ORIGINAL PAPER

A Randomized Controlled Trial Comparing the Effects of Yoga With an

Active Control on Ambulatory Blood Pressure in Individuals With
Prehypertension and Stage 1 Hypertension
Marshall Hagins, PT, PhD;1 Andrew Rundle, PH;2 Nathan S. Consedine, PhD;3 Sat Bir S. Khalsa, PhD4

From the Department of Physical Therapy, Long Island University, Brooklyn, NY;1 Department of Epidemiology, Mailman School of Public Health,
Columbia University, New York, NY;2 Department of Psychological Medicine, The University of Auckland, Auckland, New Zealand;3 and Department
of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA4

The purpose of this study was to compare the effects of respectively) but no significant within-group changes in the
yoga with an active control (nonaerobic exercise) in individ- active control group. Direct comparisons of the yoga
uals with prehypertension and stage 1 hypertension. intervention with the control group found a single blood
A randomized clinical trial was performed using two arms: pressure variable (diastolic night) to be significantly different
(1) yoga and (2) active control. Primary outcomes were 24- (P=.038). This study has demonstrated that a yoga inter-
hour day and night ambulatory systolic and diastolic blood vention can lower blood pressure in patients with mild
pressures. Within-group and between-group analyses were hypertension. Although this study was not adequately
performed using paired t tests and repeated-measures powered to show between-group differences, the size of
analysis of variance (time 9 group), respectively. Eighty- the yoga-induced blood pressure reduction appears to
four participants enrolled, with 68 participants completing justify performing a definitive trial of this intervention to test
the trial. Within-group analyses found 24-hour diastolic, whether it can provide meaningful therapeutic value for the
night diastolic, and mean arterial pressure all significantly management of hypertension. J Clin Hypertens (Greenwich).
reduced in the yoga group ( 3.93, 4.7, 4.23 mm Hg, 2014;16:54–62. ª2013 Wiley Periodicals, Inc.

Currently, almost 80 million US adults have high blood poor quality with methodological limitations. In fact, a
pressure (BP),1 with fewer than half of patients with recent American Heart Association review14 classified
hypertension having their BP controlled.2 Uncontrolled the existing evidence for the effects of yoga on BP in
hypertension is thought to be responsible for 62% of the lowest possible category for estimates of certainty
cerebrovascular disease and 49% of ischemic heart of treatment effect (class C). Many of the studies
disease events,3 and was estimated to cost the United examining the effects of yoga on BP are uncontrolled or
States $93.5 billion in health care services, medications, use nonhypertensive participants.8 Very few studies
and missed days of work in 2010.4 The cost of drugs, have controlled for important confounding factors and
drug interactions, and nonadherence with prescribed only two used ambulatory BP measures, which are
drug regimens all contribute to the high rates of known to give a more accurate estimate of treatment
uncontrolled hypertension. Alternative, less expensive effects than office measurements.13 Therefore, the
methods to reduce BP with lower risk of drug interac- purpose of the current study was to conduct and
tions, which may convey the benefit of long-term evaluate a well-controlled randomized trial comparing
adherence, are much needed. Yoga is an alternative the effects of yoga with an active control group on
health care practice that may improve BP control.5,6 The ambulatory BP in individuals with prehypertension and
number of persons who practice yoga continues to rise, stage 1 hypertension.
with current estimates indicating at least 10.4 million
people in the United States (5.1%) practice yoga.7 METHODS
BP control is one of the most studied outcomes of A randomized clinical trial of patients with prehyper-
yoga, with several reviews5,8–12 and one meta-analy- tension and stage I hypertension was performed using
sis13 suggesting that yoga is generally effective with two arms: (1) yoga and (2) active control (nonaerobic
effect sizes equivalent to other types of lifestyle exercise). Our hypothesis was that yoga practice would
interventions. Importantly, however, these reviews also provide significantly better BP reduction than the active
uniformly suggest that current studies of yoga are of control. Prior to recruitment, the study was approved by
the Long Island University (LIU’s) institutional review
board and was registered with Clinicaltrials.gov
Address for correspondence: Marshall Hagins, PT, PhD, Department of (NCTO1542359). We estimated that the study would
Physical Therapy, Long Island University, One University Plaza, Brooklyn, need 90 participants (20% expected dropout rate) to
NY, 11201 achieve an 85% power to observe a 5 mm Hg change in
E-mail: mhagins@liu.edu
systolic BP (SBP) between the two groups.15
Manuscript received: September 4, 2013; revised: October 1, 2013;
accepted: November 5, 2013 Participants were recruited through flyers, advertise-
DOI: 10.1111/jch.12244 ments, and e-mail distribution to the local community.

54 The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014

 Effects of Yoga on Hypertension | Hagins et al.

The study was described as a “stress reduction” values and 7 nighttime values were required for the data
program for hypertension. Inclusion criteria were age to be considered valid.
between 21 and 70 years; prehypertension or stage I Demographic data on race, age, sex, and height
hypertension as determined by a 24-hour ambulatory BP and weight were collected at pretest (Table I). Diet and
(ABP) reading, with SBP between 120 and 159 mm Hg physical activity were assessed preintervention and
or diastolic BP (DBP) between 80 and 99 mm Hg3; postintervention using the Block 100-Item Food
medically stable on any current medications; body mass Frequency Questionnaire (FFQ)18,19 and the Baecke
index (BMI) between 18.5 to 40 kg/m2; and English Questionnaire of Physical Activity,20 respectively.
speaking. In addition, participants were required to be Participants were encouraged to not change their diets,
available during the expected class periods (both inter- levels of physical activity, or medications during the
ventions). Exclusion criteria were current use of insulin course of the study unless advised to do so by their
or oral hypoglycemic agents; previous cardiovascular physician. At posttest, participants were asked whether
event (prior myocardial infarction, stroke, or angina they had changed medications during the course of the
pectoris); current or previous cancer diagnosis; conges- study. Participants were given access to both Internet-
tive heart failure; history of kidney disease; signs or based and paper methods of self-report for homework
symptoms of significant peripheral vascular disease; compliance. Efficacy expectations of participants for
significant comorbidities that preclude successful com- their assigned intervention were obtained after atten-
pletion of the study (eg, current fractures, Parkinson’s dance of the first treatment session using the Credibility
disease, vertigo); or current/regular yoga practitioner Expectancy Questionnaire (CEQ).21,22 Self-reported
(participated in more than 3 yoga sessions within the psychosocial measures were obtained at pretest and
past year). posttest but will be reported elsewhere.
Participants were told that the study was comparing
two potentially beneficial stress-reducing interventions. Randomization
Participants in both groups were asked to attend two Coin tosses performed by the primary investigator
55-minute classes per week for 12 weeks and to (MH) were used for sequence generation for treatment
perform 3 sessions of home practice for 20 minutes group assignment. Sequential results (eg, participant
each week as described in detail below. Participants 1 = yoga) were placed inside 90 opaque sealed enve-
received $100 for completion of all phases of the study lopes numbered in advance (eg, 1–90). Once each
including pretest and posttest measures, attendance of participant completed pretest measures (with the excep-
≥75% of the intervention sessions (18 of 24 classes), and tion of the survey regarding expectations of treatment
completion of homework logs. efficacy), he/she took the next numbered concealed
Potential participants who met initial criteria (eg, age, envelope from within a box located with the measure-
medical history, activity levels) via a phone screening ment laboratory. All outcome assessors remained
and agreed to the requirements/expectations of the blinded to assignment of intervention throughout the
study were invited to a BP screening within the Physical study. By necessity for an active intervention, partici-
Therapy Department at LIU where clinical measures of pants were not blinded to intervention assignment.
BP (eg, aneroid sphygmomanometer) were used to
determine whether the participant’s BP was in the range
of the inclusion criteria. If the clinical measures were
within the criterion range, the participant was asked to TABLE I. Baseline Characteristics by Randomized
wear an ABP device for 24 hours. After the 24-hour Group
data were evaluated, if either or both the mean 24-hour Yoga Control
SBP or DBP were within the inclusion range, partici-
pants were invited to participate in the study. Measure- Age, mean (SD), y 56.4 (9.78) 52.45 (12.19)
Women, No. (%) 33 (91.6) 25 (80.6)
ments were implemented such that no longer than
BMI, mean (SD) 30.27 (.94) 29.75 (.93)
1 month occurred between the measures and the start of
Physical activity, mean (SD)a 6.61 (2.51) 6.97 (2.25)
the intervention. Five cohorts of approximately 18
Prehypertensive (SBP 120–139 mm Hg) 23 (71.9) 25 (69.4)
participants each were enrolled across the study period.
Hypertensive (SBP >140 mm Hg) 9 (28.1) 11 (31.0)
Race or ethnicity
Measures African American 31 (86.2) 27 (84.4)
Primary outcomes were SBP and DBP and heart rate Non-Hispanic white 1 (2.7) 1 (3.1)
(HR). Twenty-four–hour ABP monitoring was per- All others 4 (11.1) 4 (12.5)
formed at pretest and posttest (“Oscar2,” Suntech
Abbreviations: BMI, body mass index; SBP, systolic blood pressure;
Medical, Morrisville, NC). This device has been vali- SD, standard deviation. Data are presented as number (percentage)
dated as per internationally recognized standards.16,17 unless otherwise indicated.
Twenty-four–hour ABP values were further categorized a
Baecke Physical Activity Survey and total of work, leisure, and sport
as day or night values using each participant’s reported
scores.
awake and sleep times. A minimum of 14 daytime

The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014 55

Effects of Yoga on Hypertension | Hagins et al.

Interventions Active Control Arm. The active control exercise class

The arms of the study were explicitly designed for was nonaerobic and consisted of a warm-up, exercises
equivalence of patient effort and time investment, (eg, “step-touch,” squats, upper extremity resistive band
investigator and instructor interaction and attention, work, abdominal strengthening), and stretching/cool
social interaction, and expectations of efficacy. Conse- down. It was designed by Tracey Rawls Martin (Assis-
quently, classes and homework requirements were tant Professor, Athletic Training and Exercise Sciences
identical in terms of length, frequency, and duration. Department, LIU). Details of the active control group
All participants were provided with printed text and can be seen in Appendix B. All active control group
photos describing the intervention, a video of the instructors had at least 2 years of experience in leading
respective intervention on DVD, and procedures for fitness classes.
recording homework compliance. Classes had similar
opportunities before, during, and after class for social Statistical Analysis
interaction. Instructors for both arms completed sepa- Analytic Strategy. Means and standard deviations for
rate 2-hour workshop sessions defining goals, approach all demographic and primary outcomes were calcu-
to participants, administrative duties, and specific lated. The primary outcomes of interest were means of
structure and physical requirement of each class. systolic and diastolic values (24-hour, day, night, mean
Instructors were provided with a standardized teacher’s arterial pressure [MAP], and “dipping” status defined
manual and a video (DVD) of the practice. Instructors as mean day less mean night values), and HR. Analysis
for both arms were trained to provide positive expec- of variance (ANOVA) and chi-square analyses were
tations to participants regarding the potential for the used for retention analysis to determine systematic
class to lower BP. variation in the factors characterizing (1) persons who
In addition, the two interventions of the study were completed ≥1 classes but were lost to follow -up from
designed to be equivalent in terms of metabolic output. (2) persons who participated fully. Separate repeated-
Our previous estimates of the metabolic output of the measures ANOVAs (time 9 group) were performed on
yoga exercises23 were used to design the level of physical activity and diet variables to determine
physical intensity of the exercises used for the active whether these factors varied across groups during the
control group. The targeted average intensity across trial. Independent t tests were performed on expecta-
the 55-minute class was 3 metabolic equivalents tion of efficacy to determine whether this factor varied
(METs) (approximately equal to a brisk pace of across groups at baseline and on measures of adherence
walking—a level considered nonaerobic. The validity at posttest (number of classes attended and homework
of this metabolic equivalence across groups during the performed). Equivalency of the interventions relative to
study was experimentally tested. A subset of partici- metabolic output was determined by independent t
pants from each arm (yoga = 9, active control = 8) tests of the mean MET values obtained with indirect
volunteered to perform his/her respective intervention calorimetry.
within the regular class period while wearing a
portable indirect calorimeter (K4b,2 Cosmed, USA, Primary Analyses. Paired t tests were used to assess
Inc, Chicago, IL).24 Estimates of metabolic output changes within group preintervention to postinterven-
(METs) were obtained from the calorimeter through tion. Separate repeated-measures ANOVAs (time 9
measures of oxygen and carbon dioxide flow through a group) were used to determine significant differences
facemask worn by participants. Measures for both relative to the intervention.
intervention arms were taken during weeks 6 to 8 of
the intervention. RESULTS
Recruitment occurred from January 2010 to March
Yoga Arm. Yoga is generally described as a practice 2012, with interventions occurring from March 2010
that incorporates 3 elements: postures, breath control, to June 2012. A large number of potential participants
and meditation.25,26 The specific yoga intervention were screened (n=459; Figure 1) to achieve 84 partic-
incorporated all 3 of these elements and was based on ipants enrolled. Sixteen (19%) were lost to follow-up
the primary (beginner) series of Ashtanga yoga origi- after completing ≥1 classes, leaving 68 participants
nally developed by Pattabhi Jois27 and as specifically who completed the trial. Baseline demographic char-
designed for this study by a long-term student of Jois: acteristics were similar in the randomized groups
Eddie Stern (Director of Ashtanga Yoga New York, (Table I) and no adverse events were reported. No
New York, NY). The program was explicitly designed participants reported changing BP medications during
to allow adaptation of poses as needed for individual the trial.
participants who were expected to be sedentary and
older and with somewhat larger body mass. Please see Retention Analysis
Appendix A for a complete description of the yoga Completion did not vary by group v2 (1)=0.81 (P=1.00)
program. All yoga instructors had a minimum of 200 and there were no differences between completers and
hours of training (Registered Yoga Teacher 200; Yoga noncompleters as a function of sex, v2 (1)=0.32, P=.45,
Alliance, Arlington, VA). race, v2 (3)=2.51, P=.47, age, F (1, 82)=0.38, P=.54,

56 The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014

 Effects of Yoga on Hypertension | Hagins et al.

FIGURE 1. Recruitment flow diagram.

BMI, F (1, 82)=0.12, P=.73, expectation scores from the METS) than the active control group (2.36, 0.49
CEQ (all P≥.12), physical activity, F (1, 77)=0.49, METS) (P<.001). Figure 2 displays the mean metabolic
P=.49, HR (P=.31) or baseline 24-hour systolic pressure requirements of both arms of the trial across a single
(P=.20). Completers did, however, have lower DBP at session.
baseline than those lost to follow-up: F (1, 82)=6.56,
P<.05. As might be expected, those who were lost to Primary Analyses
follow-up after ≥1 classes attended fewer sessions: F (1, Within Group. Results of paired t tests assessing
82)=292.83, P<.01. Mean number of classes attended within-group intervention to postintervention changes
across groups by completers was 21.91 (3.02). are described in Table II. Twenty-four–hour diastolic,
Repeated-measures ANOVAs (time 9 group) found night diastolic, and MAP were all significantly reduced
no significant differences in physical activity (P=.174) or in the yoga group ( 3.93, 4.7, 4.23 mm Hg,
diet variables (all P≥.05) across the groups during the respectively). Similarly, trends (P<.10) for the yoga
trial. Independent t tests found no significant differences group to reduce BP were seen in 24-hour SBP, day DBP,
in expectation of efficacy measures from the CEQ and night SBP. However, unlike the yoga group, the
obtained at pretest (all P≥.183). Independent t tests active control group did not demonstrate any significant
found no significant differences between groups in within group changes or trends.
number of classes attended (mean, 21.91 [3.02];
P=.749) or in minutes of homework performed (mean, Between Group. Repeated-measures ANOVAs (time
675.45 [464.39]; P=.506). 9 group) demonstrated a significant difference between
Independent t tests comparing the metabolic require- groups in preintervention to postintervention changes in
ments of the two arms found that the yoga arm required diastolic nighttime pressures (P=.038) and a trend in
significantly more energy to complete (2.79, 0.59 diastolic 24-hour pressures (P=.081). There were no

The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014 57

Effects of Yoga on Hypertension | Hagins et al.

FIGURE 2. Metabolic requirements for each intervention.

TABLE II. Results of Within-Group and Between-Group Analysis

Change Value
Time 1 Time 2 Time 1–Time 2
Within-Group Between-Group
M SD M SD M SD P Value P Value

Systolic 24 Control 133.80 9.86 133.36 18.29 .44 15.00 .868 .224
Yoga 135.53 9.79 130.68 14.99 4.84 14.54 .053
Diastolic 24 Control 80.17 7.49 79.76 11.11 .41 8.19 .778 .0814
Yoga 80.82 7.33 76.89 8.61 3.93 8.14 .006a
HR 24 Control 77.58 10.63 75.75 9.72 1.83 8.98 .258 .999
Yoga 72.34 8.25 70.5 7.04 1.83 6.79 .114
Systolic (day) Control 138.63 10.39 138.68 17.87 .052 16.18 .986 .326
Yoga 139.64 10.72 135.81 16.55 3.83 16.13 .163
Diastolic (day) Control 84.62 7.52 84.42 11.48 .19 9.90 .913 .256
Yoga 84.31 8.63 81.38 10.19 2.93 9.81 .081
HR (day) Control 80.39 11.10 79.08 10.38 1.3 9.75 .454 .905
Yoga 75.18 9.45 73.61 7.59 1.56 8.01 .248
Systolic (night) Control 122.61 12.72 121.76 19.97 .85 15.80 .764 .262
Yoga 125.14 12.06 119.96 15.05 5.17 15.70 .056
Diastolic (night) Control 69.95 10.55 69.59 12.23 .36 8.26 .807 .038a
Yoga 72.07 7.97 67.36 7.97 4.70 8.60 .002a
HR (night) Control 70.35 10.96 68.71 9.69 1.63 8.88 .306 .880
Yoga 65.59 8.29 63.65 7.07 1.93 7.63 .137
MAP 24-h Control 98.13 7.71 97.62 13.17 .51 10.23 .781 .134
Yoga 99.05 7.36 84.82 9.89 4.23 10.01 .016a
Systolic (dip) Control 13.78 10.18 14.77 8.35 .98 11.09 .620 .890
Yoga 12.29 10.84 13.66 9.72 1.37 12.08 .500
Diastolic (dip) Control 22.80 16.46 22.86 14.66 .06 15.5 .981 .370
Yoga 18.01 15.1 21.43 13.00 3.42 15.1 .183
Abbreviations: between-group, results of repeated-measures analysis of variance (time 9 group); M, mean; MAP, mean arterial pressure, diastolic +
(0.33333 9 [systolic – diastolic]); SD, standard deviation; within-group, results of paired t tests of within-group change scores Time 1 vs Time 2. Values
are expressed as means (standard deviations).
P<.05.
a

58 The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014

 Effects of Yoga on Hypertension | Hagins et al.

6
*
#
5

4
#
3

0
Yoga

Yoga

Yoga

Yoga

Yoga

Yoga
Control

Control

Control

Control

Control

Control
Systolic 24 Diastolic 24 Systolic Diastolic Systolic Diastolic
(day) (day) (night) (night)

* = significant between-group difference using repeated measures ANOVA (time x group); p < 0.05;
# = significant within group difference using paired t-tests; p < 0.05

FIGURE 3. Change value (decrease in mm Hg) from pretest to posttest.

significant differences or trends in any other variables been recommended for people with prehypertension by
(Table II). See Figure 3 for a display of BP change the national hypertension guidelines,3 and it may now
values from pretest to posttest. be appropriate to consider yoga programs, which have
no known adverse effects for participants, as an
DISCUSSION additional strategy to be considered in delaying or even
This study showed that yoga decreases BP in patients preventing the onset of hypertension in patients at risk
with very mild hypertension while the active control for this condition.
intervention (nonaerobic exercise) did not reduce BP. The mechanisms by which yoga may influence BP are
However, in direct comparisons of the yoga intervention not well understood. Figure 4 presents a previously
with the control group, only a single BP change variable suggested model of hypothesized pathways.6 Yoga may
(diastolic night) was found to be significantly different. reduce feelings of stress and increase a sense of well-
Although recruitment goals for this study were essen- being, reducing activation of the sympathetic nervous
tially met (n=84 vs goal of n=90), and effect size and system and positively altering neuroendocrine status
dropout rates were accurately estimated, the expected and inflammatory responses (see pathway 1 in
variability in BP measurements was underestimated. Figure 4). The physical practices of yoga may directly
Standard deviations are displayed in Table II and range stimulate the vagus nerve increasing parasympathetic
from approximately 9 mm Hg to 16 mm Hg. These output (see pathway 2 in Figure 4).
values are similar to some previous studies using To our knowledge, there are only 3 controlled trials
ambulatory BP monitoring28,29 but are greater than in that adequately reported BP data and have examined the
others.30 Future research will require larger sample sizes effects of yoga on individuals with hypertension using
to achieve sufficient power for comparisons with control exercise comparison groups.13 In all three studies31–33
groups. there were no significant effects of yoga when compared
The current study found that yoga decreased 24-hour with exercise. In the current study, the use of a
mean SBP and DBP by approximately 5 mm Hg and nonaerobic exercise arm was designed primarily as an
4 mm Hg, respectively. These BP reductions are consis- active control with no expectation of improvement in
tent with the values found in a recent meta-analysis of BP outcome; this was confirmed with the observation of
controlled studies examining the effect of yoga on no significant within-group changes or trends. Although
individuals with hypertension (systolic 4 mm Hg and the intent of the design was to have the active control
diastolic 4 mm Hg).13 The differences in BP reported in match the yoga arm in metabolic output, the mean
the present study are comparable to those reported for METs of the yoga arm required more energy than that
other nonpharmacologic strategies such as the Dietary of the active control group. The mean difference
Approaches to Stop Hypertension (DASH) diet, physical between treatment arms was small and unlikely to be
exercise, and salt reduction. Apart from their value in all clinically meaningful, but it did achieve statistical
patients with hypertension, these interventions have significance. Future studies attempting to balance

The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014 59

Effects of Yoga on Hypertension | Hagins et al.

FIGURE 4. Hypothesized pathways by which yoga may influence hypertension and cardiovascular risk profiles.6

treatment arms relative to metabolic output would justify performing a definitive trial of this intervention
benefit from additional efforts to develop an active to test whether it can provide meaningful therapeutic
control arm with practices more precisely aligned with value for the management of prehypertension and stage
the energy requirements of the yoga practice under 1 hypertension.
study.
Although this study is one of many that have Acknowledgments: This work was funded by the National Institute of General
Medical Sciences: 1SC3GM088049-01A1.
examined the effects of yoga on BP, it is among the
first to use rigorous methods in a randomized trial on Conflict of Interest: There are no competing financial interests in relation to the
individuals with prehypertension or stage 1 hyperten- current work.
sion. There were no significant differences between
groups on measures of physical activity, food, expecta- References
tion of efficacy, or adherence minimizing these potential 1. Roger VL, Go AS, Lloyd-Jones DM, et al. On behalf of the American
Heart Association statistics committee and stroke statistics subcom-
sources of bias. Additionally, control of potential mittee. Heart disease and stroke statistics – 2012 update: a report
sources of bias related to selection, detection, attrition, from the American Heart Association. Circulation. 2012;125:188–
197.
and reporting,34 the successful balancing of treatment 2. Gillespie C, Kuklina EV, Briss PA, et al. Vital signs: prevalence,
arms relative to duration, frequency, and social inter- treatment, and control of hypertension, United States, 1999–2002 and
action, and the use of state-of-the-art ABP monitoring 2005–2008. Morb Mortal Wkly Rep. 2011;60:103–108.
3. Chobanian AV, Bakris GL, Black HR, et al. The seventh report of the
give confidence that this type of research can be joint national committee on prevention, detection, evaluation, and
conducted in compliance with highly credible clinical treatment of high blood pressure. JAMA. 2003;289:2560–2572.
trial methodology. 4. Centers for Disease Control and Prevention Website. High blood
pressure facts. 2012. Available at: http://www.cdc.gov/bloodpressure/
Given the variability found in this study, future facts.htm. Accessed August 13, 2012.
research will require larger sample sizes to achieve 5. Okonta NR. Does yoga therapy reduce blood pressure in patients with
hypertension?: an integrative review. Holist Nurs Pract. 2012;26:
sufficient power for comparisons with control groups. 137–141.
Future research might also benefit from techniques to 6. Innes KE, Vincent HK. The influence of yoga-based programs on risk
predict which patients are most likely to positively profiles in adults with type 2 diabetes mellitus: a systematic review.
Evid Based Complement Alternat Med. 2007;4:469–486.
engage in yoga, thus making more targeted interven- 7. Birdee GS, Legedza AT, Saper RB, et al. Characteristics of yoga
tions possible. users: results of a national survey. J Gen Intern Med. 2008;23:1653–
1658.
8. Innes KE, Bourguignon C, Taylor AG. Risk indices associated with
CONCLUSIONS the insulin resistance syndrome, cardiovascular disease, and possible
protection with yoga: a systematic review. J Am Board Fam Pract.
This study demonstrates that a yoga intervention in 2005;18:491–519.
patients with mild hypertension can significantly reduce 9. Hutchinson SC, Ernst E. Yoga therapy for coronary heart disease: a
BP. Although this study was not adequately powered to systematic review. Focus Altern Complement Ther. 2003;8:144.
10. Raub JA. Psychophysiologic effects of hatha yoga on musculoskeletal
test this effect against a control group, the size of the and cardiopulmonary function: a literature review. J Altern Comple-
yoga-induced BP reduction we observed appears to ment Med. 2002;8:797–812.

60 The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014

 Effects of Yoga on Hypertension | Hagins et al.

11. Jayasinghe SR. Yoga in cardiac health (a review). EurJ Cardiovasc APPENDIX A. DESCRIPTION OF THE YOGA
Prev Rehabil. 2004;11:369–375.
12. Bussing A, Michalsen A, Khalsa SB, et al. Effects of yoga on mental PROGRAM
and physical health: a short summary of reviews. Evid Based The yoga program is based on Ashtanga (Primary
Complement Alternat Med. 2012;2012:165410.
13. Hagins M, States R, Selfe T, Innes K. Effectiveness of yoga for
Series). Conscious use of the breath, frequently cued by
hypertension: systematic review and meta-analysis. Evid Based the instructor, occurs through most of the practices. The
Complement Alternat Med. 2013;2013:649836. doi: 10.1155/2013/ intensity of the practice varies based on the participant’s
649836
14. Brook RD, Appel LJ, Rubenfire M, et al. American Heart Association capacity through the 12-week program. The expectation
Professional Education Committee of the Council for High Blood is that advances in capacity will be reflected in longer
Pressure Research CoC, Stroke Nursing CoE, Prevention, Council on holds of postures sustained with more breath cycles and
Nutrition PA: beyond medications and diet: alternative approaches to
lowering blood pressure: a scientific statement from the American more advanced postures (eg, Warrior pose).
Heart Association. Hypertension. 2013;61:1360–1383. Below represents the total available practices for this
15. van Montfrans GA, Karemaker JM, Wieling W, Dunning AJ.
Relaxation therapy and continuous ambulatory blood pressure in
program and the order to be followed for each class (eg,
mild hypertension: a controlled study. BMJ. 1990;300:1368–1372. meditation, postures, breathing, and relaxation). How-
16. Jones SC, Bilous M, Winship S, et al. Validation of the Oscar 2 ever, not all physical postures described below will
oscillometric 24-hour ambulatory blood pressure monitor according
to the international protocol for the validation of blood pressure occur in each class. Instructors will make decisions
measuring devices. Blood Press Monit. 2004;9:219–223. regarding the appropriate level of practices based on the
17. Goodwin J, Bilous M, Winship S, et al. Validation of the Oscar 2 capacity of the students present. The warm-up exercises
oscillometric 24-h ambulatory blood pressure monitor according to
the British Hypertension Society protocol. Blood Press Monit. will typically occur during the first month as students
2007;12:113–117. begin to understand the relationship of movement with
18. Block G, Woods M, Potosky A, Clifford C. Validation of a self-
administered diet history questionnaire using multiple diet records. J
breath and to have increased mobility, and may then be
Clin Epidemiol. 1990;43:1327–1335. discarded if not needed. It is expected that during the
19. Block G, Thompson FE, Hartman AM, et al. Comparison of two second month, Sun Salutation B will be implemented
dietary questionnaires validated against multiple dietary records
collected during a 1-year period. J Am Diet Assoc. 1992;92:686– and that during the third month, Warrior poses will be
693. implemented. Instructors will encourage modification of
20. Baecke JA, Burema J, Frijters JE. A short questionnaire for the postures as needed. For example, chairs and the wall
measurement of habitual physical activity in epidemiological studies.
Am J Clin Nutr. 1982;36:936–942. will be used for support for those who cannot achieve
21. Deviliya GJ, Borkovecb TD. Psychometric properties of the credibil- the standard position or who have poor balance.
ity/expectancy questionnaire. J Behav Ther Exp Psychiatry.
2000;31:73–86.
22. Smeets RJ, Beelen S, Goossens ME, et al. Treatment expectancy and Class structure
credibility are associated with the outcome of both physical and 1. Meditation 5 to 7 minutes: Upon entering, students
cognitive-behavioral treatment in chronic low back pain. Clin J Pain.
2008;24:305–315. will be asked to assume a seated position on the floor
23. Hagins M, Moore W, Rundle A. Does practicing hatha yoga satisfy or a chair and close the eyes and begin meditating.
recommendations for intensity of physical activity which improves Class begins with a led meditation focusing on the
and maintains health and cardiovascular fitness? BMC Complement
Altern Med. 2007;7:40. body and breath in month 1, the nervous system in
24. Maiolo C, Melchiorri G, Iacopino L, et al. Physical activity energy month 2, and the mind in month 3.
expenditure measured using a portable telemetric device in compar-
ison with a mass spectrometer. Br J Sports Med. 2003;37:445–447.
2. Physical postures (Asana) – 35 minutes
25. Baldwin MC. Psychological and Physiological Influences of Hatha a. Warm-up (3 9 each exercise):
Yoga Training on Healthy, Exercising Adults (Yoga, Stress, Wellness) i. Lift arms overhead while you breathe in. As
[Dissertation]. Boston, MA: Boston University; 1999.
26. Cowen VS, Adams TBP. Physical and perceptual benefits of yoga you breathe out, lower the arms.
asana practice: results of a pilot study. J Bodyw Mov Ther. ii. Bend forward as you exhale, lift up as you
2005;9:211–219. inhale, bend forward as you exhale, then lift
27. Jois P. Yoga Mala. New York: North Point Press; 1999.
28. Wittke E, Fuchs SC, Fuchs FD, et al. Association between different all the way to standing as you inhale.
measurements of blood pressure variability by ABP monitoring and iii. On all fours, or leaning against the chair or
ankle-brachial index. BMC Cardiovasc Disord. 2010;10:55.
29. Imai Y, Nagai K, Sakuma M, et al. Ambulatory blood pressure of
wall, inhale as you lift your head and arch
adults in Ohasama, Japan. Hypertension. 1993;22:900–912. your spine, exhale as you lower your head and
30. Vollmer WM, Appel LJ, Svetkey LP, et al. Comparing office-based flex your spine. You may do this on your
and ambulatory blood pressure monitoring in clinical trials. J Hum
Hypertens. 2005;19:77–82. elbows as well.
31. Niranjan M, Bhagyalakshmi K, Ganaraja B, et al. Effects of yoga and iv. Lie on your stomach. Place your hands on the
supervised integrated exercise on heart rate variability and blood floor under your shoulders. Gently lift your
pressure in hypertensive patients. JCCM. 2009;4:139–143.
32. Subramanian H, Soudarssanane MB, Jayalakshmy R, et al. Non- upper back and look to the right as you inhale.
pharmacological interventions in hypertension: a community-based Slowly look to the left and lower down as you
cross-over randomized controlled trial. Indian J Community Med.
2011;36:191–196.
exhale.
33. Saptharishi LG, Soudarssanane MB, Thiruselvakumar D, et al. b. Sun Salutation A (59)
Community-based randomized controlled trial of non-pharmacolog- c. Sun Salutation B (39) Likely to begin 2nd month
ical interventions in prevention and control of hypertension among
young adults. Indian J Community Med. 2009;34:329–334. of training
34. Higgins JP, Altman DG, Gotzsche PC, et al. Cochrane Bias Methods d. Warrior 1 and 2 (39) Likely to begin 3rd month
Group, Cochrane Statistical Methods Group. The Cochrane collab- of training
oration’s tool for assessing risk of bias in randomised trials. BMJ.
2011;343:d5928. e. Hands to feet pose (2–8 breaths)

The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014 61

Effects of Yoga on Hypertension | Hagins et al.

f. Triangle pose right and left (2–8 breaths each) and up, squat position hands on knees rotate trunk
g. Extended side angle right and left (2–8 breaths right and left, lunge position stretch calves, lean
each) forward with on leg forward, and stretch hamstrings.
h. Spread foot pose (4 types: hands on floor/hands 2. Exercises: 30 to 35 minutes: All exercises performed
behind back/elbows on floor/hands on ankles) (2– right and left: In standing: step touch side to side
8 breaths each type) hands on hips, same with “pushing arms” in rhythm,
i. Side Stretch pose (2–8 breaths) same with “swinging arms” in rhythm, rotate trunk
j. Forward Bend moving into Table Top (2–8 right and left with “pushing arms” during step touch
breaths each) side-to-side. Repeat above with alterations of entire
k. Janu Sirshasana right and left (2–8 breaths each) body relative to room—moving on diagonals, mov-
l. Butterfly pose (2–8 breaths) ing to face the rear of the room. Squats with arms
m. Shalabasana (2–8 breaths) forward. Remain in squat position and move one leg
3. Regulated Breathing – 10 Minutes backwards to touch toe on floor while extending
a. Seated cross-legged, hands clasped behind back, arms fully backwards. Continue alternating back-
head leaning forward, flexed spine (10 breaths) stepping motion while arms abduct overhead in
b. Seated cross-legged, leaning backward with hands rhythm. Participants find a partner and face each
on floor, arching spine, looking up and back with other. Using a resistive band stretched between them,
eyes (10 breaths) squat while pulling against each other with the
c. Seated cross-legged, arms straight with forearms resistive band. Face each other more closely and
resting on knees, palms upward, tips of index and perform external and internal rotation shoulder
thumbs touching, perform Victorious Breath with exercises using the resistive band. Move to hands
Abdominal Lock (10 breaths) and knees and do Cat and Cow (spinal flexion and
d. Seated cross-legged, use the thumb and little finger extension), modified push-ups (on knees as needed).
to hold the side of the nose and alternate: Move to back and do abdominal curl-ups with hands
Alternate Nostril Breathing (10 breaths each side) behind head and then curl-ups with rotations. Move
4. Relaxation – 5 minutes to side lying and do lower extremity hip abduction
a. Shavasana, lying on the back, palms upward to and adduction. Move to supine and perform pelvic
ceiling. lifts. Maintain pelvic lift and remove one foot from
floor keeping pelvis stable.
3. Stretches: 13 to 20 minutes: In supine: double knee
to chest, hold with arms. Remain in supine. Ham-
APPENDIX B. DESCRIPTION OF EXERCISE string stretch, one foot on floor, other hip flexed to
PROGRAM 90 degrees and straighten knee. Trunk stretch in
The exercise program consisted of a group of exercises supine—bend knees with feet on floor, allow knees to
commonly used within gyms and sports centers. The fall to the side achieving trunk rotation. Allow knees
goal of the program was to create a series of exercises to open up—bilateral horizontal abduction (butterfly
that could be performed by sedentary older adults and position) to stretch inner thighs. In sitting with soles
that could be varied in intensity through changing the of feet touching each other and with knees flexed
speed and number of repetitions. Instructors were approximately 90 degrees, lean forward—stretching
encouraged to modify exercises as needed for those posterior hips and back. Straighten one leg fully,
with less flexibility, strength, or endurance. Breath maintain other leg in Butterfly position and lean
control was explicitly not discussed. If questioned, forward over straight leg. In sitting butterfly posi-
participants were told to never stop breathing and to tion, rotate trunk maximally with hands on floor to
breathe in whatever way they felt most comfortable. assist. Straighten one leg out to the side and lean over
for side of trunk stretch with arm overhead reaching
Class structure: to foot. Place both legs out straight ahead in sitting
1. Warm-up: 5 to 7 minutes: In standing: Arm circles, and lean over both legs. Open up legs to as large a
shoulder circles, head/neck rotation and flexion/ “V” shape as possible. Lean forward to stretch
extension, trunk flexion/extension via rolling down hamstrings, adductors, and back.

62 The Journal of Clinical Hypertension Vol 16 | No 1 | January 2014