Increased Albuminuria and Reduced eGFR Associated with Risk of AKI

Stratification of the risk of acute kidney injury (AKI) by kidney measures, such as estimated glomerular filtration rate (eGFR) and albumin-creatinine ratio (ACR), may be more useful than factors such as age, sex, or race. That was the primary finding of a collaborative meta-analysis performed recently by Morgan E. Grams, MD, PhD, and colleagues. The researchers reported their findings in the American Journal of Kidney Diseases [2015;66(4):591-601].

There are few estimates of AKI incidence in the general population, but the complication occurs in 3.2% to 9.6% of hospital admissions, and 2.1% to 22.1% of prevalent intensive care unit patients worldwide. AKI is associated with morbidity that includes prolonged hospital stay, end-stage renal disease, earlier stages of chronic kidney disease (CKD), and short- and long-term mortality.

Patients with certain characteristics may be predisposed to AKI. For example, CKD, defined as decreased eGFR or elevated albuminuria, has been linked to risk of AKI. Further, older age, male sex, and African American race have been associated with increased risk of incident AKI. According to the researchers, most previous studies examining demographic risk factors have been limited in generalizability due to their single-cohort designs, and variations in demographic associations over the spectrum of kidney function is not known.

Dr. Grams et al. designed the current study to evaluate the associations of eGFR and albuminuria with AKI utilizing data from eight general-population cohorts (n=1,285,049 participants) and five CKD cohorts (n=79,519 participants). The researchers also sought to investigate the relative importance of age, sex, and race across the full range of eGFR and albuminuria.

The study predictors were age, sex, race, eGFR, urine ACR, and interactions. The primary outcome was hospitalization with or for AKI, using Cox proportional hazards models to estimate hazard ratios (HRs) of AKI and random-effects meta-analysis to pool results.

The 13 cohorts included in the analysis represented eight different countries: Canada, United States, Norway, Israel, Republic of Korea, Sweden, United Kingdom, and the Netherlands. During a mean follow-up of 4 years, there were 16,480 cases of AKI in the general-population cohorts. Among the general-population cohorts, the average eGFR was 90 mL/min/1.73 m2, and 7% had ACR ≥30 mg/g (or dipstick proteinuria ≥1+). Over a mean follow-up of 1 year, in the AKI cohorts there were 2087 cases of AKI. Average eGFR was 63 mL/min/1.73m2 and 55% had ACR ≥30 mg/g (or dipstick proteinuria ≥1+).

In all individual cohorts, compared with participants without AKI, those with AKI had lower average eGFRs and a greater likelihood of ACR ≥30 mg/g. Participants with AKI in the general-population cohorts were older, more often men, and more often African American. In the CKD cohorts, those associations were not seen (there was more between-study heterogeneity in the crude associations of age, sex, and race with AKI in the CKD cohorts).

In the general-population cohorts, pooled meta-analysis found the relationship between eGFR and AKI risk was nearly linear between eGFR of 90 and 15 mL/min/1.73 m2. At eGFR of 45 mL/min/1.73 m2, the risk of AKI was consistently elevated compared with eGFR of 80 mL/min/1.73 m2  (pooled HR, 3.35; 95% confidence interval [CI], 2.75-4.07; P<.001), with some variation in effect size across cohorts. Higher levels of ACR were also nearly linearly associated with higher risk of AKI. Most cohorts showed elevated risk at ACR of 300 m/g/ compared with ACR of 5 mg/g (pooled HR, 2.73; 95% CI, 2.18-3.43; P<.001).

The association of lower eGFR and higher ACR with increased risk for AKI remained in all age groups in the general-population. There was an association of older age itself with higher risk of AKI at eGFR of 80 mL/min/1.73 m2; at eGFR <60 mL/min/1.73 m2, adjusted HRs in each age category overlapped, with minimal variation across studies (overall interaction compared with reference age 55-64 years: P=.001 for ages 18-54; P<.001 for ages 65-74, and P<.001 for age ≥75 years). In CKD cohorts, there was little difference by age in the risk of AKI across the range of eGFRs and ACRs.

In the general-population cohorts, lower eGFR and higher ACR were associated with higher risk of AKI in both men and women. Men had higher risk of AKI at all levels of eGFR and ACR, but this increased risk seemed to be attenuated at lower eGFRs (but not for higher ACRs). In CKD cohorts, there was an association between higher AKI risk and male sex at eGFRs >40 mL/min/1.73 m2 and ACRs <300 mg/g. At all levels of eGFR and ACR, men had higher adjusted incidence rates than women.

Finally, at higher levels of eGFR and lower levels of ACR, there was an association between race and higher risk of AKI: African Americans had higher risk of AKI compared with whites at higher eGFRs in the general-population cohorts; however, CIs overlapped at lower eGFRs. Compared with whites, African Americans had higher AKI risk in the upper but not lower range of ACR.

Study limitations cited by the authors were only two general-population studies being able to contribute to the analyses analyzing the risk of AKI by race, and identifying AKI using diagnostic codes only.

“In conclusion, the present study demonstrates that measures of kidney function, namely, eGFR and ACR, were strongly and robustly associated with AKI risk factors across a wide range of settings. Although older age and male sex were also associated with AKI, these associations were attenuated in the presence of CKD. AKI risk stratification by kidney measures may be more useful than stratification by age, sex, or race,” the researchers said.


Takeaway Points

  1. This study to was designed to evaluate the associations of eGFR and albuminuria with AKI utilizing data from eight general-population cohorts and five CKD cohorts. The researchers also sought to investigate the relative importance of age, sex, and race across the full range of eGFR and albuminuria.
  2. In all individual cohorts, compared with participants without AKI, those with AKI had lower average eGFRs and a greater likelihood of ACR ≥30 mg/g. Participants with AKI in the general-population cohorts were older, more often men, and more often African American. In the CKD cohorts, those associations were not seen.
  3. In the general-population cohorts, lower eGFR and higher ACR were associated with higher risk of AKI in both men and women. At all levels of eGFR and ACR in the CKD cohorts, men had higher adjusted incidence rates than women.