The joint impact of elevated homocysteine and type 2 diabetes on non-healing wounds and all-cause mortality: integrated clinical and multi-omics analyses.

BACKGROUND: Hyperhomocysteinemia (HHcy) and type 2 diabetes (T2D) are recognized risk factors for non-healing wounds, yet their combined effects and underlying mechanisms remain unclear. METHODS: In this study of 8406 National Health and Nutrition Examination Survey participants (1999-2004), weighted logistic regression was used to estimate odds ratios for non-healing wounds and Cox proportional hazards models were employed to assess hazard ratios for all-cause mortality. Additive and multiplicative interactions were evaluated. Nonlinear relationships used restricted cubic splines. Mediation analyses explored inflammatory marker contributions. Bulk transcriptomic and single-cell sequencing data were integrated to identify homocysteine (Hcy)-diabetic foot ulcers (DFU) hub genes, construct an Hcy risk gene scoring system, and elucidate key mechanisms by which Hcy remodels the DFU microenvironment. RESULTS: Both T2D [odds ratio (OR): 2.42, 95% confidence interval (CI): 1.69-3.47] and HHcy (OR: 1.67, 95% CI: 1.10-2.54) independently elevated non-healing wound risk. Participants with both conditions exhibited higher odds (OR: 5.28, 95% CI: 3.20-8.70), with additive interaction (relative excess risk due to interaction: 2.83, AP: 0.54, SI: 2.95). In diabetic men, Hcy displayed a J-shaped relationship with non-healing wounds ( P for nonlinearity = 0.027), with the lowest risk around 8.9 µmol/L. In diabetic patients, inflammatory markers mediated the link, with proportions of 11.1% for red cell distribution width, 9.16% for monocyte/lymphocyte ratio, 5.28% for systemic inflammation response index, and 4.03% for neutrophil/lymphocyte ratio. Hcy drove immune-metabolic reprogramming in the DFU microenvironment by regulating key gene networks such as IL1B and MMP9, activating the MIF-CD74/CXCR4 axis to form a B cell-centric inflammatory cascade network. Both T2D and HHcy were associated with higher all-cause mortality, without joint interaction observed. Participants with a history of non-healing wounds had a higher all-cause mortality risk. CONCLUSIONS: T2D and HHcy synergistically worsen non-healing wounds, but not all-cause mortality. In diabetic males, Hcy levels displayed a J-shaped relationship with non-healing wound risk. Mechanistically, Hcy amplified systemic inflammation and drove stromal-immune crosstalk via the MIF-CD74/CXCR4 axis, reshaping the diabetic wound microenvironment.