Yoga is recommended for sleep. Against plain stretching, it did nothing extra.

Yoga is routinely recommended as a way to sleep better. The recommendation is old enough that it rarely gets examined, and the trials behind it usually compare yoga with doing nothing — which tests whether an hour of structured movement three times a week helps, not whether yoga specifically does.

A secondary analysis published in JAMA Network Open in December ran the harder comparison. It took 240 women randomised to either a three-month hatha yoga programme or a time-equivalent general physical conditioning programme, and looked at what happened to their sleep [s1].

Nothing distinguished them.

The design

The parent trial recruited ambulatory women aged 45 to 90 with daily urinary incontinence at three sites in Northern California, and randomised them to twelve weeks of twice-weekly group instruction plus once-weekly self-directed practice of pelvic-floor-specific hatha yoga, or to an equivalent-time programme of general skeletal muscle stretching and strengthening [s2]. That trial reported in Annals of Internal Medicine in August 2024 that pelvic yoga was not superior to the conditioning programme for the outcome it was designed to test: incontinence frequency fell by 2.3 episodes per day with yoga and 1.9 with conditioning, a between-group difference of −0.3 episodes per day (95% CI −0.7 to 0.0) [s2].

The December analysis is a prespecified look at sleep in the same participants, conducted from 2018 to 2022 and analysed from June to September 2024 [s1]. The design matters: the comparison arm is not rest, it is the same amount of instructed, group-based, physical activity without the yoga.

What they found

Among the 240 participants (mean age 62.0 years, SD 8.7), 133 (55.4%) had poor sleep quality at baseline, defined as a Pittsburgh Sleep Quality Index score above 5 [s1]. Of the 239 with data, 133 (55.7%) reported waking at least once a night to use the bathroom [s1].

Over three months, the mean change in PSQI score was 0.37 points (95% CI 0.78 to −0.04) in the yoga group and 0.66 points (95% CI 1.07 to 0.25) in the conditioning group — a between-group difference of 0.29 points (95% CI −0.28 to 0.86) [s1].

Wakefulness after sleep onset changed by 3.82 minutes (95% CI 8.15 to −0.52) with yoga and 6.97 minutes (95% CI 11.20 to 2.74) with conditioning, a between-group difference of 3.16 minutes (95% CI −2.84 to 9.16) [s1].

Those confidence intervals cross zero. On the trial's own reading, a three-month yoga programme did not produce superior sleep quality or reduced sleep disruption compared with general physical conditioning [s1].

The finding that was not about yoga

The analysis surfaced something with a clearer signal, and it had nothing to do with which programme anyone was assigned to.

Across all participants regardless of intervention, greater nocturnal voiding frequency tracked worse sleep quality. Estimated mean PSQI score was 5.83 (95% CI 5.35–6.32) among participants with fewer than one nocturnal voiding episode per night, and 6.66 (95% CI 6.00–7.33) among those with at least two (P = .02 for linear trend across increasing categories) [s1].

That is a modest difference on a 21-point scale, in a cross-sectional comparison within a trial, and it does not establish direction — poor sleep and night-time urination plausibly drive each other. But it points at the bladder rather than the mat.

What this does and does not settle

This is a secondary analysis of a trial designed around incontinence, in women aged 45 and over who had daily incontinence at enrolment [s1][s2]. The participants were not selected for poor sleep, although a majority had it. Any effect specific to people with diagnosed insomnia is untested here.

The intervention was also converted to videoconference delivery partway through, after the start of the COVID-19 pandemic [s1][s2]. Whether an in-person yoga class differs from a remote one on sleep outcomes is not something this trial can answer.

And a null result against an active comparator is not a null result against nothing. Both groups' PSQI scores moved in the same direction. The trial cannot say whether either programme beat sitting still, because there was no such arm.

An invited commentary published alongside the analysis framed the open question as whether sleep quality should be expected to improve in patients receiving yoga for other conditions at all [s3].

Why it matters

The commercial and clinical case for yoga as a sleep intervention rests largely on comparisons with inactive controls. This trial suggests that when the comparison holds time, instruction and group structure constant, the yoga-specific contribution to sleep is small enough that a 240-person trial could not find it.

That is a narrower claim than "yoga does not help sleep." It is the claim that what helps may be the structured movement, not the tradition attached to it — which is testable, and which most existing trials were not built to test.

What to watch

Whether trials of yoga for sleep begin routinely using time-matched active comparators rather than waitlists, and whether the nocturia signal is followed up as a treatable cause of disturbed sleep in midlife and older women.

This article is informational and is not medical advice.

Sources

• [s1] Hough E, Goldstein LA, Subak LL, et al. "Effects of Hatha Yoga vs Physical Conditioning on Sleep in Women With Urinary Incontinence: A Secondary Analysis of a Randomized Clinical Trial." JAMA Network Open, 8(12), December 2025. https://doi.org/10.1001/jamanetworkopen.2025.46499
• [s2] Huang AJ, Chesney M, Schembri M, et al. "Efficacy of a Therapeutic Pelvic Yoga Program Versus a Physical Conditioning Program on Urinary Incontinence in Women: A Randomized Trial." Annals of Internal Medicine, 177(10), published online 27 August 2024. https://doi.org/10.7326/m23-3051
• [s3] Roseen EJ, Lee S. "Should Sleep Quality Improve in Patients Receiving Yoga for Other Conditions?" JAMA Network Open, 8(12), December 2025. https://doi.org/10.1001/jamanetworkopen.2025.46414