#3617 APPLICATION OF CKD EPI 2021 EQUATION FOR ESTIMATION OF GLOMERULAR FILTRATION RATE (EGFR) TO AN ELDERLY KOREAN COHORT TO PREDICT MORTALITY

Abstract

Abstract Background and Aims A new eGFR equation using serum creatinine (CKD-EPI 2021) without race were developed by Chronic Kidney Disease Epidemiology Collaboration in 2021. We projected the changes of eGFR and CKD prevalence in a prospective cohort of elderly Koreans, using current (MDRD equation and CKD-EPI 2009) and new equation (CKD-EPI 2021) Method This study was a part of the Korean Longitudinal Study on Health and Aging, which included a randomly selected, community-based, elderly population aged 65 years or more in a city of Korea, Seong-nam. The base line study was conducted from Sep. 2005 to Sep. 2006; the follow-up study, from May 2010 to March 2012, and the final survey about mortality and renal function, at Jan. 2023. Among 1,000 subjects originally included in the study, 984 subjects with baseline serum creatinine were included in this analysis. We compared the difference of eGFR and predictability of mortality and renal survival between values calculated by MDRD, CKD-EPI 2009, and CKD-EPI 2021. Serum creatinine was expressed as the IDMS-traceable creatinine. Results At baseline study, there were 436 males (44.3%). Age was 76.0 ± 9.1 years. Levels of eGFR at baseline study were 74.2 ± 21.4 ml/min/1.73 m2 by MDRD, 72.2 ± 16.9 ml/min/1.73 m2 by CKD-EPI2009, and 76.6 ± 17.3 ml/min/1.73 m2 by CKD-EPI2021. Values subtracted eGFRs by MDRD or CKD-EPI2009 from eGFR by CKD-EPEI2021 were 2.5 ± 10.1 ml/min/1.73 m2 and 4.5 ± 1.0 ml/min/1.73 m2, respectively. Between eGFRs by CKD-EPI2009 and CKD-EPI2021, the difference of eGFR was higher in females compared to males (p<0.001), was not different between subjects aged less than 75 years and more than 75 years (p = 0.139), and lowest in subject with <45 ml/min/1.73 m2 of eGFR by CKD-EPI2009 compared to subjects with other eGFR values (p<0.001). CKD stage was improved in 241 (24.5%) participants and was not changed in 743 (75.5%) using eGFR calculated by CKD-EPI2021 instead of CKD-EPI2009. Similar trend of eGFR difference between eGFRs calculated by MDRD and CKD-EPI2021. During 8.3 ± 6.2 years, 255 (25.9%) subjects were dead and 7 (0.7%) subjects had end stage renal disease (ESRD) before death. Any eGFR was an independent risk factors to mortality estimated by Cox's hazard proportional model adjusted by related factors. AUC to estimate survival by eGFRs was not different between eGFRs by CKD-EPI2021 and CKD-EPI2009 [0.662 (0.625-0.699) vs 0.658(0.620-0.695), p = 0.758], however, different between eGFRs by CKD-EPI2021 and MDRD [0.662 (0.625-0.699) vs 0584 (0.543-0.626), p<0.001]. AUC to estimate renal survival by eGFRs was not different between eGFRs by CKD-EPI2021 and CKD-EPI2009, also. Conclusion The eGFR calculated by CKD-EPI2021 showed higher value compared to eGFR calculated by CKD-EPI2009 or MDRD. The power to estimate mortality or renal survival was not different between eGFRs by CKD-EPI2021 and CKD-EPI2009.