CATEGORY: STRESS & MIND-BODY
Sleep specialists have a first choice for chronic insomnia, and it is not a pill. Researchers in Hong Kong wanted to know whether a slow movement practice could hold up against it. Over fifteen months, it did. Insomnia-severity scores fell 9.51 points with tai chi and 10.18 points with cognitive behavioural therapy, measured on a 28-point scale, a gap of 0.68 points, across the 200 adults in a randomized trial reported in The BMJ.1 At three months, though, the talk therapy had been clearly ahead.
One Treatment Works on the Mind, the Other on the Feet
Cognitive behavioural therapy for insomnia, or CBT-I, is a structured program that retrains sleep habits along with the racing, clock-watching thoughts that keep people awake. It is what sleep specialists reach for before medication. Tai chi is a slow, flowing sequence of gentle postures and controlled breathing, often called a moving meditation. In this trial the two arrived in the same package. Both ran as one-hour group classes, twice a week for three months, 24 sessions in all.
A Coin Toss in Hong Kong, Then a Year of Waiting
Two hundred Chinese adults aged 50 and older with chronic insomnia joined the trial. A coin toss, not a preference, decided who learned which program. Sleep trouble was scored on a standard questionnaire running from 0 to 28, where a higher number means worse nights and more daytime fallout. The researchers asked a narrow question. Not whether tai chi would beat the established therapy, but whether it would land no worse than a margin they had fixed before anyone enrolled. Scores were taken when the classes finished at three months, and again a full year after that.
Slow Off the Mark, Level at the Tape
At three months tai chi had not kept up. Scores dropped 6.67 points with tai chi against 11.19 with CBT-I, and tai chi failed the no-worse test. A year later the picture had turned. The tai chi benefit kept building while the therapy group held steady, and by fifteen months the two reductions sat less than a point apart, inside the pre-set margin. CBT-I behaved like a sprinter and tai chi like a distance runner. Neither group reported any adverse events. The trial does have a boundary. Both arms got an active treatment, so it cannot say how either group would have fared with nothing at all.
Against No Treatment, Tai Chi Has Its Own Trials
That gap has been filled elsewhere. An earlier randomized trial by the same group followed 320 older adults with insomnia and measured their sleep with sensors rather than questionnaires. Tai chi beat no treatment: about 13 fewer minutes awake in bed after first dropping off, roughly two fewer awakenings a night, and the gains were still there two years on.2 A 2024 review that pooled 12 randomized trials and 1,058 community-dwelling older adults found tai chi improved sleep-quality ratings by about 2 points on a 21-point scale against control groups.3 What the new trial adds is the harder comparison.
Key Takeaways
- Tai chi trailed CBT-I badly at three months, 6.67 points of improvement against 11.19, then caught up by fifteen months: 9.51 against 10.18 on a 28-point scale.
- Both programs ran as twice-weekly group classes for three months, and the difference at fifteen months stayed inside the margin the researchers had set in advance.
- The trial had no untreated arm, so it measures tai chi against the standard therapy rather than against doing nothing.
- Earlier randomized work found tai chi improved sleep compared with no treatment in older adults, with benefits lasting up to two years.
References
- Siu PM, Yu DJ, Yu AP, et al. Tai chi or cognitive behavioural therapy for treating insomnia in middle aged and older adults: randomised non-inferiority trial. BMJ. 2025;391:e084320. link
- Siu PM, Yu AP, Tam BT, et al. Effects of Tai Chi or Exercise on Sleep in Older Adults With Insomnia: A Randomized Clinical Trial. JAMA Network Open. 2021;4(2):e2037199. link
- Lyu L, Li Y, Fan X, Seo J, Eunna O. The effect of Tai Chi exercise intervention on sleep quality in community residing older adults: a systematic review and meta-analysis. European Geriatric Medicine. 2024;15:381–396. link