Association of Cystatin C Kidney Function Measures With Long-term Deficit-Accumulation Frailty Trajectories and Physical Function Decline

Importance: It remains unclear whether cystatin C and cystatin C-based kidney function measures are associated with frailty trajectories and physical function decline. Objective: To examine the associations of cystatin C level, cystatin C estimated glomerular filtration rate (eGFRcys), and the difference between eGFRs (eGFRdiff) using cystatin C and creatinine levels with long-term deficit-accumulation frailty trajectories and physical function decline. Design, Setting, and Participants: This prospective cohort study used data from 15 949 participants in the China Health and Retirement Longitudinal Study (CHARLS) and the US Health and Retirement Study (HRS), 2 ongoing nationally representative cohort studies enrolling community-dwelling older people. Biennial surveys, known as waves, are conducted in both the CHARLS and the HRS. Seven-year data from wave 1 (May 2011 to March 2012) to wave 4 (July to September 2018) in the CHARLS and 12-year data from wave 8 (March 2006 to February 2007) to wave 14 (April 2018 to June 2019) in the HRS were assessed, with wave 1 in the CHARLS and wave 8 in the HRS serving as baseline waves. Data were analyzed from February 12 to May 20, 2022. Exposures: Baseline serum cystatin C and creatinine levels. Cystatin C eGFR and creatinine estimated GFR (eGFRcr) were calculated using the 2021 race-free equations developed by the Chronic Kidney Disease Epidemiology Collaboration. The difference between eGFRcys and eGFRcr was calculated by subtracting eGFRcr from eGFRcys. Main Outcomes and Measures: Based on 12-year follow-up data from the HRS and 7-year follow-up data from the CHARLS, a 29-item deficit-accumulation frailty index (FI) was constructed to assess frailty trajectories at each visit. Physical function decline was evaluated using repeated objective physical function measurements (grip strength and gait speed). Linear mixed models were used to examine longitudinal associations. Results: Among 15 949 older adults included in the analysis, 9114 participants were from the HRS (mean [SD] age, 66.2 [10.1] years; 5244 women [57.5%]), and 6835 were from the CHARLS (mean [SD] age, 58.4 [9.8] years; 3477 women [50.9%]). With regard to race and ethnicity, the HRS cohort included 7755 White individuals (85.1%) and 1359 individuals (14.9%) of other races and/or ethnicities (including American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Pacific Islander, and other); all participants in the CHARLS cohort were of Chinese ethnicity. Each SD increment in serum cystatin C was associated with a faster increase in FI in both the HRS cohort (β = 0.050 SD/y; 95% CI, 0.045-0.055 SD/y; P = .001) and the CHARLS cohort (β = 0.051 SD/y; 95% CI, 0.042-0.060 SD/y; P = .001). An inverse association was observed for eGFRCys (HRS cohort: β = -0.058 SD/y; 95% CI, -0.062 to -0.053 SD/y; P = .001; CHARLS cohort: β = -0.056 SD/y; 95% CI, -0.064 to -0.047 SD/y; P = .001). These associations remained after controlling for serum creatinine (β = 0.051 SD/y; 95% CI, 0.042-0.060 SD/y; P = .001) and eGFRcr (β = -0.056 SD/y; 95% CI, -0.064 to -0.047 SD/y; P = .001) in the CHARLS cohort. Similar to the results observed for eGFRcys, each SD increment in the eGFRdiff was associated with a slower increase in FI (β = -0.027 SD/y; 95% CI, -0.035 to -0.018 SD/y; P = .001) in the CHARLS cohort. Similar findings were observed for physical function decline. For example, each SD increment in serum cystatin C was associated with faster decreases in both grip strength (β = -0.006 SD/y; 95% CI, -0.008 to -0.003 SD/y; P = .001) and gait speed (β = -0.007 SD/y; 95% CI, -0.011 to -0.003 SD/y; P = .001) in the HRS cohort and faster decreases in gait speed (β = -0.017 SD/y; 95% CI, -0.027 to -0.006 SD/y; P = .002) in the CHARLS cohort. Conclusions and Relevance: In this cohort study, cystatin C, eGFRcys, and eGFRdiff were associated with long-term frailty trajectories and physical function decline among community-dwelling older people without frailty. Monitoring kidney function using cystatin C could have clinical utility in identifying the risk of accelerated frailty progression.