Estimated Glomerular Filtration Rate, Albuminuria, and Adverse Outcomes: An Individual-Participant Data Meta-Analysis

Abstract

Importance: Chronic kidney disease (low estimated glomerular filtration rate [eGFR] or albuminuria) affects approximately 14% of adults in the US.

Objective: To evaluate associations of lower eGFR based on creatinine alone, lower eGFR based on creatinine combined with cystatin C, and more severe albuminuria with adverse kidney outcomes, cardiovascular outcomes, and other health outcomes.

Design, setting, and participants: Individual-participant data meta-analysis of 27 503 140 individuals from 114 global cohorts (eGFR based on creatinine alone) and 720 736 individuals from 20 cohorts (eGFR based on creatinine and cystatin C) and 9 067 753 individuals from 114 cohorts (albuminuria) from 1980 to 2021.

Exposures: The Chronic Kidney Disease Epidemiology Collaboration 2021 equations for eGFR based on creatinine alone and eGFR based on creatinine and cystatin C; and albuminuria estimated as urine albumin to creatinine ratio (UACR).

Main outcomes and measures: The risk of kidney failure requiring replacement therapy, all-cause mortality, cardiovascular mortality, acute kidney injury, any hospitalization, coronary heart disease, stroke, heart failure, atrial fibrillation, and peripheral artery disease. The analyses were performed within each cohort and summarized with random-effects meta-analyses.

Results: Within the population using eGFR based on creatinine alone (mean age, 54 years [SD, 17 years]; 51% were women; mean follow-up time, 4.8 years [SD, 3.3 years]), the mean eGFR was 90 mL/min/1.73 m2 (SD, 22 mL/min/1.73 m2) and the median UACR was 11 mg/g (IQR, 8-16 mg/g). Within the population using eGFR based on creatinine and cystatin C (mean age, 59 years [SD, 12 years]; 53% were women; mean follow-up time, 10.8 years [SD, 4.1 years]), the mean eGFR was 88 mL/min/1.73 m2 (SD, 22 mL/min/1.73 m2) and the median UACR was 9 mg/g (IQR, 6-18 mg/g). Lower eGFR (whether based on creatinine alone or based on creatinine and cystatin C) and higher UACR were each significantly associated with higher risk for each of the 10 adverse outcomes, including those in the mildest categories of chronic kidney disease. For example, among people with a UACR less than 10 mg/g, an eGFR of 45 to 59 mL/min/1.73 m2 based on creatinine alone was associated with significantly higher hospitalization rates compared with an eGFR of 90 to 104 mL/min/1.73 m2 (adjusted hazard ratio, 1.3 [95% CI, 1.2-1.3]; 161 vs 79 events per 1000 person-years; excess absolute risk, 22 events per 1000 person-years [95% CI, 19-25 events per 1000 person-years]).

Conclusions and relevance: In this retrospective analysis of 114 cohorts, lower eGFR based on creatinine alone, lower eGFR based on creatinine and cystatin C, and more severe UACR were each associated with increased rates of 10 adverse outcomes, including adverse kidney outcomes, cardiovascular diseases, and hospitalizations.

Figure 1.. Categorical Analysis of the Associations of Estimated Glomerular Filtration Rate (eGFR) and Albuminuria With Subsequent Adverse Outcomes in the Population Based on Creatinine Alone

The numbers in the boxes reflect the adjusted hazard ratio vs the reference category. The adjustment variables appear in the legend for Figure 2. The sample sizes include individuals who are missing a measure of albuminuria. The percentile shaded the darkest green color corresponds to the proportion of cells in the grid without chronic kidney disease (eg, 6 of 35 cells with eGFR ≥60 mL/min/1.73 m² and urine albumin to creatinine ratio <30 mg/g), and the percentile shaded the darkest red color corresponds to the proportion expected to be at the highest risk for adverse outcomes (eg, 11 of 35 cells with eGFR <15 mL/min/1.73 m² and urine albumin to creatinine ratio ≥1000 mg/g).

Figure 2.. Categorical Analysis of the Associations of Estimated Glomerular Filtration Rate (eGFR) and Albuminuria With Subsequent Adverse Outcomes in the Population Based on Creatinine and Cystatin C

The numbers in the boxes reflect the adjusted hazard ratio compared with the reference category. The adjustment variables included age, sex, smoking status (current, former, never), systolic blood pressure, total cholesterol, high-density lipoprotein cholesterol, body mass index, use of antihypertensive medications, and a history of diabetes, coronary heart disease, stroke, heart failure, atrial fibrillation, peripheral artery disease, cancer, and chronic obstructive pulmonary disease when relevant. The cohorts used in these analyses are the general population and electronic health record cohorts (the chronic kidney disease [CKD] cohorts did not have sufficient individuals in the reference cells). The sample sizes include individuals who are missing a measure of albuminuria. The percentile shaded the darkest green color corresponds to the proportion of cells in the grid without CKD (eg, 6 of 24 cells), and the percentile shaded the darkest red color corresponds to the proportion expected to be at the highest risk for adverse outcomes (eg, 5 of 24 cells).

Figure 3.. Hazard Ratios for Adverse Outcomes Using a Continuous Model of Estimated Glomerular Filtration Rate (eGFR)

The diamond indicates the reference point at eGFR of 90 mL/min/1.73 m². The dots indicate that the 95% CI for the hazard ratio from this spline model does not include 1.0.