Vitamin D Binding Protein Variants Don’t Change Supplement Response in Older Adults

A large randomized trial has found that common genetic variants of the vitamin D binding protein, long suspected of shaping how individuals respond to vitamin D supplements, do not in fact alter the response to supplementation. In the Multi-Ethnic Study of Atherosclerosis Individualized Response to Vitamin D Treatment, known as the INVITe trial, older adults carrying the GC2 isoform of the vitamin D binding protein had lower baseline total serum 25-hydroxyvitamin D3 concentrations than non-carriers, yet after sixteen weeks of daily supplementation their gains in both total and free 25-hydroxyvitamin D3 were statistically indistinguishable from those of other participants. The finding, published in Frontiers in Endocrinology, challenges the idea that genotype-guided vitamin D dosing based on these variants could meaningfully personalize supplementation.

The trial enrolled 666 participants and randomized them three to one to receive either 2,000 international units of vitamin D3 or a placebo daily for sixteen weeks. Complete data from 559 participants, 414 in the supplement group and 145 on placebo, were analyzed. The participants were older adults drawn from the ongoing Multi-Ethnic Study of Atherosclerosis, a nationwide cohort originally recruited from six United States communities between 2000 and 2002. Their mean age was 72 years, 53 percent were women, and the cohort was deliberately diverse: 36 percent White, 33 percent Black, 18 percent Hispanic/Latino, and 13 percent Chinese. Adherence was high, with a median pill-count adherence of 98 percent among those with recorded counts.

What set this study apart was its measurement strategy. Total serum 25-hydroxyvitamin D3, the standard clinical marker of vitamin D status, was quantified by liquid chromatography-tandem mass spectrometry, a gold-standard method with a mean coefficient of variation of 4.3 percent. More unusually, the researchers also measured free, that is unbound, 25-hydroxyvitamin D3 directly, using a novel mass-spectrometry approach that employs size-exclusion separation to isolate the free fraction from protein-bound metabolites before quantification. Because less than one percent of circulating vitamin D metabolites are unbound, direct measurement of this tiny fraction has long been technically difficult, and most studies have relied on mathematical prediction equations instead.

The biological rationale for examining free 25-hydroxyvitamin D stems from the so-called Free Hormone Hypothesis. Roughly 85 to 90 percent of vitamin D metabolites circulate bound to vitamin D binding protein, with most of the remainder carried by albumin; only 0.02 to 0.04 percent of 25-hydroxyvitamin D is unbound. According to the hypothesis, only unbound molecules can diffuse through the plasma membrane into most target cells. Support comes from animal and rare human evidence: vitamin D binding protein knockout mice had severely low total 25-hydroxyvitamin D but showed no signs of deficiency on a vitamin D-replete diet, and a 2019 report described a woman with complete genetic deficiency of the binding protein who maintained normal calcium levels despite nearly undetectable total vitamin D metabolite concentrations.

The genetic variants under study, called GC isoforms, arise primarily from two polymorphisms in the GC gene that encodes the binding protein, rs7041 and rs4588. These produce three haplotypes, GC1f, GC1s, and GC2, and six possible diplotypes. In cross-sectional genome-wide association studies, these variants show some of the largest single-variant effects on total serum 25-hydroxyvitamin D and binding protein concentrations, with GC2 consistently associated with lower levels. Prior supplementation studies had suggested, inconsistently, that carriers of the GC2 haplotype might mount a smaller increase in total 25-hydroxyvitamin D in response to supplements, though not all studies replicated this. The researchers hypothesized that total 25-hydroxyvitamin D3 responses would differ by isoform while free responses would not.

The baseline results confirmed the well-established cross-sectional pattern. Among all participants, GC2 carriage was associated with lower total 25-hydroxyvitamin D3: the model-based mean concentration was 30.0 nanograms per milliliter in non-carriers, 26.6 in those with one GC2 allele, and 24.4 in those with two, a statistically significant trend. In practical terms, each GC2 allele corresponded to roughly 3 to 6 nanograms per milliliter less total 25-hydroxyvitamin D3. This lower total concentration persisted after sixteen weeks of supplementation. By contrast, free 25-hydroxyvitamin D3, whether directly measured or calculated, did not vary by isoform at either time point, implying that the free fraction of the metabolite was actually higher among GC2 carriers.

The supplementation outcome, however, contradicted the researchers’ initial expectation. In the vitamin D3 group, total 25-hydroxyvitamin D3 rose by a mean of about 11 nanograms per milliliter, directly measured free 25-hydroxyvitamin D3 increased by roughly 14 picograms per milliliter, and calculated free concentrations rose by about 3.6 picograms per milliliter, while placebo values remained stable or declined slightly. Formal tests of interaction between isoform carriage and treatment assignment found no evidence that GC2 or GC1s modified any of these responses. The authors had previously observed faster systemic clearance of deuterated 25-hydroxyvitamin D3 among GC2 carriers in an earlier study, and had expected this to blunt the supplementation response, but the present analysis did not support that prediction.

The study also exposed a methodological problem with calculated free hormone values. The widely used Bikle prediction equation, which estimates free 25-hydroxyvitamin D3 from total 25-hydroxyvitamin D, binding protein, and albumin concentrations, correlated only moderately with direct mass-spectrometry measurement, with a Pearson correlation coefficient of 0.62, and showed poor agreement: directly measured values averaged 34 picograms per milliliter higher than calculated ones. The authors attribute this to the equation’s assumption of uniform binding affinity across the population, an assumption violated by variant binding proteins, medications, and clinical conditions. They argue the discrepancy underscores the need for direct measurement methods and assay harmonization for free 25-hydroxyvitamin D.

The authors are careful to frame the work as exploratory. The analysis leveraged an existing randomized trial that was not specifically designed to test GC isoform effects, and the relatively small number of GC2 homozygotes, only 23 participants, limited statistical power for some comparisons. The findings apply to a multi-ethnic population of older, generally community-dwelling United States adults and may not generalize to younger populations or different health statuses. The authors also note that measured free concentrations in this study were higher than those reported in some earlier immunoassay-based and alternative mass-spectrometry studies, for reasons that remain uncertain but may reflect methodological differences.

Nevertheless, the conclusion carries practical weight. Because prior studies have not consistently shown isoform-related differences in supplementation response, and any reported differences have generally been small, these null results support the view that GC isoform effects on supplement responsiveness, if present, are modest. The authors suggest that GC isoforms may matter more for determining baseline vitamin D status than for mediating individual responsiveness to oral supplementation. They point instead to a distinct physiological pathway as a more promising target: the binding protein is required to transport vitamin D3 from the skin to the bloodstream after sun exposure, and recent UK Biobank analyses suggest the association between rs4588 and circulating 25-hydroxyvitamin D is magnified with increased time spent outdoors. Future work, they propose, should examine whether these variants shape the response to ultraviolet B exposure, and whether polygenic risk scores integrating many genetic loci might better serve efforts to personalize vitamin D dosing.

Subject of Research: The influence of vitamin D binding protein GC isoforms on total and free serum 25-hydroxyvitamin D3 responses to vitamin D3 supplementation in older adults

Article Title: Influence of vitamin D binding protein isoforms on the total and free serum 25-hydroxyvitamin D responses to vitamin D supplementation in older adults in the Multi-Ethnic Study of Atherosclerosis

Article References: Hu, X., Hoofnagle, A. N., Hsu, S., Prince, D. K., Kushnir, M. M., Wood, A. C., Sisley, S., Michos, E. D., Psaty, B. M., Shea, S. J., Rotter, J. I., Siscovick, D., Allen, N. B., Tracy, R. P., Watson, K. E., Li, X., Taylor, K. D., Kestenbaum, B. R., de Boer, I. H., & Best, C. M. (2026). Influence of vitamin D binding protein isoforms on the total and free serum 25-hydroxyvitamin D responses to vitamin D supplementation in older adults in the Multi-Ethnic Study of Atherosclerosis. Frontiers in Endocrinology, 17, Article 1910020. https://doi.org/10.3389/fendo.2026.1910020

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1910020

Keywords: vitamin D3, 25-hydroxyvitamin D, vitamin D binding protein, GC isoforms, randomized controlled trial, free 25-hydroxyvitamin D, MESA, older adults, supplementation, mass spectrometry, genetic variants, free hormone hypothesis