Fibromyalgia Exercise: Why Swimming and Liuzijue Lead

When hard training leaves you more sore, wired, and exhausted than before, exercise selection matters more than grit. In fibromyalgia, a 2026 network meta-analysis found that swimming produced the largest improvement in overall health status, while Liuzijue—a breath-focused form of qigong—produced the largest reductions in pain and sleep problems. The results are useful, but they come from one sparse evidence network rather than a final answer carved into a foam roller.

Fibromyalgia is defined partly by central sensitization: the nervous system amplifies pain signals and lowers pain thresholds. That changes the practical question. The goal is not to prove you can tolerate HIIT. It is to find a repeatable aerobic activity that improves function without setting off a symptom flare.

Swimming led for overall function

Wen Yuan and colleagues' 2026 systematic review and network meta-analysis combined 20 randomized controlled trials involving 1,196 people with fibromyalgia. It compared eight aerobic exercise modalities across four outcomes: overall health status, pain intensity, sleep quality, and depressive symptoms.

For overall health status, swimming ranked first. Compared with passive control, it reduced Fibromyalgia Impact Questionnaire (FIQ) scores by 23.36 points (mean difference −23.36; 95% CI −40.77 to −4.32). On the 0-to-100 FIQ scale, where higher scores mean more impairment, that is a clinically meaningful improvement in the estimate.

The confidence interval deserves equal billing. Its width means the true average benefit could be modestly clinically meaningful or much larger. Swimming looks like the leading option for function in this analysis, but the estimate rests on limited data rather than a large stack of direct swimming trials.

The result still makes physiological sense. Water reduces impact loading, while buoyancy can make movement more tolerable for someone with widespread pain. Those are plausible reasons swimming may suit fibromyalgia; they do not prove why it outperformed other approaches in this analysis.

Liuzijue ranked first for pain and sleep

Liuzijue combines slow, coordinated movement with diaphragmatic exhalations that are vocalized. In the same network meta-analysis, it had the largest estimated improvement in pain: a 3.69-point reduction on a 10-cm visual analog scale versus control (mean difference −3.69; 95% CI −5.37 to −2.02). That interval excludes no difference, and the point estimate exceeds the roughly 1.5-to-2.0-cm minimum clinically important difference commonly used in chronic-pain research.

Liuzijue also ranked first for sleep quality, with a mean difference of −5.29 versus control (95% CI −8.98 to −1.64). The direction indicates better sleep on the scale used by the included trials. The result was statistically significant because the confidence interval did not include zero, but the broad interval again signals uncertainty about the precise size of the benefit.

Slow movement plus controlled breathing is a reasonable fit for a sensitized nervous system. Vocalized exhalation could influence autonomic arousal, but the available evidence does not establish vagal activation—or any other specific pathway—as the reason Liuzijue worked. It also does not show that any breathwork class, generic qigong routine, or online video will reproduce the trial result.

The useful finding is specific: one network meta-analysis ranked swimming highest for overall fibromyalgia health status and Liuzijue highest for pain and sleep. It did not establish a universal dose, a guaranteed response, or a biological mechanism.

Why a ranking is not a prescription

Network meta-analysis can compare treatments that have not all been tested directly against each other. It combines direct trial comparisons with indirect ones through shared comparators. That is valuable when eight exercise modalities compete for attention, but it depends on the trials being similar enough in participants, settings, and intervention design for those indirect comparisons to hold.

Here, 20 trials were spread across eight modalities—about 2.5 trials per modality on average. Some network nodes were thin. The swimming confidence interval makes that limitation visible, and other comparisons may face similar imprecision.

The depression outcome offers the clearest warning. Walking had the highest ranking probability for depressive symptoms, yet its estimated effect versus control was not statistically significant (mean difference −5.99; 95% CI −14.67 to 2.74). A ranking can look impressive even when the underlying estimate is too imprecise to rule out no effect. Check the effect size and confidence interval before treating a league table as clinical advice.

Liuzijue has an additional validation problem. It led for pain and sleep in this analysis, but independent fibromyalgia-specific confirmation is still lacking. The published abstract also does not report a publication-bias assessment. Positive-result reporting bias is a legitimate concern for small exercise-trial literatures, especially when a modality has few studies.

Intensity does not transfer cleanly across chronic pain conditions

A 2025 multicenter randomized trial in rheumatoid arthritis found that 12 weeks of supervised high-intensity interval and strength exercise improved multidimensional fatigue and health-related quality of life in 87 participants. That is encouraging for rheumatoid arthritis, but it does not justify importing a high-intensity protocol into fibromyalgia.

Rheumatoid arthritis and fibromyalgia can both involve fatigue, pain, and reduced activity, yet they are different conditions. The fibromyalgia network meta-analysis did not establish high intensity as the driver of benefit. For a person whose symptoms worsen after exertion, intensity should be titrated to tolerance and adjusted with a clinician or physical therapist when needed.

Other chronic-pain research supports the narrower idea that exercise modality can matter. A 2021 network meta-analysis in knee osteoarthritis found that exercise programs differed in their effects on psychological well-being. A 2025 review in knee osteoarthritis found aerobic exercise improved cardiopulmonary fitness. Neither finding validates the fibromyalgia effect sizes, but both make it less credible that every form of exercise is interchangeable for people limited by musculoskeletal pain.

Sleep and cognition are relevant, but the evidence is indirect

Sleep disruption often travels with fibromyalgia, and Liuzijue's sleep result therefore matters clinically. Still, a 2025 randomized trial in 160 older women found that 12 weeks of progressive resistance training improved sleep quality, mental health, cognitive function, and functional capacity; it did not study fibromyalgia or Liuzijue. It supports the general possibility that structured exercise can improve sleep, not a claim that resistance training and Liuzijue work through the same mechanism.

Likewise, a 2024 meta-analysis in healthy middle-aged and older adults found aerobic exercise improved executive function. Fibromyalgia-related cognitive complaints are real, but healthy-aging evidence cannot tell us whether swimming or Liuzijue improves "fibro fog," or by how much.

A practical, evidence-calibrated choice

If overall function is your main target and you have safe pool access, swimming is the best-supported first option from this comparative analysis. If pain and sleep are your dominant problems, Liuzijue is the leading candidate, provided you can find instruction that lets you move slowly and stop before symptoms escalate.

The study does not provide enough information to prescribe a universal weekly frequency, session length, or intensity target. Start below the level that predictably worsens symptoms, repeat a tolerable dose, and progress gradually only if recovery remains stable. That approach is less glamorous than "push through," but chronic pain has never been impressed by motivational posters.

The bottom line is appropriately narrow: swimming and Liuzijue currently have the best comparative signals among the aerobic modalities studied for fibromyalgia. Those signals are promising, statistically significant for their reported outcomes, and still in need of replication before they become firm individualized treatment rules.