Poor sleep after a stroke is common enough to be treated as background noise, and costly enough that it should not be. In the months after an infarct or hemorrhage, half to seven in ten survivors report insomnia, broken nights, or heavy daytime sleepiness. Those nights feed fatigue and fog. They also interfere with the brain’s capacity to reorganize, raise the odds of another vascular event, and leave people with less of a life to return to.
Drugs are the usual first reach. They can work. They also sedate, raise fall risk, and add one more interaction to a list that is already long. That gap is why a 2026 systematic review in Complementary Therapies in Medicine gathered the randomized trials on traditional Chinese exercises (Tai Chi, Baduanjin, Liuzijue, Yijinjing, and related forms) and asked what those practices do to sleep, and to the anxiety and depression that so often travel with it.
Nine trials, several hundred people
The protocol was registered before the search began (PROSPERO CRD420261360688). The authors worked through PubMed, Embase, Web of Science, CINAHL, Ovid, and the three large Chinese databases, CNKI, WanFang, and VIP, from each database’s start date through April 2026. Two reviewers screened independently. A third settled disagreements.
Only randomized trials in adults with confirmed ischemic or hemorrhagic stroke counted. The intervention had to join posture, breath, and attention; treadmill work and seated meditation without movement were out. At least one sleep outcome was required. Most trials used the Pittsburgh Sleep Quality Index, a questionnaire in which a higher score means a worse night. A few used other instruments: the Epworth Sleepiness Scale, or an apnea-hypopnea index from a portable monitor.
Nine papers remained, published between 2010 and 2025. Six were in Chinese, three in English. Eight were run in China, one in Japan. Six hundred ninety people were randomized; 681 finished, which is an unusually high completion rate. Groups were small to mid-sized, 34 to 106 participants, averaging about 77. Mean ages ran from the mid-fifties to the late seventies. Five trials enrolled people in the recovery phase, two in the chronic phase, two in the acute phase.
The forms were not uniform. Three studies taught Baduanjin. Two taught Liuzijue, the six-character breathing set. Two taught Tai Chi, one of them a slimmed six-form sequence drawn from the 24-form, the other classical Yang style. One used a seven-movement Daoyin hybrid taken from Tai Chi and Baduanjin. One used modified Yijinjing, ten forms with abdominal breathing and assistance when balance required it. Programs ran from two weeks to twenty-four. Weekly practice ranged from 50 minutes to 420. Eight of the nine studies asked for more than 150 minutes a week.
Delivery was equally mixed: hospital nurses and rehabilitation physicians, professional TCM doctors, trained researchers, instructional videos, family coaching at home, WeChat check-ins. One Japanese trial met once a week for 50 minutes. A Baduanjin study in chronic stroke ran 30 minutes a day, five days a week, for half a year. That spread is useful for clinics. It is awkward for anyone hoping the literature will yield a single prescription.
Sleep scores moved, and they moved together
Seven trials could be pooled on the PSQI. Traditional Chinese exercise beat the control arms by 2.15 points (95% CI −2.65 to −1.65). Heterogeneity was low, I² = 22%. When studies wander in different directions, a pooled mean is a blur. These did not wander much.
Controls were not all the same, so the authors split them. Against other structured activity (Otago exercise, combined rehab packages, resistance and non-resistance training), the advantage was 2.20 points. Against conventional care alone, 1.71 points. Both differences were statistically significant.
Two trials stayed outside the pool because they used other sleep tools. Liu and colleagues measured daytime sleepiness with the Epworth scale after twelve weeks of Liuzijue. Chen and colleagues used a portable sleep-respiratory monitor and reported the apnea-hypopnea index in stroke survivors who also had obstructive sleep apnea. Both papers favored the exercise groups. They cannot be folded into the PSQI number, and the reviewers say so.
Anxiety and depression came from three trials, 210 people, using the Hamilton scales. Anxiety fell by 1.93 points with no statistical heterogeneity. Depression fell by 2.21 points, I² = 30%. The direction is consistent. The sample is small. GRADE therefore rates the sleep evidence as moderate and the mood evidence as low, mainly because participants cannot be blinded to a movement practice, many outcomes were self-reported, allocation concealment was uneven, and the mood analyses lack precision.
The authors’ own sentence is measured:
“Traditional Chinese Exercises could be effective adjunct therapies for improving sleep quality and alleviating psychological distress in post-stroke rehabilitation.”
Adjunct is the right word. Nothing in the paper argues that Baduanjin replaces physiotherapy, a CPAP machine, or an antidepressant. The claim is narrower: added to usual care, these practices improved how people rated their sleep, and in the few trials that measured mood, those scores improved as well.
Two points on a questionnaire
A statistically tidy forest plot does not tell you whether a person notices the difference. The reviewers reach for a benchmark from earlier sleep research: a drop of about 1.63 points on the PSQI has been treated as a change people can feel. The pooled effect here, 2.15, sits above that mark.
They also refuse to treat 1.63 as gospel for this population. That figure was not validated in stroke survivors. A person six weeks after a hemiplegic infarct is not a generic insomnia patient. Pain, nocturia, residual weakness, hospital noise, and a changed mood all shape the night. Until someone establishes a stroke-specific minimal important difference, the 2.15-point shift is encouraging and still provisional.
The co-movement of sleep and mood is at least biologically tidy. After stroke, poor sleep and psychological distress reinforce each other. Slow coordinated movement, paced breathing, and a demand for attention are a plausible way to turn down the hyperarousal that keeps people from falling or staying asleep. The paper presents that as a hypothesis, not a mapped circuit.
What the evidence will not carry
Most of the trials were conducted in mainland China. Familiarity with the forms, the skill of the instructor, and the shape of inpatient rehab all differ from clinic to clinic and from country to country. One Japanese study does not fix that.
Objective sleep data are thin. Chen’s portable monitor is the exception. Polysomnography and actigraphy barely appear. Subjective scales matter; they describe the night a person actually lived. They also swell when people know they received the “special” treatment.
Dose is unsettled. Two weeks is not six months. Fifteen minutes of Baduanjin on a ward is not a 60-minute Yijinjing class three times a week. The reviewers note that eight of nine studies exceeded 150 minutes a week, then immediately warn that this “should not be interpreted as evidence supporting a specific dosage recommendation.” That warning is the honest part of the paper.
Blinding of students and teachers is nearly impossible here. Outcome assessors can be blinded more often than they were. Anxiety and depression rest on three trials. With fewer than ten studies per outcome, funnel plots were not attempted, so publication bias remains an open question.
None of that erases the PSQI result. It sets the confidence where GRADE put it.
Clinic, home, next trial
A rehab team looking for something to offer a patient who is medically stable, sleeping badly, and already overloaded with pills has a reasonable option in a supervised traditional Chinese exercise program. The forms can be shortened, done seated or with a chair nearby, taught in a group or sent home on video. They fit beside conventional therapy rather than in place of it.
A family helping with home practice should treat medical clearance and fall risk as the first constraints, then the clock. The successful protocols in this review were regular more than heroic. Thirty minutes, most days, with someone checking the shapes, is closer to what was studied than an occasional enthusiastic hour.
Researchers have a clearer list. Larger multicenter trials. Concealed allocation. Blinded assessors. A sleep battery that includes a device, not only a questionnaire. Follow-up past the last class. Sites outside East Asia. Direct comparisons between forms, and a dose-response design that can tell a clinician whether 150 minutes a week is a threshold or an accident of current habit.
One included trial used modified Yijinjing within six months of stroke. Another used Liuzijue in people who already had obstructive sleep apnea. Those are early signs that the useful question is shifting from “do these practices help?” toward “which form, at which stage, for which kind of broken night?”
That second question is still open. The first one now has a moderate-certainty answer. Across nine randomized trials, traditional Chinese exercises improved self-reported sleep after stroke more than usual or active comparison care. The mood data, thinner and less certain, point the same way. The nights after a stroke are part of recovery. These practices have earned a place in how that part gets studied, and, with the caveats the authors themselves list, in how it gets treated.
Cao, Y., Qiu, L., Hu, X., Su, Q., & Yao, L.-Q. (2026). Effects of traditional Chinese exercises on sleep quality among post-stroke survivors: A systematic review and meta-analysis. Complementary Therapies in Medicine, 100, Article 103401.
Like what you read? Keep exploring…
If this post resonated with you, you may enjoy my new book:
The Science of Medical Qigong: How Breath, Movement, and Attention Transform the Body
This book examines the growing body of scientific research exploring how Qigong practice influences human physiology. Drawing from clinical trials and experimental studies, it explains how coordinated breathing, posture, movement, and focused attention interact with the body’s regulatory systems.
The book explores research on:
• autonomic nervous system regulation
• cardiovascular function and blood pressure
• respiratory mechanics and lung capacity
• sleep quality and recovery
• inflammatory and immune signaling
• brain function and neural plasticity
• cognitive performance and mental health
Rather than presenting Qigong through mystical language, The Science of Medical Qigong approaches the subject through physiology, neuroscience, and clinical research, offering a clear explanation of how regular practice can influence the coordination and harmony of the body’s systems.
Available now in print and Kindle formats.