Comparing Four Mind-Body Exercises for Stroke Recovery: What the Network Meta-Analysis Actually Shows

Stroke leaves many people with problems that go well beyond a single symptom. Some struggle with fine hand movements such as gripping a cup; others have difficulty walking, standing steadily, or managing basic self-care like dressing and bathing. Alongside standard physical and occupational therapy, a number of patients and families encounter mind-body practices rooted in traditional Chinese medicine, each with its own flavor of movement and breath.

Tai Ji, often practiced as a slow, continuous sequence of weight-shifting arcs, emphasizes balance and fluid transitions. Ba Duan Jin is a compact routine of eight symmetrical stretches, easy to learn and repeat. Yi Jin Jing features larger postures that emphasize tendon and bone stretching. Wu Qin Xi mimics the movements of five animals—tiger, deer, bear, ape, and bird—and is the most varied in shape and tempo. When patients ask which one suits them, personal experience alone cannot settle the question. That is why the authors pooled data from many clinical trials, aiming to give a broader answer than any single study could provide.

The analysis reported four rankings, each accompanied by a standardized mean difference, or SMD, and a 95% confidence interval, or CI. An SMD puts different outcome scales onto a common yardstick—larger values mean a bigger improvement compared with the control. A 95% CI shows the range within which the true effect likely falls; a narrower interval signals a more precise estimate.

First, for upper limb function, Ba Duan Jin ranked highest, with SMD = 1.33 and 95% CI = 0.80–1.86. Second, for lower limb function, Ba Duan Jin again came out on top, with SMD = 1.49 and 95% CI = 0.97–2.01. Third, for balance function, Wu Qin Xi ranked first, with SMD = 1.56 and 95% CI = 0.45–2.67. Fourth, for activities of daily living, Wu Qin Xi led once more, with SMD = 2.34 and 95% CI = 0.91–3.76.

Tai Ji and Yi Jin Jing did not occupy the top of any of these four rankings. The authors still reported that both produced significant therapeutic benefits compared with conventional therapies—meaning each was meaningfully better than usual care on at least some measures, even if a competitor edged ahead. The study's practical bottom line: Ba Duan Jin appears especially well-suited to upper and lower limb rehabilitation, while Wu Qin Xi appears strongest for restoring balance and daily living skills.

Rankings from a network meta-analysis describe relative position, not absolute clinical truth. Saying Ba Duan Jin ranks first for upper limb function means that, across the 46 trials included, Ba Duan Jin showed the largest improvement relative to Tai Ji, Yi Jin Jing, Wu Qin Xi, and conventional rehabilitation. It does not mean Ba Duan Jin works for everyone, works equally well at every stage of recovery, or replaces standard physiotherapy and occupational therapy.

Several technical points help in interpreting the numbers. A wider confidence interval signals greater uncertainty. Ba Duan Jin's intervals for upper limb (0.80–1.86) and lower limb (0.97–2.01) are relatively narrow, suggesting more confident estimates. Wu Qin Xi's intervals for balance (0.45–2.67) and daily living (0.91–3.76) are wider, especially the lower bound approaching zero for balance, which means the ranking for these outcomes is less certain and should be read with caution. The SMD itself is a statistical measure of how much improvement there is compared with the control; it is not a clinical cure rate or a guarantee of functional independence.

For a curious reader, the most useful takeaway is a matching exercise: if the priority is upper or lower limb movement, the current evidence points toward Ba Duan Jin; if the priority is balance and everyday self-care, the evidence tilts toward Wu Qin Xi; Tai Ji and Yi Jin Jing remain reasonable, evidence-supported options when their style, accessibility, or personal preference fits better. These choices sit alongside, not instead of, conventional rehabilitation.

The authors describe their findings as evidence-based guidance for selecting interventions tailored to specific functional deficits. That phrasing is deliberately modest: it points to relative rankings within a defined evidence base, not to cures or replacements for conventional stroke care. No exercise in this analysis is shown to reverse stroke damage, prevent recurrence, or eliminate the need for medical follow-up, and the rankings should not be interpreted that way.

The PROSPERO registration (CRD42025642353) supports transparency by letting readers see the original review plan, but registration alone does not guarantee that the conclusions are correct. The 46 trials included vary in size, quality, intervention duration, and the scales used to measure outcomes, all of which can influence pooled estimates. If larger or more rigorous trials are published in the future, the rankings may shift. For now, the most accurate framing is that Ba Duan Jin and Wu Qin Xi currently lead in their respective categories among the four exercises studied, and that all four show benefit relative to usual care—but the absolute size of that benefit, and which exercise is best for any particular individual, remain open questions.

At first glance, a 2026 network meta-analysis of stroke rehabilitation and a two-thousand-year-old medical classic seem to belong to different worlds. Yet the question they are both trying to answer is surprisingly close: when the limbs no longer respond after a stroke, what kinds of movement can help reconnect a person with ordinary life? The research team searched nine databases from inception through February 2025 and assembled 46 randomized controlled trials covering 3,096 stroke patients aged 18 and over. Inside that evidence network, four traditional Chinese mind-body exercises—Tai Ji, Ba Duan Jin, Yi Jin Jing, and Wu Qin Xi—were compared across four objective indicators: upper limb function, lower limb function, balance function, and activities of daily living.

The ranking produced a clear pattern. Ba Duan Jin placed first for both upper limb function (SMD = 1.33, 95% CI = 0.80–1.86) and lower limb function (SMD = 1.49, 95% CI = 0.97–2.01). Wu Qin Xi led the other two indicators, ranking highest for balance (SMD = 1.56, 95% CI = 0.45–2.67) and for activities of daily living (SMD = 2.34, 95% CI = 0.91–3.76). Tai Ji and Yi Jin Jing did not top any category, but each still showed significant benefit compared with conventional rehabilitation. Here, SMD (standardized mean difference) places the results of different measurement scales on a common ruler; a 95% confidence interval (CI) describes the range within which the true value is likely to fall, and a narrower interval means the estimate is more stable. Wu Qin Xi’s wider intervals—particularly for balance and daily living—signal greater uncertainty and a need for cautious reading.

These rankings are relative, not absolute. They describe which exercise leads within the body of trials assembled, not which exercise cures stroke or should replace conventional therapy. Stroke recovery is a long road, and the four movement therapies are one branch of a wider rehabilitation landscape that includes physiotherapy, occupational therapy, and ongoing medical follow-up. Readers can use the ranking as an orientation map: those most concerned with limb movement may look first at Ba Duan Jin, while those focused on balance and self-care may explore Wu Qin Xi. The map will almost certainly shift as new, higher-quality trials are published.