Abstract

Background: This study investigated the discordance between cystatin C- and creatinine-based estimates of glomerular filtration rate (eGFR) (EKFCcys/EKFCcre) to predict adverse outcomes in older adults.

Methods: Older patients with chronic kidney disease (CKD) (mean age 76.7) from 2018 to 2020 were enrolled. Cystatin C, creatinine, and clinical data were assessed at baseline and followed up until death, dialysis, or administrative censorship. The subjects were stratified according to tertiles of the EKFCcys/EKFCcre ratio. The subhazard ratios (sHR) and time ratios were measured using competing risk regression and the cause-specific accelerated failure time (CS-AFT) model. Sensitivity analysis including models with cystatin-to-creatinine (Cys/Cre) ratios was also performed.

Results: In 369 older patients, the incidence rates of mortality and dialysis were 2.7 and 7.0 per 1000 patient-months, respectively. The incidences of mortality and dialysis were higher in the lowest tertile (0.9 vs. 4.2 per 1000 patient-month and 6.1 vs. 7.0 per 1000 patient-months, respectively) than in the highest tertile group. In the competing risk regression, the lowest tertile had a higher risk of mortality (sHR [95% CI] = 6.78 [2.34–19.68], p < 0.05). In CS-AFT,  compared to the highest tertile group, the lowest tertile group exhibited an altered median survival time of 0.27 (0.11–0.68) and 0.36 (0.20–0.67), which is approximately 73% and 64% decrease in median survival time to mortality and dialysis, respectively (ps < 0.05). The results of the sensitivity tests were consistent.

Conclusion: The lowest tertile of the cystatin C-to-creatinine based eGFR ratio is a risk factor for clinical outcomes in older patients with CKD.