Comparison of immunoassays for the selective measurement of human high-molecular weight adiponectin.
Document type:
Comparative Study; Journal Article
Author(s):
Liu, D; Schuster, T; Baumann, M; Roos, M; Sollinger, D; Lutz, J; Heemann, U; von Eynatten, M
Abstract:
BACKGROUND: Adiponectin is an adipocyte-derived hormone circulating in different multimer complexes. The high-molecular-weight (HMW) complex is likely the active form of this protein and has been recognized as a risk marker for type 2 diabetes and coronary artery disease (CAD). Because quantification of HMW adiponectin by Western blot analysis is time-consuming, novel ELISAs have been developed to simplify measurements in clinical research. However, these enzyme immunoassays have not been cross-validated in larger patient groups. We evaluated 2 individual ELISA systems by comparison to Western blotting for measurement of the distribution of HMW adiponectin in healthy individuals and patients with CAD and type 2 diabetes. METHODS: We measured HMW adiponectin in 204 individuals (83 CAD patients, 81 type 2 diabetes patients, and 40 healthy controls). Correlations, range of agreement, and imprecision of HMW concentrations obtained using 2 commercial ELISAs (#1, ALPCO Diagnostics; #2, Millipore) were evaluated by comparison with quantitative Western blotting. RESULT: Adiponectin results of the ELISAs were significantly correlated with those obtained by Western blotting (both r > 0.75, P < 0.001). Deming regression and Bland-Altman analyses indicated high agreement among the 3 immunoassays. The median difference between HMW adiponectin concentrations measured by ELISA and by Western blot was +0.4 mg/L for ELISA #1 and -0.4 mg/L for ELISA #2 with 95% of value differences <3 mg/L. CONCLUSIONS: Selective measurement of HMW adiponectin by ELISA is feasible; however, individual differences among immunoassays must be considered. The evaluated ELISAs exhibit analytical characteristics that allow their use as equivalent for Western blot analysis in larger clinical and epidemiological groups.