Endocrine Abstracts

Avoiding vitamin D deficiency can be beneficial in prevention of adverse maternal, fetal and neonatal outcomes. Because of increasing level of vitamin D-binding protein (DBP) in pregnancy, the determination of free-25OHD should be required, whose adequate measure is a challenge for the analytics. Literature data are contradictory whether total-or free-25OHD fraction is more reliable taking into consideration the methods of measurement. Data of publications on free or total, bioavailable 25OHD are quite discrepant. Our study started in connection with gestational diabetes (GDM) we aimed to evaluate the correlation of 25OHD-fractions measured by different assays.

Methods: 88 sera of pregnant women (31.4 ± 5.0 years; gestational age: 26.6 ± 1.3 weeks; out of them 25 suffered from GDM; 62 cases were on 665 ± 742 IU_D3/day) were investigated. Measured biochemical markers were: calcium, t-25OHD by three methods; liquid chromatography/mass spectrometry (t-25OHD_MS) chemiluminescence immunoassay (t-25OHD_CLIA, Liaison, DiaSorin) and protein binding assay (t-25OHD_PBA, Cobas Roche), as well as albumin and DBP (immunoturbidimetry Roche and Dako, Integra, Modular) to calculate free levels (c-f-25OHD_MS; c-f-25OHD_CLIA ; c-f-25OHD_PBA). Direct measured free-25OHD (dm_f-25OHD) was determined by ELISA (DiaSource, Future Diagnostics) method. To evaluate the agreement of different methods Passing & Bablok regression and Bland & Altman analysis were used.

Results: The levels of t-25OHD_MS were significantly higher (P < 0.001) than the routinely used methods (MS: 76 ± 31 vs CLIA: 59 ± 24; PBA: 60 ± 25 nmol/l; both biases: -24%). Tests of linearity were excellent (MS-CLIA: P = 0.939 and MS-PBA P = 0.634). Mean of dm-f-25OHD level was lower compared to c-f-25OHD_MS (7.5 ± 2.1 vs 8.3 ± 3.2 pmol/L; P < 0.05; bias:-5.3% and linearity: P = 0.808).But the c-f-25OHD_CLIA and c-f-25OHD_PBA levels are lower (both: 6.5 ± 2.1 pmol/l; P < 0.01;) than dm_f-25OHD. The biases of both methods are: +18%; the linearities with dm_f-25OHD are by CLIA: P = 0.934 and by PBA method: P = 0.634. The markers of 25OHD didn’t show any significant differences between GDM and GDM-negative cases.

Conclusion: Our results confirm the acceptable dissimilarity of the routinely, frequently used methods (CLIA and PBA). They diverge from LC-MS method but show an excellent linearity with that. It should be recommended to define cut-off points depending on assay method for monitoring vitamin D supply in pregnant women. Unfortunately, we have few data to suggest to the sole measure of free-25OHD. Although our results indicate that calculated free levels are not consistent because the one underestimated, the other two overestimated the direct measured levels.