The routine chemiluminescence assay for plasma 25-hydroxyvitamin D analysis does not overestimate the prevalence of vitamin D deficiency in adolescents

Abstract

Vitamin D deficiency (VDD) is a global public health problem. Inaccurate methods for measuring plasma 25-hydroxyvitamin D (25[OH]D) may have contributed to the reported high prevalence of VDD. We hypothesized that the most commonly used assay for vitamin D status, chemiluminescence immunoassay (CLIA), underestimates 25(OH)D levels and thus overestimates VDD. Using both liquid chromatography–tandem mass spectrometry and CLIA for plasma 25(OH)D, we evaluated the prevalence of VDD in adolescents (11-16 years-old; n = 410) by both methods in a cross-sectional study. Subjects were selected from public middle schools from all the 6 Governorates of Kuwait using stratified multistage cluster random sampling. Cohen κ agreement, linear regression, and Bland-Altman plots were used to evaluate the classification of VDD by the 2 methods. VDD (25[OH]D < 50 nmol/L) was 85.9% with CLIA and 81.2% with liquid chromatography–tandem mass spectrometry. There was a good agreement between the 2 methods in classifying the study subjects as deficient, insufficient, or sufficient (κ = 85.1%, P < .001). The between-assay bias was very small with a mean percentage difference < 1% from the mean value of the 25(OH)D as assessed by the 2 methods. These data did not support our hypothesis, and we conclude that the routine methods used for plasma 25(OH)D levels have no or little impact on evaluating VDD as a public health problem or in clinical management.