False fatigue alarms / functional disorder
An over-protective nervous system generates fatigue and effort signals out of proportion to the body's actual state.
This is the idea that, after a prolonged trigger, the nervous system can get 'stuck' over-predicting threat and generating fatigue, effort, and other symptoms disproportionate to any peripheral damage — a predictive-coding or functional-disorder framing. It is the only mechanism with no confirmatory test: it is reached by ruling other things out, and its likelihood rises as organic mechanisms are excluded. This is real and treatable, and it coexists with organic drivers rather than replacing them — it is never 'all in your head.' Pacing, CBT (which helps coping and fatigue but does not assume a psychological cause), and acceptance-based approaches can help; heavily-marketed brain-retraining programs (DNRS, Lightning Process, Gupta) have weak, conflicted evidence and can do harm by implying non-recovery is the patient's fault.
Treatment options & their evidence
Graded honestly — including treatments that failed in good trials, which is worth knowing.
Cognitive behavioral therapy (CBT)
Strong evidencepromise 3 · 2 RCTsThe best-evidenced behavioral option for fatigue/coping (e.g. the ReCOVer 'Fit after COVID' RCT). It targets perpetuating factors and does NOT assume a psychological cause; note it improves subjective fatigue more than objective activity or cognition.
Caution: Must not include graded-activity escalation where PEM is present, and must avoid 'dysfunctional beliefs' framing.
Structured pacing (energy-envelope management)
Weak evidenceThe only approach recommended when post-exertional malaise is present: manage activity within an energy envelope to avoid triggering crashes. The large RECOVER-ENERGIZE pacing RCT is due to report.
Acceptance & commitment therapy (ACT)
Weak evidenceBuilds psychological flexibility and reduces symptom-related avoidance; thin Long COVID-specific evidence, mostly extrapolated from chronic-illness literature.
Caution: Acceptance framing should be paired with validation, not 'just give up.'