• Parkinsonism;
  • clinicopathological correlation;
  • autopsy;
  • diagnosis


We correlated the clinical features with pathological findings in an autopsy series of cases of incident parkinsonism. We used the medical records-linkage system of the Rochester Epidemiology Project to identify all incident cases of parkinsonism in Olmsted County, MN, for the years 1976 to 1990. Medical histories were abstracted in a standardized manner. Included in this study were those incident cases who died and underwent autopsy. Brain sections were studied with routine histology and special stainings. A neuropathologist blinded to any clinical information assigned cases to neuropathological categories. We found 364 incident cases of parkinsonism of which 235 were deceased at the time of this study; there were 39 autopsied brains available for analysis (17% of deceased cases). Of the 16 patients diagnosed pathologically with Lewy body disease, documentation indicated that 8 had an early dementia, 3 had prominent dysautonomia, and 2 had prominent ataxia. Of the 7 patients diagnosed pathologically with progressive supranuclear palsy, 4 had no documentation of supranuclear gaze palsy, and 3 had no documentation of early falls. Of the 3 patients diagnosed pathologically with multiple system atrophy, none had prominent ataxia or dysautonomia documented. Of the 5 patients with vascular disease at pathology, none had been given the clinical diagnosis of vascular parkinsonism. Of the 8 cases given the clinical diagnosis of drug-induced parkinsonism, 6 were found to have basal ganglia pathology. The autopsied cases in this study were not representative of all patients with parkinsonism, because atypical cases are more likely to come to autopsy than typical ones. Despite this selection bias, the retrospective data collection, and the small sample size, we made several observations that illustrate the difficulty in achieving an accurate antemortem diagnosis of parkinsonism. © 2002 Movement Disorder Society