Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Review
. 2015 Oct;66(4):591-601.
doi: 10.1053/j.ajkd.2015.02.337. Epub 2015 May 2.

A Meta-analysis of the Association of Estimated GFR, Albuminuria, Age, Race, and Sex With Acute Kidney Injury

Collaborators, Affiliations
Free PMC article
Review

A Meta-analysis of the Association of Estimated GFR, Albuminuria, Age, Race, and Sex With Acute Kidney Injury

Morgan E Grams et al. Am J Kidney Dis. .
Free PMC article

Abstract

Background: Acute kidney injury (AKI) is a serious global public health problem. We aimed to quantify the risk of AKI associated with estimated glomerular filtration rate (eGFR), albuminuria (albumin-creatinine ratio [ACR]), age, sex, and race (African American and white).

Study design: Collaborative meta-analysis.

Setting & population: 8 general-population cohorts (1,285,049 participants) and 5 chronic kidney disease (CKD) cohorts (79,519 participants).

Selection criteria for studies: Available eGFR, ACR, and 50 or more AKI events.

Predictors: Age, sex, race, eGFR, urine ACR, and interactions.

Outcome: Hospitalized with or for AKI, using Cox proportional hazards models to estimate HRs of AKI and random-effects meta-analysis to pool results.

Results: 16,480 (1.3%) general-population cohort participants had AKI over a mean follow-up of 4 years; 2,087 (2.6%) CKD participants had AKI over a mean follow-up of 1 year. Lower eGFR and higher ACR were strongly associated with AKI. Compared with eGFR of 80mL/min/1.73m(2), the adjusted HR of AKI at eGFR of 45mL/min/1.73m(2) was 3.35 (95% CI, 2.75-4.07). Compared with ACR of 5mg/g, the risk of AKI at ACR of 300mg/g was 2.73 (95% CI, 2.18-3.43). Older age was associated with higher risk of AKI, but this effect was attenuated with lower eGFR or higher ACR. Male sex was associated with higher risk of AKI, with a slight attenuation in lower eGFR but not in higher ACR. African Americans had higher AKI risk at higher levels of eGFR and most levels of ACR.

Limitations: Only 2 general-population cohorts could contribute to analyses by race; AKI identified by diagnostic code.

Conclusions: Reduced eGFR and increased ACR are consistent strong risk factors for AKI, whereas associations of AKI with age, sex, and race may be weaker in more advanced stages of CKD.

Keywords: Chronic Kidney Disease Prognosis Consortium; Estimated glomerular filtration rate (eGFR); acute kidney injury (AKI); acute renal failure (ARF); age; albumin-creatinine ratio (ACR); albuminuria; meta-analysis; proteinuria; race/ethnicity; renal function; sex.

Figures

FIGURE 1
FIGURE 1
Adjusted hazard ratios of acute kidney injury in the general population cohorts by level of estimated glomerular filtration rate (eGFR) and albuminuria in continuous (panels A, B, and C) and categorical (panel D) analysis. In the panel C, bold lines indicate statistical significance compared to the reference (black diamond) at eGFR 80 ml/min/1.73 m2 in the no albuminuria group, defined as urine albumin-creatinine (ACR) <30 mg/g or urine protein dipstick <1+. Stars along the x-axis represent significant pointwise interactions: the relative risk associated with a particular category of albuminuria compared to the no albuminuria category at that value of eGFR is significantly different than the corresponding relative risk seen at eGFR 80 ml/min/1.73 m2. A graph without stars would reflect parallel risk associations. Hazard ratios (HRs) are derived from meta-analyses of the general population cohorts and are adjusted for sex, race, body mass index, systolic blood pressure, total cholesterol, history of cardiovascular disease, diabetes, and smoking status. The tables represents adjusted HRs derived from categorical analysis of the general population cohorts, with bold font representing statistical significance, and color coding by risk quartile.
FIGURE 2
FIGURE 2
Adjusted hazard ratios (panel A and C) and incidence rates (panel B and D) of acute kidney injury in the general population cohorts by level of estimated glomerular filtration rate and urine albumin-creatinine ratio (ACR), within categories of age. In the panels on the left, bold lines indicate statistical significance compared to the reference (black diamond) at estimated glomerular filtration rate (eGFR) 80 ml/min/1.73 m2 or urine ACR 5 mg/g in ages 55–64 years. Stars along the x-axis represent significant pointwise interactions: the relative risk associated with a particular age category compared to the age category 55–64 years at that value of eGFR or ACR is significantly different than the corresponding relative risk seen at eGFR 80 ml/min/1.73 m2 or ACR 5 mg/g. A graph without stars would reflect parallel risk associations. Hazard ratios (HRs) are derived from meta-analyses of the general population cohorts and are adjusted for sex, race, body mass index, systolic blood pressure, total cholesterol, history of cardiovascular disease, diabetes, smoking status, and albuminuria. Tables represent adjusted HRs at eGFR 45 and 80 ml/min/1.73 m2 and ACR 300 and 5 mg/g. In the panels on the right, lines and tables depict incidence rates (IRs) per 1,000 person-years, adjusted for the same covariates.
FIGURE 3
FIGURE 3
Adjusted hazard ratios (panel A and C) and incidence rates (panel B and D) of acute kidney injury in the general population cohorts by level of estimated glomerular filtration rate (eGFR) and urine albumin-creatinine ratio (ACR), within categories of sex. In the panels on the left, bold lines indicate statistical significance compared to the reference (black diamond) at eGFR 80 ml/min/1.73 m2 or urine ACR 5 mg/g in men. Stars along the x-axis represent significant pointwise interactions: the relative risk in women compared to men at that value of eGFR or ACR is significantly different than the corresponding relative risk seen at eGFR 80 ml/min/1.73 m2 or ACR 5 mg/g. Hazard ratios (HRs) are derived from meta-analyses of the general population cohorts and are adjusted for age, race, body mass index, systolic blood pressure, total cholesterol, history of cardiovascular disease, diabetes, smoking status, and albuminuria. Tables represent adjusted HRs at eGFR 45 and 80 ml/min/1.73 m2 and ACR 300 and 5 mg/g. In the panels on the right, lines and tables depict incidence rates (IRs) per 1,000 person-years, adjusted for the same covariates.
FIGURE 4
FIGURE 4
Adjusted hazard ratios (panel A and C) and incidence rates (panel B and D) of acute kidney injury in the general population cohorts by level of estimated glomerular filtration rate (eGFR) and urine albumin-creatinine ratio (ACR), within categories of race. In the panels on the left, bold lines indicate statistical significance compared to the reference (black diamond) at eGFR 80 ml/min/1.73 m2 or urine ACR 5 mg/g in whites. Stars along the x-axis represent significant pointwise interactions: the relative risk in African Americans compared to whites at that value of eGFR or ACR is significantly different than the corresponding relative risk seen at eGFR 80 ml/min/1.73 m2 or ACR 5 mg/g. Hazard ratios (HRs) are derived from meta-analyses of the general population cohorts and are adjusted for age, sex, body mass index, systolic blood pressure, total cholesterol, history of cardiovascular disease, diabetes, smoking status, and albuminuria. Tables represent adjusted HRs at eGFR 45 and 80 ml/min/1.73 m2 and ACR 300 and 5 mg/g. In the panels on the right, lines and tables depict incidence rates (IRs) per 1,000 person-years, adjusted for the same covariates.

Comment in

Similar articles

See all similar articles

Cited by 36 articles

See all "Cited by" articles

MeSH terms

Feedback