Review The efficacy of exercise in patients with myalgic encephalomyelitis/chronic fatigue syndrome: A systematic review and meta-analysis, 2026, Zhao et al

[forestglip]

The efficacy of exercise in patients with myalgic encephalomyelitis/chronic fatigue syndrome: A systematic review and meta-analysis

Zhao, Liping; Gou, Bo; Zhang, Meng

Objective
To evaluate the efficacy of exercise training for patients with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS).

Methods
Seven databases were searched from inception to December 2025 for randomized controlled trials (RCTs) on exercise effects in ME/CFS.

A random-effects model was employed to calculate the pooled effect size, with Q, I2 statistic, and τ2 assessing a heterogeneity. Additionally, we performed moderator analyses to explore stability of results.

Results
A total of 17 RCTs (n = 1944) were included. Exercise interventions significantly improved fatigue (16 RCTs, n = 1673; standardized mean difference [SMD] = 0.85, 95% CI 0.65 to 1.06, I2 = 85.3%), sleep quality (5 RCTs, n = 806; SMD = 0.42, 95% CI 0.31 to 0.53, I2 = 41.2%), QoL (11 RCTs, n = 1171; SMD = 0.99, 95% CI 0.73 to 1.26, I2 = 89.7%), depression (9 RCTs, n = 1279; SMD = 0.56, 95% CI 0.43 to 0.69, I2 = 45.8%), and anxiety (9 RCTs, n = 1279; SMD = 0.14, 95% CI 0.04 to 0.24, I2 = 21.2%).

However, the effects on cardiopulmonary function (3 RCTs, n = 150; SMD = 0.13, 95% CI -0.11 to 0.43, I2 = 0.0%) and functional capacity (4 RCTs, n = 459; SMD = 0.09, 95% CI -0.21 to 0.38, I2 = 47.4%) were not significant. These results may be influenced by potential moderators such as intervention length, frequency, session duration, and medication use.

Crucially, there was no significant difference in acceptability or adverse events between two groups.

Conclusion
Exercise significantly improves fatigue, sleep, QoL, and depression in ME/CFS patients. However, exercise does not yield significant improvements in cardiopulmonary or functional capacity.

While not a cure, exercise is an effective strategy for symptom management, provided that individual energy limits are strictly respected to prevent harm.

Web | DOI | Journal of Psychosomatic Research | Paywall

 

[Utsikt]

No mention of the quality of the evidence is a clear red flag. They essentially found that subjective outcomes are easy to manipulate.

 

[SNT Gatchaman]

We assessed the risk of bias (RoB) in the included studies using the Cochrane RoB Tool 2.0 for RCTs. The quality of evidence for each outcome was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) method. The reliability of the evidence was rated as high, moderate, low, or very low based on five criteria: risk of bias in studies, inconsistency, indirectness, imprecision, and publication bias (online supplement 4).

The results of the RoB 2 assessment are provided in Online Supplement 3. One study raised some concerns, and 16 studies was rated as having a high risk of bias. According to GRADE criteria, the overall quality of evidence for the outcomes ranged from low to moderate. The main reason for downgrading was the risk of bias associated with allocation concealment (Online Supplement 4).

Several limitations should be recognized in this review. First, evaluating primary ME/CFS symptoms inherently requires a reliance on patient self-report. While subjective scales are clinically essential for this population, they can be susceptible to psychological states and cognitive biases. Second, most studies did not report participants' baseline activity levels, making it difficult to account for pre-existing activity as a potential confounder. Third, a notable limitation arises from the patient populations in some included Chinese studies, where the unusually high proportion of working patients (full-time or part-time) questions their representativeness and disease severity. This heterogeneity, possibly influenced by sociocultural and economic pressures, limits generalizability to typical ME/CFS cohorts.

Finally, our risk of bias assessment using the RoB 2.0 tool indicated a prevalent high risk of bias across the included studies. However, in accordance with methodological frameworks for exercise research (such as the PERSIST guidance), we must acknowledge that this predominantly stems from the physical impossibility of blinding participants and personnel in behavioral interventions. Therefore, while these methodological constraints are important to note, they represent inherent challenges of the field rather than fatal flaws that would completely invalidate the clinical value of these trials.

 

[rvallee]

A total of 17 RCTs

And still this blatant cherry-picking, meaning this comes from the most selective set of positive trials they could find using combinations of keywords and criteria. That's about the number of trials published in a low-volume year. It would be so useful to have a real, validated number of those trials, even better a database, to emphasize just how extreme the cherry-picking they have to go through to present mediocre results like this as some sort of promising anything.

The main reason for downgrading was the risk of bias associated with allocation concealment

Just absolutely laughable. This is a concern, a major one, and it's not even a top 5 concern in trials like this.

However, in accordance with methodological frameworks for exercise research (such as the PERSIST guidance), we must acknowledge that this predominantly stems from the physical impossibility of blinding participants and personnel in behavioral interventions. Therefore, while these methodological constraints are important to note, they represent inherent challenges of the field rather than fatal flaws that would completely invalidate the clinical value of these trials.

No, they are absolutely fatal flaws. The idea that just because they can't be blinded then it doesn't matter is pure nonsense. The entire point of those trials is to convince people to rate themselves as healthier than they are, then asking them how healthy they think they are. This is the absolute maximum level of bias that is possible to find in a trial design, and there are even more biases that aren't reflected in the design but still influence everything. Especially the whole damn treatment itself.

Exercise significantly improves fatigue, sleep, QoL, and depression in ME/CFS patients. However, exercise does not yield significant improvements in cardiopulmonary or functional capacity.

And this shows how laughably absurd the whole thing is. The idea that fatigue can improve with no meaningful improvements in functional capacity makes absolutely no sense, it defies reason, logic and common sense. It would be like someone making more money but having no more to spend, even accounting for all other factors. A literal mathematical impossibility, makes a joke our of primary school arithmetic.

But of those course improvements are so trivial that it becomes easy to explain it, but the only valid conclusion to this is that all of it is pure worthless bunk. Same as it was on day one, and same as it will be when the last trial gets published.