Descriptive epidemiology of Parkinson's disease through proxy measures

Abstract

Background and Objective: In preparation for analytic study we undertook to describe areas of relative excess and deficit of Parkinson's Disease (PD) in Nebraska and tested two methodologic tools for inexpensive assessment of descriptive epidemiology of PD. Methods: In lieu of large-scale population screening and diagnosis, we obtained sales information of anti-PD drugs in the state in 1988 - 1990 as well as listings of all people dying from 1984 to 1993 who had Parkinson's Disease mentioned anywhere on their death certificate. The anti-PD drug sales data are intended as a proxy far prevalence, while the death certificate data are intended as a proxy for incidence. Results: Sales divided by population over age 54 indicates where anti-PD drug sales differ from expected. We found high correlation of drug sales rates with several farming exposures. Age-adjusted death rates, however, showed a low degree of association with sales or farming variables. This may be attributable to differences in death certificate completion or in underlying incidence versus prevalence. Conclusions: These techniques provide a useful tool for delineating possible differences in incidence and prevalence. While not as accurate as full community survey with expert diagnosis, they are not as expensive, and can be followed by local cluster investigations and individual-level etiologic studies to test hypotheses resulting from the initial study.