Abstract
BACKGROUND: The correlation between various insulin resistance surrogates and frailty remains under investigation in the scientific community. METHODS: Data from NHANES (1999-2018) were used. We utilized weighted binary logistic regression, trend tests, RCS analysis, and subgroup analysis to comprehensively assess the link between METS-IR, HOMA-IR, and TyG, and frailty risk. RESULTS: The results revealed a significant positive association between high levels of METS-IR, HOMA-IR, and TyG with the risk of frailty in all models. Notably, in model 4, the highest quintile of METS-IR showed the strongest link (OR: 2.960, 95% CI: 2.219-3.949), with HOMA-IR (OR: 2.522, 95% CI: 1.927-3.301) following closely behind. Trend tests revealed a positive trend between METS-IR, HOMA-IR, and TyG with the risk of frailty (P for trend < 0.05). RCS analysis showed a linear relationship between METS-IR and the risk of frailty (P for nonlinearity > 0.05). In contrast, HOMA-IR and TyG exhibited a U-shaped nonlinear relationship (P for nonlinearity < 0.05). CONCLUSION: The research identified a linear association between METS-IR and frailty risk, whereas HOMA-IR and TyG displayed a U-shaped, nonlinear relationship pattern with the risk of frailty. Among the varying levels examined, the linkage between METS-IR and frailty was most pronounced in the top quintile.