Abstract:

Objective To investigate the application of urinary albumin determination by high performance liquid chromatography (HPLC) in patients with diabetes mellitus (DM). Methods Urine samples of 85 DM patients and 40 healthy subjects (healthy control group) were collected, and the urinary albumin concentrations were determined by HPLC and immunoturbidimetry respectively. The consistency of the results between the 2 methods was analyzed by the regression analysis and Bland-Altman analysis. Results HPLC linear regression equation was Y=2.948 2X+38.601,r²=0.994 4.The retention time (RT) of standard materials and samples were 2.42 min. Within-run coefficient of variation (CV) was 4.1%, and between-run CV was 5.6% at concentration of 40.1mg/L. Within-run CV was 4.8%, and between-run CV was 6.5% at concentration of 152.6 mg/L. The lowest detection limit was 2 mg/L. The urinary albumin concentration was significantly higher by HPLC [21.0(3.4-678.9)mg/L] than by immunoturbidimetry [8.2(2.0-442.2)mg/L] by 1.3-6.1 folds (P<0.01). There was no statistical significance between the concentrations of control group by HPLC and immunoturbidimetry(P>0.05). The regression analysis and Bland-Altman analysis demonstrated that the 2 methods were not consistent. Of all the 85 DM patients, according to microalbuminuria diagnosis standard [urinary albumin excretion rate (AER) >30 mg/24h], there was 48(56.5%) positive cases by HPLC and 27(31.8%) positive cases by immunoturbidimetry. Conclusions Immunity and no-immunity total urinary albumin can be determined by HPLC. Microalbuminuria in DM patients can be determined earlier by HPLC.

Key words: Albumin, Urine, High performance liquid chromatography, Immunochemically nonreactive albumin