Abstract
Insulin resistance (IR) has been associated with incident frailty. However, the precise association between the estimated glucose disposal rate (eGDR) and the development of frailty remains insufficiently understood. We hypothesized that a lower eGDR, indicating greater IR, would be associated with a higher risk of developing frailty and with accelerated frailty progression across cohorts. This study aimed to systematically investigate and clarify this association. Data were obtained from three prospective cohorts: the China Health and Retirement Longitudinal Study (CHARLS), the English Longitudinal Study of Ageing (ELSA), and the Health and Retirement Study (HRS). Frailty was assessed using the Rockwood Frailty Scale. Baseline eGDR, changes in eGDR, and cumulative eGDR from baseline to the most recent survey wave, as well as the frailty index for each survey year, were evaluated. Logistic and Cox regression models were used to examine the association between eGDR and incident frailty. To assess frailty progression over time, we applied linear mixed-effects models and evaluated potential nonlinear trends. In addition, K-means clustering was conducted as an exploratory analysis to describe heterogeneity in longitudinal eGDR trajectories. A total of 7,666 participants were included from CHARLS (female: 51%, mean age: 58.27 years), 4,296 from ELSA (female: 53%, mean age: 64.89 years), and 4,370 from HRS (female: 56%, mean age: 66.63 years). A significant association was observed between lower eGDR and an increased risk of frailty (hazard ratio [HR]: 0.91, 95% CI 0.88-0.94; HR: 0.90, 95% CI 0.87-0.94; and HR: 0.95, 95% CI 0.92-0.98 for CHARLS, ELSA, and HRS, respectively) and with an accelerated increase in frailty index (β: -0.005, 95% CI -0.007 to - 0.004; β: -0.009, 95% CI -0.010 to - 0.008; and β: -0.009, 95% CI -0.010 to - 0.008 for CHARLS, ELSA, and HRS, respectively). Similarly, increases in cumulative eGDR and in changes in eGDR over time were associated with lower risks of incident frailty and frailty index. Subgroup and sensitivity analyses demonstrated consistent findings across subgroups. The elbow method suggested that three clusters represented the optimal solution; however, these clusters did not differ significantly in frailty outcomes and were therefore interpreted as descriptive rather than predictive. These findings highlight the importance of early detection of eGDR reduction and the implementation of targeted interventions to mitigate incident frailty.