LAUSR

standardization was used based on the EU standard population average over the period 2011-2030. The relative annual IRs between individuals living in urban and rural areas were expressed as incidence rate ratios (IRRs). Linear regression was used to assess trends in IRRs over time. Twosided P values less than .05 were considered statistically significant. There were 26,731 new cases of PD during the 18-year period. Using a production-based categorization of regions, the rural IR increased during the follow-up period (slope 5 0.0022, 95% CI 5 0.0003-0.0040, 0.22 new cases per 100,000 people per year, P 5 .029), as did the urban IR (0.0020, 95% CI 5 0.00002-0.0041; 0.20 new cases per 100,000 people per year, P 5 .048). Using population-based categorization, the rural IR increased (0.0031, 95% CI 5 0.0013-0.0049, 0.31 new cases per 100,000 people per year, P 5 .0023), but not the urban IR (0.0016, 95% CI 5 20.0004-0.0037, P 5 .11). However, the slopes were not significantly different between rural and urban areas using population-based (P 5 .28) or production-based categorization (P 5 .93; Fig. 1). The same was true when men and women were studied separately. The results show that the age-adjusted incidence of PD continues to be higher in rural regions when compared with urban areas and that the incidence is on the rise in both rural and urban regions (except for urban incidence using population-based categorization). However, despite the large reduction in the use of agrochemicals, the rural-to-urban population migration, the increased access to piped clean water, and other potentially relevant changes in rural living, no risk reduction in rural areas has occurred as of yet. This suggests that rural environmental factors continue to affect PD epidemiology and are associated with a long prediagnostic period and/or a long latency between exposure and the start of the neurodegenerative process.