Mortality from Parkinson's disease: a population-based prospective study (NEDICES).

Most studies of mortality in Parkinson's disease have been clinical studies, yielding results that are not representative of the general population. We assessed the risk of mortality from Parkinson's disease in the Neurological Disorders in Central Spain (NEDICES) study, a prospective population-based study in which Parkinson's disease patients who were not ascertained through medical practitioners were also included. The cohort consisted of 5262 elderly subjects (mean baseline age, 73.0 years), including 81 with Parkinson's disease at baseline (1994-1995). Thirteen-year mortality was assessed. Two thousand seven hundred and one of 5262 subjects (51.3%) died over a median follow-up of 12.0 years (range, 0.04-14.8 years), including 66 of 81 subjects (81.5%) with Parkinson's disease at baseline and 2635 of 5181 subjects (50.8%) without Parkinson's disease at baseline. In an unadjusted Cox model, the hazard ratio of mortality was increased in subjects with Parkinson's disease (hazard ratio, 2.29; 95% confidence interval, 1.80-2.93; P < .001) versus subjects without Parkinson's disease (reference group). In a Cox model that adjusted for a variety of demographic factors and comorbidities, the risk of mortality remained elevated in subjects with Parkinson's disease (hazard ratio, 1.75; 95% CI, 1.32-2.31, P < .001). In additional Cox models, Parkinson's disease patients with dementia had particularly high risks of mortality (adjusted hazard ratio, 2.62; 95% CI, 1.40-4.90; P < .001). In this prospective population-based study, Parkinson's disease was an independent predictor of mortality in the elderly. Parkinson's disease patients with dementia had particularly high risks of mortality.