Parkinson's disease incidence: magnitude, comparability, time trends.

In this study, we reviewed incidence surveys of Parkinson's Disease (PD) from all over the world, published during the period 1945-1989, using reported quality criteria. In addition, we compared age-specific PD incidences from selected observations by stratified analysis. Crude incidences were described for 11 populations, and age-specific incidences for three of them: Iceland, Rochester (Minn, USA), and Turku (Finland). Effect modification by age was detected: a) by comparing incidences by age at diagnosis with incidence by age at clinical disease onset; and b) when only data on onset of disease was computed. For disease onsets, the incidences in Rochester for the period 1955-1966, and in Turku (Finland) during the interval 1968-1970, were lower than that in Iceland for the period 1958-1960: RR = 0.58 95% CI (0.41, 0.83), and RR = 0.67 95% CI (0.51, 0.87), respectively. For the Rochester population aged 40-69 years, a statistically significant 56% decrease in the incidences of Parkinsonism onsets during the period 1945-1966 was found. Validity problems in comparing PD incidences and the role of PD underdiagnosis were emphasized. We concluded that: a) stratified analysis is more suitable than standardization when comparing incidences for etiological purposes; b) the incidence of PD was highest in Iceland; and c) in Rochester, PD incidence under the age of 70 decreased with time.