Tai chi beat aerobics in the headlines, and the full study tells a fuller story

Key takeaways

• In a 2018 trial of 226 adults with fibromyalgia, tai chi improved symptom scores 5.5 points more than aerobic exercise overall, less than the 8.1 points the authors set as a change patients notice.
• One schedule, tai chi twice a week for 24 weeks, beat aerobic exercise done on the same schedule by 16.2 points.
• When the researchers re-ran their main analysis a different way, the overall advantage no longer cleared the usual bar for statistical significance, and they reported that themselves.
• The trial was registered in 2011, before anyone enrolled, and the main outcome it reported matched the one named in its protocol, published while the study was still running.
• Four questions get you through any study: who was studied, compared with what, how big the difference was, and who paid.

In March 2018, a federal health agency shared good news for people with fibromyalgia, a condition of long-lasting pain across the body. A study it helped fund found that tai chi worked as well as aerobic exercise or better³. Aerobic exercise is the non-drug treatment doctors most often prescribe for the condition³.

Two months later, a medical news site ran the headline “Tai Chi Tops Exercise as Fibromyalgia Treatment”⁴.

The study is real, carefully run and free for anyone to read¹. Read the whole thing, though, and the story gets richer.

Tai chi’s overall edge was smaller than the change the authors counted as noticeable to patients. When the authors stress-tested their math, that edge wobbled. Yet one tai chi schedule beat matched aerobic exercise by a wide margin¹.

None of that makes the headline false. It makes it a summary. This guide walks through the trial section by section, so you can read the next study yourself, whether it tests an herb, a drug or a workout.

Every paper follows the same four-part map

Nearly every health study uses the same layout, called IMRaD: Introduction, Methods, Results and Discussion. It wasn’t always so. Researchers who surveyed four leading medical journals found no IMRaD papers in 1935. By the 1980s, it was the only layout those journals used for original research⁶.

Each part has a job:

• Introduction: why the question matters and what was already known.
• Methods: the recipe. Who took part, what they got, and how results were measured.
• Results: the numbers, usually in tables.
• Discussion: what the authors think it means, including the study’s weak spots.

Above it all sits the abstract, a one-paragraph summary. Below it sit the parts many readers skip: funding, conflicts of interest, trial registration and online appendices.

You don’t have to read in order. A useful path is abstract first, then methods, then the results tables, and the discussion last. The discussion is where authors make their case, so it helps to know the facts before you hear the argument.

The tai chi paper’s introduction set up a clear gap. Two earlier 12-week trials had found tai chi helped more than wellness education or stretching. Nobody knew whether it could match aerobic exercise, or what dose worked best. The authors also noted that many patients struggle to stick with exercise programs¹.

Pin down the question: who, what, compared with what, measured how

Researchers use a memory aid called PICO to define a question. Cochrane, the international group known for evidence reviews, spells it out as population, intervention, comparator and outcome⁵.

The comparator matters most. Cochrane lists the options: a placebo or sham, no treatment, a wait list, usual care, or another active treatment⁵. A remedy that beats doing nothing has cleared a low bar. One that holds its own against a proven treatment has cleared a high one.

Here is the tai chi trial in PICO form:

Letter	Plain question	In the tai chi trial
P	Who was studied?	226 adults with fibromyalgia, average age 52, 92% women1
I	What was tested?	Yang-style tai chi, 60-minute classes, once or twice a week for 12 or 24 weeks1
C	Compared with what?	Supervised moderate aerobic exercise, 60 minutes, twice a week for 24 weeks1
O	Measured how?	Change on a symptom questionnaire scored 0 to 100, where higher is worse, after 24 weeks1

That questionnaire, the revised Fibromyalgia Impact Questionnaire (FIQR), covers pain, physical function, fatigue, morning tiredness, mood, anxiety and overall well-being³. It is filled out by patients, so it captures how people feel and function.

Notice the comparator. This was a comparative effectiveness trial, which pits two real options against each other. The trial also had five groups: one aerobic group and four tai chi groups that varied how often and how long people practiced¹.

Meet the people behind the numbers

Every study starts with a funnel. Here, 867 people were screened by phone, 272 came in for a baseline visit, and 226 were enrolled¹.

The entry rules shape who those 226 were. Volunteers had to be at least 21 and meet the American College of Rheumatology’s fibromyalgia criteria, which include widespread pain for at least three months¹. People were turned away if they had done tai chi or similar practices in the past six months, had serious illnesses or inflammatory arthritis, were pregnant, didn’t speak English, or scored low on a thinking-skills test¹.

Those who enrolled had lived with body pain for nine years on average³. About 61% were white, and everyone came through one medical center in Boston¹.

Results fit best for people like those studied. A summary for family doctors noted that most volunteers in tai chi studies for fibromyalgia have been women, so results may not apply to men¹⁰. Ask yourself: would I have qualified for this study? If not, the result is a clue for you, not a verdict.

How the groups were formed, and who knew what

The study’s statistician used software to generate the random assignments. Each one was sealed in an opaque envelope and opened only after a person had agreed to join and been found eligible¹. That second step, called allocation concealment, keeps staff from steering hopeful-looking patients into one group.

Then comes blinding. The study nurses and the staff who tested physical function didn’t know who was in which group¹. The participants did, because there is no way to hide whether you’re doing tai chi.

The authors called this design single-blind and flagged it as a limit. People who expect a practice to help may report feeling better, and the main measure was a questionnaire people filled out about themselves¹. That doesn’t sink the study. It’s a reason to treat the result as a strong clue rather than a final answer.

Read the numbers, not the adjectives

Start with a surprise: everyone got better. After 24 weeks, the aerobic group’s score dropped 9.2 points. The group doing tai chi twice a week for 24 weeks dropped 25.4 points¹.

Change within a group mixes several things: the treatment, natural ups and downs, extra attention and hope. That’s why the key number is the difference between groups.

Across all four tai chi groups combined, tai chi beat aerobic exercise by 5.5 points¹. Is that a lot?

Researchers use the minimal clinically important difference, the smallest change patients tend to notice. The authors estimated it at 8.1 points, borrowing the figure from an older version of the questionnaire¹. So the overall edge was real on average but smaller than that line.

The bigger story was dose. When tai chi and aerobic exercise were matched at twice a week for 24 weeks, tai chi came out 16.2 points ahead¹. Six months of tai chi beat three months by 9.6 points. Twice a week was not clearly better than once a week¹.

Keep group size in mind. The 16.2-point result came from comparing 36 people in the best tai chi group with 75 in the aerobic group¹. Smaller groups make estimates less precise.

How sure are we? Read the range, not just the point

Each result comes with a 95% confidence interval. For the overall comparison, it ran from 0.6 to 10.4 points¹.

In plain words, the data fit a true advantage anywhere from almost nothing to a change most patients would notice. When a range includes “barely any difference,” hold the finding loosely. Our companion article on recognizing uncertainty goes deeper.

The matched-dose range ran from 8.7 to 23.6 points¹. Even its low end cleared the 8.1-point line, which makes that result sturdier.

The paper also gives a P value, a measure of how surprising the result would be if there were no real difference. For the overall comparison, it was 0.03¹. The news story printed .005⁴. Small slips like that are one more reason to go to the source.

Count who showed up, who left and what went wrong

Retention was solid: 80% of volunteers finished the 24-week visit and 70% the one-year visit¹. Attendance was not equal. Tai chi groups came to 62% of classes; the aerobic group came to 40%¹.

That gap cuts two ways. Maybe tai chi is easier to stick with, which is a real advantage in daily life. Or the aerobic group simply got less of its treatment. The authors noted that people who skipped classes were also more likely to drop out¹.

Good papers stress-test their own results with a sensitivity analysis, which re-runs the numbers using a different reasonable method. Here, when missing data were filled in with a statistical technique called multiple imputation, the P value for the overall comparison rose to 0.08¹. That’s above the usual 0.05 cutoff. The authors reported this in their own paper, which is a mark of a careful team.

Over time, the gains faded somewhat. At one year, the matched-dose advantage was 11.1 points, and most effects were smaller than at 24 weeks¹.

Then look for harms. Of 154 reported health problems, 12 were judged related to the exercise programs, mostly minor muscle or joint complaints: 8 in tai chi and 4 in aerobics¹. None of the 10 serious events was tied to either program, and no one quit because of side effects. All groups also cut back similarly on pain relievers and some other medicines¹.

Check who paid and what was promised in advance

Turn to the end of the paper. This trial was paid for mainly by the National Institutes of Health, including its complementary and integrative health center. A U.S. Department of Agriculture agreement partly supported two authors¹.

The authors declared no financial ties to companies with a stake in the results¹. Government money doesn’t guarantee a right answer, but no one here was selling tai chi classes.

Next, check registration. The International Committee of Medical Journal Editors says trials that began on or after July 1, 2005, must be publicly registered before the first patient enrolls⁸. This one was registered on August 18, 2011, and the study began in early 2012².

The team also published its full plan, called a protocol, in January 2015. Screening had ended by then, but follow-up visits were still under way¹².

The protocol named the main outcome: change in the FIQR score at 24 weeks². That’s the outcome the final paper led with¹. The plan called for 216 volunteers, two in tai chi for every one in aerobics²; the team enrolled 226¹. When a paper’s main result doesn’t match its registered main outcome, ask why.

The newest reporting guideline for trials, CONSORT 2025, now asks every trial report to say where the protocol, statistical plan and data can be found, and to disclose funding and conflicts⁷. Our companion article on conflicts of interest shows how to check the people behind a study.

Put one study in its place

One trial is a chapter, not the whole book. The National Center for Complementary and Integrative Health (NCCIH) reports that a 2019 review of 6 studies with 657 people found tai chi eased fibromyalgia pain. The reviewers called for larger, better studies, including head-to-head tests against standard exercise⁹.

A 2020 review for family doctors rated tai chi moderately effective for quality of life in fibromyalgia, based on a small meta-analysis and this trial¹⁰.

On safety, NCCIH cites a 2019 review of 24 studies in which side effects were about as common with tai chi as with other exercise or no exercise⁹. Tai chi teachers need no license, so ask about training and experience⁹.

Claim	Evidence rating	Basis
Tai chi eases fibromyalgia symptoms	Promising but early	One 226-person trial plus a 6-study review that called for bigger trials1910
Tai chi works better than aerobic exercise	Promising but early	Overall edge below the noticeable line and shaky in a stress test; one schedule beat aerobics clearly, at one center1
Longer tai chi programs help more than short ones	Promising but early	24 weeks beat 12 weeks in one trial1
Tai chi is low risk for most adults	Strong evidence	Review of 24 studies; few related, minor problems in this trial19

Many people in natural health say, “Do your own research.” Reading the original paper is the most direct way to do it. This trial also shows that a practice rooted in Chinese tradition can be tested by fair rules and hold its own against a standard treatment¹⁹. The same questions work on a drug company’s trial and a supplement maker’s study alike.

If you live with fibromyalgia and want to try tai chi, talk with your clinician first, especially if you have other conditions. Don’t stop prescribed treatment on your own.

Your study-reading worksheet

Print this table or copy it into a notebook. Fill in the third column as you read. If you can’t fill a box, the paper may be missing something.

Part of the paper	Question to ask	Your notes	Reassuring sign
Abstract	What is the one-sentence claim?		It names the main outcome and the size of the difference
Introduction	What gap is this study meant to fill?		Earlier studies are described fairly, good and bad
Methods: people	Who got in, who was kept out, and am I like them?		Clear entry rules and a screening count
Methods: comparison	Compared with what, exactly?		An active treatment, placebo or usual care, not just “before”
Methods: fairness	Randomized? Concealed? Who was blinded?		Assessors blinded even if patients can’t be
Methods: outcome	What was the main measure, and when?		It matters to patients and matches the registry
Results: size	How big was the difference between groups?		It clears the smallest change patients notice
Results: certainty	How wide is the confidence interval?		Even the low end would matter to you
Results: losses and harms	Who dropped out, and what went wrong?		Most people finished; side effects reported by group
End matter	Who paid? Registered before enrollment?		Funding and conflicts disclosed; registry number given

Questions readers ask

The bottom line

A headline is a summary, and a summary can’t hold the whole truth. The tai chi trial was well designed and openly reported. Its overall result was more modest than the news suggested, and its best result was more striking. You see both only by reading the paper.

Use the same path every time: pin down the question, meet the people, check the comparison, read the size and range of the difference, count the losses and harms, and look at the money and the registry. It works for tai chi, turmeric and prescription drugs alike.

Related reading

• How clinical trials decide, when most treatments tested in people fail
• Drugs that quieted extra heartbeats raised deaths, a lesson in reading a study’s limits
• A 22% risk cut can mean helping 2 or 3 people in 1,000
• Who paid for the study matters, even when the science is real
• Tai chi, prayer and journaling: what research can measure and what it can’t

Sources

1. Wang C, Schmid CH, Fielding RA, et al. “Effect of Tai Chi Versus Aerobic Exercise for Fibromyalgia: Comparative Effectiveness Randomized Controlled Trial.” The BMJ, 2018;360:k851 (open access). pubmed.ncbi.nlm.nih.gov
2. Wang C, McAlindon T, Fielding RA, et al. “A Novel Comparative Effectiveness Study of Tai Chi Versus Aerobic Exercise for Fibromyalgia: Study Protocol for a Randomized Controlled Trial.” Trials, 2015;16:34. link.springer.com
3. National Center for Complementary and Integrative Health. “Tai Chi Has Similar or Greater Benefits Than Aerobic Exercise for Fibromyalgia, Study Shows.” NIH, March 22, 2018. nccih.nih.gov
4. Rheumatology Live. “Tai Chi Tops Exercise as Fibromyalgia Treatment.” MJH Life Sciences, May 21, 2018. rheum-live.com
5. McKenzie JE, Brennan SE, Ryan RE, et al. “Chapter 3: Defining the Criteria for Including Studies and How They Will Be Grouped for the Synthesis.” Cochrane Handbook for Systematic Reviews of Interventions, version 6.5, 2024. cochrane.org
6. Sollaci LB, Pereira MG. “The Introduction, Methods, Results, and Discussion (IMRAD) Structure: A Fifty-Year Survey.” Journal of the Medical Library Association, 2004;92(3):364-367 (hosted copy). stat.cmu.edu
7. Hopewell S, Chan AW, Collins GS, et al. “CONSORT 2025 Statement: Updated Guideline for Reporting Randomised Trials.” The BMJ, 2025;389:e081123. pubmed.ncbi.nlm.nih.gov
8. International Committee of Medical Journal Editors. “Clinical Trials Registration” (FAQ). ICMJE. icmje.org
9. National Center for Complementary and Integrative Health. “Tai Chi: What You Need To Know.” NIH, last updated December 2023. nccih.nih.gov
10. Wilson LM, Raza Z, Garufi LC, Keister D, Nashelsky J. “Tai Chi for Fibromyalgia.” American Family Physician, 2020;101(5) (FPIN Clinical Inquiries). aafp.org