Context:

Adequate serum 25-hydroxyvitamin D concentrations, [25(OH)D], are required for optimal bone health, and low levels are associated with chronic diseases.

Objective:

We investigated whether 41 candidate single nucleotide polymorphisms (SNPs) in vitamin D and calcium pathway genes (GC, DHCR7, CYP2R1, CYP27B1, CYP24A1, VDR, and CASR) are associated with [25(OH)D] or modify the increase in [25(OH)D] from vitamin D3 supplementation.

Design and Setting:

Baseline and year 1 [25(OH)D] measurements from a randomized controlled trial conducted at 11 clinical centers in the United States.

Participants:

A total of 1787 healthy non-Hispanic white participants aged 45–75 years.

Interventions:

Vitamin D3 (1000 IU/d), calcium carbonate (1200 mg/d elemental), both, or placebo.

Main Outcome Measures:

Genotype main effects and interactions with vitamin D3 treatment estimated using multiple linear regression.

Results:

The baseline serum [25(OH)D] was 25.4 ± 8.7 ng/mL (mean ± SD). Associations with baseline levels were discovered for SNPs in CYP24A1 (rs2209314, rs2762939) and confirmed for SNPs in GC and CYP2R1. After 1 year, [25(OH)D] increased on average by 6.1 ± 8.9 ng/mL on vitamin D3 treatment and decreased by 1.1 ± 8.4 ng/mL on placebo. The increase in [25(OH)D] due to vitamin D3 supplementation was modified by genotypes at rs10766197 near CYP2R1, rs6013897 near CYP24A1, and rs7968585 near VDR.

Conclusions:

The increase in [25(OH)D] attributable to vitamin D3 supplementation may vary according to common genetic differences in vitamin D 25-hydroxylase (CYP2R1), 24-hydroxylase (CYP24A1), and the vitamin D receptor (VDR) genes. These findings have implications for achieving optimal vitamin D status and potentially for vitamin D-related health outcomes.

Vitamin D is a prohormone whose active metabolite, 1,25-dihydroxyvitamin D [1,25(OH)2D], regulates calcium homeostasis and plays an important role in bone growth and remodeling (1). The main circulating metabolite is 25-hydroxyvitamin D or 25(OH)D, a biomarker of vitamin D status. Vitamin D metabolism is highly regulated, and variation in expression or activity of key proteins may modify its level or effects. Key metabolic enzymes include: 25-hydroxylase (CYP2R1), which converts vitamin D to 25(OH)D; 1-hydroxylase (CYP27B1), which activates 25(OH)D to 1,25(OH)2D; 24-hydroxylase (CYP24A1), which inactivates 25(OH)D and 1,25(OH)2D; and 7-dehydrocholesterol reductase (DHCR7), which shunts vitamin D precursors toward cholesterol biosynthesis. Other key components include: vitamin D-binding protein (GC), which transports circulating metabolites; and the vitamin D receptor (VDR), which binds 1,25(OH)2D to activate gene transcription and regulates vitamin D metabolism. Finally, the calcium-sensing receptor (CASR), which regulates PTH levels, is another critical part of this endocrine system controlling calcium levels and vitamin D metabolism.

Given the established risk of vitamin D deficiency for bone health and potential risks for major nonskeletal diseases (1), it is important to understand the role genetic factors play in modulating vitamin D levels. In previous research, common single nucleotide polymorphisms (SNPs) in vitamin D pathway genes have been associated with circulating [25(OH)D] in candidate gene studies (see Refs. 2 and 3) and genome-wide association studies (4, 5). A potentially important clinical question is whether these or other SNPs previously associated with other health outcomes modify the efficacy of vitamin D supplementation to increase serum [25(OH)D]. We investigated this question among participants in a randomized controlled trial of oral vitamin D3 supplementation.

Subjects and Methods

Study design and population

We analyzed associations between SNPs and serum [25(OH)D] among participants in the Vitamin D/Calcium Polyp Prevention Study, a randomized trial of vitamin D3 (1000 IU/d) and/or calcium carbonate (1200 mg/d elemental calcium) supplementation for prevention of colorectal adenomas conducted at 11 clinical centers in the United States from 2004–2013. Institutional review boards at each site approved the study, and participants provided informed consent. Participants were 45–75 years old and in good general health. Exclusion criteria included serum 25(OH)D < 12 ng/mL. At enrollment, participants provided information on demographic and lifestyle factors, physical activity (International Physical Activity Questionnaire, August 2002 short version, http://www.ipaq.ki.se), and usual dietary intake (Block Brief 2000 Food Frequency Questionnaire, http://www.nutritionquest.com). Height and weight were measured or documented by self-report. Participants agreed to avoid personal vitamin D or calcium supplements, and multivitamins without these were provided. After a blinded placebo run-in period, participants who took ≥80% of their study pills were randomized in blocks using computer-generated random numbers stratified by study center, sex, and colonoscopy interval (3 or 5 y) in a modified 2 × 2 factorial design with equal probability to vitamin D, calcium, both, or placebo (“four-arm study”). Women who wanted calcium supplementation were randomized to vitamin D only (to calcium alone, or to calcium plus vitamin D, “two-arm study”). After randomization, participants were interviewed semiannually regarding adherence, supplement use, and dietary calcium and vitamin D intake.

Measurement of [25(OH)D]

Serum [25(OH)D] was measured at enrollment and 1 year after randomization at the UCLA Center for Human Nutrition using a RIA kit (Immunodiagnostic Systems; intra-assay coefficient of variation [CV], 5.3–6.1% [26.5–151 nmol/L]; interassay CV, 8.2–7.3% [19.6–136 nmol/L]; http://www.idsplc.com). Samples from the international Vitamin D External Quality Assessment Scheme program were analyzed with every kit for quality control purposes. Repeated assay of blinded replicates from a pooled sample [average 25(OH)D, 21.6 ng/mL] indicated an intra-assay CV of 4.1%.

SNP selection and genotyping

We selected for genotyping 41 candidate SNPs in or near seven vitamin D or calcium pathway genes (GC, DHCR7, CYP2R1, CYP27B1, CYP24A1, VDR, and CASR) that were previously associated with [25(OH)D] or other health outcomes (Supplemental Table 1). Genomic DNA was isolated from buffy coat by BioServe Biotechnologies using DNAQuiK (http://www.bioserve.com). Genotyping by KBioscience used KASP technology (http://www.lgcgenomics.com); genotyping by Genome Quebec Innovation Center used Sequenom iPLEX Gold (http://www.sequenom.com) or predesigned TaqMan assays (rs228570 and rs10766197; http://www.invitrogen.com). Samples that could not be called on > four of 41 SNPs were dropped; the sample success rate was 96.1%. SNP call rates (samples successfully genotyped) ranged from 96.7 to 99.8% (median, 99%). Concordance rates among blinded replicates were 100%. All SNPs were in Hardy-Weinberg equilibrium in non-Hispanic whites (P > .05).

Statistical analyses

Only self-reported non-Hispanic whites were included in analyses to avoid spurious associations caused by population stratification. [25(OH)D] was log-transformed. Multiple linear regression was used to analyze associations of SNP genotypes with baseline [25(OH)D]. Covariates included age, sex, and season of baseline blood draw (coded as a nonordered categorical variable: 1, winter—December to February; 2, spring—March to May; 3, summer—June to August; and 4, fall—September to November). Genotype was modeled additively. Association P values were from Wald tests. Gene global P values were from likelihood ratio tests (joint contribution of all SNPs included in the model simultaneously).

Similarly, multiple linear regression was used to analyze whether SNP genotypes modified the effect of vitamin D3 supplementation on year 1 [25(OH)D]. Covariates included age, sex, season of year 1 blood draw, baseline 25(OH)D, randomized vitamin D3 treatment group (placebo or vitamin D3), and SNP genotype. Statistical adjustment for baseline [25(OH)D] removes the effect of regression to the mean (6). Genotype estimates and 95% confidence intervals (CIs) were for the association of the [SNP*vitamin D3 treatment] interaction term on year 1 [25(OH)D]. Interaction P values were obtained from Wald tests and global P values from likelihood ratio tests. To assess the impact of optimal adherence, some analyses included only participants who took ≥80% of their study pills, with no gaps >7 days and no personal vitamin D supplementation.

Due to prior evidence for associations of these candidate SNPs with circulating [25(OH)D] or other health outcomes, we did not adjust for multiple comparisons. Analysis of study treatment was intent-to-treat, except as indicated. Statistical tests were two-sided and were considered significant at P < .05. Analyses used SAS version 9.3 (SAS Institute Inc) or Stata version 12 (StataCorp).

Results

Of 1861 non-Hispanic white participants randomized, 1787 (96.1%) were included in baseline analyses and 1755 (94.3%) in year 1 analyses (Supplemental Figure 1). Baseline characteristics were similar between participants randomized to vitamin D3 treatment or placebo (Table 1). The mean baseline [25(OH)D] was 25.4 ng/mL. After 1 year of treatment, [25(OH)D] increased on average 6.1 ± 8.9 ng/mL (mean ± SD) among participants randomized to vitamin D3 and decreased 1.1 ± 8.4 ng/mL among those randomized to placebo (likely due to cessation of personal supplement use upon enrollment in the trial). The increase in year 1 [25(OH)D] due to vitamin D supplementation was not modified by randomization to calcium supplementation (Pinteraction= .94).

Table 1.

Baseline Characteristics of Study Participants by Vitamin D3 Treatment

Characteristics Placeboa Vitamin D3b P Valuec
n 890 897
Male, n (%) 571 (64) 575 (64) .98
Age at enrollment, y 58.0 ± 6.9 58.2 ± 6.8 .66
BMI, kg/m2 29.0 ± 5.2 28.8 ± 5.1 .55
Current smoker, n (%) 70 (8) 90 (10) .11
Currently drink alcohol, n (%) 602 (68) 630 (71) .17
Physical activity, MET-min/wk 3052 ± 2865 3065 ± 2950 .93
Dietary vitamin D intake, IU/d 141 ± 98 133 ± 99 .12
Dietary calcium intake, mg/d 692 ± 318 661 ± 300 .04
Multivitamin use, n (%) 504 (57) 513 (57) .85
Vitamin D supplement use, n (%) 108 (12) 133 (15) .10
Calcium supplement use, n (%) 169 (19) 182 (20) .50
Baseline serum 25(OH)D, ng/mL 25.4 ± 8.7 25.4 ± 8.3 1.00
Characteristics Placeboa Vitamin D3b P Valuec
n 890 897
Male, n (%) 571 (64) 575 (64) .98
Age at enrollment, y 58.0 ± 6.9 58.2 ± 6.8 .66
BMI, kg/m2 29.0 ± 5.2 28.8 ± 5.1 .55
Current smoker, n (%) 70 (8) 90 (10) .11
Currently drink alcohol, n (%) 602 (68) 630 (71) .17
Physical activity, MET-min/wk 3052 ± 2865 3065 ± 2950 .93
Dietary vitamin D intake, IU/d 141 ± 98 133 ± 99 .12
Dietary calcium intake, mg/d 692 ± 318 661 ± 300 .04
Multivitamin use, n (%) 504 (57) 513 (57) .85
Vitamin D supplement use, n (%) 108 (12) 133 (15) .10
Calcium supplement use, n (%) 169 (19) 182 (20) .50
Baseline serum 25(OH)D, ng/mL 25.4 ± 8.7 25.4 ± 8.3 1.00

Abbreviations: BMI, body mass index; MET, metabolic equivalent of task. Data are expressed as number of participants (percentage) or mean ± SD. Includes only non-Hispanic white participants with genotype data available.

a

Includes participants randomized to placebo (n = 336) or to calcium alone (n = 329) in the “four-arm study” and to calcium alone (n = 225) in the “two-arm study.” Counts of participants with missing data: one BMI, three alcohol, 13 activity, 48 dietary intake, two multivitamin, and three vitamin D supplement.

b

Includes participants randomized to vitamin D alone (n = 339) or to vitamin D plus calcium (n = 337) in the “four-arm study” and to vitamin D plus calcium (n = 221) in the “two-arm study.” Counts of participants with missing data: eight alcohol, 12 activity, 61 dietary intake, and one vitamin D supplement.

c

Two-sample t tests were used for continuous variables and Pearson χ2 tests for categorical variables.

Table 1.

Baseline Characteristics of Study Participants by Vitamin D3 Treatment

Characteristics Placeboa Vitamin D3b P Valuec
n 890 897
Male, n (%) 571 (64) 575 (64) .98
Age at enrollment, y 58.0 ± 6.9 58.2 ± 6.8 .66
BMI, kg/m2 29.0 ± 5.2 28.8 ± 5.1 .55
Current smoker, n (%) 70 (8) 90 (10) .11
Currently drink alcohol, n (%) 602 (68) 630 (71) .17
Physical activity, MET-min/wk 3052 ± 2865 3065 ± 2950 .93
Dietary vitamin D intake, IU/d 141 ± 98 133 ± 99 .12
Dietary calcium intake, mg/d 692 ± 318 661 ± 300 .04
Multivitamin use, n (%) 504 (57) 513 (57) .85
Vitamin D supplement use, n (%) 108 (12) 133 (15) .10
Calcium supplement use, n (%) 169 (19) 182 (20) .50
Baseline serum 25(OH)D, ng/mL 25.4 ± 8.7 25.4 ± 8.3 1.00
Characteristics Placeboa Vitamin D3b P Valuec
n 890 897
Male, n (%) 571 (64) 575 (64) .98
Age at enrollment, y 58.0 ± 6.9 58.2 ± 6.8 .66
BMI, kg/m2 29.0 ± 5.2 28.8 ± 5.1 .55
Current smoker, n (%) 70 (8) 90 (10) .11
Currently drink alcohol, n (%) 602 (68) 630 (71) .17
Physical activity, MET-min/wk 3052 ± 2865 3065 ± 2950 .93
Dietary vitamin D intake, IU/d 141 ± 98 133 ± 99 .12
Dietary calcium intake, mg/d 692 ± 318 661 ± 300 .04
Multivitamin use, n (%) 504 (57) 513 (57) .85
Vitamin D supplement use, n (%) 108 (12) 133 (15) .10
Calcium supplement use, n (%) 169 (19) 182 (20) .50
Baseline serum 25(OH)D, ng/mL 25.4 ± 8.7 25.4 ± 8.3 1.00

Abbreviations: BMI, body mass index; MET, metabolic equivalent of task. Data are expressed as number of participants (percentage) or mean ± SD. Includes only non-Hispanic white participants with genotype data available.

a

Includes participants randomized to placebo (n = 336) or to calcium alone (n = 329) in the “four-arm study” and to calcium alone (n = 225) in the “two-arm study.” Counts of participants with missing data: one BMI, three alcohol, 13 activity, 48 dietary intake, two multivitamin, and three vitamin D supplement.

b

Includes participants randomized to vitamin D alone (n = 339) or to vitamin D plus calcium (n = 337) in the “four-arm study” and to vitamin D plus calcium (n = 221) in the “two-arm study.” Counts of participants with missing data: eight alcohol, 12 activity, 61 dietary intake, and one vitamin D supplement.

c

Two-sample t tests were used for continuous variables and Pearson χ2 tests for categorical variables.

Associations with baseline [25(OH)D] for 33 SNPs are shown in Table 2 (results not shown for eight SNPs in high linkage disequilibrium; r2 > 0.95). Associations were statistically significant for SNPs in three genes: GC (rs12512631, rs4588, rs7041, rs222020, rs1155563; not shown, rs2282679 and rs3755967), CYP2R1 (rs12794714, rs10741657, rs1562902, rs10766197; not shown, rs2060793), and CYP24A1 (rs2209314, rs2762939) with per allele effect sizes ranging from −9 to +5% differences in [25(OH)D].

Table 2.

Associations of SNP Genotypes with Baseline 25(OH)D Level and Modification of the Increase in 25(OH)D Level Due to Randomized Vitamin D3 Supplementation (1000 IU/day) for 1 Year

SNP Baseline 25(OH)D Year 1 Increase in 25(OH)D
All Participants Optimally Adherente
n Estimated % difference (95% CI)a P Valueb n Estimated % difference (95% CI)c P Valued n Estimated % difference (95% CI)c P Valued
GC <.0001 .23 .39
    rs12512631 1759 4.69 (2.50, 6.92) <.0001 1728 0.47 (−3.04, 4.10) .80 1437 −0.51 (−4.20, 3.32) .79
    rs4588f 1737 −8.87 (−10.90,6.78) <.0001 1708 2.18 (−1.64, 6.15) .27 1424 3.15 (−1.02, 7.49) .14
    rs7041 1748 −6.69 (−8.57,4.76) <.0001 1717 2.42 (−1.01. 5.97) .17 1427 3.25 (−0.46, 7.10) .09
    rs222020 1755 3.21 (0.24, 6.26) .03 1726 −3.06 (−7.66, 1.77) .21 1434 −1.46 (−6.48, 3.83) .58
    rs16847015 1771 1.94 (−3.00, 7.13) .45 1740 −0.68 (−8.61, 7.94) .87 1448 0.75 (−8.03, 10.37) .87
    rs1155563 1742 −8.44 (−10.48,6.35) <.0001 1713 1.79 (−2.01, 5.75) .36 1426 2.33 (−1.77, 6.60) .27
    rs2298849 1771 1.95 (−0.70, 4.67) .15 1740 −3.59 (−7.73, 0.74) .10 1448 −2.98 (−7.40, 1.65) .20
DHCR7 .16 .05 .11
    rs12785878 1763 −2.25 (−4.51, 0.06) .06 1732 1.54 (−2.32, 5.56) .44 1442 3.04 (−1.14, 7.40) .16
    rs3829251 1754 −2.04 (−4.75, 0.75) .15 1723 −2.27 (−6.72, 2.39) .33 1433 0.01 (−4.91, 5.18) 1.00
CYP2R1 .0001 .03 .05
    rs12794714 1768 −4.74 (−6.68,2.75) <.0001 1737 −2.93 (−6.23, 0.49) .09 1446 −3.69 (−7.18,0.07) .05
    rs10741657f 1764 4.91 (2.76, 7.10) <.0001 1733 −0.80 (−4.19, 2.72) .65 1442 0.07 (−3.58, 3.85) .97
    rs1562902 1768 3.30 (1.21, 5.42) .002 1739 1.61 (−1.79, 5.13) .36 1448 0.95 (−2.65, 4.68) .61
    rs10766197 1755 −3.83 (−5.79,1.83) .0002 1724 −4.12 (−7.40,0.76) .02 1437 −4.21 (−7.66,0.63) .02
CYP27A1 .93 .23 .19
    rs703842f 1763 0.55 (−1.63, 2.78) .62 1732 0.25 (−3.36, 4.00) .89 1441 1.32 (−2.55, 5.34) .51
CYP24A1 .15 .61 0.69
    rs6013897 1766 −0.66 (−3.14, 1.88) .61 1735 −4.24 (−8.22,0.09) .04 1444 −4.86 (−9.10,0.42) .03
    rs2209314 1745 2.67 (0.19, 5.21) .03 1714 −3.88 (−7.70, 0.11) .06 1425 −3.62 (−7.66, 0.61) .09
    rs2762939 1757 2.75 (0.32, 5.23) .03 1727 −3.40 (−7.15, 0.49) .09 1437 −3.51 (−7.46, 0.62) .09
    rs4809958 1751 −1.02 (−3.75, 1.78) .47 1720 2.60 (−2.08, 7.49) .28 1434 2.71 (−2.22, 7.89) .28
    rs2244719 1756 −1.96 (−3.99, 0.12) .06 1725 1.72 (−1.78, 5.34) .34 1433 1.20 (−2.50, 5.05) .53
    rs2296241 1735 −1.24 (−3.30, 0.85) .24 1706 1.84 (−1.69, 5.49) .31 1420 1.78 (−1.98, 5.68) .36
    rs17219315 1768 −5.41 (−11.33, 0.91) .09 1737 6.00 (−4.90, 18.15) .29 1447 5.33 (−6.05, 18.09) .37
VDR .76 .23 .10
    rs7968585 1784 −1.50 (−3.49, 0.53) .15 1752 3.44 (−0.02, 7.01) .05 1458 3.15 (−0.54, 6.98) .09
    rs11574143 1758 −0.67 (−4.28, 3.09) .72 1728 −2.43 (−8.32, 3.84) .44 1441 −1.05 (−7.50, 5.85) .76
    rs731236f 1768 0.88 (−1.23, 3.04) .41 1736 −0.52 (−3.97, 3.06) .77 1443 0.06 (−3.66, 3.91) .98
    rs7975232 1769 −1.53 (−3.52, 0.50) .14 1737 2.87 (−0.57, 6.43) .10 1449 2.05 (−1.61, 5.84) .28
    rs2239179 1752 0.57 (−1.54, 2.72) .60 1721 −1.04 (−4.49, 2.53) .56 1433 −0.50 (−4.23, 3.38) .80
    rs2228570 1777 1.33 (−0.80, 3.50) .22 1745 −1.33 (−4.76, 2.23) .46 1453 −1.73 (−5.39, 2.08) .37
    rs10783219 1769 0.36 (−1.78, 2.56) .74 1739 −2.50 (−5.95, 1.09) .17 1447 −3.75 (−7.41, 0.06) .05
    rs7139166f 1772 0.02 (−2.02, 2.10) .99 1740 1.61 (−1.82, 5.17) .36 1447 3.99 (0.20, 7.92) .04
    rs11568820 1730 −0.89 (−3.37, 1.66) .49 1700 1.49 (−2.69, 5.85) .49 1417 −0.28 (−4.69, 4.33) .90
CASR .27 .41 .15
    rs1801725f 1761 −1.94 (−4.85, 1.05) .20 1731 −1.24 (−6.08, 3.84) .62 1444 −3.01 (−8.21, 2.48) .27
    rs1042636 1739 0.59 (−3.34, 4.69) .77 1711 −3.41 (−9.61, 3.22) .30 1423 1.14 (−5.83, 8.61) .76
    rs1801726 1766 1.46 (−3.64, 6.83) .58 1735 5.77 (−2.87, 15.18) .20 1444 8.69 (−0.77, 19.05) .07
SNP Baseline 25(OH)D Year 1 Increase in 25(OH)D
All Participants Optimally Adherente
n Estimated % difference (95% CI)a P Valueb n Estimated % difference (95% CI)c P Valued n Estimated % difference (95% CI)c P Valued
GC <.0001 .23 .39
    rs12512631 1759 4.69 (2.50, 6.92) <.0001 1728 0.47 (−3.04, 4.10) .80 1437 −0.51 (−4.20, 3.32) .79
    rs4588f 1737 −8.87 (−10.90,6.78) <.0001 1708 2.18 (−1.64, 6.15) .27 1424 3.15 (−1.02, 7.49) .14
    rs7041 1748 −6.69 (−8.57,4.76) <.0001 1717 2.42 (−1.01. 5.97) .17 1427 3.25 (−0.46, 7.10) .09
    rs222020 1755 3.21 (0.24, 6.26) .03 1726 −3.06 (−7.66, 1.77) .21 1434 −1.46 (−6.48, 3.83) .58
    rs16847015 1771 1.94 (−3.00, 7.13) .45 1740 −0.68 (−8.61, 7.94) .87 1448 0.75 (−8.03, 10.37) .87
    rs1155563 1742 −8.44 (−10.48,6.35) <.0001 1713 1.79 (−2.01, 5.75) .36 1426 2.33 (−1.77, 6.60) .27
    rs2298849 1771 1.95 (−0.70, 4.67) .15 1740 −3.59 (−7.73, 0.74) .10 1448 −2.98 (−7.40, 1.65) .20
DHCR7 .16 .05 .11
    rs12785878 1763 −2.25 (−4.51, 0.06) .06 1732 1.54 (−2.32, 5.56) .44 1442 3.04 (−1.14, 7.40) .16
    rs3829251 1754 −2.04 (−4.75, 0.75) .15 1723 −2.27 (−6.72, 2.39) .33 1433 0.01 (−4.91, 5.18) 1.00
CYP2R1 .0001 .03 .05
    rs12794714 1768 −4.74 (−6.68,2.75) <.0001 1737 −2.93 (−6.23, 0.49) .09 1446 −3.69 (−7.18,0.07) .05
    rs10741657f 1764 4.91 (2.76, 7.10) <.0001 1733 −0.80 (−4.19, 2.72) .65 1442 0.07 (−3.58, 3.85) .97
    rs1562902 1768 3.30 (1.21, 5.42) .002 1739 1.61 (−1.79, 5.13) .36 1448 0.95 (−2.65, 4.68) .61
    rs10766197 1755 −3.83 (−5.79,1.83) .0002 1724 −4.12 (−7.40,0.76) .02 1437 −4.21 (−7.66,0.63) .02
CYP27A1 .93 .23 .19
    rs703842f 1763 0.55 (−1.63, 2.78) .62 1732 0.25 (−3.36, 4.00) .89 1441 1.32 (−2.55, 5.34) .51
CYP24A1 .15 .61 0.69
    rs6013897 1766 −0.66 (−3.14, 1.88) .61 1735 −4.24 (−8.22,0.09) .04 1444 −4.86 (−9.10,0.42) .03
    rs2209314 1745 2.67 (0.19, 5.21) .03 1714 −3.88 (−7.70, 0.11) .06 1425 −3.62 (−7.66, 0.61) .09
    rs2762939 1757 2.75 (0.32, 5.23) .03 1727 −3.40 (−7.15, 0.49) .09 1437 −3.51 (−7.46, 0.62) .09
    rs4809958 1751 −1.02 (−3.75, 1.78) .47 1720 2.60 (−2.08, 7.49) .28 1434 2.71 (−2.22, 7.89) .28
    rs2244719 1756 −1.96 (−3.99, 0.12) .06 1725 1.72 (−1.78, 5.34) .34 1433 1.20 (−2.50, 5.05) .53
    rs2296241 1735 −1.24 (−3.30, 0.85) .24 1706 1.84 (−1.69, 5.49) .31 1420 1.78 (−1.98, 5.68) .36
    rs17219315 1768 −5.41 (−11.33, 0.91) .09 1737 6.00 (−4.90, 18.15) .29 1447 5.33 (−6.05, 18.09) .37
VDR .76 .23 .10
    rs7968585 1784 −1.50 (−3.49, 0.53) .15 1752 3.44 (−0.02, 7.01) .05 1458 3.15 (−0.54, 6.98) .09
    rs11574143 1758 −0.67 (−4.28, 3.09) .72 1728 −2.43 (−8.32, 3.84) .44 1441 −1.05 (−7.50, 5.85) .76
    rs731236f 1768 0.88 (−1.23, 3.04) .41 1736 −0.52 (−3.97, 3.06) .77 1443 0.06 (−3.66, 3.91) .98
    rs7975232 1769 −1.53 (−3.52, 0.50) .14 1737 2.87 (−0.57, 6.43) .10 1449 2.05 (−1.61, 5.84) .28
    rs2239179 1752 0.57 (−1.54, 2.72) .60 1721 −1.04 (−4.49, 2.53) .56 1433 −0.50 (−4.23, 3.38) .80
    rs2228570 1777 1.33 (−0.80, 3.50) .22 1745 −1.33 (−4.76, 2.23) .46 1453 −1.73 (−5.39, 2.08) .37
    rs10783219 1769 0.36 (−1.78, 2.56) .74 1739 −2.50 (−5.95, 1.09) .17 1447 −3.75 (−7.41, 0.06) .05
    rs7139166f 1772 0.02 (−2.02, 2.10) .99 1740 1.61 (−1.82, 5.17) .36 1447 3.99 (0.20, 7.92) .04
    rs11568820 1730 −0.89 (−3.37, 1.66) .49 1700 1.49 (−2.69, 5.85) .49 1417 −0.28 (−4.69, 4.33) .90
CASR .27 .41 .15
    rs1801725f 1761 −1.94 (−4.85, 1.05) .20 1731 −1.24 (−6.08, 3.84) .62 1444 −3.01 (−8.21, 2.48) .27
    rs1042636 1739 0.59 (−3.34, 4.69) .77 1711 −3.41 (−9.61, 3.22) .30 1423 1.14 (−5.83, 8.61) .76
    rs1801726 1766 1.46 (−3.64, 6.83) .58 1735 5.77 (−2.87, 15.18) .20 1444 8.69 (−0.77, 19.05) .07

Bold numbers indicate P ≤ .05.

a

Estimated percentage difference in baseline 25(OH)D level per variant allele compared to wild type under an additive genetic model using linear regression of log-transformed 25(OH)D, adjusting for age, sex, and season of baseline blood draw.

b

Wald test P values for single SNPs. Likelihood ratio test P values for joint contribution of all SNPs in a gene (first row).

c

Estimated percentage difference in year 1 for 25(OH)D level due to vitamin D3 supplementation per variant allele compared to wild type under an additive genetic model using linear regression of log-transformed year 1 25(OH)D, adjusting for log-transformed baseline 25(OH)D, age, sex, season of year 1 blood draw, SNP genotype, and vitamin D3 treatment assignment. Estimate is for the interaction between genotype and vitamin D3 treatment (genotype*vitamin D3 treatment) and indicates how genotype modifies the effect of vitamin D3 treatment on year 1 25(OH)D level.

d

Wald test P values for the interaction term (genotype*vitamin D3 treatment). Likelihood ratio test P values for the joint contribution of all interaction terms for all SNPs in the gene (first row).

e

Optimally adherent participants took ≥80% of their study pills, with no gaps in pill taking ≥7 days and no personal vitamin D supplementation.

f

SNPs in high linkage disequilibrium (r2 > 95%) with these SNPs are not shown (see Supplemental Table 1).

Table 2.

Associations of SNP Genotypes with Baseline 25(OH)D Level and Modification of the Increase in 25(OH)D Level Due to Randomized Vitamin D3 Supplementation (1000 IU/day) for 1 Year

SNP Baseline 25(OH)D Year 1 Increase in 25(OH)D
All Participants Optimally Adherente
n Estimated % difference (95% CI)a P Valueb n Estimated % difference (95% CI)c P Valued n Estimated % difference (95% CI)c P Valued
GC <.0001 .23 .39
    rs12512631 1759 4.69 (2.50, 6.92) <.0001 1728 0.47 (−3.04, 4.10) .80 1437 −0.51 (−4.20, 3.32) .79
    rs4588f 1737 −8.87 (−10.90,6.78) <.0001 1708 2.18 (−1.64, 6.15) .27 1424 3.15 (−1.02, 7.49) .14
    rs7041 1748 −6.69 (−8.57,4.76) <.0001 1717 2.42 (−1.01. 5.97) .17 1427 3.25 (−0.46, 7.10) .09
    rs222020 1755 3.21 (0.24, 6.26) .03 1726 −3.06 (−7.66, 1.77) .21 1434 −1.46 (−6.48, 3.83) .58
    rs16847015 1771 1.94 (−3.00, 7.13) .45 1740 −0.68 (−8.61, 7.94) .87 1448 0.75 (−8.03, 10.37) .87
    rs1155563 1742 −8.44 (−10.48,6.35) <.0001 1713 1.79 (−2.01, 5.75) .36 1426 2.33 (−1.77, 6.60) .27
    rs2298849 1771 1.95 (−0.70, 4.67) .15 1740 −3.59 (−7.73, 0.74) .10 1448 −2.98 (−7.40, 1.65) .20
DHCR7 .16 .05 .11
    rs12785878 1763 −2.25 (−4.51, 0.06) .06 1732 1.54 (−2.32, 5.56) .44 1442 3.04 (−1.14, 7.40) .16
    rs3829251 1754 −2.04 (−4.75, 0.75) .15 1723 −2.27 (−6.72, 2.39) .33 1433 0.01 (−4.91, 5.18) 1.00
CYP2R1 .0001 .03 .05
    rs12794714 1768 −4.74 (−6.68,2.75) <.0001 1737 −2.93 (−6.23, 0.49) .09 1446 −3.69 (−7.18,0.07) .05
    rs10741657f 1764 4.91 (2.76, 7.10) <.0001 1733 −0.80 (−4.19, 2.72) .65 1442 0.07 (−3.58, 3.85) .97
    rs1562902 1768 3.30 (1.21, 5.42) .002 1739 1.61 (−1.79, 5.13) .36 1448 0.95 (−2.65, 4.68) .61
    rs10766197 1755 −3.83 (−5.79,1.83) .0002 1724 −4.12 (−7.40,0.76) .02 1437 −4.21 (−7.66,0.63) .02
CYP27A1 .93 .23 .19
    rs703842f 1763 0.55 (−1.63, 2.78) .62 1732 0.25 (−3.36, 4.00) .89 1441 1.32 (−2.55, 5.34) .51
CYP24A1 .15 .61 0.69
    rs6013897 1766 −0.66 (−3.14, 1.88) .61 1735 −4.24 (−8.22,0.09) .04 1444 −4.86 (−9.10,0.42) .03
    rs2209314 1745 2.67 (0.19, 5.21) .03 1714 −3.88 (−7.70, 0.11) .06 1425 −3.62 (−7.66, 0.61) .09
    rs2762939 1757 2.75 (0.32, 5.23) .03 1727 −3.40 (−7.15, 0.49) .09 1437 −3.51 (−7.46, 0.62) .09
    rs4809958 1751 −1.02 (−3.75, 1.78) .47 1720 2.60 (−2.08, 7.49) .28 1434 2.71 (−2.22, 7.89) .28
    rs2244719 1756 −1.96 (−3.99, 0.12) .06 1725 1.72 (−1.78, 5.34) .34 1433 1.20 (−2.50, 5.05) .53
    rs2296241 1735 −1.24 (−3.30, 0.85) .24 1706 1.84 (−1.69, 5.49) .31 1420 1.78 (−1.98, 5.68) .36
    rs17219315 1768 −5.41 (−11.33, 0.91) .09 1737 6.00 (−4.90, 18.15) .29 1447 5.33 (−6.05, 18.09) .37
VDR .76 .23 .10
    rs7968585 1784 −1.50 (−3.49, 0.53) .15 1752 3.44 (−0.02, 7.01) .05 1458 3.15 (−0.54, 6.98) .09
    rs11574143 1758 −0.67 (−4.28, 3.09) .72 1728 −2.43 (−8.32, 3.84) .44 1441 −1.05 (−7.50, 5.85) .76
    rs731236f 1768 0.88 (−1.23, 3.04) .41 1736 −0.52 (−3.97, 3.06) .77 1443 0.06 (−3.66, 3.91) .98
    rs7975232 1769 −1.53 (−3.52, 0.50) .14 1737 2.87 (−0.57, 6.43) .10 1449 2.05 (−1.61, 5.84) .28
    rs2239179 1752 0.57 (−1.54, 2.72) .60 1721 −1.04 (−4.49, 2.53) .56 1433 −0.50 (−4.23, 3.38) .80
    rs2228570 1777 1.33 (−0.80, 3.50) .22 1745 −1.33 (−4.76, 2.23) .46 1453 −1.73 (−5.39, 2.08) .37
    rs10783219 1769 0.36 (−1.78, 2.56) .74 1739 −2.50 (−5.95, 1.09) .17 1447 −3.75 (−7.41, 0.06) .05
    rs7139166f 1772 0.02 (−2.02, 2.10) .99 1740 1.61 (−1.82, 5.17) .36 1447 3.99 (0.20, 7.92) .04
    rs11568820 1730 −0.89 (−3.37, 1.66) .49 1700 1.49 (−2.69, 5.85) .49 1417 −0.28 (−4.69, 4.33) .90
CASR .27 .41 .15
    rs1801725f 1761 −1.94 (−4.85, 1.05) .20 1731 −1.24 (−6.08, 3.84) .62 1444 −3.01 (−8.21, 2.48) .27
    rs1042636 1739 0.59 (−3.34, 4.69) .77 1711 −3.41 (−9.61, 3.22) .30 1423 1.14 (−5.83, 8.61) .76
    rs1801726 1766 1.46 (−3.64, 6.83) .58 1735 5.77 (−2.87, 15.18) .20 1444 8.69 (−0.77, 19.05) .07
SNP Baseline 25(OH)D Year 1 Increase in 25(OH)D
All Participants Optimally Adherente
n Estimated % difference (95% CI)a P Valueb n Estimated % difference (95% CI)c P Valued n Estimated % difference (95% CI)c P Valued
GC <.0001 .23 .39
    rs12512631 1759 4.69 (2.50, 6.92) <.0001 1728 0.47 (−3.04, 4.10) .80 1437 −0.51 (−4.20, 3.32) .79
    rs4588f 1737 −8.87 (−10.90,6.78) <.0001 1708 2.18 (−1.64, 6.15) .27 1424 3.15 (−1.02, 7.49) .14
    rs7041 1748 −6.69 (−8.57,4.76) <.0001 1717 2.42 (−1.01. 5.97) .17 1427 3.25 (−0.46, 7.10) .09
    rs222020 1755 3.21 (0.24, 6.26) .03 1726 −3.06 (−7.66, 1.77) .21 1434 −1.46 (−6.48, 3.83) .58
    rs16847015 1771 1.94 (−3.00, 7.13) .45 1740 −0.68 (−8.61, 7.94) .87 1448 0.75 (−8.03, 10.37) .87
    rs1155563 1742 −8.44 (−10.48,6.35) <.0001 1713 1.79 (−2.01, 5.75) .36 1426 2.33 (−1.77, 6.60) .27
    rs2298849 1771 1.95 (−0.70, 4.67) .15 1740 −3.59 (−7.73, 0.74) .10 1448 −2.98 (−7.40, 1.65) .20
DHCR7 .16 .05 .11
    rs12785878 1763 −2.25 (−4.51, 0.06) .06 1732 1.54 (−2.32, 5.56) .44 1442 3.04 (−1.14, 7.40) .16
    rs3829251 1754 −2.04 (−4.75, 0.75) .15 1723 −2.27 (−6.72, 2.39) .33 1433 0.01 (−4.91, 5.18) 1.00
CYP2R1 .0001 .03 .05
    rs12794714 1768 −4.74 (−6.68,2.75) <.0001 1737 −2.93 (−6.23, 0.49) .09 1446 −3.69 (−7.18,0.07) .05
    rs10741657f 1764 4.91 (2.76, 7.10) <.0001 1733 −0.80 (−4.19, 2.72) .65 1442 0.07 (−3.58, 3.85) .97
    rs1562902 1768 3.30 (1.21, 5.42) .002 1739 1.61 (−1.79, 5.13) .36 1448 0.95 (−2.65, 4.68) .61
    rs10766197 1755 −3.83 (−5.79,1.83) .0002 1724 −4.12 (−7.40,0.76) .02 1437 −4.21 (−7.66,0.63) .02
CYP27A1 .93 .23 .19
    rs703842f 1763 0.55 (−1.63, 2.78) .62 1732 0.25 (−3.36, 4.00) .89 1441 1.32 (−2.55, 5.34) .51
CYP24A1 .15 .61 0.69
    rs6013897 1766 −0.66 (−3.14, 1.88) .61 1735 −4.24 (−8.22,0.09) .04 1444 −4.86 (−9.10,0.42) .03
    rs2209314 1745 2.67 (0.19, 5.21) .03 1714 −3.88 (−7.70, 0.11) .06 1425 −3.62 (−7.66, 0.61) .09
    rs2762939 1757 2.75 (0.32, 5.23) .03 1727 −3.40 (−7.15, 0.49) .09 1437 −3.51 (−7.46, 0.62) .09
    rs4809958 1751 −1.02 (−3.75, 1.78) .47 1720 2.60 (−2.08, 7.49) .28 1434 2.71 (−2.22, 7.89) .28
    rs2244719 1756 −1.96 (−3.99, 0.12) .06 1725 1.72 (−1.78, 5.34) .34 1433 1.20 (−2.50, 5.05) .53
    rs2296241 1735 −1.24 (−3.30, 0.85) .24 1706 1.84 (−1.69, 5.49) .31 1420 1.78 (−1.98, 5.68) .36
    rs17219315 1768 −5.41 (−11.33, 0.91) .09 1737 6.00 (−4.90, 18.15) .29 1447 5.33 (−6.05, 18.09) .37
VDR .76 .23 .10
    rs7968585 1784 −1.50 (−3.49, 0.53) .15 1752 3.44 (−0.02, 7.01) .05 1458 3.15 (−0.54, 6.98) .09
    rs11574143 1758 −0.67 (−4.28, 3.09) .72 1728 −2.43 (−8.32, 3.84) .44 1441 −1.05 (−7.50, 5.85) .76
    rs731236f 1768 0.88 (−1.23, 3.04) .41 1736 −0.52 (−3.97, 3.06) .77 1443 0.06 (−3.66, 3.91) .98
    rs7975232 1769 −1.53 (−3.52, 0.50) .14 1737 2.87 (−0.57, 6.43) .10 1449 2.05 (−1.61, 5.84) .28
    rs2239179 1752 0.57 (−1.54, 2.72) .60 1721 −1.04 (−4.49, 2.53) .56 1433 −0.50 (−4.23, 3.38) .80
    rs2228570 1777 1.33 (−0.80, 3.50) .22 1745 −1.33 (−4.76, 2.23) .46 1453 −1.73 (−5.39, 2.08) .37
    rs10783219 1769 0.36 (−1.78, 2.56) .74 1739 −2.50 (−5.95, 1.09) .17 1447 −3.75 (−7.41, 0.06) .05
    rs7139166f 1772 0.02 (−2.02, 2.10) .99 1740 1.61 (−1.82, 5.17) .36 1447 3.99 (0.20, 7.92) .04
    rs11568820 1730 −0.89 (−3.37, 1.66) .49 1700 1.49 (−2.69, 5.85) .49 1417 −0.28 (−4.69, 4.33) .90
CASR .27 .41 .15
    rs1801725f 1761 −1.94 (−4.85, 1.05) .20 1731 −1.24 (−6.08, 3.84) .62 1444 −3.01 (−8.21, 2.48) .27
    rs1042636 1739 0.59 (−3.34, 4.69) .77 1711 −3.41 (−9.61, 3.22) .30 1423 1.14 (−5.83, 8.61) .76
    rs1801726 1766 1.46 (−3.64, 6.83) .58 1735 5.77 (−2.87, 15.18) .20 1444 8.69 (−0.77, 19.05) .07

Bold numbers indicate P ≤ .05.

a

Estimated percentage difference in baseline 25(OH)D level per variant allele compared to wild type under an additive genetic model using linear regression of log-transformed 25(OH)D, adjusting for age, sex, and season of baseline blood draw.

b

Wald test P values for single SNPs. Likelihood ratio test P values for joint contribution of all SNPs in a gene (first row).

c

Estimated percentage difference in year 1 for 25(OH)D level due to vitamin D3 supplementation per variant allele compared to wild type under an additive genetic model using linear regression of log-transformed year 1 25(OH)D, adjusting for log-transformed baseline 25(OH)D, age, sex, season of year 1 blood draw, SNP genotype, and vitamin D3 treatment assignment. Estimate is for the interaction between genotype and vitamin D3 treatment (genotype*vitamin D3 treatment) and indicates how genotype modifies the effect of vitamin D3 treatment on year 1 25(OH)D level.

d

Wald test P values for the interaction term (genotype*vitamin D3 treatment). Likelihood ratio test P values for the joint contribution of all interaction terms for all SNPs in the gene (first row).

e

Optimally adherent participants took ≥80% of their study pills, with no gaps in pill taking ≥7 days and no personal vitamin D supplementation.

f

SNPs in high linkage disequilibrium (r2 > 95%) with these SNPs are not shown (see Supplemental Table 1).

Next, we investigated whether these SNPs modified the efficacy of vitamin D3 treatment to increase year 1 [25(OH)D] (Table 2). Three SNPs had statistically significant interactions: rs10766197 near CYP2R1, rs6013897 near CYP24A1, and rs7968585 near VDR, with per allele effect sizes ranging from −4% to +3% differences in [25(OH)]. When these analyses were restricted to optimally adherent participants (n = 1460; 83%), three additional SNPs had statistically significant interactions (rs12794714 in CYP2R1; rs10783219 and rs7139166 in VDR) and for rs7968585 in VDR the magnitude of the interaction decreased slightly and was no longer statistically significant.

Finally, analyses used to investigate the joint effects of linked SNPs implicated haplotypes in GC, CYP2R1, CYP24A1, and VDR with differences in baseline [25(OH)D], and haplotypes in CYP2R1 and VDR with differences in the effect of vitamin D3 treatment on [25(OH)D] (Supplemental Table 2).

Discussion

We confirmed known associations between baseline [25(OH)D] and 12 SNPs in GC and CYP2R1, and we identified new associations with two SNPs in CYP24A1 (rs2209314 and rs2762939). These two SNPs have smaller effect sizes, which may explain why they weren't identified in genome-wide association studies, and have previously been associated with decreased breast cancer risk (7) or reduced coronary artery calcification (8).

In novel analyses, we observed three SNPs that significantly modified the efficacy of 1000 IU/d vitamin D3 supplementation for increasing [25(OH)D]: rs10766197 near CYP2R1, rs6013897 near CYP24A1, and rs7968585 near VDR. Two of these have previously been associated with clinical outcomes: rs6013897 with lower risk of aggressive prostate cancer (9), and rs7968585 with risks of major clinical outcomes in association with low [25(OH)D] (10). Surprisingly, many SNPs (in GC and CYP2R1) associated with baseline [25(OH)D] did not substantially modify the response to supplementation, whereas two SNPs (in CYP24A1 and VDR) not associated with baseline levels did modify response. These results suggest that different mechanisms regulate 25(OH)D derived from diet vs cutaneous synthesis or adipose stores.

We know of only two small studies that previously examined associations with response to supplementation. In a pooled analysis of three trials (n = 285), Didriksen et al (11) reported that rs10741657 (CYP2R1) was associated with a larger increase in 25(OH)D response to vitamin D 20 000 IU twice weekly plus 800 IU daily. The reason for the discrepancy with our results is unknown, but it may relate to the 6-fold higher dose of vitamin D. Also, two studies reported opposite effects of two GC SNPs; the rs4588 variant was associated with a larger proportional 25(OH)D increase from 4000 IU/d vitamin D (n = 98) (12), whereas the rs2282679 variant (high linkage disequilibrium with rs4588) was associated with a smaller increase in the Didriksen study (11). Regardless, the physiological relevance of changes in [25(OH)D] associated with GC variants is unknown because the amount of free 25(OH)D may not change substantially.

A main strength of our study is the randomized design, which maximizes internal validity by providing a consistent route and dose of vitamin D3 supplementation and minimizes differences between treatment groups, which could confound results. Participant adherence to pill taking and avoidance of personal supplementation was good. Few potential participants (3.4%) were excluded due to low baseline 25(OH)D (<12 ng/mL). Measurement of 25(OH)D was centralized with good reproducibility. We used rigorous statistical methodology to examine interactions between genotype and vitamin D3 supplementation, removing effects of regression to the mean and taking into account small decreases in [25(OH)D] among placebo participants. Finally, we focused on candidate SNPs previously associated with [25(OH)D] or other health outcomes.

The limitations of our study suggest avenues for future research. Analyses were restricted to non-Hispanic whites; other populations should be studied. We assessed only one oral dose of vitamin D3; the dosing regimen providing the optimal risk/benefit profile or [25(OH)D] for various health outcomes is unknown. To our knowledge the function of these SNPs (except missense mutations in GC and CASR) has not been investigated. A more comprehensive analysis of genetic variation in these key genes and their effects is warranted. Finally, the clinical relevance of the variants identified here should be explored further for diseases that are associated with vitamin D status.

Acknowledgments

The authors are grateful to all of the participants, clinical research coordinators, and coinvestigators in the Vitamin D/Calcium Polyp Prevention Study who made this research possible. In addition, we thank the bioinformatics and coordination staff at the Dartmouth Project Coordination Center.

This work was supported by National Institutes of Health, National Cancer Institute Grants CA159360 (to E.L.B.) and CA098286 (to J.A.B.). Study pills were provided by Pfizer Consumer Healthcare.

Clinical Trial Registration: ClinicalTrials.gov no. NCT00153816.

Disclosure Summary: All authors state that they have nothing to disclose.

Abbreviations

     
  • CASR

    calcium-sensing receptor

  •  
  • CI

    confidence interval

  •  
  • CV

    coefficient of variation

  •  
  • 1,25(OH)2D

    1,25-dihydroxyvitamin D

  •  
  • 25(OH)D

    25-hydroxyvitamin D

  •  
  • [25(OH)D]

    25(OH)D concentration

  •  
  • SNP

    single nucleotide polymorphism

  •  
  • VDR

    vitamin D receptor.

References

1.

Holick
MF
.
Vitamin D deficiency
.
N Engl J Med
.
2007
;
357
:
266
281
.

2.

McGrath
JJ
,
Saha
S
,
Burne
TH
,
Eyles
DW
.
A systematic review of the association between common single nucleotide polymorphisms and 25-hydroxyvitamin D concentrations
.
J Steroid Biochem Mol Biol
.
2010
;
121
:
471
477
.

3.

Dastani
Z
,
Li
R
,
Richards
B
.
Genetic regulation of vitamin D levels
.
Calcif Tissue Int
.
2013
;
92
:
106
117
.

4.

Wang
TJ
,
Zhang
F
,
Richards
JB
, et al. .
Common genetic determinants of vitamin D insufficiency: a genome-wide association study
.
Lancet
.
2010
;
376
:
180
188
.

5.

Ahn
J
,
Yu
K
,
Stolzenberg-Solomon
R
, et al. .
Genome-wide association study of circulating vitamin D levels
.
Hum Mol Genet
.
2010
;
19
:
2739
2745
.

6.

Vickers
AJ
,
Altman
DG
.
Statistics notes: analysing controlled trials with baseline and follow up measurements
.
BMJ
.
2001
;
323
:
1123
1124
.

7.

Yao
S
,
Zirpoli
G
,
Bovbjerg
DH
, et al. .
Variants in the vitamin D pathway, serum levels of vitamin D, and estrogen receptor negative breast cancer among African-American women: a case-control study
.
Breast Cancer Res
.
2012
;
14
:
R58
.

8.

Shen
H
,
Bielak
LF
,
Ferguson
JF
, et al. .
Association of the vitamin D metabolism gene CYP24A1 with coronary artery calcification
.
Arterioscler Thromb Vasc Biol
.
2010
;
30
:
2648
2654
.

9.

Mondul
AM
,
Shui
IM
,
Yu
K
, et al. .
Genetic variation in the vitamin D pathway in relation to risk of prostate cancer–results from the breast and prostate cancer cohort consortium
.
Cancer Epidemiol Biomarkers Prev
.
2013
;
22
:
688
696
.

10.

Levin
GP
,
Robinson-Cohen
C
,
de Boer
IH
, et al. .
Genetic variants and associations of 25-hydroxyvitamin D concentrations with major clinical outcomes
.
JAMA
.
2012
;
308
:
1898
1905
.

11.

Didriksen
A
,
Grimnes
G
,
Hutchinson
MS
, et al. .
The serum 25-hydroxyvitamin D response to vitamin D supplementation is related to genetic factors, BMI, and baseline levels
.
Eur J Endocrinol
.
2013
;
169
:
559
567
.

12.

Fu
L
,
Yun
F
,
Oczak
M
,
Wong
BY
,
Vieth
R
,
Cole
DE
.
Common genetic variants of the vitamin D binding protein (DBP) predict differences in response of serum 25-hydroxyvitamin D [25(OH)D] to vitamin D supplementation
.
Clin Biochem
.
2009
;
42
:
1174
1177
.

Supplementary data