For the purpose of patient education, disease management, and the design of upcoming clinical studies, it is critical to identify the variables predicting and driving death in PD. Newly diagnosed PD patients and healthy controls (NC) from a population-based study that underwent recurrent evaluations over a ten-year period were included in this study. We estimated survival using the Kaplan-Meier method, identified baseline risk factors for mortality using Cox proportional hazards regression models, and assessed the effect of four clinical milestones of advanced PD (visual hallucinations, recurrent falls, dementia, and nursing home placement) on mortality risk using Cox regression models with time-dependent covariates. Three independent baseline predictors of death were older age, more severe motor impairment, and the Postural Instability-gait Difficulty (PIGD) phenotype. The cumulative effect of each clinical milestone on mortality more than quadrupled the probability of death, with an HR of 10.83 (95 percent CI 4.39-26.73) in patients who experienced all four clinical milestones. Patients with PD are more likely to die, and this increased risk is disease-related and manifests early in the course of the illness. Clinical milestones showed a significantly higher risk of death later in the illness course, but motor severity and the PIGD phenotype were early risk factors for mortality. This suggests that clinical milestones may be important for clinical disease staging and prognosis.
