Abstract
Frailty is widely recognized to be associated with kidney function decline, but its longitudinal relationship with rapid estimated glomerular filtration rate (eGFR) decline remains unclear. This study examined associations between frailty measures and rapid eGFR decline using both frailty phenotype (FP) and frailty index (FI) in two large cohorts. We analyzed baseline frailty data from the China Health and Retirement Longitudinal Study (CHARLS) and Health and Retirement Study (HRS), with longitudinal trajectory analysis in HRS participants. Rapid eGFR decline was defined as annual decrease ≥4 mL/min/1.73 m(2)/year. Multiple imputation with inverse probability weighting addressed missing data and selection bias. Using FP criteria, frailty was significantly associated with higher odds of rapid eGFR decline compared to non-frail individuals (CHARLS: OR 1.93, 95% CI 1.13-3.29; HRS: OR 1.48, 95% CI 1.10-1.99). Longitudinally, progression from non-frail to pre-frail/frail status was associated with 27% higher odds, while improvement from pre-frail status was associated with 35% lower odds. Frailty improvement in baseline-frail individuals showed no significant effect. Each 1-point annual increase in FP score was associated with 91% higher odds (OR 1.91, 95% CI 1.46-2.51). FI assessment yielded consistent results with optimal population-specific thresholds (FI ≥ 0.25 for CHARLS, FI ≥ 0.32 for HRS). Reverse association analysis revealed bidirectional relationships, more pronounced among participants with baseline eGFR <60 mL/min/1.73 m(2). Both baseline frailty and frailty progression are independently associated with rapid eGFR decline, supporting the potential value of frailty assessment in identifying patients at risk for kidney function decline.