BACKGROUND: Recent in vitro and animal experiments suggest that peroxisome proliferation-activated receptor gamma (PPARɣ) agonist medications, such as antidiabetic glitazone (GTZ) drugs, are neuroprotective in models of Parkinson's disease (PD). These findings have not been tested in humans. We hypothesized that individuals prescribed GTZ drugs would have a lower incidence of PD compared to individuals prescribed other treatments for diabetes. METHODS AND FINDINGS: Using primary care data from the United Kingdom Clinical Practice Research Datalink (CPRD), we conducted a retrospective cohort study in which individuals with diabetes who were newly prescribed GTZ (GTZ-exposed group) were matched by age, sex, practice, and diabetes treatment stage with up to five individuals prescribed other diabetes treatments (other antidiabetic drug-exposed group). Patients were followed up from 1999 until the first recording of a PD diagnosis, end of observation in the database, or end of the study (1 August 2013). An incidence rate ratio (IRR) was calculated using conditional Poisson regression, adjusted for possible confounders. 44,597 GTZ exposed individuals were matched to 120,373 other antidiabetic users. 175 GTZ-exposed individuals were diagnosed with PD compared to 517 individuals in the other antidiabetic drug-exposed group. The incidence rate (IR) of PD in the GTZ-exposed group was 6.4 per 10,000 patient years compared with 8.8 per 10,000 patient years in those prescribed other antidiabetic treatments (IRR 0.72, 95% confidence interval [CI] 0.60-0.87). Adjustments for potential confounding variables, including smoking, other medications, head injury, and disease severity, had no material impact (fully adjusted IRR 0.75, 0.59-0.94). The risk was reduced in those with current GTZ prescriptions (current GTZ-exposed IRR 0.59, 0.46-0.77) but not reduced among those with past prescriptions (past GTZ-exposed IRR 0.85, 0.65-1.10). Our study only included patients with diabetes who did not have a PD diagnosis when they were first prescribed GTZ, and thus, it cannot establish whether GTZ use prevents or slows the progression of PD. CONCLUSIONS: In patients with diabetes, a current prescription for GTZ is associated with a reduction in incidence of PD. This suggests PPAR gamma pathways may be a fruitful drug target in PD.