Evaluation of automated methods for quantifying serum 25‐hydroxyvitamin D

Overview: The accepted biomarker of vitamin D nutritional status is serum 25‐hydroxyvitamin D [25(OH)D]. However, a clear understanding of vitamin D research is confounded by the variability among assays used for measuring 25(OH)D. Objectives: To compare two new automated assays with the well‐established DiaSorin radioimmunoassay (RIA) for 25(OH)D. Methodology: The 25(OH)D from human sera (n=160) was quantified using the RIA and two automated non‐radioactive immunoassays: DiaSorin Liaison â Total 25(OH)Dâ and Roche â Elecsys 25‐OHâ. Correlation was assessed with the Pearson test and agreement with the Bland‐Altman method. Results: The RIA and Liaison correlated well ( r = 0.919) and with negligible bias (bias +/â 95% limits of agreement = â 0.72 +/â 16.01nmol/L). The Roche correlated similarly with both the RIA and Liaison but with higher bias versus the RIA ( r = 0.871; bias = â 2.55 +/â 21.06 nmol/L) than the Liaison ( r = 0.861; bias = â1.72 +/â 21.34 nmol/L). Imprecision (CV%) within‐run for the RIA, Liaison and Roche was 9.66%, 6.29%, and 4.52%, and between‐run was 10.8%, 6.67%, and 9.13%, respectively. Conclusion: For quantification of serum 25(OH)D, the Liaison â Total 25(OH)Dâ demonstrated a stronger correlation and better agreement with the reference method (RIA) than the Roche â Elecsys 25‐OHâ.