Correlation Between Gut Microbiota, Metabolites, and Diabetic Complications: A Two‐Sample Mendelian Randomization Study

Given the observed links between the gut microbiota and diabetic complications in epidemiological studies, we used a two‐sample Mendelian randomization (MR) method to assess causality, considering the role of microbial metabolic pathways.

We selected genetic variants associated with gut microbiota (n = 18,340) and microbiota‐derived metabolites (n = 7824) from genome‐wide association studies. Summary statistics for diabetic nephropathy (DN), diabetic neuropathy (DNe), diabetic peripheral vascular disease (DPVD), and diabetic retinopathy (DR) were obtained from FinnGen and CKDGen genome‐wide association studies databases. In primary causality assessment, we used inverse‐variance weighted analysis supplemented with MR‐Egger and weighted median methods. We conducted sensitivity analyses (Cochran's Q, MR‐Egger intercept, MR‐Pleiotropy RESidual Sum and Outlier, leave‐one‐out) to ensure robustness.

Genetically increased genus.Eubacteriumhallii was suggestively linked to higher diabetic nephropathy risk. Eubacteriumrectale and Eubacteriumventriosum showed protective associations. The genetically increased abundance of RuminococcaceaeUCG005, Butyrivibrio, and Streptococcus at the genus level was potentially associated with a protective effect in DNe. However, a significant positive relation was found between Actinomyces and DNe. As for diabetic peripheral vascular disease, Eubacteriumrectale, Ruminococcusgnavus, and LachnospiraceaeUCG008 were risk factors, whereas Coprococcus1, Hungatella, and LachnospiraceaeUCG001 were protective factors. A significant casualty with genus.Odoribacter and Sellimonas was causally linked to diabetic retinopathy. All the above findings were robust across several sensitivity analyses. Six gut metabolites also showed suggestive associations with complications.

Our study first detected causal relationships among gut microbiota, gut metabolites, and diabetic complications. Our findings may provide new targets for treatment and offer valuable insights for further studies into the underlying mechanisms.