The combined tai chi groups improved more than aerobic exercise at 24 weeks. The same comparison was less clear at 52 weeks. A larger difference in the matched-schedule groups is a separate finding, not a promise for all pain or after stopping practice.
Study and source overview
Did the difference in fibromyalgia scores at 24 weeks last through the one-year follow-up?
- Participants
- 226 adults with fibromyalgia: 151 assigned to four tai chi groups and 75 to aerobic exercise.
- Class schedules
- Tai chi once or twice weekly for 12 or 24 weeks; aerobic exercise twice weekly for 24 weeks. Classes lasted 60 minutes.
- Primary outcome
- Change at week 24 in the Revised Fibromyalgia Impact Questionnaire (FIQR), scored 0–100. Lower scores indicate less impact.
- Follow-up
- Assessments continued to 52 weeks after randomisation; 158 of 226 participants completed that assessment. Participants were encouraged to continue activity and received monthly support calls.
Why are the combined and matched-schedule results different?
| At week 24 | Additional average FIQR improvement with tai chi | Interpretation |
|---|---|---|
| All four tai chi groups combined versus aerobic exercise | 5.5 points; 95% confidence interval 0.6–10.4. | Combines different class frequencies and durations. |
| Tai chi twice weekly for 24 weeks versus aerobic exercise on the same schedule | 16.2 points; 95% confidence interval 8.7–23.6. | A specific smaller-group comparison; it does not replace the combined primary analysis. |
Every group improved on the questionnaire. The useful comparison is the difference between groups, rather than whether one group improved from baseline. FIQR measures the impact of fibromyalgia; it does not count people cured.
The paper used 8.1 points as an estimated threshold for a clinically important difference. The combined 5.5-point estimate was smaller, with an interval extending above that threshold. Statistical significance alone does not establish a noticeable difference for each participant.
What remained at one year?
For the four tai chi groups combined versus aerobic exercise, the FIQR difference was 2.7 points at week 52 (95% confidence interval −2.3 to 7.7). The interval includes zero, so the difference was no longer clear. For the two groups with the same 24-week, twice-weekly schedule, the reported difference was 11.1 points (2.7–19.6). That subgroup result must not replace the combined result.
Participants were encouraged to continue practising after classes and received monthly telephone support. Follow-up therefore does not mean everyone stopped exercise. Only 158 people completed the one-year assessment; their outcomes cannot guarantee what happened to everyone initially enrolled.
What limits the comparison?
Participants had diagnosed fibromyalgia. The findings do not automatically apply to other pain conditions, undiagnosed symptoms or healthy beginners. The four tai chi schedules were different, so the largest difference cannot describe every schedule.
Participants knew their assigned activity. Attendance differed: about 62% of tai chi classes versus 40% of aerobic classes. Missing follow-up data were substantial, and one sensitivity analysis using imputation made the combined primary comparison no longer statistically significant. This adds uncertainty.
The paper reported no serious adverse events related to the interventions. That does not establish zero risk for every participant or every course.
What can this study help explain?
It informs a discussion about supervised exercise programmes for people with fibromyalgia. It does not test an unsupervised video course or establish how many healthy beginners will benefit.
When discussing an exercise option, compare its participants, schedule and support with those in the paper. The findings do not justify changing existing treatment on your own. Read research timeframes for the difference between a course and later follow-up.
Original source and review coverage
Wang C et al. Effect of tai chi versus aerobic exercise for fibromyalgia: comparative effectiveness randomized controlled trial. BMJ. 2018;360:k851.
DOI：10.1136/bmj.k851
Review coverage: Relevant full-text sections and Tables 2–3 accessed through the official Europe PMC XML copy: participants, schedules, FIQR comparisons at 24 and 52 weeks, attendance, missing data and adverse events. We did not reanalyse participant data.
By First Chi; source checked .The paper was published in 2018; the website update date does not replace that date.
