Capillary blood samples from 103 infants were analyzed for plasma 25(OH)D using an enzyme immunoassay (EIA, Octeia, IDS Ltd.) and radioimmunoassay (RIA, DiaSorin). Plasma 25(OH)D3, C-3 epimer of 25(OH)D3 (3-epi-25(OH)D3) and 24,25-dihydroxyvitamin D (24,25(OH)2D3) were measured on the same samples using LC-MS/MS. To establish whether plasma 24,25(OH)2D3 or 3-epi-25(OH)D3 interferes with these immunoassay results, the zero 25(OH)D calibrator from each assay kit was spiked with increasing amounts of 24,25(OH)2D3 or 3-epi-25(OH)D3.
Classifying infants below the common vitamin D status targets of 50 nmol/L and 75 nmol/L respectively, 58% and 99% fell below using the RIA, 19% and 56% with the EIA and 31% and 76% with LC-MS/MS. Compared to LC-MS/MS, both immunoassays showed poor Bland-Altman limits of agreement for 25(OH)D concentrations (RIA: limits of agreement 鈭?#xA0;27 to + 13%; EIA: 鈭?#xA0;12 to + 41%), and mountain plots (folded cumulative distribution) depicted significant skew and bias. Spiked 24,25(OH)2D3 concentrations, but not 3-epi-25(OH)D3, appeared as > 100% of known values on the EIA but not on the RIA thus, suggesting that the EIA may cross-react with 24,25(OH)2D3 to a greater extent than 3-epi-25(OH)D3.
Two common immunoassays resulted in very different classifications of vitamin D status possibly related to the interference of other vitamin D metabolites. Based on these data, LC-MS/MS assessment of vitamin D status is recommended in young infants (4-6 weeks of age).