A cross-sectional screening model using routine health indicators including the TyG index and uric acid for MASLD in rural elderly populations

To develop a screening model for metabolic dysfunction-associated steatotic liver disease (MASLD) using routine health indicators in rural older adults, and to evaluate the incremental screening value of the triglyceride-glucose (TyG ) index and uric acid. This cross-sectional study analyzed health-examination data collected from 1 January to 31 December 2024 involving 3,020 rural older adults (≥ 60 years) in Putian City, Fujian Province. MASLD was defined according to the 2024 Chinese guideline, incorporating documented diabetes, antihypertensive medication, and lipid-lowering medication into the cardiometabolic criteria. A prespecified core model (sex, age, ALT, AST, GGT, and serum creatinine) was fitted to the full cohort, with uric acid and the TyG index added separately and jointly to evaluate incremental screening value. Internal validation was performed using 1,000 bootstrap resamples. Logistic regression, ROC curves, calibration assessments, and decision curve analysis were used for model development and evaluation. Among 3,020 participants, 792 (26.2%) met the concurrent MASLD definition. The TyG index showed the strongest independent association with MASLD (OR = 2.823 per 1-unit increase, 95% CI 2.422–3.290, P < 0.001); uric acid was also independently associated (OR = 1.041 per 10 µmol/L, 95% CI 1.029–1.054, P < 0.001). The final model achieved an optimism-corrected AUC of 0.783 (95% CI 0.764–0.800). Calibration assessment revealed local miscalibration (Hosmer-Lemeshow P < 0.001), though calibration intercept (− 0.001) and slope (1.002) were near-ideal; the Brier score was 0.155 versus 0.193 for the null model (Brier skill score = 0.197). The TyG index alone demonstrated moderate discrimination (AUC = 0.736, 95% CI 0.716–0.756), with a sample-specific optimal cutoff of 8.675; MASLD prevalence increased progressively from 7.7% in the lowest TyG quartile to 49.9% in the highest (P for trend < 0.001). Uric acid alone showed limited discrimination (AUC = 0.590, 95% CI 0.567–0.613), and no clinical cutoff is proposed. An exploratory two-marker model containing only TyG and uric acid achieved an AUC of 0.743 (95% CI 0.723–0.762) with a Brier score of 0.167. The TyG index provided greater incremental screening information for MASLD than uric acid in this rural older adult cohort. The final model demonstrated moderate discrimination (optimism-corrected AUC = 0.783) using only routine health indicators, suggesting potential utility for preliminary case finding in primary care. Furthermore, the TyG + uric acid model may serve as a resource-appropriate alternative in settings where liver and kidney function tests are not routinely available. However, residual miscalibration and the absence of external validation limit immediate clinical implementation; the model should be considered a preliminary screening tool requiring prospective and external validation.