Background Liquid chromatography-tandem mass spectrometry (LC-MS/MS) has become the gold standard for the measurement of serum 25-hydroxyvitamin D (25(OH)D) levels instead of the conventional method, radioimmunoassay (RIA). However, there was no study that compared RIA and LC-MS/MS for measuring serum 25(OH)D levels in infants and their mothers. The aim of this study was to assess the agreement of RIA and LC-MS/MS for measuring the serum levels in infants and postpartum women. Methods This study enrolled 70 preterm infants, 113 term infants (134 samples), and 120 postpartum women. Serum concentration of 25(OH)D was measured by RIA and LC-MS/MS. We evaluated the correlation between RIA and LC-MS/MS. Also, we evaluated the bias between RIA and LC-MS/MS using Bland-Altman analysis. Results Sixty percent of preterm infants had serum 25(OH)D levels below the lower limit of quantification (LOQ) (4 ng/mL) and 90% of them were classified as vitamin D deficient. The serum 25(OH)D levels measured by RIA were significantly correlated with those measured by LC-MS/MS in all groups. According to the Bland-Altman plot, the serum 25(OH)D levels of infants measured by RIA had constant positive bias (mean±standard deviation [SD] [95% confidence interval, CI], preterm: +4.8± 2.4 ng/mL [4.2-5.4], term: +5.8±4.0 [5.1-6.5]) and proportional bias (preterm: r=0.44, p<0.01, term: r=0.50, p<0.01) compared with LC-MS/MS. The serum 25(OH)D levels of postpartum women measured by RIA had constant positive bias compared with LC-MS/MS, but no proportional bias was found. Conclusions RIA demonstrated falsely high 25(OH)D levels when used for infants and postpartum women.
Keywords: 25-hydroxyvitamin D; infants; liquid chromatography tandem mass spectrometry; radioimmunoassay.