Abstract
OBJECTIVE: To investigate the incidence of advanced cardiovascular-kidney-metabolic (CKM) syndrome and its associated risk factors in patients with early-onset type 2 diabetes mellitus (T2DM). METHODS: This cross-sectional study enrolled 1830 T2DM patients attending Shanghai Seventh People's Hospital (July 2019-May 2025). Participants were stratified into early-onset (diagnosis age <40 years; n=509) and non-early-onset (n=1321) cohorts. Advanced CKM was defined as stages 3-4 per American Heart Association (AHA) criteria. Comparative analysis, restricted cubic spline (RCS) modeling, binary logistic regression, and receiver operating characteristic (ROC) curves were employed to characterize advanced CKM distribution and determinants. RESULTS: Advanced CKM incidence was significantly lower in the early-onset group (31.2%, 159/509) versus the non-early-onset group (60.6%, 801/1321) (P < 0.001). Among patients with ≤10 years' disease duration, early-onset individuals exhibited a markedly lower incidence (19.95%, 80/401) compared to non-early-onset counterparts (53.46%) (P<0.001). With disease duration >10 years, the early-onset group incidence rose to 68.64% (79/108), converging with the non-early-onset group (76.59%; P = 0.08). Binary logistic regression identified independent risk factors for advanced CKM in early-onset T2DM: urine albumin-to-creatinine ratio (UACR; OR = 1.077, 95% CI: 1.046-1.110), blood urea nitrogen (BUN; OR = 1.202, 95% CI: 1.005-1.436), and diabetes duration (OR = 1.102, 95% CI: 1.060-1.145). Protective factors included subcutaneous fat area (OR = 0.995, 95% CI: 0.991-0.999) and antihypertensive medication use (OR = 0.374, 95% CI: 0.199-0.702). The ROC model incorporating these predictors demonstrated an AUC of 0.850 (95% CI: 0.812-0.888) for advanced CKM, with 84.3% sensitivity and 76.8% specificity. CONCLUSION: Early-onset T2DM patients exhibit a lower incidence of advanced CKM than non-early-onset individuals, though risk escalates substantially with prolonged diabetes duration. UACR, BUN, and diabetes duration are independent risk factors, while greater subcutaneous fat area and antihypertensive therapy confer protection. The derived prediction model may facilitate early clinical intervention.