Background: Urinary albumin is the main parameter employed to diagnose diabetic nephropathy (DN). The exclusion of bacteriuria has been recommended at the time of DN diagnosis. This approach has been debated and information on this suggestion in patients with diabetes is scarce. The present case-control study was conducted to investigate the interference of bacteriuria in the interpretation of urinary albumin measurements in random urine samples of diabetic patients.
Methods: Urinary albumin concentration (UAC) was measured in random urine samples twice in diabetic patients with and without bacteriuria (> or =10(5) colony-forming units/mL). Cases (n = 81) were defined as patients who had baseline UAC measurement in the presence of bacteriuria and had the second UAC measured in a sterile urine sample. Controls (n = 80) had the two UAC measured in sterile urine specimens.
Results: Baseline UAC was not different between case [15.4 (1.5-2148) mg/L] and control groups [14.2 (1.5-1292) mg/L; P = 0.24], nor was the proportion of patients with normo-, micro- and macroalbuminuria. In cases, UAC measurements in the presence of bacteriuria and in sterile urine specimens were not different [15.4 (1.5-2148) versus 13.7 (1.5-2968) mg/L; P = 0.14)], nor was the proportion of normo- (51.9% versus 61.5%), micro- (40.7% versus 32.1%) and macroalbuminuria (7.4% versus 6.4%; P = 0.46). In the control group, UAC values were also not different in the two urine samples: [14.2 (1.5-1292) versus 9.7 (1.5-1049) mg/L, P = 0.22].
Conclusions: The presence of bacteriuria does not interfere significantly with urinary albumin measurements and its exclusion is not necessary to diagnose DN.