Title: Diabetes and Parkinson's Disease: A Hazard Ratio Meta-Analysis with Global and Sex-Specific Projections to 2050
Authors: Jamir Pitton Rissardo, Ana Leticia Fornari Caprara, and Ian M. Walker
Conference: 2026 MDS-PAS, Houston, TX
Objective
To estimate Parkinson’s disease (PD) risk in individuals with diabetes mellitus (DM) using hazard ratios (HRs) from population-based studies and assess variation by sex, region, and diabetes type.
Background
DM and PD may share mechanisms such as insulin resistance and neuroinflammation, but the magnitude and consistency of this association remain unclear.
Design/Methods: A systematic review and meta-analysis of population-based cohort studies reporting HRs for PD among individuals with DM. Eligible studies used time-to-event models (Cox regression) and provided subgroup data by region (Asia, Europe, America), sex, and diabetes type (T1DM, T2DM, PreDM). Random-effects meta-analysis (DerSimonian-Laird) was applied. Publication bias was assessed via trim-and-fill and fail-safe N. Sex-specific projections of PD cases attributable to DM (2025–2050) were modeled using normalized population data and linear growth assumptions.
Results
Thirty studies were included. Pooled log(HR) for PD in DM was 1.29 (95% CI: 1.21–1.37), with a prediction interval of 0.85–1.72. Regional estimates: Asia 1.25 (1.08–1.42), Europe 1.37 (1.08–1.66), America 1.17 (1.11–1.22). Sex-specific log(HRs) were higher in women [1.41 (1.26–1.56)] than men [1.25 (1.10–1.41)] in European cohorts. By diabetes type, T2DM showed the strongest association [log(HR) 1.17–1.42], PreDM had a smaller effect (1.01–1.29), and T1DM was based on one study [1.38 (0.86–1.90)]. Heterogeneity was low (I² < 1%). Publication bias was minimal (adjusted log(HR) 1.36; fail-safe N > 11,000). Projections showed PD cases attributable to DM increasing from 12.3 to 21.4 per 100,000 in Asia, 15.1 to 24.8 in Europe, and 10.7 to 18.2 in America by 2050, with steeper increases in women.
Conclusion
Diabetes is associated with a modest but significant increase in PD risk, varying by geographic distribution, sex, and diabetes type. Using log(HRs) provides time-dependent insights, enhancing clinical relevance for risk stratification.
Citation
Rissardo JP, Caprara ALF, Walker I. Diabetes and Parkinson's Disease: A Hazard Ratio Meta-Analysis with Global and Sex-Specific Projections to 2050. Mov Disord Clin Pract 2026;13(S1):S67–S68. doi: 10.1002/mdc3.7047.
Figure 1. Forest plot showing pooled hazard ratios linking T2DM to Parkinson’s disease across Asian, European, and American cohorts, with subgroup and overall fixed‑effect estimates
Figure 2. Forest plot comparing hazard ratios for Parkinson’s disease in men and women with T2DM across Europe and Asia, showing subgroup and overall fixed‑effect estimates.
Figure 3. Forest plot showing pooled hazard ratios for Parkinson’s disease in prediabetes and type 1 diabetes across Asian and European cohorts, including subgroup and overall fixed‑effect estimates.
Figure 4. Funnel plot with confidence contours and Egger regression assessing publication bias across included studies examining diabetes and Parkinson’s disease, showing effect‑size symmetry and small‑study effects.
Figure 5. Projected Parkinson’s disease cases attributable to T2DM, T1DM, and prediabetes from 2025–2050 across Asia, Europe, and America, normalized per 100,000 population.
Figure 6. Projected Parkinson’s disease cases attributable to T2DM, T1DM, and prediabetes by sex from 2025–2050 across Asia, Europe, and America, normalized per 100,000 population.