to undergraduate norms (Grant et al., 2007). While mindful of the discrepant time parameters used, SIP-R scores were lower than those among treatment-seeking alcoholics (Feinn et al., 2003) but similar to undergraduate drinkers (Stahlbrandt, Johnsson, & Berglund, 2007). 3.1. Multilevel models 3.1.1. Anxiety sensitivity When predicting ASI-3 scores, there was a significant quadratic time ∗ group interaction, with a medium-large effect size (Table 2). The WLC showed a small linear reduction in AS over time while the CBT group had a quadratic change consisting of a sharp linear reduction in AS from pre-treatment to eight weeks that was maintained to 12 weeks. 3.1.2. Drinking motives When predicting coping-with-anxiety motives, there was a significant linear time ∗ group interaction, with a small-medium effect size (Table 2). The CBT group's coping-with-anxiety motives declined in a linear fashion over time while the WLC group's did not change. No significant interactions were found for the remaining motives. 3.1.3. Problem drinking We found a significant linear time ∗ group interaction of moderate effect size for physical alcohol-related problems3 and a marginally significant linear time ∗ group interaction for intrapersonal alcoholrelated problems (Table 2). For both, the CBT, but not WLC, group's problems decreased over time. 3.2. Mediation analysis We investigated whether the time ∗ group interaction predicted reduced AS, which in turn predicted decreases in alcohol variables. Because there should ideally be direct effects when testing mediation, we only tested AS as a mediator for coping-anxiety motives and physical alcohol-related problems. The indirect effect of the linear time ∗ group interaction on copinganxiety motives through AS was significant, 95% CI [− 0.20, − 0.01], PM = .35. That is, the treatment reduced AS, which induced reductions in coping-anxiety motives. However, the indirect effect of the linear time ∗ group interaction on physical alcohol-related problems through AS was not significant, 95% CI [− 0.02, 0.005], PM = .15. Exploratory analyses examined whether changes in coping-anxiety motives mediated the linear time ∗ group interaction for physical alcohol-related problems. We found a significant indirect effect, 95% CI [− 0.03, − 0.01], PM = .29, suggesting therapy-induced reductions in alcohol-related problems were mediated by reductions in coping-anxiety motives. 3 At pre-treatment, the CBT group had a significantly higher SIP-R physical score than the WLC. We conducted a supplemental analysis after removing outliers from the CBT group. (We removed the three highest scores at pre-treatment, which were from two participants who did not complete eight or 12 week assessment measures and one participant who reported alcohol use problems in the past month but no alcohol use.) After removing these outliers, pre-treatment scores were no longer different between groups and the time ∗ group interaction for SIP-R physical remained significant. 22 J.V. Olthuis et al. / Addictive Behaviors 46 (2015) 19–24 Table 1 Means and standard deviations for study variables. Measure Group Pre M (SD) 1. ASI-3 WLC PT 2. MDMQ-R social WLC PT 3. MDMQ-R WLC enhancement PT 4. MDMQ-R WLC coping-anxiety PT 5. MDMQ-R WLC coping-depression PT 6. MDMQ-R WLC conformity PT 7. SIP-R physical WLC PT 8. SIP-R interpersonal WLC PT 9. SIP-R intrapersonal WLC PT 10. SIP-R impulse WLC control PT 11. SIP-R social WLC responsibility PT 8 week M (SD) 12 week M (SD) r 36.83 (13.67) 31.31 (13.71) 28.56 (13.16) .74 39.93 (13.50) 23.57 (13.44) 24.54 (14.71) 2.36 (0.90) 2.45 (1.05) 2.45 (1.13) .73 2.64 (1.16) 2.56 (1.07) 2.37 (1.04) 2.09 (1.02) 1.88 (0.92) 1.82 (1.05) .84 2.19 (1.07) 2.08 (1.04) 1.93 (1.08) 1.94 (0.95) 1.75 (0.88) 1.86 (1.07) .84 2.37 (1.14) 2.08 (1.03) 1.98 (1.16) 1.35 (0.48) 1.31 (0.52) 1.23 (0.51) .70 1.83 (1.03) 1.60 (0.81) 1.60 (0.99) 1.25 (0.34) 1.30 (0.52) 1.19 (0.32) .84 1.39 (0.72) 1.18 (0.35) 1.23 (0.55) 0.36 (0.91) 0.48 (0.89) 0.41 (0.98) .61 1.23 (1.95) 0.87 (1.52) 0.52 (1.42) 0.21 (0.66) 0.16 (0.58) 0.22 (0.66) .57 0.68 (1.80) 0.39 (1.67) 0.32 (1.25) 0.74 (1.33) 0.45 (0.93) 0.78 (1.36) .76 1.50 (2.55) 0.96 (1.61) 0.72 (1.43) 0.54 (1.02) 0.26 (0.63) 0.38 (0.75) .34 0.93 (1.72) 0.39 (0.99) 0.44 (0.87) 0.46 (1.00) 0.35 (0.84) 0.28 (0.81) .46 0.80 (1.95) 0.74 (1.48) 0.52 (1.48) Note. r = 12 week test–retest reliability in the WLC. WLC = waiting list control; PT = phone therapy; MDMQ-R = Modified Drinking Motives Questionnaire—Revised; SIP-R = Short Inventory of Problems — Recent. All means above are raw means (i.e., transformed data is not described above). 4. Discussion This study investigated the alcohol outcomes of an AS-targeted CBT intervention. Findings showed a significant time ∗ group interaction when predicting high AS suggesting that the intervention (vs. WLC) significantly reduced high AS. Results also revealed a significant linear time ∗ group interaction in predicting coping-with-anxiety drinking motives. Simple slopes showed a significant linear reduction in coping-with-anxiety motives in the CBT group but not in the WLC. This unique interaction for coping-with-anxiety motives reflects the intervention's specificity. Changes in coping-with-anxiety motives may have stemmed directly from treatment-related reductions in AS, as treatment-related AS reductions significantly mediated changes in coping-with-anxiety motives. Reducing coping-with-anxiety motives is important, as they have been linked to alcohol-related problems (Grant et al., 2007; Martens et al., 2008). No significant interactions were found for the remaining drinking motives. Though a brief form of this intervention led to reduced conformity motives (Watt et al., 2006), this may not have emerged here due to sample differences (college students vs. community sample) or the use of measures with different factor structures (DMQ; Cooper, 1994 vs. MDMQ-R). A significant linear time ∗ group interaction emerged for physical alcohol-related problems as did a marginal linear time ∗ group interaction for intrapersonal problems. For both, simple slopes revealed a linear reduction for the CBT, but not WLC, group. These two subscales in particular may reflect AS physical and psychological concerns, respectively. Unexpectedly, given the interpersonal nature of AS social concerns, no interactions emerged for interpersonal problems. However, because AS is associated with social phobia (Norton, Cox, Hewitt, & McLeod, 1997), high AS individuals might avoid situations in which such social alcohol-related problems would arise. No interaction emerged for impulsivity problems, unsurprising, given that high AS is not associated with externalizing problems (Olatunji & Wolitzky-Taylor, 2009), nor for social responsibility problems, which may not be a concern for high AS individuals. Overall, the reductions in physical and intrapersonal alcohol-related problems mirror previous work showing a reduction in problem drinking after AS-targeted interventions (Watt et al., 2006). While we did not find a mediating role of AS for alcohol-related problems, coping-withanxiety motives in turn mediated the reduction in physical alcoholrelated problems. This aligns with prior research linking drinking to cope and alcohol-related problems (Kuntsche, Stewart, & Cooper, 2008; Martens et al., 2008). Overall, the results suggest “chained mediation” (Taylor, MacKinnon, & Tein, 2008) where the intervention led to reductions in AS which in turn led to reductions in coping-with-anxiety motives which ultimately led to reductions in problem drinking. Results should be considered in light of several limitations. First, we relied on self-report measures vulnerable to inaccurate reporting. Second, participants were selected for high AS as opposed to drinking-related behaviors, limiting variability in our sample; the intervention should be investigated among high AS problem drinkers. Third, our sample size (N = 80) and missing data may have limited our ability to detect small effects. Fourth, we compared the intervention to a waiting list rather than another active intervention; while this was a suitable starting point for the present study, it will be informative to compare the intervention to other active interventions in the future. Fifth, given the skew and poor reliability of some of the SIP-R subscales (Table 1), SIP-R results should be interpreted with caution. Finally, it is important to note that the treatment-related changes in drinking motives and alcohol-related Table 2 Hierarchical linear modeling results. Time × group interaction 11ASI-3 MDMQ-R social MDMQ-R enhance MDMQ-R cop-anx MDMQ-R cop-depa MDMQ-R conforma SIP-R physicala SIP-R intera SIP-R intraa SIP-R impulsea SIP-R sociala Intervention group only Control group only Lin B Lin tdf Quad B Quad tdf Lin B Lin tdf Quad B Quad tdf Lin B Lin tdf Quad B Quad tdf dGMA-raw −4.35 −0.15 −0.03 −0.15 −0.01 −0.01 −0.05 −0.03 −0.03 −0.01 0.00 −2.98109⁎ −1.7676 −0.3876 −2.2176⁎ 2.57 – – – – – – – – – – 2.88109⁎⁎⁎ – – – – – – – – – – −7.82 −0.08 −0.08 −0.17 −0.02 −0.01 −0.04 −0.03 −0.04 −0.03 −0.01 −5.8045⁎⁎⁎ −1.3847 −1.5447 −3.2247⁎⁎ 2.89 – – – – – – – – – – 3.4745⁎ – – – – – – – – – – −3.39 0.06 −0.07 −0.02 −0.01 0.00 0.00 0.00 0.00 −0.02 −0.01 −4.5164⁎⁎⁎ 1.2067 −2.1467⁎ −0.7067 −2.1267⁎ 0.29 – – – – – – – – – – 0.6564 – – – – – – – – – – 0.77 −0.43 −0.09 −0.43 −0.27 −0.38 −0.64 −0.48 −0.34 −0.14 0.00 −0.8376 −1.1576 −2.6377⁎ −1.5477 −1.7577+ −0.6177 0.0977 −1.9447+ −1.9547+ −2.8447⁎⁎ −1.7647 −2.2647⁎ −1.6247 −0.8047 −0.8867 0.3366 −0.1966 −0.3766 −1.2566 −1.3466 Note. When quadratic models are not reported, the quadratic slope with fixed slopes and random intercepts was not significant and so the linear model with random slopes and random intercepts is reported instead. SIP-R = Short Inventory of Problems — Recent. dGMA-raw = Cohen's d adapted for use with growth curve models. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001. + p ≤ .08. a Scores have been log10 transformed to address skew. J.V. Olthuis et al. / Addictive Behaviors 46 (2015) 19–24 problems noted in the present study are not directly linked to changes in absolute levels of drinking. Future research is needed to investigate the impact of the intervention on drinking quantity and frequency, particularly among a sample with patterns of problematic alcohol use. Nevertheless, our findings have important clinical implications. Results presented here and elsewhere (Olthuis et al., 2014) show that the present intervention reduced AS, coping-with-anxiety drinking motives, physical and intrapersonal alcohol-related problems, as well as panic, social anxiety, and posttraumatic stress symptoms. Taken together, these findings suggest that an AS-targeted intervention may address an underlying vulnerability contributing to comorbid anxiety and alcohol problems. Role of funding sources Funding for this study was provided by the Marvin Burke Endowment, Faculty of Medicine, Dalhousie University. Dr. Olthuis' work was supported by a Canadian Institutes of Health Research Vanier Canada Graduate Scholarship. The funders had no role in the study design, collection, analysis, or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication. Contributors Drs. Olthuis, Watt, and Stewart designed the study, wrote the protocol, and carried out the study. Dr. Mackinnon conducted the statistical analyses. All authors contributed to and have approved the final manuscript. Conflict of interest Drs. Olthuis and Mackinnon declare that they have no conflicts of interest. Drs. Watt and Stewart are authors of the published manual on which the study intervention is based. Acknowledgments The authors wish to thank the psychologists and clinical psychology graduate students who conducted assessment and treatment for the present study and the research assistants who helped with data collection. References Allan, N. P., Albanese, B. J., Norr, A. M., Zvolensky, M. J., & Schmidt, N. B. (2015). Effects of anxiety sensitivity on alcohol problems: Evaluating chained mediation through generalized anxiety, depression, and drinking motives. Addiction, 110, 260–268. http:// dx.doi.org/10.1111/add.12739. Conrod, P. J., Pihl, R. O., Stewart, S. H., & Dongier, M. (2000). Validation of a system of classifying female substance abusers based on personality and motivational risk factors for substance abuse. Psychology of Addictive Behaviors, 14, 243–256. http://dx.doi. org/10.1037//0893-164X.14.3.243. Cooper, M.L. (1994). Motivations for alcohol use among adolescents: Development and validation of a four-factor model. Psychological Assessment, 6, 117–128. http://dx. doi.org/10.1037/1040-3590.6.2.117. DeMartini, K.S., & Carey, K.B. (2011). The role of anxiety sensitivity and drinking motives in predicting alcohol use: A critical review. Clinical Psychology Review, 31, 169–177. http://dx.doi.org/10.1016/j.cpr.2010.10.001. Feingold, A. (2009). Effect sizes for growth-modeling analysis for controlled clinical trials in the same metric as for classical analysis. Psychological Methods, 14, 43–53. http:// dx.doi.org/10.1037/a0014699. Feinn, R., Tennen, H., & Kranzler, H.R. (2003). Psychometric properties of the short index of problems as a measure of recent alcohol-related problems. Alcoholism: Clinical and Experimental Research, 27, 1436–1441. http://dx.doi.org/10.1097/01.ALC.0000087582. 44674.AF. First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W. (1997). Structured clinical interview for DSM-IV axis I disorders: Clinician version. Arlington, VA: American Psychiatric Publishing, Inc. Garson, G.D. (2013). Fundamentals of hierarchical linear and multilevel modeling. In G.D. Garson (Ed.), Hierarchical linear modeling: Guide and applications (pp. 3–26). Thousand Oaks, CA: Sage Publications, Inc. Graham, J.W. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60, 549–576. http://dx.doi.org/10.1146/annurev.psych.58. 110405.085530. Grant, V.V., Stewart, S.H., O'Connor, R.M., Blackwell, E., & Conrod, P.J. (2007). Psychometric evaluation of the five-factor Modified Drinking Motives Questionnaire — Revised in undergraduates. Addictive Behaviors, 32, 2611–2632. http://dx.doi.org/10.1016/j.addbeh. 2007.07.004. Gueorguieva, R., & Krystal, J.H. (2004). Move over ANOVA: Progress in analyzing repeated-measures data and its reflection in paper published in the archives of general psychiatry. Archives of General Psychiatry, 61, 310–317. http://dx.doi.org/10.1001/ archpsyc.61.3.310. Johnson, K.A., Stewart, S.H., Steeves, D., & Zvolensky, M.J. (2012). Prospective evaluation of the effects of anxiety sensitivity and state anxiety in predicting acute nicotine withdrawal symptoms during smoking cessation. Psychology of Addictive Behaviors, 26, 289–297. http://dx.doi.org/10.1037/a0024133. 23 Kuntsche, E., Stewart, S.H., & Cooper, M.L. (2008). How stable is the motive-alcohol use link? A cross-national validation of the Drinking Motives Questionnaire Revised among adolescents from Switzerland, Canada, and the United States. Journal of Studies on Alcohol and Drugs, 69, 388–396. Lejuez, C.W., Zvolensky, M.J., Daughters, S.B., Bornovalova, M.A., Paulson, A., Tull, M.T., et al. (2008). Anxiety sensitivity: A unique predictor of dropout among inner-city heroin and crack/cocaine users in residential substance use treatment. Behaviour Research and Therapy, 46, 811–818. http://dx.doi.org/10.1016/j.brat.2008.03.010. MacDonald, A.B., Baker, J.M., Stewart, S.H., & Skinner, M. (2000). The effects of alcohol on the response to hyperventilation of participants high and low in anxiety sensitivity. Alcoholism: Clinical and Experimental Research, 24, 1656–1665. http://dx.doi.org/10. 1111/j.1530-0277.2000.tb01967.x. Martens, M.P., Neighbors, C., Lewis, M.A., Lee, C.M., Oster-Aaland, L., & Larimer, M.E. (2008). The roles of negative affect and coping motives in the relationship between alcohol use and alcohol-related problems among college students. Journal of Studies on Alcohol and Drugs, 69, 412–419. McNally, R.J. (1996). Anxiety sensitivity is distinguishable from trait anxiety. In R.M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 214–227). New York, NY: Guilford Press. Miller, W., Tonigan, J.S., & Longabaugh, R. (1995). The Drinker Inventory of Consequences (DrInC): An instrument for assessing adverse consequences of alcohol use. Project MATCH Monograph Series, NIH Publication No. 95–3911, Vol. 4, Rockville, MD: M. Mattson & L. Marshall. Mroczek, D.K., & Griffin, P.W. (2007). The use of growth-curve modeling in estimating stability and change in well-being over time. In A.D. Ong, & M.M. van Dulmen (Eds.), Oxford handbook of methods in positive psychology (pp. 467–476). New York, NY: Oxford University Press. Norton, G.R. (2001). Substance use/abuse and anxiety sensitivity: What are the relationships? Addictive Behaviors, 26, 935–946. http://dx.doi.org/10.1016/S03064603(01)00244-1. Norton, G.R., Cox, B.J., Hewitt, P.L., & McLeod, L. (1997). Personality factors associated with generalized and non-generalized social anxiety. Personality and Individual Differences, 22, 655–660. http://dx.doi.org/10.1016/S0191-8869(96)00243-7. Olatunji, B.O., & Wolitzky-Taylor, K.B. (2009). Anxiety sensitivity and the anxiety disorders: A meta-analytic review and synthesis. Psychological Bulletin, 135, 974–999. http://dx.doi.org/10.1037/a0017428. Olthuis, J.V., Watt, M.C., Mackinnon, S.P., & Stewart, S.H. (2014). Telephone-delivered CBT for high anxiety sensitivity: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 82, 1005–1022. http://dx.doi.org/10.1037/a0037027. Otto, M.W., Safren, S.A., & Pollack, M.H. (2004). Internal cue exposure and the treatment of substance use disorders: Lessons from the treatment of panic disorder. Journal of Anxiety Disorders, 18, 69–87. http://dx.doi.org/10.1016/j.janxdis.2003.07.007. Peterson, R.A., & Reiss, S. (1992). Anxiety sensitivity index manual (2nd ed.). Worthington, OH: International Diagnostic Services. Preacher, K.J., Curran, P.J., & Bauer, D.J. (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling and latent curve analysis. Journal of Education and Behavioral Statistics, 31, 437–448. http://dx.doi.org/10. 3102/10769986031004437. Preacher, K.J., & Selig, J.P. (2012). Advantages of Monte Carlo confidence intervals for indirect effects. Communication Methods and Measures, 6, 77–98. http://dx.doi.org/10. 1080/19312458.2012.679848. Reiss, S. (1991). Expectancy model of fear, anxiety, and panic. Clinical Psychology Review, 11, 141–153. http://dx.doi.org/10.1016/0272-7358(91)90092-9. Reiss, S., & McNally, R.J. (1985). The expectancy model of fear. In S. Reiss, & R.R. Bootzin (Eds.), Theoretical issues in behavior therapy (pp. 107–122). New York: Academic Press. Reiss, S., Peterson, R., Taylor, S., Schmidt, N., & Weems, C.F. (2008). Anxiety Sensitivity Index consolidated user manual: ASI, ASI-3, and CASI. Worthington, OH: IDS Publishing Corp. Schmidt, N.B., Buckner, J.D., & Keough, M.E. (2007). Anxiety sensitivity as a prospective predictor of alcohol use disorders. Behavior Modification, 31, 202–219. http://dx.doi. org/10.1177/0145445506297019. Stahlbrandt, H., Johnsson, K.O., & Berglund, M. (2007). Two-year outcome of alcohol interventions in Swedish university halls of residence: A cluster randomized trial of a brief skills training program, twelve-step-influenced intervention, and controls. Alcoholism: Clinical and Experimental Research, 31, 458–466. http://dx.doi.org/10. 1111/j.1530-0277.2006.00327.x. Stewart, S.H., Peterson, J.B., & Pihl, R.O. (1995). Anxiety sensitivity and self-reported alcohol consumption rates in university women. Journal of Anxiety Disorders, 9, 283–292. http://dx.doi.org/10.1016/0887-6185(95)00009-D. Stewart, S.H., Samoluk, S.B., & MacDonald, A.B. (1999). Anxiety sensitivity and substance use and abuse. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 287–319). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Stewart, S.H., Zvolensky, M.J., & Eifert, G.H. (2001). Negative-reinforcement drinking motives mediate the relation between anxiety sensitivity and increased drinking behavior. Personality and Individual Differences, 31, 157–171. http://dx.doi.org/10. 1016/S0191-8869(00)00213-0. Taylor, A.B., MacKinnon, D.P., & Tein, J. (2008). Tests of the three-path mediated effect. Organizational Research Methods, 11, 241–269. http://dx.doi.org/10.1177/ 1094428107300344. Taylor, S., Zvolensky, M.J., Cox, B.J., Deacon, B., Heimberg, R.G., Ledley, D.R., et al. (2007). Robust dimensions of anxiety sensitivity: Development and initial validation of the Anxiety Sensitivity Index — 3. Psychological Assessment, 19, 176–188. http://dx.doi. org/10.1037/1040-3590.19.2.176. Tull, M.T., Schulzinger, D., Schmidt, N.B., Zvolensky, M.J., & Lejuez, C.W. (2007). Development and initial examination of a brief intervention for heightened anxiety sensitivity among heroin users. Behavior Modification, 31, 220–242. http://dx.doi.org/10.1177/ 0145445506297020. 24 J.V. Olthuis et al. / Addictive Behaviors 46 (2015) 19–24 Watt, M.C., & Stewart, S.H. (2008). Overcoming the fear of fear: How to reduce anxiety sensitivity. Oakland, CA: New Harbinger. Watt, M.C., Stewart, S.H., Birch, C., & Bernier, D. (2006). Brief CBT for high anxiety sensitivity decreases drinking problems, relief alcohol outcome expectancies, and conformity drinking motives: Evidence from a randomized controlled trial. Journal of Mental Health, 15, 683–695. http://dx.doi.org/10.1080/09638230600998938. Zack, M., Poulos, C.X., Aramakis, V.B., Khamba, B.K., & MacLeod, C.M. (2007). Effects of drink-stress sequence and gender on alcohol stress response dampening in high and low anxiety sensitive drinkers. Alcoholism: Clinical and Experimental Research, 31, 411–422. http://dx.doi.org/10.1111/j.1530-0277.2006.00322.x. Zvolensky, M.J., Bernstein, A., Cardenas, S.J., Colotla, V.A., Marshall, E.C., & Feldner, M.T. (2007). Anxiety sensitivity and early relapse to smoking: A test among Mexican daily, low-level smokers. Nicotine & Tobacco Research, 9, 483–491. http://dx.doi.org/ 10.1080/14622200701239621.