Causal effects of diabetes and alcohol intake on Dupuytren’s disease: a two-sample Mendelian randomization study in UK Biobank and FinnGen

Abstract

Background: Dupuytren's disease (DD) is associated with metabolic and lifestyle traits. However, these associations remain uncertain because exposures frequently cluster, and conventional studies are vulnerable to residual confounding and reverse causation. Methods: A two-sample Mendelian randomization (MR) analysis using genetic instruments for diabetes, obesity-related, smoking, and alcohol-related traits from genome-wide association studies of large European ancestry was conducted. Summary statistics of DD were obtained from UK Biobank and FinnGen and combined using meta-analysis. Multiplicative random-effects inverse variance weighted MR was used as the primary method, with complementary sensitivity analyses to test robustness. Results: Genetic liability for diabetes was associated with a higher risk of DD (meta-analyzed odds ratio [OR], 1.0157; 95% confidence interval [CI], 1.0070-1.0245; P<0.001). Genetically predicted alcohol intake was also positively associated with DD (alcohol intake OR, 1.0150; 95% CI, 1.0087-1.0212; P<0.001; drinks per week OR, 1.0062; 95% CI, 1.0015-1.0110; P=0.0103). However, we found no convincing evidence that obesity traits or smoking intensity had causal effects on DD. Sensitivity analyses were directionally consistent, the instruments were adequately strong, and there was no substantial evidence of horizontal pleiotropy for the main associations. Although the observed effect sizes were modest, consistent with the multifactorial nature of DD, supporting directional risk associations rather than large individual-level effects. Conclusion: This MR study supported diabetes liability and higher alcohol intake as causal risk factors for DD, whereas genetically proxied obesity and smoking traits showed no clear causal effects. These findings suggest that better glycemic management and lower alcohol exposure may help reduce the risk of DD.