Urinary albumin to osmolality ratio predicts 24-hour urine albumin excretion in diabetes mellitus.

BACKGROUND Urinary albumin to creatinine ratio (ACR) in a single urine sample has been proposed to provide an estimate of microalbuminuria by adjusting for variability in urine concentrations. We hypothesized that adjusting the urine albumin concentration of single-void specimens for actual urine osmolality (urinary albumin to osmolality ratio [AOR]) may provide a more accurate estimate of 24-hour urine albumin excretion rates (AERs). METHODS Patients with diabetes mellitus (DM; n = 136) had urinary concentrations of albumin, glucose, and creatinine and osmolality measured on single-void samples, and albumin levels, on 24-hour samples. Microalbuminuria is defined as an AER between 30 and 300 mg/d. RESULTS Correlation between AOR on single-void samples and AER on 24-hour samples (r = 0.87; P < 0.001) was similar to that between ACR and AER (r = 0.88; P < 0.001). Using a cutoff value of 18.4 mg/kg/mOsm x 10(2) (18.4 mg/mmol x 10(2)) for AOR resulted in a sensitivity and specificity of 82% and 86% in detecting microalbuminuria, respectively. The area under the curve (AUC) for AOR was 0.89. Using a cutoff value of 15.0 mg/g (1.7 mg/mmol) for ACR resulted in a sensitivity and specificity of 85% and 85% in detecting microalbuminuria, respectively. The AUC for ACR was 0.90. The ability of AOR to predict AER was maintained at varying degrees of glycosuria (glucose < 100 mg/dL [<5.5 mmol/L]; r = 0.77; 100 to 750 mg/dL [5.5 to 42 mmol/L]; r = 0.85; and >750 mg/dL [>42 mmol/L]; r = 0.92). CONCLUSION Urinary AOR correlates closely with 24-hour microalbuminuria determination, and the correlation is not appreciably affected by glycosuria. Thus, AOR can be used as an alternative test to ACR in the assessment of microalbuminuria in the population with DM.