Abstract
The negative impacts of stress on antiretroviral therapy (ART) adherence and the many stressors faced by citizens returning from incarceration support the need for stress-reduction interventions for this population. We hypothesized that 37 returning citizens living with HIV and substance use problems randomized to a 12-session weekly yoga intervention would experience improved ART adherence, lower viral loads, and lower heart rate and blood pressure, compared to 36 people randomized to treatment as usual (TAU). We found that ART adherence increased for yoga participants from 65% at baseline to 75% at three-months, and decreased for TAU participants from 77% at baseline to 66% at three-months [F (1,69)=0.89, p< 0.35]. There were no significant changes for viral load, heart rate or blood pressure. The lack of statistically significant effects may have been due to a small sample size and enrollment of people largely in HIV treatment. Studies with larger sample sizes and participants exhibiting low ART adherence are warranted to better understand yoga’s impact.
Keywords: yoga, antiretroviral therapy adherence, HIV, substance use, reentry
Introduction
Prisons in the United States have about five times the number of people living with HIV than the general population (Centers for Disease Control and Prevention, 2012; Maruschak, 2015). While comprehensive HIV care is often provided in prisons and jails (Beckwith, Zaller, Fu, Montague, & Rich, 2010; Meyer et al., 2014); upon release, treatment rates decline for returning citizens (Iroh, Mayo, & Nijhawan, 2015). For example, among returning citizens living with HIV who were taking antiretroviral therapy (ART) medication at the time of their release from a Texan prison, only 30% refilled their ART medications within 60 days of release (Baillargeon et al., 2009), putting them at high risk of adverse immunological and clinical outcomes (Paterson et al., 2000; Sethi, Celentano, Gange, Moore, & Gallant, 2003) and of transmitting HIV (Castilla et al., 2005; Porco et al., 2004). With declines in ART adherence post-incarceration, it is unsurprising that reentry as compared with incarceration is associated with poorer virological and immunological outcomes for people living with HIV (Iroh et al., 2015; Stephenson et al., 2005).
Factors related to interruptions in ART adherence during reentry include homelessness, drug use and difficulties accessing medical care (Clements-Nolle et al., 2008; Swan, 2015). Further, the persistent stress experienced during reentry (Western, Braga, Davis, & Sirois, 2015) may affect ART adherence in multiple ways. Stress on its own has been directly associated with lower ART adherence (Ammassari et al., 2002; French, Tesoriero, & Agins, 2011; Leserman, Ironson, O’Cleirigh, Fordiani, & Balbin, 2008). Stress also may precipitate drug use (Sinha, 2001, 2008). Substance use, particularly substance use to cope with stress, is associated with lower rates of ART adherence (Arnsten et al., 2002; Gonzalez, Mimiaga, Israel, Bedoya, & Safren, 2013). Therefore, interventions that address stress management may be key to support ART adherence. One study demonstrated an association between reduced stress and substance use and improved medication adherence, highlighting the promise of such interventions (French et al., 2011). If effective, stress- and substance use-reduction interventions could improve virological, immunological and clinical outcomes.
Yoga is one such stress-reduction intervention (Li & Goldsmith, 2012; Macy, Jones, Graham, & Roach, 2015) that holds potential for people living with HIV (Agarwal, Kumar, & Lewis, 2015; Bonadies, 2004; Cade et al., 2010). Yoga facilitates relaxation through spiritual, physical and mental practices. In the United States, hatha yoga is the predominant type practiced, focusing on physical postures, and to a lesser extent breathing and meditation.
We hypothesized that returning citizens living with HIV and substance use problems randomized to a 12-week hatha yoga intervention (yoga) would experience improved ART adherence and viral suppression; and reduced heart rate and blood pressure as compared with those randomized to treatment as usual (TAU).
Methods
All study activities took place at an HIV/AIDS service provider in Philadelphia, PA, which includes services targeted towards returning citizens. Case managers referred English-speaking adults age 18 or older with HIV infection released from prison or jail in the previous 12 months who were thought to have a substance use disorder (assessed at screening with the Texas Christian University Drug Screen II; Institute of Behavioral Research, 2007). Self-referral via a study flier was permitted.
Design
We conducted a secondary analysis of a randomized trial of yoga’s effect on stress and substance use (Wimberly, Engstrom, Layde, & McKay, 2018). We randomized participants to either yoga or TAU at four recruitment periods, each three and a half months apart. The multiple recruitment periods ensured that there were at least eight people to randomize to a yoga class at any one time.
Treatment conditions.
Those randomized to the experimental arm received 12 sessions of a weekly 90-minute hatha yoga class, guided by a curriculum developed by the principal investigator. The curriculum was comprised of yogic philosophy, breathing exercises, physical postures and meditation and was adapted from standard practices based on interviews conducted with students from yoga classes offered to returning citizens. At weeks four and eight of the intervention, participants received a handout detailing yoga poses and meditations for self-study. Yoga classes were conducted (Wednesdays from 6-7:30 PM for the first cohort and Mondays from 12-1:30 PM for the second, third and fourth cohorts) when drop-in services were closed to ensure participant privacy and avoid non-study participants attempting to join classes.
A 30-year old female, certified yoga instructor who had practiced yoga regularly for nine years, taught Vinyasa and Yin yoga for six years, and had worked in a public health capacity at a jail, taught the first cohort. A 29-year old male, certified yoga instructor who had practiced yoga regularly for 12 years and taught Vinyasa, restorative and therapeutic yoga for six years, taught the second, third and fourth cohorts. Instructors were encouraged to adjust the curriculum based on the needs of the students, and provided written and verbal feedback regarding curriculum adjustments. Because the yoga protocol designed for this study was deemed too difficult to implement for many of the students who were new to yoga, the teachers added more simple ‘restorative’ poses that could be held for prolonged periods along with more preparatory stretches. The teacher in the first cohort taught all restorative poses in classes six and nine. The teacher in the second through fourth cohorts incorporated restorative poses in each class.
Both those randomized to yoga and TAU received standard of care including some or all of the following reentry services: case management, recreational or General Educational Development classes, and free healthcare.
Study Procedures
Assessments.
The following assessments were conducted at baseline and after completion of the yoga intervention, approximately three-months post-baseline.
Primary outcome: Antiretroviral Therapy Adherence.
ART adherence was defined as the percentage of days that a person refilled their ART medications in the previous 90 days. Prescription refill as a measure of medication adherence has been validated among people with chronic disease (Steiner, Koepsell, Fihn, & Inui, 1988) and people living with HIV (Gross et al., 2006; Grossberg & Gross, 2007). While this approach is flawed if refill data and medication compliance are discrepant (Barai et al., 2017), research has demonstrated that prescription refill data has stronger associations with virological response than self-report (Grossberg & Gross, 2007).
Participants identified their pharmacies, which were contacted to determine when they had refilled their ART medication in the 90 days prior to randomization. If they had been incarcerated within 90 days prior to enrollment, only pharmacy data after release were included. For example, if a participant had been released from prison 30 days prior to randomization and they had filled a one-month supply immediately after release, they were considered 100 % adherent. Medication adherence for an index drug was computed by dividing the number of days of prescribed doses by the number of days between first fill and last fill of the defined interval (pills prescribed per day/days between refills) × 100% (Grossberg, Zhang, & Gross, 2004). For example, three refills of a 30-day ART supply over 90 days was calculated as 90 days supply/90 days=100% adherence (Grossberg et al., 2004). The index drug was either the fixed dose combination of three antiretrovirals or if not on a fixed dose combination, then the integrase inhibitor or non-nucleoside reverse transcriptase inhibitor, or protease inhibitor.
Secondary outcomes.
Plasma HIV RNA (viral load) was assessed at the health center affiliated with the service provider with a lower limit of quantification of <50 copies/ml (“undetectable”). Viral load changes of one log or greater from baseline to three-months were considered clinically relevant (Dehovitz et al., 2000).
Four blood pressure readings, two taken on each arm, were measured with an automatic sphygmomanometer and averaged (Carlson, Speca, Faris, & Patel, 2007). Participants were asked to relax and sit quietly during the readings.
Demographic information, experience with yoga, and HIV history was collected.
Randomization.
The principal investigator generated the randomization scheme on Random.org and stratified by incarceration in jail or prison. Participants opened their own treatment assignment envelope. The principal investigator and a research assistant, not blinded to treatment assignment, completed the baseline interviews.
Data Analysis
Analysis followed the intent to treat principle. Participants who were withdrawn after randomization because they did not have an HIV diagnosis were excluded from analysis. Participants who had no pharmacy refill data either due to erroneous pharmacy information or death were considered to have zero adherence. Missing follow-up data were considered missing at random. We completed sensitivity analyses using multiple imputation with ten imputed data sets for outcomes that had greater than ten percent loss-to-follow-up rates, in order to address potential bias (Bennett, 2001).
Viral load.
We compared the number of individuals in the yoga versus TAU arms with a > 1 log increase and a > 1 log decrease in viral load from baseline to three-months using Fisher exact tests.
We compared the number of individuals in the yoga versus TAU arms in the following categories: no change in detectable status from baseline to three-months, undetectable at baseline and detectable at three-months, detectable at baseline and undetectable at three-months using a Fisher exact test.
ART adherence, heart rate and blood pressure.
Four separate analyses of covariance (ANCOVA; SAS PROC GLM) tests compared yoga and TAU on the following continuous dependent variables: 1.) ART medication adherence (percentage of ART pills refilled in the 90 days of non-incarceration from baseline to three-months), 2.) heart rate (beats per minute) at three-months, 3.) systolic blood pressure (mm Hg) at three-months, and 4.) diastolic blood pressure (mm Hg) at three-months. To take into account group differences, the baseline value of the respective dependent measure was used as a covariate in each of the latter analyses.
Cohen’s d effect size was calculated for statistically significant and nearly statistically significant associations. To do so, the difference between the yoga and TAU groups’ change in the dependent variable from baseline to three-months (based on the predicted value from the ANCOVA analysis) was divided by the pooled standard deviation of the dependent variable at baseline (Friedmann et al., 2008).
The target sample size of the parent study was 80, with 80% power and an effect size of 0.85 that was based on the primary outcome of change in stress, measured by the Perceived Stress Scale, between the experimental and control groups (Wimberly et al., 2018).
Results
Sample characteristics
One hundred eight people were screened. Seventy-five people were enrolled. Two people were withdrawn because they did not have an HIV diagnosis and were excluded from analyses. One person withdrew because he attended one class and reported that the class was too slow and aggravated a back injury (see Figure 1). Most participants had an undetectable viral load at baseline (see Table 1). Regarding pre-intervention ART medications, three yoga participants reported that they were not taking ART medications while all TAU participants reported that they were. One yoga participant with HIV maintained an undetectable viral load off ART medications (i.e., an “elite controller”) and was excluded as ineligible, leaving 72 participants eligible for adherence analysis.
Figure 1.
Consort diagram
Table 1.
Baseline characteristics of participants
| Yoga | Treatment as usual | |||
|---|---|---|---|---|
| (n = 37) | (n = 36) | |||
| Gender, n (%) | ||||
| Male | 29 | (78.38) | 21 | (58.33) |
| Female | 7 | (18.92) | 14 | (38.89) |
| Transgender | 1 | (2.70) | 1 | (2.78) |
| Race/Ethnicity, n (%) | ||||
| Black | 31 | (83.78) | 26 | (72.22) |
| White | 2 | (5.41) | 4 | (11.11) |
| Multiple races | 2 | (5.41) | 4 | (11.11) |
| Latino | 2 | (5.41) | 2 | (5.56) |
| Age in years, M (SD) | 43.30 | (10.59) | 45.61 | (10.33) |
| Income, M (SD) | $780.24 | (956.70) | $655.31 | (589.99) |
| Years of Education, M (SD) | 12.18 | (2.03) | 10.97 | (1.84) |
| Marital Status, n (%) | ||||
| Single | 32 | (86.49) | 31 | (86.11) |
| Married | 2 | (5.41) | 3 | (8.33) |
| Engaged/ committed relationship | 3 | (8.11) | 2 | (5.56) |
| Employment, n (%) | ||||
| Receive disability | 15 | (40.54) | 19 | (52.78) |
| Unemployed | 18 | (48.65) | 16 | (44.44) |
| Work | 4 | (10.81) | 1 | (2.78) |
| Years since HIV diagnosis, M (SD) | 15.70 | (8.49) | 13.08 | (7.96) |
| ART prescription refills in previous 90 days (M %) (SD) | 65.45 | (34.82) | 76.89 | (50.34) |
| Undetectable viral load, n (%) | 24 | (66.67) | 23 | (63.89) |
| Detectable viral load, n (%) | 9 | (25) | 11 | (30.56) |
| Currently taking ART, n (%) | ||||
| Yes | 34 | (91.89) | 36 | (100.00) |
| No | 3 | (8.11) | 0 | (0.00) |
| Yoga | TAU | |||
| ART when incarcerated, n (%) | ||||
| Yes | 26 | (70.27) | 30 | (83.33) |
| No | 11 | (29.73) | 6 | (16.67) |
| ART disbursement when incarcerated, n (%) | ||||
| Directly observed therapy | 18 | (69.23) | 24 | (80.00) |
| Keep on person | 8 | (30.77) | 6 | (20.00) |
| Recent incarceration, n (%) | ||||
| Jail | 30 | (81.08) | 30 | (83.33) |
| Prison | 7 | (18.92) | 6 | (16.67) |
| Days since release, M (SD) | 177.03 | (100.53) | 145.64 | (122.73) |
| Months last incarcerated, M (SD) | 16.28 | (35.16) | 9.06 | (23.75) |
| Most problematic drug, n (%) | ||||
| Crack cocaine | 15 | (40.54) | 14 | (38.89) |
| Heroin | 7 | (18.92) | 8 | (22.22) |
| Alcohol | 6 | (16.22) | 5 | (13.89) |
| Powdered cocaine | 5 | (13.51) | 5 | (13.89) |
| Heroin & crack/cocaine | 0 | (0.00) | 2 | (5.56) |
| Marijuana & K2 | 1 | (2.70) | 1 | (2.78) |
| MDMA | 1 | (2.70) | 0 | (0.00) |
| Methamphetamine | 1 | (2.70) | 0 | (0.00) |
| Heroin & alprazolam | 0 | (0.00) | 1 | (2.78) |
| Crack cocaine & alcohol | 1 | (2.70) | 0 | (0.00) |
| Previous yoga classes, n (%) | ||||
| None | 24 | (64.86) | 26 | (72.22) |
| 1-5 classes | 9 | (24.32) | 9 | (25.00) |
| 1-3 years (1-4 classes per month) | 4 | (11.11) | 1 | (2.78) |
Pharmacy refill data at three-months was collected for 34 (94%) of yoga participants and 35 (97%) of TAU participants. Two yoga participants provided erroneous pharmacy information. They were considered to have zero adherence in the analysis. One TAU participant died during the course of the intervention. They were considered to have zero adherence in the analysis. Viral load tests at three-months were completed for 27 (73%) of yoga participants and 31 (86%) of TAU participants. Twenty-eight (78%) of yoga participants and 32 (89%) of TAU participants completed the three-month heart rate and blood pressure assessment. During the course of the study, seven yoga participants and nine TAU participants were re-incarcerated. Yoga participants on average attended about four classes (median: 3.5, range: 0-12). Eight (22%) participants attended zero classes, seven (19%) attended nine-twelve classes and 22 (59%) attended one-eight classes (see Table 2 for reasons for non-attendance).
Table 2.
Reasons for non-attendance to yoga classes
| Self-reported reason | Number of classes missed |
|---|---|
| Incarceration | 48 |
| Illness | 34 |
| Employment | 24 |
| Medical or Parole Appointments | 10 |
| Inpatient substance use treatment | 8 |
| Transportation | 4 |
| Caretaking for ill family member | 3 |
| Overslept | 2 |
| Forgot | 1 |
| Involvement in a fight | 1 |
Viral load
There was no difference between the yoga and TAU groups in regard to viral load status changes from baseline to three-months (p < 0.63, see Table 3). There were no differences between the yoga and TAU groups in regard to number of people who had increases in viral load of > log 1 (p < 0.24) or decreases in viral load > log 1 (p < 0.99). Four yoga participants and five TAU participants had a decrease > 1 log from baseline to three-months. No yoga participants and three TAU participants had an increase > 1 log from baseline to three-months. Multiple imputation analyses did not alter results for viral load (data not shown).
Table 3.
Counts and percentages of viral load status at baseline and three-months among participants who completed both viral load assessments
| Yoga (n=27) | TAU (n=31) | |
|---|---|---|
| No change from baseline to three-months, n (%) | 23 (85.19) | 25 (78.13) |
| Undetectable at baseline & detectable at three-months, n (%) | 0 (0.00) | 2 (6.45) |
| Detectable at baseline & undetectable at three-months, n (%) | 4 (14.81) | 4 (12.90) |
ART medication adherence
ART medication adherence was twelve percentage points lower for yoga (65%, SD= 35) than TAU (77%, SD= 50) at baseline. At three-months, the ANCOVA model estimated that ART adherence increased for yoga to 75%, 95% CI [62, 87] and decreased for TAU to 66%, 95% CI [54, 78]. The mean change difference from baseline to three-months between yoga and TAU represented a medium effect size of d=0.46. After controlling for baseline ART medication adherence, yoga did not have a statistically significant effect on subsequent ART medication adherence at three-months, [F (1,69)=0.89, p < 0.35]. Multiple imputation analyses were not completed for ART medication adherence because there was less than 10 % missing data.
Heart rate, and blood pressure
After controlling for the respective baseline variable, yoga did not have a significant effect at three-months on heart rate [F (1,57)=0.69, p < 0.41], systolic blood pressure [F (1,57)=0.07, p < 0.79], nor diastolic blood pressure [F (1,57)=0.00, p < 0.98, See Table 4]. Multiple imputation analyses did not alter results for heart rate or blood pressure.
Table 4.
Baseline values and three-month ANCOVA estimates of heart rate and blood pressure: Yoga versus TAU
| Baseline value M(SD) |
ANCOVA estimate M, 95% CI |
|
|---|---|---|
| Heart rate | ||
| Yoga | 75 (10) | 81 [77, 85] |
| TAU | 79 (11) | 79 [75, 83] |
| Systolic blood pressure | ||
| Yoga | 129 (12) | 129 [124, 134] |
| TAU | 125 (18) | 128 [123, 133] |
| Diastolic blood pressure | ||
| Yoga | 86 (12) | 88 [84, 92] |
| TAU | 86 (12) | 88 [84, 91] |
Adverse Events
In the first cohort, a study participant experienced minor back pain after the fourth yoga class, reporting that he had over-stretched. In response, the yoga instructor offered specific adjustments and poses for back pain. The pain subsided and had resolved 10 days after that class. No other adverse effects were detected or reported.
Discussion
This study found no statistically significant treatment effect of yoga on ART medication adherence, viral load, heart rate, or blood pressure. Despite the lack of statistical significance, ART adherence increases in the yoga group compared with ART adherence decreases in TAU suggest the potential of yoga to impact improved ART adherence. Further, adverse effects were rare and reversible, indicating yoga is a safe intervention for this population.
Our findings contribute to a nascent body of literature that has found reductions in stress-related outcomes among people living with HIV who practice yoga, including the outcomes of anxiety, pain, utilization of pain medication (Bonadies, 2004), perceived stress (Agarwal et al., 2015), and blood pressure (Cade et al., 2010). We add to this literature by being among the first studies to look at the effect of yoga on ART adherence. Our study also addresses limitations of prior research including very small sample sizes (Agarwal et al., 2015; Bonadies, 2004) and lack of experimental design (Bonadies, 2004).
The lack of statistically significant findings in our study may be due to several factors. First, the sample size for the study was chosen for an effect on stress, not ART adherence. Given the large variance in ART adherence, the sample size may have been too small to find an effect on ART adherence. In addition, the sample was largely ART adherent at baseline, potentially resulting in a ceiling effect. Similarly, the majority of participants were virally suppressed at baseline and thus it was unlikely to see a change in three months. This finding corresponds with Cade et al’s (2010) study that found no immunological or virological changes among participants randomized to yoga who were largely virally suppressed at baseline. To see viral load status changes, future studies should consider enrolling samples with larger sample sizes and enroll participants who have lower levels of ART adherence, detectable loads at baseline and implement interventions of longer duration.
Regarding heart rate and blood pressure, we found no significant treatment effect of yoga. These findings are discordant with Cade et al. (2010), who found that individuals living with HIV and cardiovascular disease risk factors who participated in a yoga intervention experienced greater decreases in blood pressure than a standard of care treatment group. These conflicting results may be due to differences in each of the study’s intervention intensity and attendance rate. While our intervention included twelve 90-minute weekly yoga classes, Cade et al.’s (2010) included twenty 60-minute yoga classes two-three times per week. In our study, participants attended an average of four classes, while in Cade et al.’s (2010) study, participants attended an average of 33 classes. Therefore, it may be that yoga interventions need to offer more class sessions than our study and boost attendance rates in order to find statistically significant changes in heart rate and blood pressure.
Our findings may also demonstrate that yoga does not impact ART adherence. Returning citizens face a number of challenges to maintaining ART adherence and a once-a-week yoga intervention may be insufficient to counter these challenges. Additionally, with the low yoga class attendance rates, we may have an incomplete picture of the potential effects of yoga. It may be that with improved class attendance levels, yoga could impact ART adherence.
Strengths & limitations
Strengths of this study include the randomized design that presumably controlled for other unmeasured factors that might contribute to a change in outcomes, such as motivation or additional resources. Further, the intervention was implemented with a population that is in great need because of the very stressful time of reentry and the decline of HIV care during reentry. By collecting pharmacy prescription refill data we had very high rates of ART medication adherence follow-up data.
In addition to the previously mentioned limitations, the TAU group did not receive an intervention of equal intensity as the yoga group. Therefore, we cannot rule out the extra time and attention received in the yoga intervention as a causal mechanism, as opposed to the actual content of the yoga classes. In addition, while recruiting from the service provider helped this study meet its eligibility criteria and maintain low loss-to-follow-up rates, such a recruitment strategy neglected people with HIV in reentry who are most in need of HIV care. Finally, monetary reimbursement for the assessments may have encouraged participation by people who otherwise would be uninterested in yoga. However, reimbursement is important to acknowledge participants’ time and effort. Further, yoga class attendance was not compensated, and so it is unlikely that reimbursement affected class attendance rates.
Conclusion
This study found evidence that yoga was a safe intervention which may have an effect on important outcomes like ART adherence among returning citizens living with HIV and substance use problems, although no statistical associations between yoga and TAU were found. Future studies that explore the effect of yoga on ART adherence in reentry are warranted, with studies incorporating larger sample sizes, and more frequent yoga sessions. The absence of adverse effects suggest that yoga is not contra-indicated for this population, but more studies are needed to elucidate if and how it might work. It is also important to consider that reentry is a particularly trying time that requires multiple intensive interventions and support. Yoga, an individual intervention that may not be attractive to everyone, is insufficient on its own to address the many challenges of reentry, but it is one potential support.
Footnotes
Ethics Statement:
All study procedures were in accordance with the ethical standards of the institutional review board of the service provider from which the study sample was recruited and a local university. All participants provided verbal agreement to the screening assessment, and written informed consent. Participants received $35 at baseline, $40 at three-months, and an additional $10 if they completed all of the substance use assessments. Tokens for public transportation travel were provided for all study related activity.
Contributor Information
Alexandra S. Wimberly, UM School of Social Work, 525 West Redwood Street, Baltimore, MD 21201 USA.
Robert Gross, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA.
Molly Layde, US Peace Corps, Masasi, Mtwara, Tanzania.
References
- Agarwal RP, Kumar A, & Lewis JE (2015). A pilot feasibility and acceptability study of yoga/ meditation on the quality of life and markers of stress in persons living with HIV. Journal of Alternative and Complementary Medicine, 21(3), 152–158. 10.1089/acm.2014.0112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ammassari A, Trotta MP, Murri R, Castelli F, Narciso P, Noto P, … Antinori A (2002). Correlates and predictors of adherence to highly active antiretroviral therapy: Overview of published literature. Journal of Acquired Immune Deficiency Syndromes, 31, S123–S127. [DOI] [PubMed] [Google Scholar]
- Arnsten JH, Demas PA, Grant RW, Gourevitch MN, Farzadegan H, Howard AA, & Schoenbaum EE (2002). Impact of active drug use on antiretroviral therapy adherence and viral suppression in HIV-infected drug users. Journal of General Internal Medicine, 17(5), 377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Baillargeon J, Giordano TP, Rich JD, Wu ZH, Wells K, Pollock BH, & Paar DP (2009). Accessing antiretroviral therapy following release from prison. JAMA, 301(8), 848–857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barai N, Monroe A, Lesko C, Lau B, Hutton H, Yang C, … Monroe A (2017). The association between changes in alcohol use and changes in antiretroviral therapy adherence and viral suppression among women living with HIV. AIDS and Behavior, 21(7), 1836–1845. 10.1007/s10461-016-1580-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Beckwith CG, Zaller ND, Fu JJ, Montague BT, & Rich JD (2010). Opportunities to diagnose, treat, and prevent HIV in the criminal justice system. Journal of Acquired Immune Deficiency Syndromes, 55, S49–S55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bennett DA (2001). How can I deal with missing data in my study? Australian and New Zealand Journal of Public Health, 25(5), 464–469. [PubMed] [Google Scholar]
- Bilderbeck AC, Farias M, Brazil IA, Jakobowitz S, & Wikholm C (2013). Participation in a 10-week course of yoga improves behavioural control and decreases psychological distress in a prison population. Journal of Psychiatric Research, 47(10), 1438–1445. [DOI] [PubMed] [Google Scholar]
- Bonadies V (2004). A yoga therapy program for AIDS-related pain and anxiety: Implications for therapeutic recreation. Therapeutic Recreation Journal, 38(2), 148–166. [Google Scholar]
- Cade WT, Reeds DN, Mondy KE, Overton ET, Grassino J, Tucker S, … Yarasheski KE (2010). Yoga lifestyle intervention reduces blood pressure in HIV-infected adults with cardiovascular disease risk factors. HIV Medicine, 11, 379–388. 10.1111/j.1468-1293.2009.00801.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carlson LE, Speca M, Faris P, & Patel KD (2007). One year pre – post intervention follow-up of psychological, immune, endocrine and blood pressure outcomes of mindfulness-based stress reduction (MBSR) in breast and prostate cancer outpatients. Brain, Behavior, and Immunity, 21, 1038–1049. 10.1016/j.bbi.2007.04.002 [DOI] [PubMed] [Google Scholar]
- Castilla J, del Romero J, Hernando V, Marincovich B, Garcia S, & Rodreiguez C (2005). Effectiveness of highly active antiretroviral therapy in reducing heterosexual transmission of HIV. Journal of Acquired Immune Deficiency Syndromes, 40(1), 96–101. [DOI] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention. (2012, December). New HIV infections in the United States. Retrieved April 30, 2015, from http://www.cdc.gov/nchhstp/newsroom/docs/2012/hiv-infections-2007-2010.pdf
- Clements-Nolle K, Marx R, Pendo M, Loughran E, Estes M, & Katz M (2008). Highly active antiretroviral therapy use and HIV transmission risk behaviors among individuals who are HIV infected and were recently released from jail. American Journal of Public Health, 98(4), 661–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dehovitz JA, Kovacs A, Feldman JG, Anastos K, Young M, Cohen M, … Greenbkatt RM (2000). The relationship between virus load response to highly active antiretroviral therapy and change in CD4 cell counts: A report from the women’s interagency HIV study. The Journal of Infectious Diseases, 182, 1527–1530. [DOI] [PubMed] [Google Scholar]
- French T, Tesoriero J, & Agins B (2011). Changes in stress, substance use and medication beliefs are associated with changes in adherence to HIV antiretroviral therapy. AIDS and Behavior, 15(7), 1416–28. [DOI] [PubMed] [Google Scholar]
- Friedmann PD, Rose JS, Swift R, Stout RL, Millman RP, & Stein MD (2008). Trazodone for sleep disturbance after alcohol: A double-blind placebo-controlled trial. Alcoholism: Clinical and Experimental Research, 32(9), 1652–1660. 10.1111/j.1530-0277.2008.00742.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gonzalez A, Mimiaga MJ, Israel J, Bedoya CA, & Safren SA (2013). Substance use predictors of poor medication adherence: The role of substance use coping among HIV-infected patients in opioid dependence treatment. AIDS and Behavior, 17, 168–173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gross R, Yip B, Re VL III, Wood E, Alexander CS, Harrigan PR, … Hogg RS (2006). A simple, dynamic measure of antiretroviral therapy adherence predicts failure to maintain HIV-1 suppression. Journal of Infectious Diseases, 194(8), 1108–1114. [DOI] [PubMed] [Google Scholar]
- Grossberg R, & Gross R (2007). Use of pharmacy refill data as a measure of antiretroviral adherence. Current HIV/AIDS Reports, 4, 187–191. [DOI] [PubMed] [Google Scholar]
- Grossberg R, Zhang Y, & Gross R (2004). A time-to-prescription-refill measure of antiretroviral adherence predicted changes in viral load in HIV. Journal of Clinical Epidemiology, 57, 1107–1110. 10.1016/j.jclinepi.2004.04.002 [DOI] [PubMed] [Google Scholar]
- Institute of Behavioral Research. (2007). Texas Christian University Drug Screen II. Fort Worth: Texas Christian University, Institute of Behavioral Research; Retrieved from ibr.tcu.edu [Google Scholar]
- Iroh PA, Mayo H, & Nijhawan AE (2015). The HIV care cascade before, during, and after incarceration: A systematic review and data synthesis. American Journal of Public Health, 105(7), e5–e16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leserman J, Ironson G, O’Cleirigh C, Fordiani JM, & Balbin E (2008). Stressful life events and adherence in HIV. AIDS Patient Care and STDs, 22(5), 403–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li AW, & Goldsmith C-AW (2012). The effects of yoga on anxiety and stress. Alternative Medicine Review, 17(1), 21–35. [PubMed] [Google Scholar]
- Macy RJ, Jones E, Graham LM, & Roach L (2018). Yoga for trauma and related mental health problems: A meta-review with clinical and service recommendations. Trauma, Violence and Abuse, 19(1), 1–23. 10.1177/1524838015620834 [DOI] [PubMed] [Google Scholar]
- Maruschak LM (2015). HIV in prisons, 2001-2010. (NCJ 238877). Washington, DC: Bureau of Justice Statistics. [Google Scholar]
- Meyer JP, Cepeda J, Wu J, Trestman RL, Altice FL, & Springer SA (2014). Optimization of Human Immunodeficiency Virus treatment during incarceration: Viral suppression at the prison gate. JAMA Internal Medicine, 174(5), 721–729. 10.1001/jamainternmed.2014.601 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paterson DL, Swindells S, Mohr J, Brester M, Vergis EN, Squier C, … Singh N (2000). Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Annals of Internal Medicine, 133(1), 21–30. 10.7326/0003-4819-133-1-200007040-00004 [DOI] [PubMed] [Google Scholar]
- Porco TC, Martin JN, Page-Shafer KA, Cheng A, Charlebois E, Grant RM, & Osmond DH (2004). Decline in HIV infectivity following the introduction of highly active antiretroviral therapy. AIDS, 18(1), 81–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sethi AK, Celentano DD, Gange SJ, Moore RD, & Gallant JE (2003). Assocation between adherence to antiretroviral therapy and human immunodeficency virus drug resistance. Clinical Infectious Diseases, 37(8), 1112–1118. Retrieved from http://www.jstor.org/stable/4483627 [DOI] [PubMed] [Google Scholar]
- Sinha R (2001). How does stress increase risk of drug abuse and relapse? Psychopharmacology, 158(4), 343–59. [DOI] [PubMed] [Google Scholar]
- Sinha R (2008). Chronic stress, drug use, and vulnerability to addiction. Annals of the New York Academy of Sciences, 1141, 105–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Steiner JF, Koepsell TD, Fihn SD, & Inui TS (1988). A general method of compliance assessment using centralized pharmacy records: Description and validation. Medical Care, 26(8), 814–823. Retrieved from http://www.jstor.org/stable/3765465 [DOI] [PubMed] [Google Scholar]
- Stephenson BL, Wohl DA, Golin CE, Tien H-C, Stewart P, & Kaplan AH (2005). Effect of release from prison and re-incarceration on the viral loads of HIV-infected individuals. Public Health Reports, 120(1), 84–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Swan H (2015). Different patterns of drug use and barriers to continuous HIV care post-incarceration. Journal of Drug Issues, 45(1), 38–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Western B, Braga AA, Davis J, & Sirois C (2015). Stress and hardship after prison. American Journal of Sociology, 120(5), 1512–47. [DOI] [PubMed] [Google Scholar]
- Wimberly AS, Engstrom M, Layde M, & McKay JR (2018). A randomized trial of yoga for stress and substance use among people living with HIV in reentry. Journal of Substance Abuse Treatment, in press. [DOI] [PMC free article] [PubMed] [Google Scholar]