Unmasking cardiovascular risk beyond BMI in type 2 diabetes and MASLD: an interpretable multiorgan score for lower-BMI patients

<title>Abstract</title> <p>Background Cardiovascular risk in patients with type 2 diabetes mellitus (T2DM) with metabolic dysfunction-associated steatotic liver disease (MASLD) may be underestimated when body mass index (BMI) is used as the dominant marker of metabolic severity. Lower BMI does not preclude hepatic fibrosis, vascular injury, kidney dysfunction, or systemic vulnerability. We aimed to develop and internally validate an interpretable multiorgan score for cardiovascular risk stratification in lower-BMI patients with T2DM-MASLD. Methods In this retrospective hospital-based cohort study, adults with T2DM, MASLD, and BMI &lt; 27.0 kg/m² were identified from the hospital information system of Gansu Provincial Hospital, China. Patients were divided before modelling into training and locked internal-validation cohorts at a 7:3 ratio. Candidate predictors were screened within the training cohort only. A parsimonious ridge-logistic model was developed as the primary model, converted into an additive clinical score, and evaluated by discrimination, precision–recall performance, Brier score, calibration, decision-curve analysis, and training-derived risk strata. Results Among 1466 lower-BMI patients with T2DM-MASLD, 240 had documented cardiovascular disease (CVD) by the administrative study cut-off. The training and validation cohorts included 1026 and 440 patients, respectively, with identical CVD rates of 16.4%. The final A2 score retained eight routinely available predictors: diabetes duration, carotid plaque number, mean carotid intima–media thickness, age, fibrosis-4 index, estimated glomerular filtration rate, LDL cholesterol, and albumin. In locked internal validation, A2 achieved an area under the receiver operating characteristic curve of 0.792, average precision of 0.420, Brier score of 0.116, calibration intercept of − 0.307, and calibration slope of 0.849. Training-derived low-, intermediate-, and high-risk strata separated observed CVD burden in validation, with event rates of 2.9%, 12.0%, and 33.3%; corresponding rates in the BMI &lt; 25 kg/m² subgroup were 1.3%, 10.8%, and 36.8%. Conclusions In lower-BMI T2DM-MASLD, cardiovascular burden was not captured by BMI category alone. A routine, interpretable multiorgan score provided stable internal risk stratification and may support earlier cardiovascular evaluation beyond BMI-defined obesity categories, pending external and prospective validation.</p>