Cognitive fatigue has a neural signature that standard assessment cannot detect

Cognitive fatigue is one of the most disabling features of post-infectious illness, and one of the least accepted. Standard clinical evaluation treats it as subjective. Neuropsychological testing fails to capture it because it measures performance, not the cost of producing that performance. A 2026 study by Inderyas and colleagues in the Journal of Translational Medicine used 7 Tesla fMRI to image what is actually happening in the brain during cognitive fatigue in ME/CFS and long COVID — and what they found is the same thing 7T imaging has shown in autonomic testing: the abnormalities are there. The standard tools are not looking in the right place.

What 7 Tesla Imaging Shows During Cognitive Fatigue

The study used a Stroop colour-word paradigm — a sustained executive task that progressively depletes cognitive resources — divided into a Pre phase and a Post phase to capture connectivity patterns before and after fatigue had set in. At 7 Tesla field strength, subcortical circuits that lower-field clinical scanners typically miss become resolvable. In long COVID participants, the Pre phase showed significantly reduced functional connectivity between the nucleus accumbens and cerebellar vermis 3. The nucleus accumbens governs the subjective experience of cognitive effort as worthwhile — the signal that this task is worth sustaining attention on. When accumbens connectivity is reduced, this signal is attenuated. The patient is not choosing not to try. The neural signal that drives sustained trying is not being generated at normal strength. Once fatigue had set in during the Post phase, long COVID participants showed increased prefrontal-hippocampal connectivity — the brain shifting toward a memory-based compensatory strategy as executive resources deplete, a reorganization that consumes additional resources in an already fatigued system. In ME/CFS, the finding was in different circuitry: increased connectivity between the cuneiform nucleus and the medulla — brainstem structures coordinating arousal, autonomic cardiovascular regulation, and motor control. The cognitive effort was imposing demands on a system already managing impaired autonomic integration at the brainstem level. That impairment was visible. The standard evaluation did not see it because the standard evaluation does not look there.

Why Neuropsychological Testing Misses This

Neuropsychological testing measures cognitive output: how many items completed, how fast, how accurately. It does not measure what the neural system expends to produce that output. A patient can score within normal limits on a standardized cognitive battery while the brain connectivity required to generate that performance is visibly disrupted at 7 Tesla. The test result says normal. The imaging says the system is dysregulated under the conditions the test demands. These are different measurements of different things. The fact that one appears normal does not mean the other is absent — it means the one that appears normal is not measuring the right variable. Patients who have been told their cognitive symptoms lack objective correlates have been told something that is only true under a narrow definition of "objective" — one that excludes imaging their brains work abnormally hard not to fail.

What This Means for Patients Told Their Cognitive Symptoms Are Not Measurable

The Inderyas findings are one more piece of evidence in the same direction the autonomic research has been pointing for decades: the upstream problem is real, it is measurable, and the standard clinical workup is not designed to find it. The tools that would detect it exist. The 7 Tesla scanner exists. The 7T functional connectivity analysis exists. What does not yet exist is a clinical pathway that deploys these tools for patients with post-infectious cognitive fatigue, the way that transcranial Doppler and capnography do not yet exist in standard autonomic testing despite being validated for it. The abnormalities are present in these patients' brains during cognitive effort. Standard assessment returns them normal because standard assessment is measuring something else. That is a measurement gap. It is not a finding of no pathology.