The excretion of small quantities of urinary albumin (microalbuminuria) may predict renal failure in diabetes. The measurement of microalbuminuria with radioimmunoassays has been based on 24-h, overnight, and 3- to 4-h collections. To determine whether single-void urine samples can be used to estimate 24-h excretion, we compared the results of 24-h outpatient urine collections with single-void samples corrected for creatinine from diabetic and nondiabetic subjects.
The overall correlation of single-void sample results expressed as microgram albumin per milligram creatinine with 24-h excretion (mg/24 h) was excellent (r = .82, P < .001). More important, in the diabetic patients the sensitivity and specificity of detecting 24-h microalbuminuria in the abnormal range were at least 94 and 96%, respectively. Single-void urine specimens adjusted for creatinine discriminate between normal and abnormal levels of microalbuminuria, as determined in 24-h urine collection, with high specificity and sensitivity.
