Comparison of low eGFR prevalence and prediction for mortality using 2009 and 2021 CKD-EPI equations in Mexican adults

Abstract

Accurate estimation of glomerular filtration rate (eGFR) is essential for identifying and managing chronic kidney disease (CKD). The CKD-EPI 2021 equation removed the race coefficient from the 2009 version, but its impact in Mexican populations remains unclear. Here, we compared eGFR category prevalence, and predictive performance between the CKD-EPI 2009 and 2021 creatinine-based eGFR equations, as well as the prognostic relevance of reclassification in eGFR categories using the 2021 equation in Mexicans. We evaluated 25,110 adults ≥20 years from the 2016-2023 cycles of the Mexican National Health and Nutrition Survey (ENSANUT) to estimate national low eGFR and eGFR category prevalence using both equations. We also assessed 5-year and 10-year risk of all-cause, cardiovascular, and kidney-related mortality in 142,884 adults from the Mexico City Prospective Study (MCPS) using Cox proportional hazards and Fine & Gray regression models. In ENSANUT 2023, prevalence of eGFR <60mL/min/1.73m2 was lower with CKD-EPI 2021 (2.9%, 95%CI 1.56-4.24%) compared to the 2009 equation (3.6%, 95%CI 1.99-5.21). Use of the 2021 equation resulted in upward eGFR reclassification in 6.52% (95%CI 4.07-8.97) of adults ≥20 years, particularly among older adults and those with hypertension or diabetes, yielding a reduction in 486,532 adults identified with eGFR <60mL/min/1.73m² compared to the 2009 equation. In MCPS, despite both equations showing similar C-statistics, the 2021 equation showed slightly improved predictive performance for 5-year and 10-year mortality outcomes. The 2021 equation reclassified 8.3% of participants to higher eGFR categories, and reclassification was associated with decreased risk of all-cause, cardiovascular, and kidney-related mortality, particularly for participants reclassified upward from G3a-G5 categories. The CKD-EPI 2021 equation yields lower prevalence of low eGFR but leads to prognostically relevant eGFR category reclassification compared to the 2009 equation. Our findings support the implementation of the 2021 equation for population health monitoring in Mexico without compromising prognostic utility.

Keywords: CKD-EPI; Chronic kidney disease; Mexico; eGFR; mortality.

Figure 1.

Prevalence of eGFR categories and CKD based on low eGFR in Mexico using the 2021 CKD-EPI creatinine equation in ENSANUT 2016-2023 data. (A) G1, (B) G2, (C) G3a, G3b, G4, and G5, (D) Overall CKD prevalence and CKD prevalence stratified by (E) sex, (F) age category, (G) diabetes, (H) hypertension, and (I) obesity status. G1 normal or high (≥90 mL/min per 1.73 m²), G2 mildly decreased (60–89 mL/min/1.73 m²), G3a mildly to moderately decreased (45–59 mL/min/1.73 m²), G3b moderately to severely decreased (30–44 mL/min/1.73 m²), G4 severely decreased (15-29 mL/min/1.73 m²), and G5 kidney failure (<15 mL/min/1.73 m²). eGFR: Estimated glomerular filtration rate, CKD: Chronic Kidney Disease, CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration.

Figure 2.

eGFR distribution and comparison of CKD prevalence using either 2021 or 2009 CKD-EPI creatinine equations in ENSANUT 2016-2023 data. (A) eGFR distribution, (B) Comparison of overall CKD prevalence using both equations, (C) Comparison of CKD prevalence in men, (D) age category, (E) diabetes status, and (F) hypertension status. eGFR: Estimated glomerular filtration rate, CKD: Chronic Kidney Disease, CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration.

Figure 3.

eGFR distribution and all-cause mortality risk prediction performance of eGFR estimated with either 2021 or 2009 CKD-EPI creatinine equation in the Mexico City Prospective Study. (A) Comparison of eGFR distribution in MCPS, (B) Comparison of HRs estimations across continuous eGFR levels, (C) 5-year discrimination assessment of eGFR estimated with either 2009 or 2021 CKD-EPI creatinine equation, (D) 10-year discrimination assessment of eGFR estimated with either 2009 or 2021 CKD-EPI creatinine equation. In both C and D, the gray line represents the baseline model without the eGFR covariate. All models were stratified by age category and adjusted for sex, place of residence, educational level, smoking, diabetes, history of cardiovascular disease, systolic blood pressure, and total cholesterol. To this baseline model, GFR estimated with either the 2021 or 2009 CKD-EPI creatinine equation was added. eGFR: Estimated glomerular filtration rate, CKD: Chronic Kidney Disease, CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration.

Figure 4.

Kaplan-Meier cumulative incidence curves adjusted using inverse probability of treatment weighting (IPTW) by age, sex, community of origin, educational level, smoking, diabetes, prior cardiovascular disease, systolic blood pressure and total cholesterol for all-cause, cardiovascular and kidney-related mortality in MCPS participants. Panels A-C compare participants reclassified from G2 to G1 and Panels D-F participants reclassified from G3a-G5 upward using the 2021 eGFR creatinine-based CKD-EPI equation compared to non-reclassified participants in the same categories for risk of these outcomes.