Abstract
Sleep disturbances are both prodromes and potential risk factors for Parkinson's disease (PD), yet associations between long-term sleep patterns and PD remain unclear. We analyzed data from two online cohorts-PPMI-Online (n = 15,905) and Fox Insight (n = 1929)-to examine associations between self-reported sleep duration trajectories across life stages and PD risk and age at onset (AAO). Latent class growth analysis identified distinct trajectories, linear and logistic regression examined their associations with PD risk and AAO, adjusting for demographics, lifestyle, and comorbidities. In PPMI-Online, sleep duration was generally stable in early adulthood with divergence in midlife. Compared to stable sleepers (7-8 h/day), individuals with midlife reductions (6-7 to ≤5-6 h/day: OR = 1.90; 7-8 to ≤6-7 h/day: OR = 1.64) and persistent short sleep (≤6 h/day: OR = 1.41) had higher PD risk, independent of comorbidities. Similar patterns were observed in probable prodromal PD. Persistent short sleep and sleep decline were also linked to earlier AAO (up to -4.23 years) across cohorts. These findings suggest that midlife sleep reduction may signal early PD, while chronic short sleep may represent a modifiable risk factor, highlighting the need for prospective studies to explore early detection and prevention potential.