Triglyceride glucose–body mass index versus conventional lipid ratios for newly detected, ultrasound-defined hepatic steatosis compatible with MASLD: a retrospective cohort study in community-dwelling older adults with type 2 diabetes and hypertensio

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is common in older adults with type 2 diabetes and hypertension, yet simple predictors from routine examinations remain insufficiently evaluated. We compared the triglyceride glucose–body mass index (TyG-BMI) with TyG-WC and six lipid ratios for newly detected, ultrasound-defined hepatic steatosis compatible with MASLD, testing robustness to outcome definition and differential detection. Methods: All 898 participants had type 2 diabetes and hypertension and attended three annual examinations (2023–2025); the 453 free of baseline steatosis with complete covariates formed the analytic cohort. The primary estimand was cumulative risk over the fixed window to 2025, estimated by logistic regression with covariates excluding the components of TyG-BMI, supported by Cox and complementary log–log models. The primary outcome was ever-positive; last-visit-positive and persistent-positive were sensitivity analyses. Results: Steatosis was newly detected in 203/453 participants (44.8%; a detection rate, not incidence). TyG-BMI showed the largest estimated association (OR per 1-SD 2.03, 95% CI 1.62–2.54; standardized risk ratio 1.35, 1.26–1.47) and the highest observed single-index discrimination (AUC 0.682, 0.632–0.731). After Holm correction, it discriminated better than TyG, LDL/HDL-C, NHHR and AIP, but not TyG-WC or UHR, and held under all three outcome definitions (1.82–2.03). It did not discriminate significantly better than BMI alone (ΔAUC 0.011, 95% CI −0.017 to 0.040; DeLong P=0.45). Incremental value over a BMI-containing model was small (apparent AUC 0.680 to 0.692; IDI 0.015, 0.001–0.045; continuous NRI confined to non-events). Entering BMI and TyG separately performed equally well. A cases-only reclassification attenuated the association monotonically by construction; a symmetric version left it intact. Read reversion fell with body mass overall but only within ungraded reads, leaving misclassification direction undetermined. Conclusions: BMI, recorded at every visit, captured most of the available discrimination, and TyG-BMI added no demonstrable value beyond it, so no case for adding it arises here. TyG-BMI had the highest observed discrimination and beat most, though not all, of the conventional ratios after correction, and remains the candidate where a single composite is preferred. Discrimination was moderate; any expanded role would require added value over BMI and external validation. Keywords: TyG-BMI; lipid ratios; ultrasound-detected hepatic steatosis; insulin resistance; type 2 diabetes; hypertension; older adults; cohort study; community chronic-disease management. Preprint notice: This manuscript is a preprint, has not undergone peer review, and has been submitted for publication.