SEARCH

SEARCH BY CITATION

Keywords:

  • VITAMIN D BINDING PROTEIN (DBP);
  • PTH;
  • VITAMIN D CONCENTRATION

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Vitamin D insufficiency, as measured by 25-hydroxyvitamin D (25[OH]D) levels, has been associated with important health outcomes. The majority of vitamin D in circulation is bound to vitamin D–binding protein (DBP) and albumin, and recent genetic studies have demonstrated that serum DBP is a major determinant of 25(OH)D concentrations in adults. The impact of circulating DBP levels on vitamin D's biologic action, is unclear, but is of particular relevance to vitamin D epidemiology, because a lack of control for DBP levels could strongly influence the association of vitamin D with disease. Serum parathyroid hormone (PTH) levels can act as a biological readout of 25(OH)D activity. We therefore assessed the relationship between serum total and free 25(OH)D and PTH with and without adjusting for DBP, in 2073 subjects of European descent. Total 25(OH)D levels correlated positively (r = 0.19, p = 1.8 × 10−17) with DBP, whereas the free 25(OH)D correlated negatively (r = −0.14, p = 5.0 × 10−12). Total and free 25(OH)D levels correlated negatively with PTH (r = −0.29, p = 1.3 × 10−39; r = −0.26, p = 1.9 × 10−33, respectively). Including age, body mass index (BMI), sex, estimated glomerular filtration rate, calcium, and season of blood draw as covariates, total 25(OH)D levels were significantly associated with log-transformed PTH (lnPTH) levels (linear term: β = −0.010, p < 0.0001, squared term: β = 0.00004, p < 0.0001) and this association was not changed by adjusting for DBP. These findings provide evidence that in a largely vitamin D–sufficient cohort, the biological effect of vitamin D on PTH levels is mainly independent of DBP concentration. Accordingly, this study may provide useful information for studies investigating the relationship between vitamin D, DBP, and disease. © 2014 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

There is growing interest in the study of vitamin D due to its association with various diseases of public health importance, including cancer, autoimmune, infectious, metabolic, and heart diseases.[1-4] In almost all studies, total serum levels of the most abundant circulating vitamin D metabolite, 25-hydroxyvitamin D (25[OH]D), are taken as the best index of vitamin D sufficiency.

Vitamin D–binding protein (DBP) is the principal protein carrier of hydroxylated vitamin D metabolites in serum. Approximately 85% to 90% of the 25(OH)D and 1,25-dihydroxyvitamin D (1,25[OH]2D) are bound to DBP and 10% to 15% are bound to albumin, leaving less than 1% of vitamin D metabolites in a free form.[5] DBP has its own metabolic cycle and thus DBP is not just a simple intravascular transporter from liver to kidney.[6, 7]

Many hormones in circulation bind to proteins; however, the unbound levels of hormone often represent the amount of hormone available for biological activity. (For example, although the total levels of thyroid hormone measure both bound and free forms, only the free forms are biologically active and therefore of clinical interest.) Current laboratory assays measure total circulating 25(OH)D, and do not distinguish between free and bound forms. The biologic importance of the 25(OH)D could therefore depend strongly upon the relative abundance of the free form when compared with the total, which is largely a function of DBP levels.

Consequently, it is not clear if the association of vitamin D with the aforementioned diseases is effectively captured by measuring total circulating 25(OH)D, or whether the free proportion more strongly influences disease status. Recent genetics studies clearly demonstrate that genetic variation in DBP is a major determinant to 25(OH)D levels.[8] Also, it has been shown that changes in the serum concentrations of DBP can have a large effect on serum total 25(OH)D concentrations. Under most circumstances, however, they do not alter free 25(OH)D concentrations.[9] Therefore, the action of 25(OH)D may be depend upon the concentration of free 25(OH)D and, if this is the case, then the association of 25(OH)D with common diseases would have to be reconsidered after controlling for DBP levels.

In this study, we assessed whether DBP levels influence the relationship between 25(OH)D and a measure of its biologic activity, serum PTH levels.

Subjects and Methods

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Data for this study derives from the Canadian Multicentre Osteoporosis Study (CaMos), an ongoing population-based prospective cohort study of 9423 community-dwelling women and men aged 25 years and older at baseline (1995–1997), randomly selected from 9 urban regional centers across Canada.[10] At study entry all participants were interviewed to assess vitamin D intake, osteoporosis, and fracture-related risk factors, and vitamin D–related traits were measured. Interviews and measurements were repeated at 5 and 10 years after enrolment to reassess risk factors. A similar approach was used to recruit a random sample of Canadians (n = 1001) between 16 and 24 years of age, called the “CaMos youth cohort,” with reassessment at year 2 follow-up.[11] The current study included women and men ≥16 years old with their most recent blood results. A full discussion of the 25(OH)D levels in the CaMos population has been reported.[12]

Serum total 25(OH)D and PTH were measured using the Liaison (DiaSorin, Stillwater, MN, USA) autoimmunoanalyzer, using chemiluminescent immunoassay technology. The details of the assay have been described.[13] Serum DBP concentration was measured by an automated immunoturbidimetric assay (Dako, Glostrup, Denmark) configured for the Roche Modular P System (Roche Diagnostics, Laval, Quebec, Canada). The imprecision of the DBP mass assay was 2.2% at 275 mg/L (Dako Low QC) and 1.4% at 365 mg/L (in-house quality control). The assay was linear within the calibration range (28–544 mg/L) and the measurable range could be expanded with sample dilution. We have previously demonstrated good agreement between immunoturbidimetric values and dynamic binding capacity estimations (r = 0.86, p < 0.0001).[14] The reference range for normal serum calcium levels is 2.13 to 2.60 mmol/L with 97.1% of the participants falling within that range but none had calcium levels in the critical range (<1.75 or >3 mmol/L). The normal reference range for albumin is 34 to 48 g/L. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration formula.[15] Free levels of 25(OH)D were calculated using the following equation which has previously showed a significant correlation between the measured free fraction of 25(OH)D and its calculated value (r = 0.92, p < 0.0001)[16]:

  • display math(1)

Statistical methods

Using Stata (version: 64/MP10.1, College Station, TX, USA), we calculated nonparametric correlations between 25(OH)D, DBP, and PTH levels. We used log-transformed PTH (lnPTH) as the dependent variable and total and free 25(OH)D as explanatory variables. Linear regression models were constructed with polynomials of second degree (quadratic) in total and free 25(OH)D to explain PTH. Both sets of models (total and free 25[OH]D) were controlled for sex, age, and body mass index (BMI) at the time of the blood draw, serum calcium levels, estimated glomerular filtration rate (eGFR), and season of blood draw (categorical), with and without inclusion of DBP as a covariate. Models were stratified by sex and 25(OH)D levels, using the cut points of <50 nmol/L, ≥50 nmol/L and <75 nmol/L, and ≥75 nmol/L, in secondary analyses. In additional analyses we examined the effect of inclusion of albumin concentrations on the relationship between DBP, 25(OH)D, and PTH. Finally, although PTH is a proximal readout of 25(OH)D activity, we also assessed, independently of the aforementioned covariates, whether DBP influenced the relationship between 25(OH)D and serum calcium, another biochemical parameter influenced by 25(OH)D levels.

Results

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

A total of 2073 subjects participated in this study (69.5% women). The age range of the population-based participants was 16 to 98 years with mean ± SD of 65.7 ± 15.4 years and the mean 25(OH)D concentrations was 68.9 ± 24.7 (range, 10.0–212.0) nm/L. The characteristics of the study participants are presented in Table 1. Notably, 22% (n = 455) of the participants had 25(OH)D levels <50 nmol/L whereas 39% (n = 811) of the participants had values >75 nmol/L.

Table 1. Characteristics of the Study Participants (n = 2073)
Categorical variablesn%
Sex (women)142868.9%
Season of blood draw
Winter40719.6
Spring61629.7
Summer51624.9
Fall53425.8
Continuous variablesMean (SD)Minimum1st Q3rd QMaximum
  1. Q = quartile; DBP = vitamin D binding protein; BMI = body mass index; 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone; eGFR = estimated glomerular filtration rate.

Age (years)65.72 (15.4)16.1359.5176.1097.79
DBP (mg/L)369.6 (50.1)216.8338.6394.6759.0
BMI (kg/m2)27.49 (5.4)15.5723.7830.0874.85
25(OH)D (nm/L)68.95 (24.7)9.9852.3982.82212.05
PTH (pg/mL)62.42 (30.6)10.6044.5072.90380.00
Albumin (g/L)44.18 (2.6)32.3042.5046.0051.90
Free-25(OH)D (pmol/L)14.32 (5.1)2.1210.9517.1447.30
Calcium (mmol/L)2.38 (0.11)2.062.312.452.89
eGFR (mL/min)78.8 (19.1)23.667.190.7147.8

The Spearman correlation showed that the total 25(OH)D levels correlated positively (r = 0.19, p = 1.8 × 10−17) with DBP, whereas the free 25(OH)D correlated negatively with DBP (r = −0.14, p = 5.0 × 10−12). The total 25(OH)D and free 25(OH)D levels negatively correlated with PTH (r = −0.29, p = 1.3 × 10−39; r = −0.26, p = 1.9 × 10−33, respectively).

Although total 25(OH)D concentrations were significantly associated with lnPTH levels (linear term: β = −0.010, p < 0.0001, squared term: β = 0.00004, p < 0.0001), these associations were little changed after adjusting for the DBP level (Table 2). These results were unchanged when albumin was included as a covariate (results not shown). Using the same model but with free 25(OH)D, instead of total 25(OH)D, lnPTH concentrations were associated with free 25(OH)D (linear term: β = −0.051, p < 0.0001, squared term: β = 0.001, p < 0.0001) levels independently of DBP (Table 2). Moreover, stratifying the analysis by sex, or strata of 25(OH)D levels, did not change the results of the association of total 25(OH)D and free 25(OH)D with lnPTH, adjusted for DBP levels (Table 2, Supplemental Table S1 [women only], and Supplemental Table S2 [men only]). We next assessed whether DBP influenced the relationship between 25(OH)D and serum calcium. We found that the relationship between 25(OH)D and serum calcium was not strongly influenced by inclusion of DBP in the model. Beta (SE) for the effect of 25(OH)D and (25[OH]D)[2] on serum calcium excluding DBP as a covariate was 2.4 × 10−4 (0.0001) and −2.3 × 10−6 (8.9 × 10−7), respectively. Beta (SE) for the effect of 25(OH)D and (25[OH]D)2 on serum calcium including DBP as a covariate was 2.1 × 10−4 (0.0001) and −2.3 × 10−6 (8.8 × 10−7), respectively.

Table 2. Results of the Association Between lnPTH and 25(OH)D and Free 25(OH)D Including Age, Sex, BMI, Season, eGFR, and Serum Calcium as Covariates, Stratified by 25(OH)D Levels
 Excluding DBPIncluding DBP
 βp95% CIβp95% CI
  1. lnPTH = log-transformed parathyroid hormone; 25(OH)D = 25-hydroxyvitamin D; BMI = body mass index; eGFR = estimated glomerular filtration rate; DBP = vitamin D binding protein; CI = confidence interval.

Both genders
Model with total 25(OH)D
Total 25(OH)D−0.010<0.0001−0.013 to −0.008−0.010<0.0001−0.013 to −0.008
[Total 25(OH)D]20.000038<0.00010.000023 to 0.0000520.000038<0.00010.000023 to 0.000053
DBP   0.000310.081−0.00004 to 0.00067
Model with free 25(OH)D
Free 25(OH)D−0.051<0.0001−0.062 to −0.040−0.052<0.0001−0.063 to −0.041
[Free 25(OH)D]20.000905<0.00010.0005823 to 0.0012270.000912<0.00010.000590 to 0.001233
DBP   −0.000450.012−0.00080 to −0.00010
25(OH)D < 50 nmol/L (n = 455)
Model with total 25(OH)D
Total 25(OH)D−0.0210.118−0.047 to 0.005−0.0200.125−0.047 to 0.006
[Total 25(OH)D]20.0001530.424−0.000223 to 0.0005290.0001480.438−0.000227 to 0.000524
DBP   0.000550.182−0.00026 to 0.00135
Model with free 25(OH)D
Free 25(OH)D−0.0840.078−0.178 to 0.009−0.0780.107−0.173 to 0.017
[Free 25(OH)D]20.0026090.388−0.003328 to 0.0085470.0020430.515−0.004112 to 0.008198
DBP   −0.000330.489−0.00125 to 0.00060
50 ≤ 25(OH)D < 75 nmol/L (n = 870)
Model with total 25(OH)D
Total 25(OH)D0.0470.191−0.023 to 0.1170.0450.203−0.025 to 0.115
[Total 25(OH)D]2−0.00040.143−0.000970 to 0.00014047−0.0004060.152−0.000961 to 0.000150
DBP   0.000340.222−0.00021 to 0.00090
Model with free 25(OH)D
Free 25(OH)D0.0100.877−0.112 to 0.132−0.0100.870−0.137 to 0.116
[Free 25(OH)D]2−0.0010790.636−0.005551 to 0.003393−0.0005940.797−0.005130 to 0.003941
DBP   −0.000470.213−0.00121 to 0.00027
25(OH)D ≥ 75 nmol/L (n = 748)
Model with total 25(OH)D
Total 25(OH)D−0.0110.019−0.020 to −0.002−0.0110.019−0.020 to −0.002
[Total 25(OH)D]20.0000390.058−0.000001 to 0.0000780.0000380.059−0.000001 to 0.000078
DBP   −0.000020.936−0.00060 to 0.00055
Model with free 25(OH)D
Free 25(OH)D−0.0370.026−0.070 to −0.004−0.0570.002−0.093 to −0.021
[Free 25(OH)D]20.0006280.075−0.000063 to 0.0013180.0009660.0110.000224 to 0.001709
DBP   −0.000830.017−0.00151 to −0.00015

The nonlinear regression term can be interpreted to indicate an association between 25(OH)D and PTH varies by absolute level of 25(OH)D. Thus, at 30 nmol/L, each 1-nmol/L increase in 25(OH)D is associated with a 0.8% decrease in PTH, whereas, at 100 nmol/L, each 1-nmol/L increase in 25(OH)D is associated with a 0.2% decrease in PTH. Above 125 nmol/L there is no longer a negative association between 25(OH)D and PTH. Similarly, the association between free 25(OH)D and PTH varies by absolute level of free 25(OH)D. At 10 pmol/L, each 1-pmol/L increase in free 25(OH)D is associated with a 3.2% decrease in PTH, whereas, at 20 pmol/L, each 1-pmol/L increase in free 25(OH)D is associated with a 1.4% decrease in PTH. Above 28 pmol/L there was no longer a negative association between free 25(OH)D and PTH.

Discussion

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Although considerable controversy persists as to the definition of vitamin D insufficiency, it is now thought to be pandemic. Several health problems, such as malignancies,[17] fractures, and cardiovascular disease,[3] have been associated with low vitamin D status. Currently, total serum levels of 25(OH)D, are taken as the “gold-standard” measure of vitamin D insufficiency. Considering that more than 99% of the vitamin D in circulation is protein-bound, it is important to assess whether the biological effect of vitamin D is dependent on DBP levels—the major 25(OH)D binding protein.

In this study, we evaluated the effect of DBP concentration on the action of vitamin D by examining the relationship between concentrations of 25(OH)D and serum PTH—a biologic indicator of 25(OH)D activity—while controlling for DBP levels. Our results show that the correlation between total vitamin D and PTH is little changed in magnitude after controlling for DBP levels. This suggests that DBP levels do not influence the relation between 25(OH)D and PTH. Similar results were demonstrated for free 25(OH)D and PTH.

Previous studies with targeted deletion of DBP in mice showed that with complete absence of DBP, PTH levels were no different than controls in animals without vitamin D toxicity or prolonged vitamin D deficiency (levels of 25[OH]D less than 15 nmol/L). These findings support our observations in healthy adults, indicating that the relation between PTH and 25(OH)D is little affected by DBP status until the extremes of DBP levels and of vitamin D intake are reached.[18]

Previous studies have identified a interaction between 25(OH)D and DBP in disease states, such as pancreatic cancer.[19] Although our results do not support the notion that DBP influences the relationship between 25(OH)D and its more proximal outcome, PTH, our data does not permit direct comment on the potential impact of DBP on disease outcomes, such as cancer.

Recently, we, and others, described genetic variants that influence 25(OH)D in general populations through the SUNLIGHT consortium, involving 33,868 subjects.[8] One of the genetic polymorphisms in the human genome having a large effect on 25(OH)D levels is at the GC locus, which encodes DBP (we note that this polymorphism has an effect on 25[OH]D concentrations, but there is no clear evidence that it changes DBP binding affinity). Since this polymorphism (rs2282679) has an effect on allele frequency, which is 70% in CaMos and 73% in the HapMap CEU (Haplotype Map in northern and western European ancestry) population, it is unlikely that a different allele frequency distribution in CaMos has led to results that are not generalizable to other European populations.

Serum 1,25(OH)2D was not measured in our study. According to conventional endocrinology, serum 1,25(OH)2D is the negative regulator of circulating PTH levels; however, a previous study[20] reported no correlation between serum 1,25-(OH)2D and serum PTH. A second more recent study found a weak positive correlation between serum 1,25(OH)2D and PTH,[21] and a very recent study has reported that serum 1,25(OH)2D was positively associated with 25(OH)D but was not associated with PTH.[13] In the latter report,[13] 25(OH)D was negatively associated with PTH study as we and others have shown. This lack of correlation may be because serum 1,25(OH)2D circulates at much lower levels than 25(OH)D, has a shorter half-life in serum, and its production and clearance is rapidly and finely regulated, or because local 1,25(OH)2D levels produced in the parathyroid from circulating 25(OH)D, and not serum 1,25(OH)2D, are the more important determinants of circulating PTH. Irrespective of the precise mechanism, serum 1,25(OH)2D does not appear to correlate with serum PTH, which is in contrast to circulating 25(OH)D.

Several physiological conditions, such as pregnancy and childhood,[22] and several disease states (liver or kidney disease) have been associated with variation in DBP levels. The association between DBP levels and total vitamin D and free vitamin D has been evaluated in the individuals with these conditions. A study by Bikle and colleagues[23] found that both the total and the free 1,25(OH)2D concentrations were raised in pregnant subjects. However, in their patients with liver disease, the free 1,25(OH)2D concentrations were normal despite an evident decrease in total 1,25(OH)2D concentrations. This study assessed the correlation between measured free 1,25(OH)2D and DBP. The 1,25(OH)2D active metabolite has a concentration which is three orders of magnitude lower than 25(OH)D and has a shorter half-life, thus making serum 1,25(OH)D concentrations a less reliable biomarker for vitamin D status. Moreover, these authors did not evaluate the effect of DBP alterations and any consequent effects on free 1, 25(OH)2D concentrations on any biological outcome. Powe and colleagues[24] showed that free and bioavailable 25(OH)D were more strongly correlated with bone mineral density than total 25(OH)D. However, this correlation did not assess the effect of DBP on an outcome more immediately associated with 25(OH)D, such as PTH. Consistent with our study in adults, Carpenter and colleagues[22] studied a cohort of 762 infants and toddlers 6 to 36 months of age, and reported that both free and total 25(OH)D correlated inversely with serum PTH, but that DBP genotype did not appreciably affect the correlation. Their study group, however, was highly admixed and showed strong background effects for ethnicity.

We note that the mean concentration of DBP in our cohort is higher than previously reported values using a variety of techniques.[25, 26] We believe that the assay is unlikely to be the sources of discrepancy, because the quality control checks, both internal and external, demonstrate the validity of this assay and the overall coefficient of variation (CV) in duplicates is extremely low. The imprecision of the DBP mass assay was 2.2% at 275 mg/L (Dako Low QC) and 1.4% at 365 mg/L (in-house quality control). Moreover, the relative change in DBP levels in our cohort per allele at the genetic polymorphism, rs2282679, is highly consistent with previous studies that report the relative change in DBP per allele at the T436K polymorphism, which is in tight linkage disequilibrium with the rs2282679 polymorphism (r2 = 0.85). Thus, whereas the absolute DBP levels were higher in our cohort, we do not find that variation within the cohort was relatively different than other studies, when compared to an external reproducible variable (genotype at rs2282679). We do, however, point out that the beta estimate of the relationship between DBP and 25(OH)D or PTH in our cohort may be different than other studies and it may therefore may differ significantly.

In conclusion, our findings provide evidence that the biological activity of serum 25(OH)D, as assessed by serum PTH, is largely independent of DBP concentrations in healthy adults. However, it should be noted that DBP status is also predictive of total and free 1,25-(OH)2D concentrations.[7] Whether or not there are DBP-dependent effects on 1,25(OH)2D status that impact biological responses, such as PTH, in some pathophysiologic states has not been addressed here. Nevertheless, we suggest that controlling for DBP levels is likely to be of no additional benefit in epidemiological studies assessing the relationship between 25(OH)D and diseases that do not affect DBP metabolism directly.

Acknowledgements

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

This work was supported by grants from the Canadian Foundation for Innovation, the Canadian Institutes of Health Research (CIHR), Fonds de la recherche en sante du Québec, Ministère du Développement Economique, Innovation et Exportation du Québec, the Lady Davis Institute and the Jewish General Hospital. Funding to Suneil Malik, and to David E. C. Cole, Lei Fu, and Betty Wong for the DBP assay was jointly sponsored by Dairy Farmers of Canada and the Public Health Agency of Canada. Drs. Richards and Dastani are supported by the CIHR. We thank all study participants, volunteers, and study personnel that made this consortium possible.

Authors' roles: ZD, designed, performed the analysis and wrote the first draft of the manuscript; CB performed the analysis for the revision, assisted in the interpretation of results, and helped to write the manuscript; LL helped to analyze data; BYW and LF collected the data; SM, DG, DEC, and JBR initially designed the study, collected the data, assisted in the interpretation of results, and edited the manuscript. All authors also approved the final version.

References

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
jbmr2042-sm-0001-SupTabs-S1.doc113KSupplementary Tables.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.