One Form Of Exercise Ranked Best For Sleep Across 30 Trials. Walking Finished Less Than A Point Behind

Ask four research teams which exercise does the most for sleep, and you’ll get three different answers.

A 2025 analysis from Harbin Sport University picked yoga. Two others picked Pilates, and a fourth picked aerobic exercise. Even the yoga result came with an asterisk: walking finished less than a point behind, close enough that the study’s own statistics couldn’t separate the two. Understanding why the rankings disagree starts with how yoga reached the top in the first place.

How Did Yoga Reach The Top?

By probability. The analysis, published in Sleep and Biological Rhythms by researchers at Harbin Sport University in China, pooled 30 randomized trials of people with sleep disturbances. It compared seven kinds of exercise: yoga, walking, resistance exercise, combined aerobic and resistance training, dance, aerobic exercise, and traditional Chinese exercises such as tai chi and qigong.

Sleep was measured with the Pittsburgh Sleep Quality Index. This survey asks people to rate their overall sleep quality, how long it takes them to fall asleep, how long they sleep, how much of their time in bed is spent asleep, how often their sleep is disturbed, whether they use sleep medication, and how much trouble they have functioning during the day.

The researchers used a method called network meta-analysis, which lets scientists compare exercises that were never tested head to head. Suppose one trial pits yoga against no exercise and another pits walking against no exercise. The method uses that shared comparison point to estimate how yoga and walking stack up against each other.

A ranking tool called SUCRA then scored each exercise from 0 to 100 based on how likely it was to land near the top. Yoga scored 82.1, walking 81.2, and resistance exercise (strength training) 66.3. These are ranking scores, not percentages of improvement.

The underlying comparisons tell a more modest story. Of the seven exercise types, only yoga and walking clearly improved sleep compared with no exercise, and yoga also outperformed aerobic exercise. Resistance exercise, based on just two trials, did not clearly beat doing nothing. The comparisons of yoga against walking and against resistance exercise were inconclusive.

The authors urged caution because of the limited number of studies and the particular traits of people with sleep disturbances. They wrote that “high-quality research is needed to confirm these findings.”

Was The Highest-Ranked Routine Tested As One Program?

No. The researchers ranked exercise type, weekly frequency, session length, intensity and program length separately, then combined the top finisher from each category.

Twice-weekly exercise scored 83.3, sessions of no more than 30 minutes scored 85.1, high-intensity activity scored 90.3, and programs lasting eight to 10 weeks scored 92.0. Each of those rankings pooled trials of all exercise types. Only one trial, a resistance-training study, was rated high intensity. None of the four yoga trials reported intensity or combined all the other winning elements. The authors note that no trials have compared different exercise prescriptions head-to-head in this group, so the routine remains a hypothesis that needs direct testing.

The trials also covered a wide mix of people, from older adults with sleep complaints to patients with cancer, rheumatoid arthritis, Parkinson’s disease, heart failure and obstructive sleep apnea. The authors advise clinicians to weigh each patient’s age, sleep disorder and exercise habits before prescribing a routine.

Why Did Other Reviews Find Different Winners?

Because rankings shift with the studies, people and outcomes included.

A BMC Public Health network meta-analysis published in April 2025 pooled 86 randomized trials involving 7,276 adults ages 18 to 75. It counted daytime sleepiness and insomnia scales alongside the Pittsburgh index. Pilates ranked first at 91.7, and yoga came last of six exercise types at 30.1. Yet every exercise type improved sleep compared with no exercise, and the authors found no statistically significant differences among most of them. Most comparisons were rated moderate to low in evidence quality. A lopsided funnel plot (a chart researchers use to spot missing studies) suggested possible publication bias or small-study effects.

A 2024 review in the Journal of Clinical Sleep Medicine, by Beijing Sport University researchers, pooled 28 randomized studies and 3,460 middle-aged and older adults. Twenty-five of the 28 studies focused on older adults. Comparing aerobic exercise, resistance exercise, combined training and yoga, it ranked aerobic exercise first for the overall Pittsburgh score at 93.2. Yoga led for four parts of the questionnaire: sleep disturbances, sleep efficiency, sleep duration and daytime dysfunction.

A separate 2024 analysis in Preventive Medicine, by another Beijing Sport University team, covered 27 randomized trials involving 2,142 healthy adults. Pilates ranked highest, and Pilates, yoga and traditional Chinese exercises all improved sleep compared with no exercise.

What Does The Evidence Actually Support?

Yoga is a reasonable option, with limits. The 30-trial analysis did not show that everyone should swap walking or weight training for high-intensity yoga, and it did not address when during the day to exercise.

For someone choosing an exercise to stick with, the ranking matters less than the headline suggests. Every review found that some forms of exercise improved sleep compared with doing nothing, but they disagreed about which worked best. The Preventive Medicine authors suggested tailoring exercise to what each person enjoys and needs.

This article is for informational purposes only and is not medical advice. People with sleep disorders or other health conditions should talk with a doctor before starting a new exercise program, especially a high-intensity one.