Diabetes preceding Parkinson's disease onset. A case-control study

Objective: To assess the association between diabetes preceding Parkinson's disease (PD) and PD. Methods: PD individuals were matched to PD free individuals randomly selected from people in the same municipality as the cases. Occurrence of diabetes preceding PD onset among cases and controls was assessed through a structured questionnaire. Information regarding current and past medical treatment and other variables was also collected. We used univariate and multivariate logistic models to calculate crude and adjusted odds ratios (OR). Covariates are adjusted for included education, smoking habit, alcohol and coffee consumption. Results: 318 PD individuals (165 women, 153 men) and 318 matched controls were included in the study. PD patients had a mean age at interview of 66.7 years. Mean age at PD onset was 60.8 years and mean PD duration 5.9 years. We found an inverse association between PD and diabetes preceding PD onset in all groups stratified by gender, age at PD onset, body mass index (BMI), smoking habit, alcohol and coffee consumption. Multivariate analysis yielded the same findings after controlling for the variables (adjusted OR 0.4; 95% CI, 0.2-0.8). Conclusions: Our findings provide additional support for a potential link between diabetes and PD