Abstract
OBJECTIVE: The triglyceride-glucose (TyG) index is a validated marker of insulin resistance (IR) and predictor of cardiovascular outcomes. However, the prognostic utility of integrating TyG with body mass index (BMI) as the TyG-BMI index in elderly patients with severe heart failure (HF) and type 2 diabetes mellitus (T2DM) remains unestablished. We aimed to evaluate associations between TyG-BMI and all-cause mortality at multiple time points in this high-risk cohort. METHODS: This retrospective cohort study analyzed 4,523 elderly patients (aged >65 years) with severe HF and T2DM from the MIMIC-IV database. Participants were stratified into TyG-BMI quartiles (Q1-Q4) at ICU admission. Primary outcomes were 60-, 90-, 180-, and 365-day all-cause mortality. Associations were assessed using Kaplan-Meier analysis, Cox proportional hazards models, and restricted cubic splines (RCS). RESULTS: The cohort (mean age 72.79 ± 7.84 years; 41.5% male) demonstrated graded mortality reductions with increasing TyG-BMI quartiles. Compared to Q4, Q1 (lowest TyG-BMI) had significantly higher mortality at 90 days (58.70% vs. 48.45%; p = 0.008) and 365 days (80.54% vs. 73.91%; p < 0.001), with similar 60-day trends (58.79% vs. 39.34%; p = 0.059). Adjusted Cox models confirmed progressively lower mortality risk in higher quartiles (365-day HR for Q4 vs. Q1: 0.74, 95% CI: 0.68-0.93). Subgroup analyses demonstrated a consistent inverse TyG-BMI-mortality association across all strata (age, cardiac function, comorbidities), with pronounced risk reduction in HFrEF (LVEF ≤40%; all-timepoint HR >1, p<0.05) and patients without prior myocardial infarction (365-day aHR 0.69 vs. 0.81 with infarction). RCS analysis identified nonlinear thresholds (TyG-BMI = 148.73 for 60-day; 163.38 for 365-day mortality), below which each unit increase conferred greater protective effects. CONCLUSION: Lower TyG-BMI independently predicted increased short-, intermediate-, and long-term mortality in elderly patients with severe HF and T2DM. This composite index-integrating metabolic (TyG) and nutritional (BMI) dimensions-provides practical risk stratification, particularly within identified threshold ranges.