RESUMO
OBJECTIVE: In elderly patients, age-related, disease-related, and drug-related factors are associated with chronic kidney disease (CKD). Little is known about which factors are the best predictors for CKD in elderly patients. MATERIALS AND METHODS: The study was based on 784 patients aged 75 years or older for whom the clinical and serum creatinine on admission to our hospital were available. Impaired renal function, including CKD and transient renal insufficiency, was defined as a non-indexed glomerular filtration rate (GFR) below 60 mL/min. A logistic regression model was developed for predictors of CKD and was internally validated using bootstrapping. RESULTS: Approximately 61% of the patients, who had CKD (46%) and transient renal insufficiency (15%), had a non-indexed GFR < 60 mL/min. Synergistic use of 3 drugs potentially impairing renal function, diuretics, ACE-I/ARB, and NSAIDs (odds ratio (OR), 4.66; 95% confidence interval (CI), 1.48 - 17.7, p = 0.012) was a significantly associated factor for CKD in a multivariate logistic regression analysis. Age (OR 1.56, 95% CI 1.04 - 2.33, p = 0.03), female gender (OR 1.58, 95% CI 1.04 - 2.39, p = 0.03), any prescription ACE-I/ARB either alone or in combinations with diuretics or NSAIDs (OR 2.74, 95% CI 1.83 - 4.13, p = 0.0001), and proteinuria (OR 1.98, 95% CI 1.27 - 3.10, p = 0.003), were included as the best model for CKD. The area under the curve (AUC) of the best model and the bootstrapping validation were 0.68 and 0.71, respectively. CONCLUSION: Given the widespread use of ACE-I/ARB for elderly patients, our findings suggest that caution is needed when they are prescribed because of the possibility of the patient developing CKD.