Vitamin D3 for Life

Live Enlightened

Article

Impact of DBP variability on free vs total vitamin D: Full Study

Previous related pageHow Vitamin D3 WorksNext related pageImpact of DBP variability on free vs total vitamin D: Study Summary
Study at a glance

How does vitamin D-binding protein affect free and total blood measurements?

This review examines how vitamin D-binding protein affects blood measurements of 25-hydroxyvitamin D. Total measurements include amounts attached to proteins and amounts that are free. Free measurements concern only the amount not attached to proteins. The review compares studies in people and laboratory methods.

Did inherited protein differences affect free and total vitamin D equally?

A 12-month vitamin D3 trial included 472 people. Before treatment, two groups with different inherited protein forms differed by about 30.2% in total 25-hydroxyvitamin D. The difference in directly measured free amounts was only 9.0% and was not statistically clear. The participants were almost entirely White, and included people with raised blood sugar below the diabetes range.

Why does the method used to measure vitamin D-binding protein matter?

Some binding-protein tests produced large differences linked to the protein forms people inherited and their ancestry that other methods did not reproduce. Errors in these measurements can make calculated free amounts inaccurate. Calculations also depend on uncertain assumptions about how tightly the protein holds vitamin D. Ancestry cannot substitute for measuring the protein.

Do free and total vitamin D measurements usually agree?

In a separate study of 1,057 men, total and directly measured free amounts were strongly related. Liver disease and a rare complete lack of binding protein showed that this relationship can change substantially. These unusual cases cannot be applied to everyone with common inherited differences.

Is measuring free vitamin D always better than measuring the total?

Free testing has not been shown to be better in every setting. Total measurements remain useful. Similar free amounts in blood do not prove identical effects in the body’s tissues.

Read the full article below ↓

Vitamin D-binding protein (DBP) sits directly between vitamin D physiology and vitamin D measurement. Because most circulating 25-hydroxyvitamin D [25(OH)D] is protein-bound, a change in DBP can alter the measured total pool without producing an equivalent change in the unbound fraction. But DBP variability has another complication: apparent differences in the protein can be created or exaggerated by the assay used to measure it. The evidence therefore has to separate biological variation in DBP concentration or genotype from analytical variation in DBP measurement, and then ask how each changes the relationship between total, calculated free and directly measured free 25(OH)D. See Study Summary.

Total and free 25(OH)D measure related but different circulating pools

Total 25(OH)D includes the DBP-bound, albumin-bound and unbound fractions. Directly measured free 25(OH)D isolates the very small unbound fraction. Calculated free 25(OH)D attempts to estimate that fraction from total 25(OH)D, DBP, albumin and assumed binding constants. The three measures are therefore mathematically and physiologically related, but they are not interchangeable. The carrier system creating those pools is described more broadly in Vitamin D Transport Proteins.

The distinction is developed conceptually in Free vs Total Vitamin D Explained. For DBP variability, the key point is that a total concentration can change because the size of the carrier reservoir changes. A calculated free value can also change because the DBP input or binding constant is wrong. Direct measurement avoids some of those assumptions, but it introduces its own assay characteristics and does not automatically become the superior clinical measure in every setting.

A supplementation trial shows that common DBP phenotypes affect total more than directly measured free 25(OH)D

Sollid and colleagues analysed 472 participants from a 12-month randomised trial of 20,000 IU vitamin D3 per week or placebo. They compared total 25(OH)D, directly measured free 25(OH)D, calculated free 25(OH)D and DBP across common DBP phenotypes while also examining sex, age, body mass index, season and supplementation response.

At baseline, total 25(OH)D and DBP concentrations were lower in the Gc2/Gc2 phenotype than in groups carrying Gc1S. The contrast became much smaller when the investigators measured free 25(OH)D directly. The difference between Gc2/Gc2 and Gc1S/Gc1S was approximately 30.2% for total 25(OH)D but about 9.0% for directly measured free 25(OH)D, and the free difference was no longer statistically significant.

That result is evidence for a real physiological distinction, but not for a universal rule that free 25(OH)D is always better. The phenotype groups differed in DBP concentration, so a larger difference in the total DBP-bound pool than in the unbound fraction is exactly what transport physiology would predict. The study also found that vitamin D supplementation increased vitamin D measures without materially changing serum DBP, and that the increases did not differ significantly across phenotype groups.

The trial population was almost entirely Caucasian and included people with prediabetes, while phenotype group sizes were unequal. Those boundaries matter. The study is strong evidence that common DBP phenotype can influence the relationship between total and free measures in this cohort; it is not a universal calibration table for every ancestry or clinical state. The broader role of inherited variation in vitamin D systems is a separate topic in Vitamin D Genetics.

Calculated free vitamin D is only as reliable as the DBP value used in the calculation

The attraction of calculated free 25(OH)D is obvious: if total 25(OH)D, albumin and DBP are known, an equation can estimate the unbound fraction without a dedicated free assay. The weakness is equally important. The calculation assumes that the measured DBP concentration is correct and that the binding constants applied to DBP variants are appropriate. Error in either component becomes error in the calculated free value.

This is why DBP assay architecture became scientifically important rather than merely a laboratory detail. Earlier work had supported a simple narrative in which Black Americans had much lower DBP concentrations and therefore similar calculated bioavailable vitamin D despite lower total 25(OH)D. That interpretation depended heavily on a monoclonal DBP assay. Later studies tested the protein with methods that did not rely on the same antibody recognition.

Mass spectrometry revealed genotype-sensitive monoclonal assay bias

Hoofnagle and colleagues quantified DBP by targeted mass spectrometry in genotyped samples and compared those results with immunoassay measurements. The large genotype pattern seen with the monoclonal assay was not reproduced by the mass-spectrometric method. The most plausible explanation was differential antibody recognition of common DBP isoforms rather than a comparably large biological difference in protein concentration.

This matters because GC variants are distributed differently across populations. If a monoclonal antibody recovers one common isoform less efficiently, a population in which that isoform is more frequent can appear to have lower DBP even when an orthogonal method does not show the same difference. When that underestimated DBP value is then inserted into an equation for free or bioavailable vitamin D, the analytical bias is propagated into a derived physiological quantity.

The issue is explored more broadly in Variability in Vitamin D Measurements. Here, its importance is specific: measurement error in the binding protein can mimic a biological mechanism and change the apparent relationship between total and free vitamin D.

An independent CKD cohort reproduced the method-dependent genotype effect

Denburg and colleagues compared two ELISA approaches with mass spectrometry in a genotyped chronic kidney disease cohort. The genotype dependence was substantially stronger with the monoclonal assay than with the polyclonal assay or mass-spectrometric measurement. This independent method comparison strengthened the conclusion that the assay itself could create variant-dependent recovery.

The study also demonstrated why the problem extends beyond DBP concentration. Bioavailable or free vitamin D calculations use the measured DBP value. When the monoclonal assay produced genotype-dependent DBP differences, derived bioavailable vitamin D values changed accordingly. The apparent physiological conclusion was therefore partly a property of the analytical pipeline. This is one concrete example of the broader cautions collected in Limitations of Vitamin D Blood Tests.

Chronic kidney disease is a relevant but non-neutral context. Protein handling, mineral metabolism and the clinical population differ from a healthy community sample. The finding should therefore be used for what it directly establishes: genotype-sensitive assay behaviour and its consequences for derived measures. It should not be used as proof that every DBP phenotype has the same biological effect in every population.

Direct free measurement changed the interpretation of ancestry-associated DBP differences

Schwartz and colleagues examined 1,057 men from the United States, United Kingdom and The Gambia. They measured free 25(OH)D directly and assessed DBP with proteomic, monoclonal and polyclonal methods. Total 25(OH)D correlated strongly with directly measured free 25(OH)D, with a Spearman correlation of approximately 0.84. This is an important counterweight to any claim that total 25(OH)D is generally disconnected from free vitamin D.

The DBP result depended sharply on method. With the monoclonal assay, mean DBP in participants of African ancestry appeared roughly 50% lower than in White participants. With polyclonal assays and proteomic methods, that large ancestry-associated difference was not present. Calculated free 25(OH)D based on polyclonal DBP measurements correlated strongly with directly measured free 25(OH)D, with correlations around 0.80 to 0.83.

The interpretation therefore changed when the measurement problem was removed. In these cohorts, free 25(OH)D reflected total 25(OH)D rather than showing a hidden preserved free pool that erased the lower total concentration in African American participants. This does not make ancestry a biological mechanism. It demonstrates the opposite lesson: race or ancestry cannot be used as a substitute for genotype, protein concentration or assay performance.

Direct measurement and calculation answer different methodological questions

The contrast between direct and calculated free 25(OH)D is not simply a contest between a good method and a bad one. Direct measurement asks an assay to quantify the unbound fraction itself. Calculation asks several measured variables and binding assumptions to reproduce that fraction mathematically. A direct assay therefore avoids propagation of a biased DBP concentration into the equation, but its own calibration, analytical performance and biological interpretation still matter.

Sollid’s trial illustrates the distinction because all three forms of information were available in the same cohort. Calculated free values were higher than the directly measured values, while the phenotype-associated differences were reduced when free 25(OH)D was assessed. That combination warns against treating a calculated value as though it were a direct observation of the free compartment. It is a model-derived quantity whose assumptions must remain visible.

Orthogonal measurement changes how confidently a biological explanation can be made

The assay-bias literature is methodologically important because several independent measurement traditions were brought to the same biological question. A genotype pattern that is large with one monoclonal antibody but much smaller with polyclonal antibodies and targeted proteomics is evidence against explaining the full difference as protein biology. The disagreement between methods is itself scientific information: it identifies the measurement step as a plausible source of the apparent phenotype.

This is why the later studies do more than replace one number with another. They change the causal interpretation. Once orthogonal methods no longer reproduce a large DBP deficit, a narrative built on that deficit loses its evidential foundation. The appropriate repair is not to deny all GC-related variation, but to distinguish biological concentration differences that survive method comparison from analytical differences that depend on how the protein is detected.

Binding-affinity assumptions add a second uncertainty to calculated free vitamin D

Even a perfectly measured DBP concentration would not remove every uncertainty from calculated free 25(OH)D. The equation also requires binding constants. Historical literature has sometimes treated large isoform-specific affinity differences among Gc1f, Gc1s and Gc2 as settled universal constants. The locked evidence does not support that level of certainty. Direct affinity studies and later syntheses have reported smaller or absent differences, leaving the magnitude of common isoform-affinity effects disputed.

That uncertainty should narrow the claim rather than erase genotype biology. Common GC phenotypes can be associated with different DBP concentrations and different total 25(OH)D. What cannot be asserted securely is that one fixed affinity hierarchy universally explains those differences or can be inserted into every free-vitamin calculation without qualification.

Liver disease shows a genuine low-binding-protein state in which total and free values separate

Bikle and colleagues studied people with liver disease, a setting in which DBP and albumin can be genuinely reduced. Compared with controls, patients had lower total 25(OH)D and a higher percentage free fraction, while absolute free 25(OH)D was similar. Total 25(OH)D correlated positively with DBP and albumin, whereas the percentage free fraction correlated inversely with those binding proteins.

This is physiologically different from an assay artefact. Here the binding-protein environment itself is altered by disease. The finding demonstrates that total concentration can fall because the carrier pool is reduced while the unbound concentration is relatively preserved. It also shows why a result can have a different interpretation when the protein-binding context is abnormal.

Liver disease is not equivalent to ordinary healthy-population variation. The study should not be used to claim that every person with a low total 25(OH)D has adequate free vitamin D. Instead, it establishes a boundary condition: substantial changes in binding proteins can decouple total and free measures.

Rare complete DBP deficiency pushes the same principle to an extreme

The rare human GC-deletion case provides an extreme natural experiment. With essentially no detectable DBP, total vitamin D metabolite concentrations were extraordinarily low, yet mineral homeostasis and vitamin D function were comparatively preserved. This is not a model of a common polymorphism, but it confirms that the total circulating pool is strongly dependent on the carrier reservoir.

Placed beside the common-phenotype and liver-disease evidence, the case creates a graded conceptual picture. Common phenotype differences can shift total more than directly measured free 25(OH)D. Acquired low-binding states can produce a clearer separation. Complete DBP absence can collapse total concentrations almost entirely. The magnitude differs, but the direction of the physiological principle is coherent.

Status, measurement and biological effect must remain separate questions

The evidence does not support choosing one number and treating it as a universal measure of vitamin D action. Total 25(OH)D remains strongly related to directly measured free 25(OH)D in ordinary populations and remains useful as a status marker. Free measurement can become informative when DBP concentration is materially altered or when a research question specifically concerns the unbound fraction. Calculated free values require particular caution because both DBP measurement and binding assumptions enter the estimate.

This is why Vitamin D Status vs Vitamin D Effect is a distinct reader question. A circulating concentration describes one layer of the system. Tissue effect also depends on delivery, uptake, metabolism, receptor signalling and local context. Similar free values cannot prove identical tissue action, just as different total values cannot automatically prove proportional differences in biological effect.

Evidence boundaries

  • Free 25(OH)D is not established as a universally superior replacement for total 25(OH)D.
  • Common DBP phenotypes can alter total and free measures differently, but their effects are smaller and more context-dependent than complete DBP deficiency.
  • Monoclonal DBP assays can show genotype-sensitive recovery; those results must not be interpreted as pure biology when orthogonal methods disagree.
  • Race or ancestry is not a biological surrogate for GC genotype, DBP concentration, DBP affinity or vitamin D availability.
  • Large universal isoform-specific DBP affinity constants remain insufficiently secure for unqualified use.
  • Liver disease and rare GC deletion demonstrate altered-binding states but cannot be generalised directly to common healthy-population variation.
  • Similar circulating free 25(OH)D does not establish identical tissue-level vitamin D action.

Primary evidence and publication details

Effects of vitamin D binding protein phenotypes and vitamin D supplementation on serum total 25(OH)D and directly measured free 25(OH)D. Sollid ST, Hutchinson MYS, Berg V, et al. European Journal of Endocrinology. 2016;174:445-452. doi:10.1530/EJE-15-1089.

Vitamin D-Binding Protein Concentrations Quantified by Mass Spectrometry. Hoofnagle AN, Eckfeldt JH, Lutsey PL. New England Journal of Medicine. 2015;373:1480-1482.

Comparison of Two ELISA Methods and Mass Spectrometry for Measurement of Vitamin D-Binding Protein: Implications for the Assessment of Bioavailable Vitamin D Concentrations Across Genotypes. Denburg MR, Hoofnagle AN, Sayed S, et al. Journal of Bone and Mineral Research. 2016;31:1128-1136.

Free 25-Hydroxyvitamin D: Impact of Vitamin D Binding Protein Assays on Racial-Genotypic Associations. Schwartz JB, et al. Journal of Clinical Endocrinology and Metabolism. 2016;101:2226-2234.

Free 25-hydroxyvitamin D levels are normal in subjects with liver disease and reduced total 25-hydroxyvitamin D levels. Bikle DD, Halloran BP, Gee E, Ryzen E, Haddad JG. Journal of Clinical Investigation. 1986;78:748-752.

Vitamin D-Binding Protein Deficiency and Homozygous Deletion of the GC Gene. Henderson CM, Fink SL, Bassyouni H, et al. New England Journal of Medicine. 2019;380:1150-1157.

Related reading

Original sources

Source

Effects of vitamin D binding protein phenotypes and vitamin D supplementation on serum total 25(OH)D and directly measured free 25(OH)D

Published in Sollid ST, Hutchinson MYS, Berg V, et al. European Journal of Endocrinology (2016).

Publication details

Effects of vitamin D binding protein phenotypes and vitamin D supplementation on serum total 25(OH)D and directly measured free 25(OH)D. Sollid ST, Hutchinson MYS, Berg V, et al. European Journal of Endocrinology. 2016;174:445-452. doi:10.1530/EJE-15-1089.

Read more →
Source

Comparison of Two ELISA Methods and Mass Spectrometry for Measurement of Vitamin D-Binding Protein: Implications for the Assessment of Bioavailable Vitamin D Concentrations Across Genotypes

Published in Denburg MR, Hoofnagle AN, Sayed S, et al. Journal of Bone and Mineral Research (2016).

Publication details

Comparison of Two ELISA Methods and Mass Spectrometry for Measurement of Vitamin D-Binding Protein: Implications for the Assessment of Bioavailable Vitamin D Concentrations Across Genotypes. Denburg MR, Hoofnagle AN, Sayed S, et al. Journal of Bone and Mineral Research. 2016;31:1128-1136.

Read more →