Novel obesity and metabolic indices better predict non-alcoholic fatty liver disease in elderly T2DM patients: evidence from cross-sectional and longitudinal analysis.

Objective: The associations between obesity- and metabolism-related indices and the risk of non-alcoholic fatty liver disease (NAFLD) in elderly patients with type 2 diabetes mellitus (T2DM) remain unclear. This study aimed to investigate these associations and assess their predictive value for NAFLD in this high-risk population. Methods: A total of 789 elderly T2DM patients recruited between 2020 and 2022 were included in the cross-sectional analysis, and 382 patients without NAFLD were followed in the longitudinal cohort for a median of 25.37 months. Binary logistic regression and Cox models were used to assess associations between obesity- and metabolism-related indices and NAFLD risk. Kaplan-Meier curves, restricted cubic spline (RCS) models, subgroup analysis, and receiver operating characteristic (ROC) curves were applied to explore these relationships further. Results: In the cross-sectional analysis, all obesity- and metabolic-related indices were significantly and positively associated with NAFLD risk, with odds ratio (OR) ranging from 1.014 (95% CI: 1.010-1.018) for LAP to 3.288 (95% CI: 2.414-4.533) for WHtR. RCS analysis revealed significant nonlinear associations for LAP, MetS scores, VAI, CMI, METS-IR, and ABSI. In the cohort analysis, 67 participants developed NAFLD, with an incidence rate of 8.35 per 100 person-years. Baseline LAP (HR = 3.10, 95% CI: 1.48-6.51), and MetS scores (HR = 4.26, 95% CI: 1.99-9.11) were independently associated with increased risk of incident NAFLD. Subgroup analysis demonstrated consistent positive associations across most subgroups. Time-dependent ROC analysis showed that LAP had the highest AUCs at 24 months (AUC = 0.725). Conclusion: The findings from cross-sectional and cohort studies collectively supported that MetS score and LAP may be the most effective predictive indicators for the risk of NAFLD among Chinese elderly T2DM Patients.