Two sources of gain, one moving boundary: a hypothesis for pathological fatigue

Pathological fatigue is the most common disabling symptom in neurological disease and occurs across conditions with no shared pathology. It is poorly predicted by lesion site or disease burden and corresponds neither to performance decrement nor to peripheral physiology. Its neural correlates, now established by large meta-analyses of resting-state imaging and electrophysiology, are diffuse, diagnostically non-specific and inconsistent in direction, with hyperconnectivity and hypoconnectivity reported in the same networks. This review argues that the pattern treated as the field's central embarrassment is the expected signature of altered neural gain, the slope of the neural input–output function set by ascending neuromodulatory tone: raising gain makes strongly driven circuits respond more and weakly driven ones less, producing opposite-signed change within one brain. Pain research made the structurally identical move in defining central sensitisation. Drawing together fatigue neuroimaging, the psychophysics of conscious access, exercise physiology, sensory attenuation, disease mechanisms and pharmacological trials, I argue that a single global gain factor is insufficient, because it fixes the shape of the channel gain profile so that severity and phenotype cannot vary independently; patients matched on severity nonetheless differ in profile. A second, channel-specific source is therefore required, and I specify the designs under which the two are separately identifiable. Because gain sets which afferent signals reach the threshold for global broadcast, I propose that fatigue is the experience of an outward displacement of the boundary of conscious access, with the common factor setting severity and the channel profile setting phenotype, and state the measurements that would falsify each step.