Background: Urinary albumin is now measured by high-performance liquid chromatography (HPLC) which also detects albumin missed by traditional immunochemical methods. A predictive effect of HPLC-detected albuminuria on mortality has just been reported in the AusDiab study, measuring albuminuria with HPLC after 7 years of -80 degrees C storage. However, there are already some data suggesting that HPLC-detected albuminuria is affected by -80 degrees C storage. We aimed to measure changes in HPLC-detected albuminuria after 2.5 years and find the factors which may be responsible for this alteration.
Methods: Urinary albumin was measured by the US Food and Drug Administration approved HPLC Accumin kit. Total free sulfhydryl groups (TFSG) of urine samples were measured by Ellman's reagent.
Results: We found a significant 24% average decrease in HPLC-detected albuminuria and a correlation between the magnitude of decrease and urinary pH. We found a correlation between changes of urinary albumin dimeric to monomeric ratio of stored urine and pH; however, only changes of monomeric form were found to be significant. A correlation was also found between the TFSG of fresh urine samples and pH. Less TFSG could be detected, and a correlation between TFSG and pH was absent in stored urine.
Conclusions: We conclude that measurement of albuminuria by HPLC in long-term -80 degrees stored urine gives unreliable results. Decrease of HPLC-detected albuminuria is pH-dependent and may be due to the reducing capacity of urine. Prospective studies need to decide whether the predictive properties of HPLC-detected albuminuria decrease during longterm storage.