Breaking the Freeze: The Role of Cognitive Function in Freezing of Gait in Parkinson's Disease

Freezing of gait (FOG) is a highly disabling, poorly dopa-responsive symptom of Parkinson's disease (PD) that becomes increasingly prevalent with disease progression and is one of the main contributors to falls and loss of independence. Although FOG has long been viewed as a motor phenomenon, converging evidence shows that executive impairment-particularly deficits in attention, task-switching, inhibition, and visuospatial processing-is strongly implicated in its pathophysiology. Yet the field lacks a coherent framework explaining how these cognitive processes contribute to FOG and how they should be targeted in therapy. This narrative review synthesizes current evidence on the neural mechanisms linking cognitive dysfunction and FOG. Literature was identified through targeted searches of neuroimaging, behavioral, and rehabilitation studies in PD with and without FOG. Findings consistently demonstrate altered activity and connectivity within corticostriatal and corticolimbic circuits in freezers, including inefficient hyperactivation of the frontal, prefrontal, and posterior parietal cortices, and abnormal coupling between the ventral striatum, precuneus, and amygdala. These patterns suggest that cognitive networks become overrecruited yet insufficient to compensate for impaired motor automaticity, especially under dual-task demands or emotional load. Robust cognitive reserve can serve as compensation, whereas cognitive impairment can contribute to FOG in situations with high cognitive load. Freezers show disproportionate deficits in inhibitory control, visuospatial processing, and task-switching, which correlate with gait initiation failures and FOG severity. Cognitive training, particularly dual-task and other motor-cognitive interventions, shows promising yet variable effects on gait, executive performance, and FOG; however, the field lacks clarity about which cognitive and executive domains are causally involved and which merely reflect compensatory strain. FOG-specific mechanistic models integrating motor, executive, and limbic dysfunction are needed to guide individualized cognitive training and optimize therapeutic outcomes for people with PD.