Background: Albuminuria is a predictor of disease progression in patients with chronic kidney disease (CKD). However, the ability of proteinuria parameters measured at various time periods to predict renal outcomes is unclear.
Method: This observational cohort study included 165 non-diabetic hypertensive CKD patients who took olmesartan medoxomil. We measured the albuminuria at five different time points (0, 2, 4, 26, and 38 months) and the mean levels. The mean albuminuria levels were calculated during 0-4 months, 0-26 months, and 0-38 months. The renal outcome was defined as a decline in eGFR ≥ 40% during the entire study period.
Result: The albuminuria at five different time points and the mean albuminuria levels were independent risk factors for a worse renal outcome after adjusting for age, sex, and estimated glomerular filtration rate (eGFR) at enrollment and were able to predict the renal outcome, although the performance of the estimation tended to be more effective using the mean albuminuria level at the 38-month follow-up time point. The risk of a decline in eGFR ≥ 40% was increased by 1.690-folds [95% CI 1.110-2.572, P = 0.014] per 500 mg/day increase in the mean albuminuria at 38 months. With a cut-off value of 897 mg/day for mean albuminuria at 38 months after treatment, a decline in eGFR ≥ 40% was predicted with a sensitivity of 88.9% and specificity of 81.3%. The ability of albuminuria to predict a renal event at different measurement points does not differ in CKD patients.
Conclusion: The time-averaged albuminuria cut-off of 900 mg/day during the 3-year follow-up period showed high sensitivity and specificity for predicting a decline in eGFR ≥ 40% in CKD patients, although the albuminuria at different measurement points did not predict a worse renal outcome.