Trial Report Effects of Tai Chi on fatigue and resting-state brain activity in chronic fatigue syndrome: a randomized controlled fMRI study, 2026, Wang et al

[forestglip]

Effects of Tai Chi on fatigue and resting-state brain activity in chronic fatigue syndrome: a randomized controlled fMRI study

Wang, Bin; Zhang, Xiaodong

Background
Chronic fatigue syndrome (CFS) is a complex disorder characterized by persistent fatigue, sleep disturbance, and cognitive impairment. It substantially impairs both physical and mental health. Tai Chi, a traditional Chinese mind–body exercise, may improve symptoms in patients with CFS. However, its neural mechanisms remain unclear. This study investigated the effects of Tai Chi on spontaneous brain activity in patients with CFS using resting-state functional magnetic resonance imaging (rs-fMRI) and amplitude of low-frequency fluctuations (ALFF).

Methods
Sixty patients with CFS were randomly assigned in a 1:1 ratio to either a Tai Chi group or a health education group; 57 participants completed both fMRI scans and fatigue assessments and were included in the complete-case analysis. Both groups underwent a 12-week intervention consisting of three 60-min sessions per week. Adherence was monitored using weekly logs. Fatigue severity was assessed before and after the intervention using the Multidimensional Fatigue Inventory (MFI-20).

Resting-state fMRI data were acquired at both time points. Preprocessing included motion correction, normalization to Montreal Neurological Institute space, and spatial smoothing. ALFF was calculated to quantify regional spontaneous neural activity. Correlation analyses were performed to assess associations between changes in ALFF and MFI-20 scores. Harms were monitored throughout the intervention period.

Results
Fifty-seven participants were included in the complete-case analysis (Tai Chi, n = 29; health education, n = 28). Compared with health education, Tai Chi produced greater reductions in MFI-20 total score from baseline to week 12 (between-group change difference, −9.50 points; 95% CI, −12.31 to −6.70; Cohen's d = −1.83; P < 0.001). Greater improvements were also observed in overall fatigue, physical fatigue, mental fatigue, activity reduction, and interest reduction, with between-group differences ranging from −1.26 to −2.44 points.

In imaging analyses, Tai Chi was associated with increased ALFF in prefrontal, hippocampal/striatal, and supplementary motor regions and decreased ALFF in occipital and temporal regions, using voxel-level P < 0.001 and cluster-level family-wise error correction. Exploratory brain–behavior analyses showed that increases in ALFF in the left medial superior frontal gyrus and left dorsolateral superior frontal gyrus were associated with greater reductions in MFI-20 scores, whereas decreases in ALFF in the left inferior occipital gyrus were associated with fatigue improvement.

No falls, injuries, severe post-exertional malaise, or intervention-related medical events were reported through weekly participant logs or supervised-session monitoring.

Conclusion
A 12-week Tai Chi intervention was associated with greater fatigue improvement than health education and with region-specific changes in resting-state spontaneous brain activity in patients with CFS. Tai Chi was associated with reduced fatigue and altered spontaneous brain activity in prefrontal, hippocampal, motor, and sensory-related regions in patients with CFS. These findings suggest that Tai Chi may modulate fatigue-related brain function, which may contribute to improvements in fatigue severity.

Future studies should replicate and extend existing network-level findings using prespecified, adequately powered connectivity and mediation analyses with active exercise comparators and longer-term follow-up.

Chinese clinical trial registration https://www.chictr.org.cn/showproj.html?proj=223860 , ChiCTR2400082268; March 26, 2024.

Web | DOI | PDF | Frontiers in Neuroscience | Open Access

 

[Utsikt]

As usual, the control condition was nowhere near adequate:

The health education group controlled for disease-related information exposure and online contact but did not match the Tai Chi group for physical activity, movement training, instructor feedback, or exercise-related expectancy. Accordingly, this comparator was designed to control for disease-related information exposure and study contact rather than to isolate the specific effects of Tai Chi from those of light physical activity, motor learning, instructor attention, or exercise-related contextual effects.

The changes in the subjective outcomes are not that impressive when you look at the graphs:

For some reason, they cropped the Y axis for F.

The adverse event logs did not capture PEM:

No adverse events were reported through the weekly participant logs or observed during supervised sessions. Specifically, no falls, injuries, dizziness requiring intervention discontinuation, or severe worsening described by participants as post-exertional malaise was reported through these monitoring procedures. However, because a validated PEM-specific instrument and scheduled delayed post-session assessments were not used, these findings indicate an absence of reported severe events under the study monitoring procedures rather than definitive evidence that no PEM occurred. Within these limitations, the intervention was generally well-tolerated by the participants who completed the study.

There is no data on general activity levels, so we have no idea how much the participants had to sacrifice to complete the programme.

The limitations section is quite comprehensive, but it begs the question of why they did not do any better with designing this study.

First, the diagnosis of CFS relied partly on self-reported symptoms, introducing potential variability.

Second, participant blinding was not feasible, and the health education comparator did not match the Tai Chi intervention for physical activity, instructor feedback, motor-skill learning, or exercise-related expectancy. Therefore, the observed between-group differences cannot be attributed specifically to the mind–body components of Tai Chi and may partly reflect non-specific exercise, attention, or expectancy effects.

Third, because the primary clinical outcome, the MFI-20, was entirely self-reported, participants' awareness of treatment allocation could have influenced symptom reporting and may have inflated the apparent between-group benefit. Although the fMRI analyses provide a concurrently measured biological outcome and were performed using coded identifiers, the imaging findings do not eliminate expectation bias in the self-reported outcome and should not be interpreted as proof that the neural changes mediated fatigue improvement.

Fourth, the absence of a healthy control group limits interpretation of the observed ALFF changes as normalization vs. compensatory processes.

Fifth, the analyses were based on complete cases with usable paired fMRI data; therefore, attrition and imaging-quality exclusions may have introduced selection bias.

Sixth, the present manuscript reports the 12-week fatigue and ALFF outcomes, whereas other protocol-listed cognitive and follow-up outcomes are not reported here, the present manuscript should therefore be interpreted as a focused ALFF report rather than the complete trial report.

Seventh, ALFF captures local spontaneous activity but does not characterize network-level interactions, and the brain-behavior correlations were exploratory and should be interpreted cautiously.

Eighth, PEM was not assessed using a validated PEM-specific instrument or scheduled assessments during the hours and days following each intervention session. Because PEM may have delayed onset, the weekly logs and supervised-session monitoring may have underdetected delayed or transient post-exertional symptom exacerbation.

 

[rvallee]

The limitations section is quite comprehensive, but it begs the question of why they did not do any better with designing this study.

Seems pretty simple: why? Why would anyone bother doing any better? It's an industry where slop rules, it's what all the cool kids are doing and all the money is going.

There is not a single incentive to do rigorous trials of non-pharmaceutical anything. In fact it's a bad career move. The goal is to give false hope and pretend like this is all legitimate. Bad trials fulfill that requirement, there's nothing more to it. Bad trials get rewards. Good trials are pretty much a non-starter. Too expensive. Too unwieldy. Also guaranteed to have a predictable result: nope, doesn't work, wasting time here.

This is why I don't bother with people who think it's a good idea to have good high quality trials so that it can shut up some people about useless treatments. It never works. They always cheat, because cheating is the ends and the means.

 

[Sean]

There is not a single incentive to do rigorous trials of non-pharmaceutical anything. In fact it's a bad career move. The goal is to give false hope and pretend like this is all legitimate. Bad trials fulfill that requirement, there's nothing more to it. Bad trials get rewards.

Yep, the incentive (or disincentive) structure has become perverse in the extreme. Competence, diligence, honesty, and moderation are harshly punished.