Precision dynamics of predictive coding in functional neurological disorder

Abstract Functional neurological disorder (FND) presents disabling symptoms that fluctuate, migrate across systems and yet routinely show preserved structural integrity—features that can frustrate diagnosis and patient education. Crucially, these symptoms are genuine and reflect altered brain regulation across multiple systems rather than tissue damage or conscious control. This paper offers a clinically usable account of these puzzles by reframing FND as a disorder of precision control within predictive coding. The brain’s confidence (precision) in its own predictions is modelled as a dynamic quantity that can surge with arousal, settle back towards baseline and slowly recalibrate over longer periods. Placing this mechanism in a four-level hierarchy (affective, interoceptive, proprioceptive, spinal) shows how arousal can temporarily make an unhelpful expectation dominate, suppress corrective feedback, and produce motor, sensory, cognitive or visceral symptoms—often in combination. The same dynamics explain rule-in positive signs such as distractibility, entrainment and give-way weakness, and clarify the lived experience of reduced agency without implying wilfulness or malingering. For clinicians, the framework provides an intuitive way to explain symptoms to patients (‘a temporary gain miscalibration that overweights predictions and underweights sensory feedback’), and to justify why physiotherapy (amplifying proprioceptive feedback), psychotherapy (reducing arousal) and mindfulness or biofeedback (raising the effective threshold for arousal-driven ‘gain spikes’) can all help by recalibrating that weighting. The framework also yields concrete, falsifiable tests, linking bedside observation to measurable biomarkers and offering ways to refine it further. By centring dynamic precision, the account unifies diagnosis, communication and treatment planning, and provides a mechanistic foundation for precision-guided care in FND.