Gait Rather Than Cognition Predicts Decline in Specific Cognitive Domains in Early Parkinson's Disease

Background: Dementia is significant in Parkinson's disease (PD) with personal and socioeconomic impact. Early identification of risk is of upmost importance to optimize management. Gait precedes and predicts cognitive decline and dementia in older adults. We aimed to evaluate gait characteristics as predictors of cognitive decline in newly diagnosed PD. Methods: One hundred and nineteen participants recruited at diagnosis were assessed at baseline, 18 and 36 months. Baseline gait was characterized by variables that mapped to five domains: pace, rhythm, variability, asymmetry, and postural control. Cognitive assessment included attention, fluctuating attention, executive function, visual memory, and visuospatial function. Mixed-effects models tested independent gait predictors of cognitive decline. Results: Gait characteristics of pace, variability, and postural control predicted decline in fluctuating attention and visual memory, whereas baseline neuropsychological assessment performance did not predict decline. Conclusions: This provides novel evidence for gait as a clinical biomarker for PD cognitive decline in early disease.