A Systematic Review Investigating the Nonparticipation Rate and Recruitment Challenges in Yoga Trials for Addictive Disorders

Abstract

Yoga has been studied as an effective alternative therapy for substance use disorders, demonstrating its ability to reduce stress and addictive behaviors while improving self-control and self-esteem. However, not everyone may be willing or able to engage in yoga for various reasons. Furthermore, there is a lack of systematic reviews and meta-analyses of the existing research on this topic. The objective is to explore the nonparticipation rate in yoga intervention randomized controlled trials related to addictive disorders, and to explore the reasons for nonparticipation in yoga trials related to addictive disorders. We searched Google Scholar, CENTRAL, and PubMed for peer-reviewed published RCTs that incorporated yoga interventions in any format of at least one session. Data were extracted using standard procedures, and the risk of bias was evaluated. Pooled analysis and sensitivity analysis were done. About 54% of potential participants were not included. The pooled effect size was 0.544 (95% confidence interval: 0.301–0.787), with an I² = 99.797. Key barriers included failure to meet inclusion criteria, such as medical conditions, mental health issues, and specific health conditions such as posttraumatic stress disorder, depression, and heart disease. Participant refusals and lack of interest in randomization or the study itself were common reasons for nonparticipation. Scheduling conflicts were also frequently reported as a significant obstacle. Two broad categories for noninclusion were meeting exclusion criteria, and not interested/declined to participate. Understanding reasons for noninclusion can lead to more efficient and effective recruitment efforts in future trials and also possibly guide the design of future studies to minimize dropout rates and ensure a representative sample.

Introduction

Yoga is a mind–body exercise of Indian origin that has been shown to enhance mental and physical health and lessen perceived stress and anxiety. Hatha yoga, the most common form practiced in the West, combines physical postures (asanas), breath control (pranayama), and meditation (dhyana) to promote relaxation.[1] Yoga has been used to treat stress, depression, anxiety, and childhood disorders[2,3,4,5,6,7] and is also effective in substance use disorders.[8,9] Particularly, yoga has been researched as a successful alternative therapy for substance use disorders and has been shown to lower stress and addictive behaviors while enhancing self-control and self-esteem.[10]

Research on yoga’s impact on substance use is limited, but it is often used as a complementary approach in treatment programs. Studies suggest that yoga may help reduce substance use; for instance, a randomized trial found that Hatha yoga and group therapy both led to decreases in drug use and illegal behavior among individuals in methadone treatment.[1] Substance use disorders typically have a prolonged course, characterized by relapses and remissions. It aids Yoga may help during detoxification and relapse prevention by reducing stress and managing triggers. Yoga breathing exercises help lower heart rate and blood pressure during detoxification.[11,12] Regular yoga practice improves emotional regulation, reducing relapse risks. Yoga has an impact on how people with substance use disorders process emotions and how aroused they are. When a person with substance use disorder is subjected to cues or an unpleasant emotion, they often relapse. Regular yoga practice lowers stress levels and improves a person’s ability to manage negative emotions.[13]

Not everyone is inclined or able to practice yoga for various reasons. Literature highlights that common barriers to participation across the general population[14] and individuals with conditions such as cardiovascular disorders,[15] gastrointestinal issues,[16] and cancer[17] include strict eligibility criteria, limited awareness, accessibility challenges, skepticism about efficacy, time constraints, physical limitations, fear of side effects, personal disinterest, financial barriers, and logistical issues. In those with chronic illnesses, fatigue and low energy are particularly significant obstacles.[18,19]

By understanding the factors which contribute to nonparticipation in yoga-related trials, it can also help us identify and address the barriers to enrolling in such trials. Analyzing previous studies, we can further identify effective recruitment strategies and develop new approaches to enhance participation rates. This review highlights how recruitment challenges may impact the generalizability of study findings. Addressing these challenges can improve the external validity of yoga trials, making the results more applicable to the general population. A systematic review synthesis such as this can guide future research efforts and contribute to the academic literature on yoga interventions for mental health and addiction.

This systematic review aims to explore the nonparticipation rate in yoga intervention randomized controlled trials related to addictive disorders and also aims to explore the reasons for nonparticipation in yoga trials related to addictive disorders.

Methods

Search strategy

This review covered the trials that were published from the beginning to July 2024. Google Scholar, CENTRAL, and PubMed were the databases that were employed for methodical search. Boolean search strategy was employed using the keywords: Yoga AND (addiction OR alcohol OR smoking OR nicotine OR “use disorder” OR dependence). The search process was carried out systematically through several steps. First, the predefined search string was entered into each database and the results were sorted by relevance. Relevant articles were then shortlisted based on their titles and abstracts. For larger databases such as Google Scholar, filtering was continued until approximately three to four out of every 10 titles per page were identified as relevant. In contrast, all articles retrieved from Cochrane and PubMed were screened in full, and inclusion, as the number of records in these databases was more manageable. Following this identification step, eligibility assessment, including screening, the subsequent PRISMA stages were completed to arrive at the final set of studies included in the analysis. Hand searches and contacting of authors were not done. The study protocol was uploaded to open science framework prior to searches with the link: Https://osf.io/xcq7f/?view_only=93fd4ea8c88b4df185fca0378de55c28, Study selection: This review included only randomized controlled trials (RCTs). The selection criteria for studies included peer-reviewed published RCTs that incorporated yoga interventions in any format of at least one session. Studies were considered eligible if they reported the number of participants screened or approached for the study, as well as the number of individuals randomized. Exclusion criteria ruled out studies where yoga was one of the components of a multi-component intervention, but not the yoga intervention condition being studied as a separate arm. In addition, trials that may have included individuals with addictive disorders, but where individuals with addictive disorders comprised a separate group/arm in the trial were excluded. This systematic review targeted studies involving participants diagnosed with addiction or substance use disorders, defined according to any classification system, including those addressing behavioral addiction. Interventions considered included any form of yoga, without restrictions on the type, duration, or length of practice, provided there was at least one yoga group. Studies were included if they incorporated one or more control groups, which could consist of either passive or active controls. No restrictions were placed on the outcomes reported, allowing for the inclusion of studies regardless of the specific results.

The primary outcome of this systematic review was the nonparticipation rate in yoga trials targeting addiction. Nonparticipation rate was defined as the proportion of individuals who met the eligibility criteria for the study but either opted not to join or were excluded before the trial began. The secondary outcome included assessing the causes of nonparticipation. Data extraction focused on several details from the included studies: Author and publication year, country of origin, study setting (e.g., outpatient or emergency, etc.), type of addictive disorder studied, type of yoga, trial duration, number of yoga sessions, comparative or control groups, the number of participants approached, and the number ultimately randomized. Additional extracted information included reasons for noninclusion when reported, incentives or compensation offered for participation, and demographic characteristics of the approached participants, such as mean age, percentage of females, and employment status.

Risk of bias

Risk of bias (RoB) revised Cochrane risk-of-bias tool was used for assessment of the quality of the included randomized controlled trials.

Statistical analysis

The data extracted from the included studies were presented in tabular format and a qualitative synthesis was done of the characteristics of the studies and the intervention and control groups. Pooled analysis (meta-analysis) of the inclusion rate (i.e., the proportion of the participants randomized among the potential participants approached for inclusion in the yoga trial) was done. Random effects model were used due to the potential heterogeneity among the studies. Der-Simonian Laird method of pooled analysis (meta-analysis) was used. I² test of heterogeneity was used to formally assess the heterogeneity. Analysis was performed using OpenMetaAnalyst software (Open Source Software, Brown University, Providence, RI, USA).[20] For the additional outcome, the nonparticipation rate ascribed to different reasons (for example, age outside the eligible range, or nonparticipation because of the reason of noninterest in yoga) was also pooled. Subgroup analyses were conducted based on the availability of studies. It included differences by types of yoga, substance use disorders and behavioral addictions, country of origin (especially economic group), duration of intervention, number of yoga sessions, and type or amount of incentive offered to participants. Sensitivity analyses included leave-one-out meta-analysis to examine the influence of outliers.

Results

This systematic review included 11 RCTs that presented data on nonparticipation in yoga trials in addictive disorders from diverse geographical locations, including six studies from the USA, two from India, one from Nepal, one from China, and one from Sweden. As shown in Table 1, among the 11 studies reviewed, five by Bock et al., in 2012;[21] Reddy et al., in 2014;[22] Jeffries et al., in 2020;[23] Gunillasdotter et al., in 2022[24] and Wimberly et al., in 2018[25] utilized Hatha yoga, with Gunillasdotter et al., in 2022[24] also incorporating Ashtanga, Les Mills Body Balance, Yin Yoga, and Yin Release. Three studies done by Gaihre et al., in 2023;[26] Zhuang et al., in 2013;[27] Lotfalian et al., in 2020[28] had focused on a combination of breathing and relaxation exercises. Bock et al., in 2019[29] used Iyengar yoga as the intervention, while Chauhan and Negi, in 2024[30] included a mix of sukshma vyayama, suryanamaskar, shavasana, asanas, pranayama, aum chanting, and meditation. Sarkar et al., in 2017[31] implemented Kapalbhati and Anulom Vilom techniques.

Table 1.

Studies included in systematic review

Author year		Country		Yoga type		Yoga session week		Total yoga session		Duration trial (min)		Comparison
Bock et al., 2012[21]		USA		Vinayasa hatha		8		16		60		Wellness
Reddy et al., 2014[22]		USA		Kripalu-based hatha yoga		4		12		75		Assessment only
Zhuang et al., 2013[27]		China		Breathing training, meditation, asanas (postures done with awareness), and relaxation techniques		24		120		50		Routine hospital care
Sarkar et al., 2017[31]		India		Kapalbhati and anulom vilom		1		1		15		Brief quit advice
Wimberly et al., 2018[25]		USA		Hatha yoga		3		12		90		Treatment as usual
Bock et al., 2019[29]		USA		Iyengar		8		16		60		Wellness
Jeffries et al., 2020[23]		USA		Hatha yoga (intervention is a video recording)		1		1		30		Wellness control
Lotfalian et al., 2020[28]		USA		Ujjayi breathing		1		2		20		Cognitive strategy group and no treatment group
Gunillasdotter et al., 2022[24]		Sweden		Ashtanga and hatha yoga, les mills body balance, yin yoga and yin release		12		84		60		Aerobic exercise and treatment as usual
Gaihre et al., 2023[26]		Nepal		Combination of surya namaskar, kapalabhati, asana, pranayama, meditation and relaxation technique		12		72		90		Physical exercise
Chauhan and Negi, 2024[30]		India		Combination of sukshma vyayama, suryanamaskar, shavasana, asanas, pranayama, aum chanting, meditation		8		56		60		Treatment as usual

Yoga sessions ranged from 1 week to 24 weeks. The total number of sessions also varied widely, from a single session in brief interventions like those studied by Sarkar et al., in 2017[31] in India to 120 sessions reported by Zhuang et al., in 2013[27] in China. Session durations ranged from brief 15-min[31] practices to 90-min sessions by Wimberly et al., in 2018[25] and Gaihre et al., in 2023.[26]

The studies included diverse control groups. Bock et al., in 2012[21] and in 2019[29] in the USA, investigated yoga as part of wellness programs, comparing its impact against general wellness activities. Gaihre et al., in 2023[26] compared its outcomes with those from physical exercise. In China, Zhuang et al., in 2013[27] examined yoga combined with routine hospital care. Sarkar et al., in 2017[31] in India, employed Kapalbhati and Anulom Vilom, comparing their effects against brief quit advice. Gunillasdotter et al., in 2022,[24] Chauhan and Negi, in 2024,[30] and Wimberly et al., in 2018[25] kept the treatment as usual.

As indicated in Table 2, the final sample size ranged from 38 to 1,213 individuals (median sample size was 80, cumulative sample size was 2148). Yoga-specific groups within these studies included 20–611 participants. Participants’ mean ages ranged from 25.2 to 54 years. Gender composition varied considerably, with some studies involving only male participants and others exclusively female partcipants. Employment status among yoga groups, when reported, ranged from 10.8% to 80%.

Table 2.

Studies included in systematic review

Author year		Sample size approached		Sample final		Sample yoga		Reason noninclusion		Mean age yoga		Percentage fem yoga		Percentage emp yoga
Bock et al., 2012[21]		184		55		32		Medical condition: 58%, Major depression: 47%, Scheduling conflicts: 18% Smoking <5 cigarettes per day: 15%		43.8 (9.4)		100		59
Reddy et al., 2014[22]		96		38		20		Not met for PTSD criteria: 48.27% Did not complete screening: 27.58% Recently attended yoga classes: 12.06% Unable to attend classes: 6.89% Did not agree to be randomised: 3.44% Outside age eligibility: 1.72%		45.5 (12.1)		100		NA
Zhuang et al., 2013[27]		90		81		37		Declined: 3.33% Liver diseases: 3.33% Severe heart diseases: 2.22% Severe bone metastases: 1.11%		29.13 (7.69)		100		24.3
Sarkar et al., 2017[31]		1412		1213		611		Declined: 16.15% Ineligible due to severely mentally challenged: 0.08% Did not complete baseline assessment: 0.08%		45.2 (12.8)		22.6		71.7
Wimberly et al., 2018[25]		102		75		38		Did not meet inclusion criteria: 81.48% Scheduling conflicts: 14.81% Declined to participation: 3.70%		43.3 (10.59)		18.9		10.8
Bock et al., 2019[29]		1065		255		113		Not eligible: 54.4% Declined to participate: 36.5% Eligible but unable to schedule: 7.2% Failed to complete screener: 1.9%		46.1 (12.0)		59.3		58.4
Jeffries et al., 2020[23]		867		55		25		Smoking <5 cigarettes per day: 21.61% Unable to safely exercise: 19.16% Unwilling to attend appointments: 17.57% Practiced yoga once per week or more during past year: 14.76% Using NRT or smoking cessation counseling: 6.50% Recently decreased smoking by at least half: 4.74% Smoked for <1 year: 2.10% Pregnant: 1.40% Low motivation to change smoking: 1.40% Under age 18 years: 1.23% Planning to change psychiatric treatment during study: 0.17%		NA		32		64
Lotfalian et al., 2020[28]		113		60		20		Uninterested/scheduling conflicts: 26.54 Ineligible due to insufficient smoking rate: 15.9% Due to insufficient smoking duration: 14.15% Health problem: 0.88% Seeking treatment for smoking: 0.88%		39.55 (12.78)		50		NA
Gunillasdotter et al., 2022[24]		472		140		46		Met exclusion criteria: 46.38% Declined to participate: 26.50% Did not meet inclusion criteria: 19.27% Other reasons (missed visit, not in Stockholm during intervention period, participating in other trial): 7.83%		54 (13)		61		80
Gaihre et al., 2023[26]		110		96		48		Not meeting inclusion criteria: 71.4% Refused to participate: 28.5%		25.2 (6.4)		0		NA
Chauhan and Negi, 2024[30]		100		80		40		Did not meet inclusion criteria: 20%		NA		0		NA

NA: Not available, PTSD: Posttraumatic stress disorder, NRT: Nicotine replacement therapy

As depicted in Table 2, reviewed studies identified several common reasons for nonparticipation in yoga-related trials. Key barriers included failure to meet inclusion criteria, such as medical conditions, mental health issues, and specific health conditions such as PTSD, depression, and heart disease.[21,22,27] Scheduling conflicts were also frequently reported as a significant obstacle.[21,23,25] Participant refusals and lack of interest in randomization or the study itself were common reasons for nonparticipation.[31,26] Previous yoga experience and low smoking rates were also cited as reasons for exclusion.[23,28]

Risk of bias

Table 3 summarizes the RoB across the included studies. Five trials scored 3, two trials scored 4, and four trials scored 5, as detailed in Table 3.

Table 3.

Risk of bias in the studies

Authors		Random sequence		Allocation concealment		Blinding of participants/personnel		Blinding of outcome assessment		Incomplete outcome data		Jadad score
Bock et al., 2012[21]		+		+		−		?		+		4
Reddy et al., 2014[22]		+		+		?		?		+		3
Zhuang et al., 2013[27]		+		+		−		?		+		4
Sarkar et al., 2017[31]		+		+		+		+		+		5
Wimberly et al., 2018[25]		+		+		?		?		+		3
Bock et al., 2019[29]		+		+		+		+		+		5
Jeffries et al., 2020[23]		+		?		−		?		+		3
Lotfalian et al., 2020[28]		+		+		?		?		+		3
Gunillasdotter et al., 2022[24]		+		+		?		?		+		3
Gaihre et al., 2023[26]		+		+		+		+		+		5
Chauhan and Negi, 2024[30]		+		+		+		+		+		5

+: Attribute present, −: Attribute absent, ?: Unclear

Effect sizes

Figure 1 presents the forest plot of the included studies. The pooled effect size was 0.544 (95% CI: 0.301–0.787), with an I² value of 99.797, indicating high heterogeneity.

Figure 1.

Forest plot of included studies

Sensitivity analyses

The leave-one-out sensitivity analysis demonstrated robustness of the pooled effect size, with consistent estimates across all recalculations [Figure 2]. The overall pooled proportion was 0.544 (95% CI: 0.301–0.787, P < 0.001). When individual studies were excluded, the recalculated effect sizes remained stable, ranging from 0.509[27] to 0.593[23], with all confidence intervals overlapping the overall estimate. Removal of one of the studies did not lead to substantially divergent results.

Figure 2.

Leave-one-out forest plot

Discussion

This systematic review aimed to investigate and understand the nonparticipation rate in yoga intervention across substance use disorder and addictive behaviors. Understanding participation and engagement in such mind–body interventions is crucial, as yoga has been shown to reduce craving, improve emotion regulation, and enhance treatment adherence in addiction recovery settings.[10,13]

More than half of the participants were not included in the trials. A wide range of individuals declined to participate which ranged as low as 3.33% to as high as 36.5%. The key reasons contributing to nonparticipation included – two broad categories: (a) failure to meet inclusion criteria or exclusion criteria (medical comorbidities, mental health conditions, prior yoga exposure, low levels of substance use), and (b) disinterest in participation/declined to participate which were either due to (lack of time, lack of perceived need for yoga, scheduling conflicts and unwillingness to undergo randomization). Similar findings were noted across yoga studies in substance use disorders, where Bock et al. and Reddy et al. reported that refusal to participate was mainly linked to disinterest in random allocation or logistical constraints.[21,22] These findings parallel those reported in the broader yoga-intervention literature. Studies across psychiatric illnesses such as depression, anxiety disorders, schizophrenia and medical conditions such as diabetes, hypertension, chronic pain, and cancer have consistently shown relatively high dropout rates ranging from 15% to 35%, despite demonstrated therapeutic benefits of yoga-based interventions.[2,19,32,33,34,35] Logistic barriers, including travel distance and scheduling difficulties, have been the most frequently cited causes for discontinuation, followed by domestic responsibilities, lack of motivation, skepticism about efficacy, and physical health constraints.[2,19,32,33,34] In a systematic review of randomized controlled trials, Cramer et al. reported dropout rates of 20%–25% across yoga interventions for both psychiatric and medical illnesses, with logistic and time-related factors predominating.[2] Similarly, Mishra et al. observed comparable attrition patterns in a diabetes-prevention yoga trial, where scheduling difficulties, domestic responsibilities, and doubts regarding efficacy were major contributors.[19] Baspure et al. evaluated barriers to yoga participation among patients with schizophrenia and found that more than half of the eligible participants (56.8%) declined to participate, primarily due to distance from the center (37%), lack of accompaniment (11%), busy schedules (9%), and unwillingness to attend daily sessions (5%), rather than issues related to research design or randomization.[32] These studies provided an important insight into the challenges faced by individuals while undergoing yoga as an intervention across general medical conditions and psychiatric disorders. At the same time, the current review highlighted some additional attitudinal barriers unique to addiction settings.

Studies such as Reddy et al. and Bock et al. noted explicit refusal to be randomized, whereas others reported a lack of motivation or perceived irrelevance of yoga.[22,29] Sarkar et al., in an add-on yoga trial for individuals with substance use disorders, found that only 24% of approached patients consented, with leading reasons for nonparticipation being lack of time (73.2%), logistical challenges (52%), and lack of familiarity with yoga (33%).[36] Refusal to randomization, low motivation, and perceived irrelevance often stem from a mix of autonomy issues, practical constraints, and cultural or cognitive perceptions about yoga’s role and value in clinical and lifestyle contexts.

The findings from both Indian and international studies consistently highlight yoga’s beneficial role as an adjunct in addiction treatment. As per Indian studies, participants with opioid dependence who received yoga-based interventions showed significant reductions in craving intensity, anxiety, and withdrawal symptoms, along with improvements in sleep quality, emotional regulation, and overall psychological well-being compared to those receiving standard care alone.[37,38] Similarly, international research supports these observations. Khanna and Greeson found that yoga and mindfulness practices enhance self-awareness, stress tolerance, and emotional stability, factors that are critical for preventing relapse.[8] Hallgren et al. demonstrated that yoga as an adjunct to conventional therapy for alcohol dependence led to improved mood, decreased alcohol consumption, and greater engagement with treatment programs.[39]

Hence, it is important to address these barriers and enhance participation. Motivational ambivalence can also play an important role in nonparticipation among patients suffering from addictive disorders. Addressing these barriers requires motivational interviewing-based strategies, collaborative involvement to enhance autonomy, and interventions that build confidence and readiness for change among individuals struggling with motivational ambivalence. The motivational ambivalence and autonomy issues are, therefore, more prominent in addiction, even though both populations face similar practical challenges. Unstable housing situations, demanding jobs, and fluctuating motivation are common challenges for people in recovery, and they may lower their desire to engage in structured behavioral interventions.

For the advancement of research, it is crucial to address areas requiring attention and prioritize strategies that enhance participation, ultimately benefiting patients. In light of this, we would attempt to make some conclusions from this review and meta-analysis. First, this systematic review included data from 11 RCTs across diverse geographical locations, examining nonparticipation in yoga trials for addictive disorders. Most studies were conducted in specific regions, such as the USA and India. The findings highlight multiple barriers, including strict inclusion criteria, scheduling conflicts, lack of interest, and prior yoga experience, which restrict participant enrolment. Second, the studies varied widely in intervention types, session durations, and control groups, reflecting heterogeneity in methodologies. To enhance the inclusivity and effectiveness of future research, it is essential to address these challenges by adapting eligibility criteria, improving accessibility, and fostering participant interest through tailored interventions and awareness programs. Furthermore, recognizing noninclusion in trial designs as an unavoidable aspect of research emphasizes the importance of respecting individuals’ liberty, autonomy, and personal values within the healthcare context.

There are some limits to the generalizations that can be drawn from the present review. Most studies were conducted in specific regions, such as the USA and India, limiting the applicability of findings to global populations. Cultural differences in attitudes toward yoga and addiction treatment may influence both participation rates and outcomes, potentially reducing generalizability. Reasons for nonparticipation were inconsistently reported and varied widely, creating gaps in understanding the barriers and potential biases in recruitment. Some studies included only male or female participants, further restricting the generalizability of findings. The review process has certain limitations. These include the use of a limited selection of search engines, restricting the search to studies published in English, the absence of a formal assessment to ensure inter-rater consistency during the search process, the lack of subgroup analysis based on study quality/sample size and the lack of assessment of publication bias.

Conclusion

Taking all these limitations into consideration, it can be said that there is a need to conduct large-scale pragmatic, multicentric trials. By understanding reasons for noninclusion, the findings can lead to more efficient and effective recruitment efforts in future trials and also guide the design of future studies to minimize dropout rates and ensure a representative sample. This includes addressing potential biases and making studies more accessible and appealing to a broader population.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

The work was funded in part by the grant received by Dr Siddharth Sarkar (DST SATYAM/2020/457 (G)).