Does Qigong Actually Do Anything? Reading the Evidence Honestly
研究之實
A practitioner-facing look at what the research literature on qigong and tai chi is generally able to support, where it is weak, and how to read a study without being sold to.
There is a real research literature on qigong and tai chi. It is larger than sceptics assume and much weaker than enthusiasts claim. Here is how to read it without being sold anything.
What the literature is like
A search for randomised trials of qigong or tai chi returns a large number of studies, concentrated in a few areas: older adults and falls, chronic conditions such as fibromyalgia and chronic obstructive pulmonary disease, cancer-related fatigue and quality of life, sleep quality, and symptoms of anxiety and depression.
Trials tend to be small — tens rather than hundreds of participants — and short, often eight to twelve weeks. Systematic reviews and meta-analyses exist across most of these areas, including Cochrane reviews. Their conclusions are strikingly consistent in tone: some evidence of benefit, low to moderate certainty, high heterogeneity, more and better trials needed.
Why it is hard to study
- Blinding is close to impossible. A participant always knows whether they spent twelve weeks doing slow movement in a group. That opens the door to expectation effects that no amount of statistics closes.
- The control group problem. Compared with doing nothing, almost any structured group activity looks good. The meaningful comparison is against an active control — stretching, walking, a social group — and far fewer trials use one.
- "Qigong" is not one intervention. One trial's twelve weeks of Baduanjin and another's of a bespoke sequence are different things. Pooling them produces heterogeneity that is hard to interpret.
- Dose is rarely standardised. Frequency, duration, session length and instructor quality all vary and are often under-reported.
- Publication and regional bias. A large share of trials come from a small number of countries where these practices are culturally endorsed, and null results are systematically less likely to be published in any field.
The absence of good evidence is not evidence of absence — but it is also not permission to claim a benefit.
A reasonable working principle for reading this literature
Where the evidence is comparatively strong
Balance and falls in older adults is the clearest area. Tai chi in particular has been studied repeatedly for fall prevention, the trials are relatively numerous, the outcome is objective and hard to fake, and the finding has been consistent enough that tai chi appears in fall-prevention guidance from several public health bodies.
This is worth dwelling on, because it illustrates what a defensible claim looks like: a specific population, a measurable outcome, replication across multiple trials, and adoption into guidance by bodies with no stake in the tradition.
Where it is weaker than advertised
| Claim you will see | Actual state of the evidence |
|---|---|
| "Boosts the immune system" | Some small studies report changes in immune markers. Marker changes are not the same as fewer illnesses, and the clinical relevance is unclear. |
| "Reduces inflammation" | Similar. Biomarker findings are inconsistent and study sizes are small. |
| "Treats depression" | Trials often report improvement in depressive symptoms, generally with low certainty and rarely against an active control. Not a substitute for treatment. |
| "Improves sleep" | One of the more consistently reported outcomes, though usually via self-report questionnaires rather than objective measures. See Qigong and Sleep. |
| "Lowers blood pressure" | Some evidence of modest reductions. Effect sizes are small and should not displace medical management. |
| "Cures" anything | No. Nothing in this literature supports a curative claim for any condition. |
Reading a study without being taken in
- 01Check the control group first
Before the results, before the abstract. Compared with a waiting list, almost anything works. Compared with an equivalent amount of stretching or walking, much less does.
- 02Check the sample size
Under about fifty participants, treat the result as a signal worth following up rather than a finding.
- 03Check whether the outcome is objective
Timed balance tests and blood pressure are harder to influence by expectation than "how well did you sleep this week?"
- 04Check the intervention description
If the paper does not say which set, for how long, how often, and taught by whom, nobody can replicate it.
- 05Prefer reviews to single trials
A single small positive trial means very little. A systematic review that grades its certainty means considerably more.
- 06Notice who is summarising
Press releases, wellness sites and app marketing pages — this one included — have an interest. Go to the review.
What we are willing to say
That qigong is a low-intensity physical activity with a good safety profile, that practitioners commonly report improvements in sleep and in general tension, that the evidence for balance and fall risk in older adults is reasonably good, and that everything beyond that is currently promising rather than established.
That is a less impressive list than most marketing for this kind of practice. It has the advantage of being defensible, and it does not require anyone to discover later that they were misled.