Effect of Vitamin D Supplementation on Free and Total Vitamin D

A Comparison of Asians and Caucasians

Jaya Sujatha Gopal-Kothandapani; Lucy Faith Evans; Jennifer S. Walsh; Fatma Gossiel; Alan S. Rigby; Richard Eastell; Nick J. Bishop


Clin Endocrinol. 2019;90(1):222-231. 

In This Article


We found that serum total 25OHD was low in young men of White Caucasian and Asian ethnic origin during winter in Sheffield and increased similarly in both groups following administration of 150 000 units of vitamin D3, but that measured free 25OHD increased more in those of Asian origin. We found a relationship between both baseline and mean DBP concentration and incremental change in measured free 25OHD according to ethnicity. As mean DBP rose, the incremental increase in free 25OHD reduced in Caucasians and increased in Asians. This suggests that the binding affinity of DBP may vary with ethnic origin.

There is a biochemical basis for the differences in affinity of the different DBPs for vitamin D metabolites. Polymorphisms in the Gc gene (codominant alleles) give rise to three major electrophoretic variants of Gc (Gc1f, Gc1s and Gc2), which differ by amino acid substitutions as well as glycosylation.[15] The amino acid sequence of the three isoforms 1f, 1s and 2 differ at positions 152, 311, 416 and 420.[16] The position 416 difference is functionally important as this site contains a branched trisaccharide that in turn is linked to sialic acid (1f), mannose (1s) or no sugar residue (2). Such post-translational modifications are likely to affect the properties of the isoforms. Indeed, the relative affinity constants differ, so that if 1f is assigned a value of one, then the values for 1s and 2 are 0.536 and 0.321, respectively.[11,4]

Calculated estimates of free 25OHD were found to be significantly different from direct measurements and showed a systematic bias. According to the free hormone hypothesis,[5] the biological action of 25OHD is exerted by its freely available form (<1% of the total), not by the total circulating amount which comprises DBP-bound (85%-90%) and albumin-bound (10%-15%) forms plus free. There is support also for the concept of bioavailable 25OHD,[17] comprising the albumin-bound and free fractions, suggesting that measurement of free or bioavailable 25OHD concentration may provide a more meaningful marker of vitamin D function than total.[6]

The movement of DBP-bound 25OHD to bioavailable or free forms likely depends in part on the concentration of DBP and its binding affinity for 25OHD. Both are known to vary significantly both by and within ethnic groups. A recent report by Yao et al has also demonstrated significantly lower levels of DBP (P < 0.001) measured using a monoclonal assay (165.3 ± 90.4 μg/mL) in comparison with the polyclonal assay (418.7 ± 99.0 μg/mL) in a Chinese population.[7] Powe et al[17] in their cross-sectional study reported that Black Americans have lower total 25OHD and DBP resulting in similar concentrations of calculated bioavailable 25OHD compared with white counterparts . Bouillon et al[8] showed similar results in Black Gambians compared with White Caucasians using the same DBP assay. In contrast, when they measured DBP using a polyclonal assay (as used here), in the same cohort, they did not find any difference between the groups . Similarly, Aloia et al[18] reported identical concentrations of DBP between US Black people and White people, using a polyclonal assay . Recently, Nielson et al[19] compared the DBP assays used in the MrOS and MRC cohorts[20] and characterized the molecular forms of DBP.[21] The authors reported that the difference in DBP levels between the Africans and White people identified using a monoclonal assay disappeared when measured using polyclonal or proteomic methods.[21] This contrasts with our results; we found a significant difference in DBP levels between the Asians and White Caucasians using the polyclonal assay. Lower DBP levels have previously been documented in carriers of two Gc2 alleles.[22–24] We found no clear effect of genotype either at baseline or following intervention on serum DBP, irrespective of ethnicity. As our population size is small, the results in relation to DBP and genotype should be considered preliminary and thus need confirming in a larger population.

We found directly measured serum free 25OHD at baseline in Asians and Caucasians to be very similar despite lower total 25OHD in Asians. We hypothesized originally that the lower DBP concentrations found in the Asians in our cohort were likely to be the reason for their comparable levels of free 25OHD. Following intervention with vitamin D3, no change in DBP levels was observed in either Asians or White Caucasians when compared to baseline, indicating that the serum DBP concentration is not altered by single-dose supplementation. Our findings agree with those of with Sollid, where no effect of vitamin D3 on serum DBP concentration was shown in an interventional trial (20 000 IU D3 weekly for a year) in Caucasians.[24]

We found a higher increment in measured free 25OHD concentration in Asians following vitamin D3 supplementation, despite a similar increment in total 25OHD concentration and no significant change in DBP concentration. In addition, the incremental increase in Caucasians' serum free 25OHD was inversely related to both baseline and mean DBP, whereas in Asians, it was not. This implies that DBP affinity for vitamin D and its metabolites may be a key factor in the response of Asians to vitamin D treatment. The lack of a clear relationship with DBP genotype could imply that additional factors may be at play, or may be due to the small sample size.

Alzaman et al[25] compared the differences in total and free 25OHD levels between Black and White Americans following daily D3 supplementation (2000 IU or 4000 IU) or placebo for a total of sixteen weeks in nearly 200 diabetic subjects. The authors found similar and dose-proportionate increases in both total and free 25OHD in both groups. Sollid et al studied the relationships between serum total and free 25OHD (both directly and by calculation) in relation to age, sex, BMI, season and DBP genotype and their inter-relationship with the weekly administration of 20 000 units of D3/placebo for a year in nearly 500 individuals. The authors found that serum DBP was not affected by vitamin D supplementation. They demonstrated that age, sex and DBP concentration did not affect increment in vitamin D parameters, following supplementation.[24]

If serum free 25OHD increases disproportionately in some ethnic groups following vitamin D administration, is this a problem? There is another step beyond 25-hydroxylation in producing the biologically active form of vitamin D (1,25(OH)2D), and this step is highly regulated to protect against hypercalcaemia. The finding here that PTH did not decrease as expected in White Caucasians as both free and total serum 25OHD increased is puzzling. However, PTH did decrease in the Asian group; this suggests that the greater increase in free 25OHD in Asians may have broader biological significance in terms of calcium metabolism, perhaps reflecting increased calcium absorption. Asians had lower baseline total serum 25OHD and are more likely (as a population group) to receive treatment. The concern would be therefore that significantly increasing calcium absorption might have unexpected and undesirable consequences in an ethnic group already at higher risk for cardiovascular and renal disease.

In the method comparison between the calculated and directly measured serum free 25OHD using fixed affinity constant for the DBP genotype, there was a positive but not significant systematic bias at baseline. Postsupplementation, the systematic bias was negative and also nonsignificant. We cannot explain this finding. This may be due to small sample size (Figure 3A,B,C and D).