Yue · Medicine 2016 · systematic review and meta-analysis of cohort studies · n=7 studies (1,761,632 participants)

Risk of Parkinson Disease in Diabetes Mellitus: An Updated Meta-Analysis of Population-Based Cohort Studies.

Cited 187 times in the scientific literature.

Level 3 - non-randomized controlled study

Meta-analysis of non-randomized prospective and retrospective cohort studies

PubMed 27149468 · doi:10.1097/MD.0000000000003549 · record verified 2026-08-27

What was done

Authors conducted a systematic search of PubMed and Embase through October 2015 to identify population-based cohort studies examining the relationship between diabetes mellitus and incident Parkinson disease. Adjusted risk estimates were combined using a random-effects meta-analysis. Publication bias was evaluated using funnel plots, Begg and Egger tests, and Duval and Tweedie trim-and-fill analysis. Subgroup analyses were performed across study quality, geographic region, study design, sample size, publication year, diabetes duration, and baseline age.

What was found

The meta-analysis included 7 population-based cohort studies covering 1,761,632 individuals. Diabetes mellitus was associated with a statistically significant increase in the pooled adjusted relative risk of Parkinson disease (RR 1.38, 95% CI 1.18 to 1.62, P < 0.001). The risk increase was present in females (RR 1.50, 95% CI 1.07 to 2.11, P = 0.019) and males (RR 1.40, 95% CI 1.17 to 1.67). Application of the trim-and-fill method slightly attenuated but maintained the association (RR 1.31, 95% CI 1.09 to 1.57, P = 0.015). Subgroup analyses by study quality, region, and duration showed similar risk elevations.

Why it matters

This updated synthesis of large population-based cohorts provides clear evidence that diabetes is an independent risk factor for developing Parkinson disease. It supports further investigation into shared metabolic and neurodegenerative pathways.

Limits

The findings derive entirely from observational cohort designs, which cannot establish causality or fully exclude residual confounding. The total number of included studies was relatively small (seven cohorts), and the abstract does not report specific diagnostic validation methods for Parkinson disease or details on glycemic control and anti-diabetic medication use.

Cited by