Background: Microalbuminuria is an independent risk factor for cardiovascular morbidity and mortality in the general population. Standard immunochemical urinary albumin assays detect immunoreactive albumin, whereas high-performance liquid chromatography (HPLC) detects both immunoreactive and immunounreactive albumin.
Methods: Using data from the Australian Diabetes, Obesity, and Lifestyle cohort study of randomly selected community-based Australian adults, spot urine samples were tested for albuminuria (spot urine albumin-creatinine ratio [ACR]: normal, < 30 mg/g; microalbuminuria, 30 to 300 mg/g; and macroalbuminuria, > 300 mg/g) by using both immunonephelometry (IN) and HPLC (n = 10,010).
Results: Bland-Altman analysis showed significant bias, with a greater ACR by means of HPLC, particularly at lower levels of ACR. Mean ACR was 15.8 mg/g (95% confidence interval [CI], 12.3 to 19.2) by means of IN compared with 30.0 mg/g (95% CI, 27.0 to 35.0) by means of HPLC. The prevalence of microalbuminuria was 4 times greater by means of HPLC compared with IN (20% versus 5.5%). In all demographic and comorbid subgroups associated with microalbuminuria, the prevalence of microalbuminuria increased by 2 to 4 times. A total of 1,743 subjects (17.4%) classified as normoalbuminuric by means of IN were reclassified as microalbuminuric by means of HPLC. Using multivariate logistic regression, women, patients with untreated and treated hypertension, and those with impaired glucose tolerance or diabetes were associated significantly with a change in category from normoalbuminuric to microalbuminuria by means of HPLC.
Conclusion: HPLC measures significantly more urinary albumin within the normoalbuminuria and microalbuminuria range, resulting in a significant increase in prevalence of microalbuminuria. Longitudinal studies are needed to determine whether the extra individuals identified by means of HPLC are at increased risk for developing hard clinical outcomes (renal and cardiovascular).