INTRODUCTION
Psychological stress is a critical determinant of mental and physical health. Chronic psychological stress is associated with emotional difficulties, cardiovascular responses, altered autonomic activity and changes in physiological stress regulation. Because psychological and physiological responses to stress are tightly interconnected, interventions that influence both domains are of vital importance for health promotion and preventive care.
Yoga is a traditional mind body practice that typically combines physical postures, controlled breathing, meditation, relaxation and attentional regulation. Unlike conventional physical exercise, yoga incorporates both physical and psychological components and this multi-componential nature have led to increased scientific interest in yoga as an intervention for psychological stress and physiological regulation. The autonomic nervous system is one of the major physiological systems related to stress adaptation. Heart rate variability (HRV) is a widely used non-invasive measure of cardiac autonomic regulation. Changes in HRV can provide information regarding the balance and flexibility of autonomic cardiac control. Therefore, HRV has become an important physiological outcome in yoga research.
One randomized controlled trial (RCT) conducted amongst mental-health professionals examined the effects of a 12-week yoga program on work-related stress, stress adaptation and autonomic activity. The study specifically measured HRV before, during and after the intervention and therefore provides direct evidence linking psychological stress with autonomic outcomes [1]. Psychological outcomes have also been studied in healthy adults. Szaszkó et al. carried out an eight-week RCT of Hatha yoga amongst healthy yoga novices. The study reported reductions in self-reported stress and stress reactivity and an increase in mindfulness, although significant intervention-related changes in state or trait anxiety were not found [2]. A subsequent RCT looked at whether Hatha yoga could influence both subjective stress and biological stress markers. 98 participants were randomized to an eight-week Hatha yoga intervention or a wait-list control. Subjective momentary stress was reduced in the yoga group, whereas diurnal salivary cortisol and alpha-amylase did not significantly change [3]. This shows that psychological and biological stress responses may not necessarily change in parallel. Cardiovascular outcomes have also been studied extensively. Wolff et al. investigated a 12-week home-based Kundalini yoga program in adults with hypertension and assessed blood pressure, quality of life, stress, anxiety and depression [4]. Hagins et al. looked at yoga against an active control in adults with prehypertension or stage-1 hypertension and used ambulatory blood pressure as the primary physiological outcome [5]. Similarly, Cohen et al. investigated a 12-week Iyengar yoga intervention in yoga-naïve adults with prehypertension or stage-1 hypertension [6]. Dhungana et al. carried out a multicentre RCT amongst hypertensive adults in Nepal. Participants underwent a health-worker led yoga intervention consisting of structured training followed by home-based yoga practice. The study demonstrated reductions in systolic and diastolic blood pressure compared with the wait-list control [7]. Psychological and emotional outcomes have been studied in people with hypertension. Tolbaños Roche et al. examined yoga, pranayama and meditation interventions in individuals with essential arterial hypertension and looked at self-regulation and emotional symptomatology [8]. There is also early evidence of randomized trials. Patel and North looked at yoga relaxation with biofeedback in hypertensive patients and compared the intervention with general relaxation. Both groups showed reductions in blood pressure with a greater reduction in the yoga/biofeedback group [9]. More recently, studies have looked at yoga-based breathing practices in clinical populations. Yadav et al. conducted a RCT amongst hospitalized adults with COVID-19 and assessed depression, anxiety, stress and fear of COVID-19. The yoga-based breathing group showed better psychological outcomes than the control group after the 10-day intervention [10]. Taken together, these studies suggest that yoga may influence multiple components of stress regulation. However, the evidence is heterogeneous as different studies have used different yoga styles, intervention durations, participants and outcome measures.
Therefore, our present systematic review focuses on RCT evidence examining psychological and physiological outcomes associated with yoga in adults.
Objective
The objective of the study was to assess the effect of yoga on psychological stress and physiological or autonomic regulation in adults as randomized controlled trials. The outcome of interest included measures of perceived stress, anxiety, depression, heart rate variability (HRV), blood pressure, cortisol, and alpha-amylase as well as the methodological limitations and research gaps.
Review Question
Among adults, what are the effects of yoga-based interventions on psychological stress and physiological/autonomic regulation compared with no intervention, usual care, wait-list or active control conditions?
METHODOLOGY
This systematic review was conducted to examine the effects of yoga-based interventions on psychological stress and physiological regulation in adults. Randomized controlled trials involving adults aged 18 years or older were considered eligible. Interventions included Hatha, Kundalini, Iyengar, integrated yoga, pranayama, yoga-based breathing and relaxation practices. Studies were included when they compared yoga with usual care, wait-list, active control or other non-yoga interventions and reported at least one relevant psychological or physiological outcome.
A systematic search was conducted using databases including PubMed/MEDLINE, Scopus, Web of Science, PsycINFO, SPORTDiscus and the Cochrane Library. Search terms combined yoga, stress/mental health, autonomic regulation/physiological outcomes, and randomized controlled trial terms. Studies such as reviews, meta-analyses, observational studies, case reports and non-randomized studies were excluded.
Data were extracted on author, year, sample characteristics, yoga intervention, duration, comparator, psychological outcomes, physiological outcomes and main findings. Due to differences in populations, interventions and outcome measures, findings were synthesized narratively. Risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB 2) framework.
| Component | Description |
|---|---|
| Population | Adults aged ≥18 years, including healthy adults, occupational groups, and adults with cardiovascular or other health conditions |
| Intervention | Yoga-based interventions, including Hatha yoga, Kundalini yoga, Iyengar yoga, integrated yoga, pranayama, yoga breathing, yoga relaxation, and yoga with meditation |
| Comparator | No intervention, usual care, wait-list, active exercise, general relaxation, enhanced usual care, or health education |
| Outcomes | Psychological: perceived stress, stress reactivity, anxiety, depression, emotional symptoms, psychological wellbeing, mindfulness. Physiological: HRV, autonomic activity, systolic/diastolic blood pressure, ambulatory blood pressure, cortisol, and alpha-amylase |
| Inclusion Criteria | Exclusion Criteria |
|---|---|
| Adults aged ≥18 years | Narrative reviews |
| Randomized controlled trials (RCTs) | Systematic reviews and meta-analyses |
| Yoga-based intervention | Case reports |
| Control or comparison group included | Editorials and commentaries |
| At least one relevant psychological or physiological outcome | Conference abstracts without sufficient data |
| Outcomes related to stress regulation | Non-randomized studies |
| Sufficient information on intervention and outcomes | Studies without relevant psychological or physiological outcomes |
| Human studies | Studies not meeting the predefined eligibility criteria |
| Study | Sample/Population | Yoga intervention | Duration | Comparator | Psychological outcomes | Physiological outcomes | Main finding |
|---|---|---|---|---|---|---|---|
| Lin et al. [1] | Mental-health professionals | Yoga | 12 weeks | Control | Stress, stress adaptation | HRV | Psychological stress and autonomic regulation were assessed together |
| Szaszkó et al. [2] | Healthy adults/yoga novices | Hatha yoga | 8 weeks | Wait-list | Stress, stress reactivity, mindfulness, anxiety | Physiological measures | Reduced perceived stress and stress reactivity |
| Szaszkó et al. [3] | Healthy adults | Hatha yoga | 8 weeks | Wait-list | Momentary stress | Cortisol, alpha-amylase | Reduced momentary stress; no clear change in diurnal biomarkers |
| Wolff et al. [4] | Adults with hypertension | Kundalini yoga | 12 weeks | Usual care | Stress, anxiety, depression, QoL | BP | Psychological and cardiovascular outcomes were evaluated |
| Hagins et al. [5] | Prehypertension/stage-1 hypertension | Yoga | Study-specific | Active control | — | Ambulatory BP | Evaluated yoga against an active comparator |
| Cohen et al. [6] | Prehypertension/stage-1 hypertension | Iyengar yoga | 12 weeks | Enhanced usual care | — | 24-h BP | Examined changes in ambulatory/24-h BP |
| Dhungana et al. [7] | Hypertensive adults | Integrated yoga | 3 months | Wait-list | — | SBP, DBP | BP reductions reported following yoga |
| Tolbaños Roche et al. [8] | Essential hypertension | Yoga + pranayama + meditation | Study-specific | Control | Emotional symptoms, self-regulation | BP | Psychological and cardiovascular outcomes improved in selected measures |
| Patel & North [9] | Hypertensive adults | Yoga relaxation + biofeedback | 6 weeks | General relaxation | Relaxation | BP | BP reduction observed |
| Yadav et al. [10] | Hospitalized COVID-19 adults | Yoga-based breathing | 10 days | Control | Depression, anxiety, stress, fear | — | Psychological outcomes improved following breathing intervention |
RESULTS
Characteristics of Included Studies
The identified randomized-controlled trials involved different adult populations, including healthy adults, mental-health professionals, patients with hypertension or prehypertension, and hospitalized adults. The authors used different yoga interventions, including Hatha yoga, Kundalini yoga, Iyengar yoga, integrated yoga, yoga breathing, pranayama, meditation, and yoga relaxation [1–10].
The interventions varied between studies, ranging from one-year programs to three-month interventions. The control group received either wait-list, standard, or enhanced care, and active treatments such as general relaxation [1–10]. The outcomes were also diverse, including psychological (stress, stress reactivity, anxiety, depression, and emotional symptoms), mindfulness and psychological wellbeing, physiological (heart rate variability (HRV), blood pressure (BP)), and biomarkers (cortisol and salivary alpha-amylase) [1–10].
Psychological Stress
The identified RCTs provide consistent evidence on the efficacy of yoga interventions in reducing subjective psychological stress. For example, Lin et al. [1] evaluated a 12-week yoga intervention among mental-health professionals and found that it reduced stress adaptation and improved HRV. The study is notable because it integrates the outcome of interest (i.e., stress) and HRV as a physiological marker.
Similarly, Szaszkó et al. [2] found that eight weeks of Hatha yoga practice reduced self-reported stress and stress reactivity while enhancing mindfulness among healthy adults. However, the intervention had minimal impact on state and trait anxiety [2]. In another trial, Szaszkó et al. [3] evaluated the efficacy of eight weeks of Hatha yoga on subjective and objective markers of stress. The intervention reduced subjective momentary stress but had no effect on diurnal salivary cortisol and alpha-amylase levels [3]. The finding highlights the differential effect of yoga interventions on subjective and objective measures of stress.
Anxiety and Depression
The evidence for the effect of yoga interventions on anxiety and depression was mixed. Wolff et al. [4] conducted a home-based Kundalini yoga trial among adults with hypertension and found that it reduced stress, anxiety, and depression. Notably, the study provides empirical evidence on the efficacy of yoga interventions in reducing both psychological and cardiovascular risks. Similarly, Yadav et al. [10] evaluated a yoga breathing program among hospitalized COVID-19 patients and found that it reduced depression, anxiety, and stress.
However, Tolbaños Roche et al. [8] found that yoga breathing reduced emotional symptoms while improving self-regulation among hypertensive adults with essential hypertension. The study adds to the growing body of evidence that yoga interventions are effective in reducing emotional symptoms and enhancing self-regulation.
Autonomic Regulation
Studies examining the effect of yoga interventions on autonomic regulation used the same methodology, which involved measuring HRV. For example, Lin et al. [1] integrated HRV as an outcome of interest in their trial, which evaluated the effect of a 12-week yoga intervention on stress adaptation among mental-health professionals. Indeed, the study is among the few trials that examine the effect of yoga interventions on both subjective psychological stress and HRV.
However, only a few studies have evaluated the effect of yoga interventions on HRV. As a result, the evidence base for the effect of yoga interventions on each of the following outcomes – RMSSD, SDNN, HF and LF – is limited.
Blood Pressure
Blood pressure was the most commonly examined cardiovascular outcome. For example, Cohen et al. [6] evaluated the effect of Iyengar yoga on 24-hour BP among adults with prehypertension or stage-1 hypertension. Similarly, Hagins et al. [5] examined the effect of a yoga intervention on ambulatory BP while comparing two active control groups.
Wolff et al. [4] conducted a home-based Kundalini yoga trial among adults with hypertension and found that it reduced systolic and diastolic BP. On the other hand, Dhungana et al. [7] evaluated a three-month health-worker led yoga intervention and found that it reduced systolic and diastolic BP among hypertensive adults. Patel and North [9] conducted an early RCT on the effect of yoga relaxation and biofeedback on hypertension and found that it reduced BP.
Overall, the evidence on the effect of yoga interventions on BP is consistent but mixed. Most studies found that yoga interventions had a significant effect on lowering BP, but the effect size varied between the trials.
Biological Stress Markers
Few studies have evaluated the effect of yoga interventions on biomarkers. For example, Szaszkó et al. [3] examined the effect of Hatha yoga on subjective stress and salivary cortisol and alpha amylase levels. The intervention had a significant effect on reducing self-reported momentary stress but had no effect on salivary cortisol and alpha amylase levels [3]. The findings demonstrate the need to measure both subjective and objective markers of stress in future trials.
Risk of Bias Interpretation
The main challenge in terms of methodology across the evidence base was the challenges in blinding participants and instructors in yoga interventions, while the psychological outcomes are vulnerable to self-report and expectation effects. The results such as blood pressure, HRV and salivary biomarkers can be measured objectively, although it is likely that procedures and timing can add additional variability.
Therefore, the evidence should not be interpreted based on if the study was randomized. Randomization, allocation concealment, missing data, outcome measurement and selective reporting should all be taken into account when interpreting the certainty of the evidence.
| Study | Randomization process | Deviations from intended intervention | Missing outcome data | Outcome measurement | Selective reporting | Overall consideration |
|---|---|---|---|---|---|---|
| Lin et al. [1] | Some methodological information required | Potential concern due to behavioural intervention | Assess from full text | Psychological outcomes self-reported; HRV objective | Requires protocol comparison | Some concerns |
| Szaszkó et al. [2] | Randomized design reported | Participant blinding difficult | Requires full-text assessment | Psychological outcomes mainly self-reported | Requires protocol comparison | Some concerns |
| Szaszkó et al. [3] | Randomized design reported | Blinding difficult | Requires full-text assessment | Biomarkers objectively measured | Requires protocol comparison | Some concerns |
| Wolff et al. [4] | Randomized trial | Behavioural intervention cannot be fully blinded | Requires full-text assessment | BP objective; psychological scales self-reported | Requires protocol comparison | Some concerns |
| Hagins et al. [5] | Randomized controlled design | Active intervention | Requires assessment | Ambulatory BP objective | Requires protocol comparison | Some concerns |
| Cohen et al. [6] | Randomized controlled design | Yoga participants unlikely to be blinded | Requires assessment | BP objective | Requires protocol comparison | Some concerns |
| Dhungana et al. [7] | Multicentre randomized design | Intervention adherence may vary | Requires assessment | BP objective | Requires protocol comparison | Some concerns |
| Tolbaños Roche et al. [8] | Randomized design | Blinding difficult | Requires assessment | Mixed subjective/objective outcomes | Requires protocol comparison | Some concerns |
| Patel & North [9] | Older randomized trial | Blinding difficult | Historical reporting limitations | BP objective | Protocol unavailable/limited | Some concerns / unclear |
| Yadav et al. [10] | Randomized controlled design | Behavioural intervention | Requires assessment | Psychological outcomes largely self-reported | Requires protocol comparison | Some concerns |
DISCUSSION
The present systematic review analysed the randomized controlled trial evidence exploring the effects of yoga on the psychological and physiological aspects of stress. Overall, the results demonstrated the beneficial effects of yoga interventions on psychological stress, while the physiological responses were more mixed. The most consistent finding was the reductions in the measures of perceived or momentary stress following yoga interventions. Szaszkó et al. [2] found reduced levels of perceived stress and stress reactivity following Hatha yoga, while another trial demonstrated reduced momentary stress following the similar general approach [3]. Lin et al.’s study [1] also included analysis of stress adaptation as well as HRV, thus demonstrating the importance of the simultaneous assessment of the psychological and autonomic aspects. Regarding the effects of yoga interventions on the symptoms of anxiety and depression, the results were partially consistent. While some studies demonstrated the beneficial effects of yoga on anxiety, depressive symptoms or overall emotional symptomatology [4,8,10], others failed to demonstrate statistically significant improvements in all of the psychological outcomes [2]. Possible explanations concern the differences in the participants’ pre-intervention levels of psychological stress, the duration of the interventions, the styles of yoga and the utilized outcome measures. An important observation was the distinction between the effects of yoga interventions on the subjective and objective domains of stress. Szaszkó et al. [3] found that the participants who practiced yoga exercises experienced reductions in terms of momentary stress without notable between-group differences in terms of diurnal salivary cortisol and alpha-amylase. This indicates that the alleviation of the perceived levels of stress does not necessarily correspond to changes in specific endocrine biomarkers. Future intervention studies should therefore consider exploring the effects of yoga on multiple domains of the stress response. The evidence on the effects of yoga on the autonomic regulation of stress was largely positive, but the relatively small number of studies prevents more conclusive remarks. The only study that directly integrated an assessment of heart rate variability as a part of the intervention [1] demonstrated the connection between the alleviation of psychological stress and improvements in several indices of autonomic regulation. Nevertheless, the relatively small number of trials that utilized HRV measurements prevents further detailed conclusions.
Finally, the assessment of the effects of yoga interventions on blood pressure was more consistent across studies, despite the small sample sizes. Multiple trials explored the effects of yoga interventions on patients diagnosed with hypertension or prehypertension [4–7,9], while the findings indicated mixed effects on systolic and diastolic blood pressure. These results were partially consistent with those of the previous trials, but the heterogeneity of the findings can be explained by the differences in the baseline values of participants’ blood pressure, the differences in the interventions and the comparison groups. The present discussion also highlights the importance of the utilization of active control trials. The comparison of yoga interventions to a wait-list or usual-care group can demonstrate the effects of enhanced physical activity, attention and relaxation, but they fail to isolate the unique effects of yoga. Active control trials are therefore necessary in order to isolate the unique effects of yoga interventions. The evidence from the reviewed studies also suggests the presence of a potential mind body pathway in which the alleviation of psychological stress contributes to reductions in autonomic and cardiovascular responses to stress. However, such a mechanism has yet to be empirically established, as the majority of the reviewed RCTs did not explore potential mediation effects. Finally, the heterogeneity of the interventions is another important limitation to consider. The reviewed studies utilized a variety of interventions, ranging from Hatha, Kundalini and Iyengar styles to integrated interventions. Given the substantial differences between the aforementioned approaches, it is evident that yoga interventions cannot be considered interchangeable. Overall, the reviewed studies demonstrated largely positive effects of yoga interventions on the psychological aspects of stress as well as mixed effects on physiological regulation. Therefore, the research on the effects of yoga should be expanded in order to include the interventions that utilize standardized protocols and simultaneously assess multiple domains of stress.
LIMITATIONS
There are several limitations to this systematic review. First, there was substantial heterogeneity among the included randomized controlled trials in terms of participant characteristics, health status, yoga style, intervention duration, frequency of sessions and comparator conditions. Second, psychological outcomes were often assessed using self-report questionnaires, which may be biased by participant expectations, intervention preferences and a lack of participant blinding. Third, physiological outcomes were not consistently measured across the trials. HRV, blood pressure, cortisol and alpha amylase were assessed in different subsets of trials. Fourth, HRV measurement procedures may vary with regard to recording duration, body position, respiratory control, time of day and analytical parameters, all of which can affect HRV findings. Fifth, biological stress markers such as cortisol show substantial circadian variation, and differences in sampling time, diet, sleep, and physical activity can influence results. Sixth, several studies were conducted in specific clinical or occupational populations, and this may limit generalizability to the broader adult population.
Seventh, yoga interventions are difficult to blind; participants usually know whether they have received yoga or not, which can lead to expectation and performance biases. Finally, the present evidence synthesis should not be considered a quantitative estimate of treatment effect since the included trials showed substantial variability in terms of study design and interventions. A formal meta-analysis would only be appropriate if outcome variables and intervention characteristics were found to be comparable.
FUTURE DIRECTIONS
Future studies should aim to improve methodological consistency in randomized controlled trials investigating the effects of yoga and stress regulation. First, studies should develop standardized reporting of yoga interventions, including details on yoga style, components, the duration of each session, the frequency of sessions per week, the qualifications of instructors, home practice and adherence. Second, future trials should simultaneously assess psychological and physiological outcomes. Using validated stress and anxiety scales in combination with HRV, blood pressure and biological markers would provide a deeper insight into the mind body response to yoga. Third, HRV should be more consistently used as an objective marker of autonomic regulation. Future studies should explicitly describe recording conditions, respiratory rate, duration of recordings and indices such as RMSSD, SDNN and frequency-domain analysis. Fourth, future RCTs should explore whether changes in psychological stress mediate physiological responses.
One possible model is:
Yoga → reduced psychological stress → improved autonomic regulation → improved cardiovascular regulation.
Formal mediation analysis could determine whether psychological adaptation acts as an intermediary mechanism linking yoga participation with physiological outcomes. Fifth, future research should include longer follow-up periods to determine whether the psychological and physiological improvements are sustained after supervised intervention periods. Sixth, studies should compare yoga with active exercise, relaxation and other mind body interventions. Such comparisons could determine whether the observed benefits are specific to yoga or reflect broader effects of physical activity and relaxation. Seventh, researchers should explore dose response relations by comparing different frequencies and durations of yoga practice. Finally, future trials should use adequately powered sample sizes, pre-registered protocols, transparent outcome reporting and modern risk-of-bias procedures to improve confidence in the evidence.
CONCLUSION
This systematic review of randomized controlled trials suggests that yoga interventions can influence both psychological stress and some physiological markers of stress. The most consistent finding across trials was a reduction in perceived or momentary stress, with mixed results for other psychological outcomes such as anxiety and depression. Physiological evidence primarily points to an impact on blood pressure and autonomic nervous system function, with limited evidence for an effect on heart rate variability. It is also notable that psychological and physiological stress responses do not always track together, as evidenced by several trials in which subjective stress was reduced without changes in objective measures such as salivary cortisol and alpha-amylase. The current findings highlight the need to consider multiple domains in future research on the effects of yoga and stress, including psychological, autonomic, cardiovascular, and endocrine responses. Randomized controlled trials are needed that employ standardized interventions, report adequately on allocation and other biases, and include sufficiently powered analyses of psychological and physiological outcomes. Furthermore, future trials should include comprehensive assessments of the mechanisms by which yoga may influence stress, including mediation analyses and dose–response evaluations. Yoga interventions appear to be a promising mind–body method for the investigation of the effects of stress and physiological regulation, but further research is needed to better understand these relationships.