Yoga research on sleep quality includes controlled trials, subgroup analyses, and objective measurements. The evidence is mixed by population and outcome: self-reported sleep quality often improves, while some insomnia scales and actigraphy outcomes do not show significant changes. The findings below retain each study’s measurement period, population, comparison group, and source-reported statistic.
Contents
- Overall controlled-study evidence
- Older adults and PSQI outcomes
- Insomnia and sleep onset
- Objective sleep measurements
- Cancer and breast-cancer findings
- Dose and response patterns
Overall controlled-study evidence
The 2019 review The effect of yoga on sleep quality and insomnia in women with sleep problems: a systematic review and meta-analysis included 19 yoga sleep studies and 1,832 participants. Across 19 randomized controlled trial outcomes, yoga improved sleep problems with a standardized mean difference (SMD) of −0.327 (95% CI −0.506 to −0.148; P<0.001). The combined analysis had moderate statistical heterogeneity: I²=58.287%, Q=43.152, and P=0.001. These figures describe the review’s 2019 search period, not a current update. Source: 2019 systematic review
The strongest pooled result was found in studies using the Pittsburgh Sleep Quality Index (PSQI). Sixteen randomized trials reported better sleep quality with yoga than with control, with SMD −0.54 (95% CI −0.89 to −0.19; P=0.003). By contrast, three randomized trials using the Insomnia Severity Index (ISI) found no significant effect on insomnia severity: SMD −0.13 (95% CI −0.74 to 0.48; P=0.69). The difference matters because PSQI and ISI measure related but distinct subjective outcomes.
| Outcome or subgroup | Studies | Reported result |
|---|---|---|
| Overall sleep problems | 19 randomized outcomes | SMD −0.327; P<0.001 |
| PSQI sleep quality | 16 randomized trials | SMD −0.54; P=0.003 |
| ISI insomnia severity | 3 randomized trials | SMD −0.13; P=0.69 |
| Actigraphy sleep efficiency | 2 trials | SMD 0.85; P=0.26 |
| Actigraphy total sleep time | 2 trials | SMD −0.06; P=0.59 |
The same review found no significant yoga effect in two actigraphy trials for sleep efficiency, SMD 0.85 (95% CI −0.56 to 2.26; P=0.26), or total sleep time, SMD −0.06 (95% CI −0.26 to 0.13; P=0.59). It also reported no serious adverse effects across the 19 included studies, although two studies reported nonserious events.
Subgroup results were not uniformly positive. Seven breast-cancer randomized trials found no significant PSQI improvement, SMD −0.15 (95% CI −0.31 to 0.01; P=0.50). Four trials involving peri- or postmenopausal participants likewise found no significant PSQI improvement, SMD −0.31 (95% CI −0.95 to 0.33; P=0.34). These are subgroup findings from the 2019 review and should not be generalized to every yoga program or every person with sleep problems.
Older adults and PSQI outcomes
An elderly-home randomized trial assigned 62 participants to yoga and 58 to a waitlist control. The intervention delivered 60-minute yoga sessions daily for one month, followed by daily practice with weekly one-hour supervised sessions in months two and three. The study reported a six-month intervention and follow-up period. Source: Effects of yoga intervention on sleep and quality-of-life in elderly: A randomized controlled trial
The trial found a significant yoga-by-time interaction for total sleep quality after adjustment for education; corrected significance was set at P<0.01. When both groups were analyzed together, however, there was no overall time effect on total PSQI sleep quality: F=0.30, df=1,117, P=0.584. At the six-month endpoint, the yoga group was significantly better than the waitlist control on total sleep quality after Bonferroni correction at P<0.01.
A separate 65-person elderly cross-sectional comparison reported mean PSQI of 3.771±0.3623 among regular yoga practitioners and 8.000±0.4315 among non-yoga participants. The yoga group averaged 5.26 years of regular practice, while the comparison group had no regular yoga exposure. The study assessment period found less sleep disturbance among yoga participants (P<0.0001), as well as shorter sleep latency and better subjective sleep quality (P<0.0001 for each). Source: Impact of long term Yoga practice on sleep quality and quality of life in the elderly
Sleep-medication use was lower among the elderly yoga participants in that comparison (P<0.05), while daytime dysfunction did not differ significantly. Because this was a cross-sectional comparison, its figures describe differences between groups during the study assessment period; they do not establish that years of yoga caused every observed difference.
Insomnia and sleep onset
In an eight-week randomized trial of chronic primary sleep-onset insomnia, 20 adults per group received either Kundalini yoga or active sleep-hygiene instruction. The yoga group had greater total sleep time, with between-group d=0.95 (P=0.002), and greater sleep efficiency, d=1.36 (P<0.001). Sleep-onset latency was lower with yoga, d=−1.16 (P<0.001). Pre-sleep arousal did not change significantly, d=−0.30 (P=0.59). Source: Treatment of chronic primary sleep onset insomnia with Kundalini yoga
The six-month follow-up supplied additional threshold results. At least 80% of yoga participants reported average sleep-onset latency under 30 minutes and sleep efficiency above 80%. More than 50% reduced their ISI scores by at least eight points at both treatment end and follow-up. These percentages apply to the yoga participants in that trial, not to all people practicing yoga.
A smaller 2018–2019 sleep-lab pilot randomized 18 adults with insomnia symptoms: nine to Yoga Nidra and nine to control. Baseline PSQI averaged 9±3 in the Yoga Nidra group and 10±4 in control, with both averages above the study’s poor-sleep threshold. During the second sleep-lab visit, seven of nine Yoga Nidra participants fell asleep during practice, with average sleep-onset latency of nine minutes. Sleep efficiency during the practice averaged 39±33%. Source: A Closer Look at Yoga Nidra—Early Randomized Sleep Lab Investigations
Of the seven Yoga Nidra participants who fell asleep, three—or 43%—did not wake when the practice ended. In the same pilot, 78% (7/9) fell asleep during practice. These are sleep-lab observations from a very small group, and the practice-period sleep efficiency is not equivalent to a full-night sleep-efficiency estimate.
Objective sleep measurements
Actigraphy provides a different perspective from questionnaire scores. In a four-week cancer-survivor randomized controlled trial, yoga produced greater actigraphy-measured improvement in wake after sleep onset than standard care (P<0.01). Yoga also produced greater improvement in actigraphy-measured sleep efficiency than standard care (P<0.01). The largest wake-after-sleep-onset reductions occurred among survivors with at least 60 minutes of baseline wake after sleep onset, while the greatest sleep-efficiency gains occurred among those whose baseline efficiency was 60% or lower. Source: Multicenter, Randomized Controlled Trial of Yoga for Sleep Quality Among Cancer Survivors
A six-week fibromyalgia pilot found that PSQI sleep-disturbance scores fell from 11.3±4.5 to 10.0±4.40 after yoga, reported as a 12% change (P=0.027). Actigraphy sleep efficiency changed from 87.1±7.6% to 85.5±9.9%, which was not significant (P=0.59). Higher week-six actigraphy sleep efficiency correlated with greater reduction in PSQI sleep disturbance (rho=−0.40; P=0.03) and with fatigue reduction (rho=−0.395; P=0.034). Source: Impact of daily yoga-based exercise on pain, catastrophizing, and sleep amongst individuals with fibromyalgia
The Yoga Nidra sleep-lab pilot also found no significant visit-to-visit sleep-efficiency difference: change was −7±25% after Yoga Nidra versus −2±35% in control (P=0.73). Together, these actigraphy findings illustrate why subjective improvement and device-measured change may not move in parallel.
Cancer and breast-cancer findings
The cancer-survivor randomized trial enrolled 410 participants. Ninety-six percent were female, the mean age was 54 years, and 75% had breast cancer. The yoga program used two 75-minute sessions per week for four weeks. Both groups began with mean global PSQI scores of 9.0, above the clinical criterion of 8.0 used in that report. Source: Multicenter, Randomized Controlled Trial of Yoga for Sleep Quality Among Cancer Survivors
The yoga group’s pre-to-post sleep-quality effect size was d=0.62, compared with d=0.37 in standard care. Yoga participants reduced sleep-medication use by 21% per week, while standard-care participants increased use by 5% per week. In the post-intervention survey, 90% of cancer survivors rated yoga useful for improving sleep quality. All participants who received yoga said they would recommend it to other survivors with sleep problems, and 63% would highly recommend it.
A separate breast-cancer pilot delivered eight weekly 60-minute group yoga sessions. From baseline to one-month follow-up, the study reported a PSQI group-by-time interaction of F=3.975 (P=0.010; partial eta²=0.110) and a sleep-disturbance interaction of F=4.116 (P=0.017; partial eta²=0.114). Source: Effects of a tailor-made yoga program on upper limb function and sleep quality in women with breast cancer
These breast-cancer findings differ from the nonsignificant breast-cancer subgroup in the 2019 women’s meta-analysis. The studies used different programs, samples, outcome timing, and designs, so their results should remain separate rather than being treated as one pooled estimate.
Dose and response patterns
The 2019 women’s meta-analysis reported several differences by measurement and population. PSQI-based studies had Hedges g=−0.369, compared with g=0.031 for studies using other sleep instruments (P=0.002). Non-breast-cancer studies had Hedges g=−0.522, compared with −0.148 in breast-cancer studies (P=0.001). Studies without peri- or postmenopausal participants had Hedges g=−0.419, compared with −0.084 in studies that included those participants (P=0.003). Source: 2019 systematic review
Meta-regression found a positive association between total class hours and significant sleep results (P=0.003). Mean age was associated with effect differences (P=0.003), with younger study samples more likely to show significant results. Sample size was also associated with effect differences (P=0.032), with smaller studies more likely to show significant results. These associations describe patterns across studies; they do not specify a universally effective number of classes or hours.
Study duration was not a significant moderator of sleep effects in the meta-regression (P=0.20). Results also differed by region: American-region studies had Hedges g=−0.123 versus −0.844 in other regions (P<0.001). The combination of heterogeneity, subgroup differences, and measurement-specific results supports a careful reading of yoga-for-sleep statistics: the clearest signals in this evidence set concern reported sleep quality and selected sleep-onset outcomes, while objective measures and some clinical subgroups remain less consistent.