Skip to main content
NCBI home page
International Journal of Molecular Epidemiology and Genetics logoLink to International Journal of Molecular Epidemiology and Genetics
. 2014 Feb 17;5(1):31–46.

Genetic variants and non-genetic factors predict circulating vitamin D levels in Hispanic and non-Hispanic White women: the Breast Cancer Health Disparities Study

Wei Wang 1,2, Sue Ann Ingles 4,5, Gabriela Torres-Mejía 6, Mariana C Stern 4, Frank Z Stanczyk 5, Gary G Schwartz 7, David O Nelson 1, Laura Fejerman 8, Roger K Wolff 9, Martha L Slattery 9, Esther M John 1,3
PMCID: PMC3939005  PMID: 24596595

Abstract

Genome-wide association studies (GWAS) have identified common polymorphisms in or near GC, CYP2R1, CYP24A1, and NADSYN1/DHCR7 genes to be associated with circulating levels of 25-hydroxyvitamin D [25(OH)D] in European populations. To replicate these GWAS findings, we examined six selected polymorphisms from these regions and their relation with circulating 25(OH)D levels in 1,605 Hispanic women (629 U.S. Hispanics and 976 Mexicans) and 354 non-Hispanic White (NHW) women. We also assessed the potential interactions between these variants and known non-genetic predictors of 25(OH)D levels, including body mass index (BMI), sunlight exposure and vitamin D intake from diet and supplements. The minor alleles of the two GC polymorphisms (rs7041 and rs2282679) were significantly associated with lower 25(OH)D levels in both Hispanic and NHW women. The CYP2R1 polymorphism, rs2060793, also was significantly associated with 25(OH)D levels in both groups. We found no significant associations for the polymorphisms in the CYP24A1. In Hispanic controls, 25(OH)D levels were significantly associated with the rs12785878T and rs1790349G haplotype in the NADSYN1/DHCR7 region. Significant interactions between GC rs2282679 and BMI and between rs12785878 and time spent in outdoor activities were observed. These results provide further support for the contribution of common genetic variants to individual variability in circulating 25(OH)D levels. The observed interactions between SNPs and non-genetic factors warrant confirmation.

Keywords: Circulating levels, Hispanics, genetic polymorphisms, SNPs, genotype-phenotype correlation, vitamin D

Introduction

Vitamin D deficiency is a common health problem worldwide [1] and has been implicated in a wide range of diseases beyond bone disease, e.g., diabetes, cardiovascular disease, multiple sclerosis, and some cancer types, with the evidence being most consistent for colorectal cancer ([2] and reviewed in [1,3,4]).

Known determinants of an individual’s vitamin D status, which is measured by circulating 25-hydroxyvitamin D (25(OH)D) levels, include sunlight exposure, and vitamin D from diet and supplements. Twin studies and family-based studies suggest that genetic factors also contribute to a significant proportion of the inter-individual variability in vitamin D status [5-7].

There have been three published genome-wide association studies (GWAS) of vitamin D status (circulating 25(OH)D levels and vitamin D insufficiency) in adults (http://www.genome.gov/gwastudies/). The two GWAS conducted in populations of European ancestry found significant associations with several single nucleotide polymorphisms (SNPs) located within or near vitamin D metabolic pathway genes [8,9]. These genes include the GC gene (encoding vitamin D binding protein, formerly known as group-specific component), the CYP2R1 gene (encoding 25-hydroxylase, an enzyme that catalyzes the conversion of vitamin D to 25(OH)D), the CYP24A1 gene (encoding 24-hydroxylase, a key enzyme in vitamin D inactivation), and the NADSYN1/DHCR7 gene region (encoding nicotinamide adenine dinucleotide synthetase and 7-dehydrocholesterol reductase, respectively. The latter converts 7-dehydrocholesterol to cholesterol and thus modulates the availability of 7-dehydrocholesterol in the skin). In the third GWAS, conducted in U.S. Hispanics, polymorphisms in these four genes or genomic regions were not associated with 25(OH)D levels at a genome-wide significance level [10]. However, that study had a much smaller sample size and used a family-based design, which typically has reduced power due to overmatching on genotype [11,12].

The associations between polymorphisms in the GC, CYP24A1, and CYP2R1 genes and circulating 25(OH)D have been supported by several candidate gene studies [13-18]. However, with a few exceptions [16-19], those studies have focused primarily on populations of European descent. The NADSYN1/DHCR7 region was replicated in a candidate gene study of a Chinese population [18]. Further replication is needed in other populations, including Hispanics who are at a greater risk of vitamin D deficiency than non-Hispanic Whites [20].

The aim of this study was to replicate the GWAS-identified polymorphisms in Mexican women, and U.S. Hispanic and non-Hispanic White (NHW) women living in California. Hereafter, the Mexican and U.S. Hispanic women are referred to as Hispanics. In addition, in a subset of study participants, we assessed the interactions between the SNPs and non-genetic determinants of circulating 25(OH)D.

Materials and methods

Study population

This analysis from the Breast Cancer Health Disparities Study [21] includes participants from two population-based case-control studies who completed an in-person interview and provided a blood or mouthwash sample. All study participants signed written informed consent. The study was approved by the institutional review board of the Cancer Prevention Institute of California and each participating institution in Mexico.

The San Francisco Bay Area Breast Cancer Study (SFBCS) [22,23] included women aged 35-79 years from the San Francisco Bay Area diagnosed with a first primary histologically confirmed invasive breast cancer between April 1997 and April 2002; controls were identified by random-digit dialing and frequency-matched to cases based on the expected race/ethnicity and 5-year age distribution. A total of 1,105 cases and 1,318 controls completed the interview and 93% of cases (750 Hispanics and 276 NHWs) and 92% of controls (916 Hispanics and 298 NHWs) provided a blood or mouthwash sample.

The Mexico Breast Cancer Study (MBCS) [24,25] included women aged 28-74 years from Mexico. Eligible cases were diagnosed with either a new histologically confirmed in situ or invasive breast cancer between January 2004 and December 2007 at 12 participating hospitals from three main health care systems; controls were randomly selected from the catchment area as the cases and frequency matched to cases based on 5-year age distribution, membership in health care institution, and place of residence. A total of 1,000 cases and 1,074 controls completed the interview. Blood was collected from 85% of cases and 93% of controls.

Data collection

Both studies administered similar structured questionnaires in English or Spanish to collect information on demographic background, lifestyle factors, menstrual and reproductive history, supplement use, and other breast cancer risk factors. The interview also included measurements of weight, height, hip and waist circumferences.

Additionally, the SFBCS collected information on sunlight exposure, including lifetime residential history; lifetime history of time spent in outdoor activities (occupational or recreational physical activities, outdoor chores, walking and bicycling for transportation); as well as sun avoidance behaviors. Usual dietary vitamin D intake in the reference year (defined as the calendar year before diagnosis for cases or before selection into the study for controls) was assessed using a modified Block Food Frequency Questionnaire [26,27]. Annual average ultraviolet (UV) exposure at the residence at the time of diagnosis (cases) or selection into the study (controls) was estimated using the GIS (geographical information systems)-based ANUSPLIN model, as described elsewhere [28]. Several summary variables of residential UV exposure were created: mean at the point residence; and mean, median, minimum and maximum, and standard deviation (SD) within a 20 km radius of the residence. Skin pigmentation was measured using a Minolta Chromameter, with measures ranging from 0 (perfect black) to 100 (perfect white), and constitutive pigmentation was calculated as the average of two measurements taken at the upper inner arm, an area usually not exposed to the sun [23]. For 206 Hispanic cases, no skin pigmentation measurements were taken.

A second short questionnaire was administered at the time of blood collection (except for 206 Hispanic cases), with a median time between interview and blood draw of 10 months for cases and 9 months for controls. This questionnaire updated information on current weight, menstrual and reproductive history, and outdoor activities during six months prior to blood collection.

Measurement of circulating 25(OH)D

In the MBCS, blood samples were processed within 1 hour of blood draw and serum was stored at -70°C until 25(OH)D measurement by liquid chromatography-tandem mass spectrometry at Quest Diagnostics Nichols Institute (San Juan Capistrano, CA). The intra-assay and inter-assay coefficients of variation (CVs) were 6.4-12.6% and 8.0-14.4%, respectively. The assay sensitivity was 4 ng/mL (1 ng/mL=2.496 nmol/L). 25(OH)D measurements were available for a random sample of 574 cases and 649 controls.

In the SFBCS, blood samples were transported to the laboratory on ice, processed within 3 hours of blood draw, and stored at -70°C until 25(OH)D analysis. Serum total 25(OH)D in controls was measured using chemiluminescent immunoassay from Nichols Institute Diagnostics (San Clemente, CA). The intra-assay and inter-assay CVs were 3.0-4.5% and 6.4-14.5%, respectively. The assay sensitivity was 4 ng/mL. For cases, plasma 25(OH)D level was measured using an iodinated RIA kit from DiaSorin (Stillwater, MN) with a two-step procedure [29]. The intra-assay and inter-assay CVs were 8.6-12.5% and 8.2-11.0%, respectively. The assay sensitivity was 1.5 ng/mL. 25(OH)D measurements were available for 670 cases (440 Hispanics and 230 NHWs) and a subset of 358 controls (223 Hispanics and 135 NHWs).

Genotyping

DNA was extracted from either blood or mouthwash samples. Whole Genome Amplification (WGA) was applied to the mouthwash-derived DNA samples prior to genotyping. Six SNPs were selected from among those that passed the genome-wide significance level (p<10-8) in the two prior GWAS (described above). Redundant SNPs (r2 of at least 0.7 in HapMap-MEX panels) were excluded, with priority given to SNPs that were identified by both GWAS (rs2282679 and rs7041 in GC; rs12785878 and rs1790349 in DHCR7/NADSYN1; rs2060793 in CYP2R1; rs6013897 in CYP24A1) (Table 1). The six SNPs and 104 Ancestral Informative Markers (AIMs) were genotyped using a custom Illumina GoldenGate 1536 SNP OPA (Illumina Inc, San Diego, CA) [21]. Overall, a genotyping call rate of 99.93% was attained (99.65% for whole-genome amplified samples). For the 132 duplicate samples that were included as an internal quality control, the concordance rate was 99.996% as determined by 193,297 matching genotypes among sample pairs [21].

Table 1.

Linkage Disequilibrium (Measured by r2) Between the GWAS-identified SNPs and Those Selected in the Study

Nearby gene SNPs1 p value reported in GWAS LD in Hapmap-CEU panel LD in Hapmap-MEX panel
GC rs2282679 2.0x10-30 [9]; 1.9x10-109 [8]
rs7041 4.1x10-22 [9]; 6.31x10-59 [8] r2=0.6 with rs2282679 r2=0.2 with rs2282679
rs3755967 2.4x10-75 [8] r2=1.0 with rs2282679 r2=1.0 with rs2282679
rs17467825 6.7x10-74 [8] r2=1.0 with rs2282679 Not available
rs1155563 3.8x10-25 [9]; 2.4x10-73 [8] r2=0.8 with rs2282679 r2=0.7 with rs2282679
rs2298850 2.0x10-71 [8] r2=0.95 with rs2282679 r2=0.8 with rs2282679
DHCR7/NADSYN1 rs12785878 2.1x10-27 [8]
rs7944926 9.0x10-16 [8] r2=1.0 with rs12785878 r2=1.0 with rs12785878
rs12800438 2.5x10-15 [8] r2=1.0 with rs12785878 r2=0.96 with rs12785878
rs3794060 3.4x10-15 [8] r2=1.0 with rs12785878 r2=1.0 with rs12785878
rs4945008 4.6x10-15 [8] r2=0.95 with rs12785878 Not available
rs4944957 8.7x10-15 [8] r2=1.0 with rs12785878 r2=0.96 with rs12785878
rs1790349 1.8x10-6 [9] r2=0.5 with rs12785878 r2=0.1 with rs12785878
rs3829251 8.8x10-7; 3.4x10-9 in combined meta-analysis [9] r2=0.9 with rs1790349 r2=0.7 with 1790349
rs11234027 3.4x10-3; 3.4x10-9 in combined meta-analysis [9] r2=0.9 with rs1790349 r2=0.7 with 1790349
CYP2R1 rs2060793 2.9x10-5; 2.9x10-17 in combined meta-analysis [9]; 1.7x10-11 [8]
rs10741657 3.3x10-20 [8] r2=1.0 with rs2060793 Not available
rs1993116 2.9x10-17; 2.9x10-17 in combined meta-analysis [9] r2=0.98 with rs2060793 r2=0.9 with rs2060793
rs12794714 1.8x10-9 [8] r2=0.5 rs2060793 r2=0.7 with rs2060793
rs10500804 2.7x10-9 [8] r2=0.5 with rs2060793 r2=0.9 with rs2060793
rs7116978 5.0x10-9 [8] r2=0.8 with rs2060793 r2=0.9 with rs2060793
CYP24A1 rs6013897 7.2x10-10 [8]
Open in a new tab
1

SNPs that were selected for genotyping in the present study are highlighted in bold.

Statistical analysis

Because different measurement methods were used between the MBCS and SFBCS and between cases and controls in the SFBCS, circulating 25(OH)D was standardized according to the mean and SD in each of the following groups: MBCS controls, MBCS cases, SFBCS Hispanic controls, SFBCS Hispanic cases, SFBCS NHW controls and SFBCS NHW cases, with standardized 25(OH)D calculated as [measured 25(OH)D minus group mean] divided by group SD.

Hardy-Weinberg equilibrium (HWE) was tested for each control group using the Pearson goodness-of-fit statistic. Multiple testing-adjusted HWE p-value was obtained using the Benjamini and Hochberg procedure in SAS MULTTEST [30]. Multiple linear regression models were fit to examine associations between each SNP and standardized 25(OH)D levels. SNPs were analyzed assuming a co-dominant model. An additive model was also assessed. Individual ancestry was computed using the STRUCTURE program [31,32] based on 104 AIMs [21]. Percent Native American ancestry according to a two-founding population model was included as a continuous covariate in all regression models, in addition to age at blood draw (continuous), season at blood draw (spring: March 21-June 20; summer: June 21-September 20; fall: September 21-December 20; December 21-March 20), and body mass index (BMI) [weight (kg)/height2 (m), continuous]. Associations between SNPs and 25(OH)D levels were analyzed separately in Hispanic controls, Hispanic cases, NHW controls and NHW cases. No significant differences in associations by disease status were detected in either Hispanics or NHWs, and we therefore pooled cases and controls within each ethnic group. Secondary analyses were performed on the non-standardized levels of 25(OH)D within each of the following groups: Mexican controls, Mexican cases, U.S. Hispanic controls, U.S. Hispanic cases, NHW controls and NHW cases. A generalized linear model framework with natural log link and a gamma distribution was applied. Results on the genotype-25(OH)D associations were similar (data not shown).

Haplotype analysis was performed for the two GC SNPs and the two SNPs in the NADSYN1/DHCR7 region. Estimated haplotype frequencies and individual haplotype probabilities were obtained using the TagSNP software [33]. Gene-gene interactions between polymorphisms in or near the GC, CYP2R1 and CYP24A1 genes and the NADSYN1/DHCR7 region were examined by stratified analysis. Statistical interactions were assessed by including cross-product terms in regression models.

Associations between 25(OH)D levels and BMI, total vitamin D intake, residential UV exposure, outdoor activity during the 6 months prior to blood draw, and constitutive skin pigmentation were analyzed in the SFBCS only. Gene-environment interactions were examined between the significant non-genetic predictors of 25(OH)D levels and the SNPs.

Of the 2,251 women with measured 25(OH)D levels, 184 (153 from the MBCS and 31 from the SFBCS) were excluded due to missing genotype data. We also excluded 108 women with missing data on covariates that were adjusted for in the analysis. The final analysis included 976 women from the MBCS and 983 from the SFBCS, for a total of 1,605 Hispanics and 354 NHWs.

Results

Characteristics of study participants

Participant characteristics are summarized in Table 2. Overall, vitamin D deficiency (defined as ≤20 ng/mL as in [34]) was present in 44% of the study participants and was more prevalent in breast cancer cases than in controls in both the MBCS and SFBCS. Hispanic cases and controls from the SFBCS both had a higher prevalence of vitamin D deficiency than their NHW counterparts. In the SFBCS, Hispanic women had darker constitutive skin pigmentation and spent less time in outdoor activities than NHW women. The SFBCS Hispanics had higher dietary vitamin D intake than NHWs, but were less likely to take vitamin D supplements and therefore had lower total vitamin D intake.

Table 2.

Circulating 25(OH)D Levels and Covariates by Study, Ethnicity and Case-control Status1

Mexico Breast Cancer Study (MBCS) San Francisco Bay Area Breast Cancer Study (SFBCS)
Controls N=564 Cases N=412 Hispanic controls N=215 NHW controls N=130 Hispanic cases N=414 NHW cases N=224 P value5
Vitamin D deficient (<20 ng/mL), % 27.3 47.3 52.0 27.7 63.5 48.2 <0.01
Age at blood draw (years), mean (SD) 50.3 (8.9) 52.6 (9.7) 52.8 (11.6) 56.8 (14.0) 59.5 (11.2) 61.1 (12.3) <0.01
Body mass index at blood draw (kg/m2), mean (SD)2 30.7 (5.4) 29.8 (5.9) 29.7 (5.5) 26.8 (6.3) 29.5 (8.8) 27.2 (5.1) <0.01
Percent Native American ancestry, mean (SD) 0.71 (0.2) 0.66 (0.2) 0.46 (0.2) 0.05 (0.1) 0.42 (0.2) 0.05 (0.1) <0.01
Time between interview and blood draw (months), median 10 7 18 5
Dietary vitamin D intake (IU/day), mean (SD) 283.0 (216.0) 248.3 (154.6) 254.0 (150.7) 247.3 (146.7) 0.09
Total vitamin D intake (IU/day), mean (SD) 426.0 (385.1) 509.8 (346.6) 341.4 (252.5) 503.2 (370.6) <0.01
Outdoor activities (hours/week), mean (SD)3 3.5 (4.6) 5.0 (4.8) 3.2 (4.6) 4.0 (3.7) <0.01
Constitutive skin pigmentation, mean (SD)4 34.7 (4.3) 39.6 (4.2) 34.5 (3.9) 38.6 (3.4) <0.01
Open in a new tab
1

N=1959 women with measured 25(OH)D levels, after excluding those with missing covariates, including season at blood draw, age at blood draw, body mass index (BMI), and genetic ancestry.

2

For 190 SFBCS Hispanic cases who had missing BMI at blood draw, BMI in the reference year was used instead.

3

For 190 SFBCS Hispanic cases, no information was collected on outdoor activities within the 6 months prior to blood collection.

4

Skin pigmentation measured at the upper inner arm, not taken for the 190 SFBCS Hispanic cases and missing for another 13 women.

5

P values are for chi square tests for categorical variables and for ANOVA tests comparing mean values for continuous variables.

The genes, corresponding SNPs and their minor allele frequencies (MAFs) in controls are presented in Table 3. No significant departure of observed genotype frequencies from HWE expectations was found for any of the SNPs [false discovery rate (FDR) p-value>0.05], except for rs12785878 in Mexican controls.

Table 3.

Genes, SNPs, and Allele Frequencies in Controls, by Study and Ethnicity

Nearby gene SNP Chromosome location Region Major/Minor Allele MAF1 in MBCS2 controls HWE p-value3 MAF in SFBCS4 Hispanic controls HWE p-value3 MAF in SFBCS NHW controls
HWE p-value3
GC rs7041 4q12-q13 coding G/T 0.50 0.04 (0.21) 0.47 0.78 (0.93) 0.44 0.58 (0.87)
GC rs2282679 4q12-q13 intron A/C 0.19 0.50 (0.84) 0.21 0.05 (0.21) 0.30 0.68 (0.89)
NADSYN1 rs12785878 11q13.4 intron T/G 0.56 <0.001 (0.01) 0.48 0.24 (0.47) 0.31 0.89 (0.94)
NADSYN1/DHCR7 rs1790349 11q13.4 Intergenic A/G 0.17 0.13 (0.34) 0.20 0.04 (0.21) 0.15 0.19 (0.44)
CYP2R1 rs2060793 11p15.2 5’ upstream G/A 0.36 0.69 (0.89) 0.38 0.84 (0.94) 0.38 0.51 (0.84)
CYP24A1 rs6013897 20q13 intergenic TT/A 0.42 0.08 (0.23) 0.34 0.97 (0.97) 0.21 0.07 (0.23)
Open in a new tab
1

Minor allele frequency (MSF).

2

Mexico Breast Cancer Study.

3

Hardy-Weinberg Equilibrium (HWE) p-value from a 1df chi-square test (false discovery rate p-value in parenthesis).

4

San Francisco Bay Area Breast Cancer Study.

Associations between circulating 25(OH)D levels and SNPs

The associations between the SNPs and standardized 25(OH)D levels are shown in Table 4 for Hispanic and in Table 5 for NHW controls and cases. In both ethnic groups, associations did not differ significantly between cases and controls for any of the six SNPs, thus cases and controls were combined. In both groups, the less common alleles at the two GC loci (rs7041 and rs2282679) were associated with significantly lower 25(OH)D levels, and the A allele of the CYP2R1 polymorphism (rs2060793) was significantly associated with higher 25(OH)D levels. The magnitudes of the associations were similar for Hispanic and NHW women. Overall, the three SNPs individually accounted for 0.6-3.5% of the inter-individual variability in circulating 25(OH)D levels, with the r2 varying between Hispanic and NHW women and between women with or without breast cancer (data not shown).

Table 4.

Associations Between Circulating 25(OH)D Levels and Genotypes in Hispanic Cases and Controls from the Mexico Breast Cancer Study and the San Francisco Bay Area Breast Cancer Study

Cases and Controls Combined Controls Cases
Genotypes N 25(OH)D (ng/mL) Mean (SD) Adjusted beta (SE)1 Adjusted P value1 N 25(OH)D (ng/mL) Mean (SD) Adjuste beta (SE)1 Adjuste P value1 N 25(OH)D (ng/mL) Mean (SD) Adjuste beta (SE)1 Adjuste P value1 P interaction 2
rs7041
    GG 419 21.8 (7.7) 216 23.6 (7.0) 203 19.8 (7.9)
    GT 764 20.6 (7.3) -0.15 (0.06) 0.01 362 22.2 (7.3) -0.20 (0.08) 0.02 402 19.1 (7.0) -0.09 (0.09) 0.29
    TT 422 19.4 (7.2) -0.33 (0.07) <10-4 201 21.3 (7.3) -0.34 (0.10) 0.0004 221 17.8 (6.7) -0.32 (0.10) 0.001
Per copy of T allele -0.17 (0.03) ptrd<10-4 -0.17 (0.05) ptrd=4x10-4 -0.16 (0.05) ptrd=0.001 0.73
rs2282679
    AA 1005 21.4 (7.7) 516 23.1 (7.5) 489 19.6 (7.4)
    AC 529 19.5 (6.8) -0.23 (0.05) <10-4 227 21.4 (6.2) -0.28 (0.08) 0.0002 302 18.2 (6.9) -0.18 (0.07) 0.01
    CC 71 16.7 (6.4) -0.56 (0.12) <10-4 36 17.7 (7.4) -0.65 (0.17) 0.0001 35 15.8 (5.0) -0.43 (0.18) 0.02
Per copy of C allele -0.25 (0.04) ptrd<10-4 -0.30 (0.06) ptrd<10-4 -0.19 (0.06) ptrd=0.002 0.21
rs12785878 185 22.3 (6.6) 189 19.4 (7.0)
    TT 374 20.8 (6.9) 353 22.6 (7.8) 0.09 (0.09) 0.33 410 18.9 (7.4) -0.07 (0.09) 0.42
    TG 763 20.6 (7.8) 0.01 (0.06) 0.87 241 22.0 (7.0) -0.05 (0.10) 0.59 227 18.6 (6.9) -0.10 (0.10) 0.31
    GG 468 20.4 (7.1) -0.07 (0.07) 0.29 185 22.3 (6.6) 189 19.4 (7.0)
Per copy of G allele -0.04 (0.03) ptrd=0.26 -0.03 (0.05) ptrd=0.50 -0.05 (0.05) ptrd=0.32 0.86
rs1790349
    AA 1100 20.8 (7.5) 527 22.6 (7.2) 573 19.2 (7.3)
    AG 454 20.0 (7.3) -0.11 (0.06) 0.05 227 21.7 (7.4) -0.07 (0.08) 0.34 227 18.3 (6.8) -0.14 (0.08) 0.07
    GG 51 20.6 (6.9) -0.07 (0.14) 0.63 25 23.3 (6.2) 0.01 (0.20) 0.95 26 18.0 (6.6) -0.13 (0.20) 0.52
Per copy of G allele -0.08 (0.05) ptrd=0.08 -0.05 (0.07) ptrd=0.49 -0.11 (0.07) ptrd=0.08 0.39
rs2060793
    GG 618 19.9 (6.8) 315 21.5 (6.4) 303 18.3 (6.9)
    GA 747 21.0 (7.8) 0.20 (0.05) 0.0002 356 23.0 (7.7) 0.25 (0.08) 0.001 391 19.3 (7.4) 0.15 (0.08) 0.05
    AA 240 20.8 (7.6) 0.21 (0.08) 0.005 108 22.6 (7.9) 0.21 (0.11) 0.05 132 19.4 (7.2) 0.20 (0.10) 0.05
Per copy of A allele 0.13 (0.04) ptrd=0.004 0.14 (0.05) ptrd=0.005 0.11 (0.05) ptrd=0.03 0.72
rs6013897
    TT 614 20.7 (7.9) 294 22.4 (7.7) 320 19.1 (7.8)
    TA 736 20.7 (7.3) 0.005 (0.05) 0.93 351 22.4 (7.3) 0.02 (0.08) 0.84 385 19.0 (7.0) -0.02 (0.08) 0.77
    AA 255 20.1 (6.4) -0.14 (0.07) 0.06 134 21.9 (6.1) -0.10 (0.10) 0.35 121 18.2 (6.1) -0.19 (0.11) 0.09
Per copy of A allele -0.06 (0.04) ptrd=0.11 -0.04 (0.05) ptrd=0.46 -0.08 (0.05) ptrd=0.14 0.66
Open in a new tab
1

Using standardized 25(OH)D levels as the dependent variable; multiple regression models adjusted for age at blood draw (continuous), body mass index (BMI, at blood draw if available; if not, BMI at interview for controls and BMI in the reference year for cases, continuous), season of blood draw (4 categories), and genetic ancestry (continuous).

2

Interaction between the genotypes and case-control status.

Table 5.

Associations Between Circulating 25(OH)D Levels and Genotypes in non-Hispanic White Cases and Controls from the San Francisco Bay Area Breast Cancer Study

Cases and Controls Combined Controls Cases
Genotypes N 25(OH)D (ng/mL) Mean (SD) Adjusted beta (SE)1 Adjusted P value1 N 25(OH)D (ng/mL) Mean (SD) Adjuste beta (SE)1 Adjuste P value1 N 25(OH)D (ng/mL) Mean (SD) Adjuste beta (SE)1 Adjuste P value1 P interaction 2
rs7041
    GG 115 24.5(10.5) 42 26.8 (11.0) 73 23.2 (10.1)
    GT 177 22.5(8.4) -0.27(0.12) 0.02 61 25.0 (10.1) -0.22 (0.19) 0.24 116 21.2 (7.1) -0.29 (0.15) 0.05
    TT 62 23.1(8.5) -0.30(0.15) 0.05 27 27.5 (7.6) -0.15 (0.25) 0.56 35 19.8 (7.7) -0.42 (0.20) 0.04
Per copy of T allele -0.17 (0.07 ptrd=0.02 -0.09 (0.12) ptrd=0.45 -0.23 (0.10) ptrd=0.02 0.32
rs2282679
    AA 182 23.9(9.9) 62 26.4 (10.6) 120 22.6 (9.4)
    AC 144 22.6(8.5) -0.28(0.11) 0.01 57 26.1 (9.6) -0.22 (0.18) 0.22 87 20.3 (6.9) -0.32 (0.14) 0.02
    CC 28 22.6(7.0) -0.24(0.20) 0.22 11 23.9 (7.8) -0.58 (0.32) 0.07 17 21.8 (6.5) -0.06 (0.25) 0.81
Per copy of C allele -0.19(0.08) ptrd=0.02 -0.26 (0.13) ptrd=0.06 -0.16 (0.10) ptrd=0.12 0.62
rs12785878
    TT 185 23.2 (9.0) 62 26.0 (9.5) 123 21.8 (8.4)
    TG 145 23.0 (9.1) -0.06 (0.11) 0.58 55 25.0 (9.7) -0.14 (0.18) 0.42 90 21.8 (8.5) 0.003 (0.14) 0.98
    GG 24 25.0 (11.2) 0.02 (0.21) 0.92 13 30.7 (11.8) 0.56 (0.30) 0.06 11 18.3 (5.5) -0.49 (0.31) 0.11
Per copy of G allele -0.02 (0.08) ptrd=0.77 0.11 (0.13) ptrd=0.39 -0.10 (0.11) ptrd=0.36 0.17
rs1790349
    AA 263 23.4 (9.2) 95 26.0 (10.0) 168 21.9 (8.3)
    AG 82 22.8 (9.4) -0.04 (0.12) 0.72 30 26.0 (9.9) 0.07 (0.21) 0.75 52 20.9 (8.6) -0.11 (0.16) 0.48
    GG 9 24.8 (8.3) -0.01 (0.33) 0.98 5 28.1 (9.5) 0.16 (0.44) 0.72 4 20.6 (4.6) -0.21 (0.50) 0.68
Per copy of G allele -0.03 (0.10) ptrd=0.76 0.07 (0.16) ptrd=0.65 -0.11 (0.14) ptrd=0.43 0.37
rs2060793
    GG 140 22.5 (9.1) 51 26.9 (10.2) 89 19.9 (7.3)
    GA 162 23.3 (8.8) 0.13 (0.11) 0.22 58 24.8 (9.0) -0.17 (0.18) 0.62 104 22.5 (8.7) 0.30 (0.14) 0.03
    AA 52 25.3 (10.2) 0.38 (0.16) 0.02 21 27.7 (11.6) 0.12 (0.25) 0.34 31 23.7 (9.0) 0.52 (0.21) 0.01
Per copy of A allele 0.18 (0.07) ptrd=0.02 0.01 (0.12) ptrd=0.91 0.27 (0.10) ptrd=0.005 0.09
rs6013897
    TT 227 23.4 (9.3) 85 25.7 (10.1) 142 22.1 (8.5)
    TA 112 23.2 (8.8) 0.001 (0.11) 0.99 36 28.8 (7.8) 0.23 (0.19) 0.23 76 20.5 (8.1) -0.12 (0.14) 0.39
    AA 15 21.9 (10.7) -0.21 (0.26) 0.42 9 19.2 (12.7) -0.50 (0.33) 0.14 6 26.0 (5.2) 0.45 (0.42) 0.29
Per copy of A allele -0.04 (0.09) ptrd=0.63 -0.05 (0.14) ptrd=0.74 -0.02 (0.13) ptrd=0.85 0.92
Open in a new tab
1

Using standardized 25(OH)D levels as the dependent variable; multiple regression models adjusted for age at blood draw (continuous), body mass index (BMI, at blood draw if available; if not, BMI at interview for controls and BMI in the reference year for cases, continuous), season of blood draw (4 categories), and genetic ancestry (continuous).

2

Interaction between the genotypes and case-control status.

We did not observe significant associations between 25(OH)D levels and SNPs rs12785878 and rs1790349 in the NADSYN1/DCRH7 gene region or SNP rs6013897 near CYP24A1, for either Hispanic women (Table 4) or NHW women (Table 5).

The two GC SNPs were in weak LD in Hispanics (r2=0.24) and remained statistically significant after mutual adjustment (β=-0.19 for the rs7041 TT genotype, p for trend=0.02; β=-0.44 for the rs2282679 CC genotype, p for trend<10-4), suggesting independent associations with 25(OH)D levels for the two SNPs. In NHW women, a stronger LD (r2=0.50) between the two SNPs and a much smaller sample size precluded the examination of the independent associations for the two SNPs. The GC haplotype analysis did not provide better prediction than the single SNP analysis in either Hispanics or NHWs (data not shown).

The two SNPs in the NADSYN1/DHCR7 region had r2=0.14 and 0.19 in Hispanic and NHW women, respectively. Compared to the most common rs12785878T_rs1790349A haplotype, the rare rs12785878T_rs1790349G haplotype was significantly associated with 25(OH)D levels in Hispanic, but not NHW, control women. The difference in associations between Hispanic and NHW controls was statistically significant (p=0.02). Associations did not differ significantly by case-control status (Table 6).

Table 6.

Associations Between Circulating 25(OH)D Levels and Haplotypes in the NADSYN1/DHCR7 Region, by Ethnicity and Case-control Status

Cases and Controls Combined Controls Cases
Ethnicity Haplotype Frequency Adjusted beta (SE)1 Adjusted P value1 Frequency Adjuste beta (SE)1 Adjuste P value1 Frequency Adjuste beta (SE)1 Adjuste P value1 P interaction 2
Hispanic T_A 0.47 ref 0.46 ref 0.48
G_A 0.36 -0.02 (0.04) 0.58 0.36 -0.03 (0.05) 0.52 0.35 -0.004 (0.06) 0.95 0.37
G_G 0.17 -0.08 (0.05) 0.1 0.18 -0.04 (0.07) 0.53 0.17 -0.13 (0.08) 0.09 0.64
T_G 0.001 -1.41 (0.70) 0.05 0.002 -1.52 (0.70) 0.03 0.0004 7.76 (23.72) 0.74 0.71
NHW T_A 0.72 0.68 ref 0.74
G_A 0.14 0.01 (0.11) 0.90 0.17 0.13 (0.15) 0.38 0.13 -0.06 (0.15) 0.69 0.39
G_G 0.13 -0.07 (0.11) 0.55 0.14 0.05 (0.18) 0.80 0.12 -0.14 (0.15) 0.33 0.30
T_G 0.01 0.34 (0.37) 0.36 0.009 0.69 (0.70) 0.33 0.01 0.10 (0.45) 0.83 0.48
Open in a new tab
1

Using standardized 25(OH)D levels as the dependent variable; multiple regression models adjusted for age at blood draw (continuous), body mass index (BMI, at blood draw if available; if not, BMI at interview for controls and BMI in the reference year for cases, continuous), season of blood draw (4 categories), and genetic ancestry (continuous).

2

Interaction between the haplotypes and case-control status.

No significant difference by self-reported ethnicity was observed for the associations between 25(OH)D levels and any of the six SNPs, nor did the associations differ significantly by estimated genetic ancestry (data not shown).

Hispanics and NHWs were combined in the analyses of gene-gene interactions to increase statistical power. No significant interactions between the GC SNPs and the CYP2R1 rs2060793 were found, nor did they interact significantly with rs6013897 near CYP24A1 or the SNPs in the NADSYN1/DHCR7 region (data not shown).

Non-genetic predictors of circulating 25(OH)D levels

Associations between 25(OH)D levels and potential non-genetic predictors in women from the SFBCS are shown in Table 7. Various summary measures of residential UV exposure were examined individually and all provided similar model prediction [based on Aikaike Information Criterion (AIC)]. Mean UV levels within a 20 km radius of the residence was included in the final model. Models with BMI or weight also had similar AIC scores and therefore only associations with BMI are presented. Among cases and controls combined, we observed a significant negative association for BMI (p for trend=0.001), and positive associations for total vitamin D intake (p=0.001) and recent outdoor activities (p=0.001). Results were similar in the analysis of standardized 25(OH) levels (data not shown). Associations did not differ significantly between cases and controls, although the association for outdoor activities was significant only in controls. In cases, neither stage at diagnosis or time interval between diagnosis and blood collection were significant predictors of 25(OH)D levels. Adding them in the model did not result in discernible changes in the associations for other predictors (data not shown). Lighter constitutive skin pigmentation was significantly associated with higher 25(OH)D levels in Hispanic controls, but lower levels in NHW controls. Constitutive skin pigmentation was not a significant predictor of standardized 25(OH)D levels (data not shown).

Table 7.

Associations Between Circulating 25(OH)D Levels and Non-genetic Factors in Cases and Controls from the San Francisco Bay Area Breast Cancer Study

Cases and Controls Combined Controls Cases
Exposure N 25(OH)D (ng/mL) Mean (SD) Adjusted beta1,2 Adjusted P value1,2 N 25(OH)D (ng/mL) Mean (SD) Adjuste Beta1 Adjuste P value1 N 25(OH)D (ng/mL) Mean (SD) Adjuste Beta3 Adjuste P value3 P interaction 4
Body mass index (kg/m2)
    <25 244 23.7 (9.6) 103 25.8 (10.2) 141 22.2 (8.8)
    25-29.9 254 20.3 (8.1) -0.09 (0.04) 0.02 116 21.7 (8.7) -0.04 (0.06) 0.58 138 19.2 (7.4) -0.10 (0.05) 0.04
    ≥30 278 17.9 (7.8) -0.18 (0.04) <10-4 119 18.5 (8.3) -0.18 (0.06) 0.003 159 17.5 (7.4) -0.17 (0.05) 2x10-4
ptrd<10-4 ptrd=0.002 ptrd=2x10-4 0.57
Total vitamin D intake from diet and supplements (IU/day)
    <215 242 18.7 (8.6) 106 19.1 (8.9) 136 18.4 (8.3)
    215-499.9 236 19.7 (8.2) 0.05 (0.04) 0.15 99 21.4 (9.5) 0.15 (0.06) 0.01 137 18.4 (6.9) -0.003 (0.05) 0.96
    ≥500 298 22.7 (9.0) 0.14 (0.04) 5x10-4 133 24.2 (9.4) 0.17 (0.06) 0.003 165 21.4 (8.6) 0.12 (0.05) 0.01
ptrd=0.001 ptrd=0.04 ptrd=0.01 0.43
Residential UV exposure (Wh/m2)5
    <4757. 270 20.8 (9.3) 135 22.3 (10.4) 135 19.3 (7.7)
    4757.1-4827.9 245 21.2 (8.6) -0.01 (0.04) 0.74 85 22.8 (9.7) -0.04 (0.06) 0.50 160 20.3 (7.8) 0.01 (0.05) 0.78
    ≥4828.0 261 19.6 (8.4) 0.001 (0.04) 0.97 118 20.5 (8.0) -0.003 (0.06) 0.96 143 18.9 (8.8) 0.02 (0.05) 0.65
ptrd=0.98 ptrd=0.93 ptrd=0.65 0.40
Recent outdoor activities (hrs/wk)
    0 211 18.0 (8.0) 94 18.4 (8.9) 117 17.7 (7.2)
    0.1-3 156 20.2 (8.8) 0.07 (0.04) 0.09 53 22.2 (10.2) 0.11 (0.08) 0.14 103 19.1 (7.8) 0.03 (0.05) 0.63
    ≥3 409 22.0 (8.9) 0.12 (0.04) 0.001 191 23.4 (9.1) 0.19 (0.06) 8x10-4 218 20.7 (8.5) 0.07 (0.05) 0.14
ptrd=0.001 ptrd=0.001 ptrd=0.13 0.09
Constitutive skin pigmentation (NHWs only)
    Dark 46 24.6 (9.7) 13 34.4 (9.6) 33 20.7 (6.5)
    Intermediate 103 23.7 (8.2) -0.06 (0.07) 0.39 39 25.3 (8.7) -0.34 (0.13) 0.01 64 22.6 (7.8) 0.06 (0.08) 0.43
    Light 190 22.7 (9.5) -0.09 (0.06) 0.17 72 24.8 (9.6) -0.36 (0.12) 0.002 118 21.4 (9.2) 0.01 (0.07) 0.86
ptrd=0.17 ptrd=0.01 ptrd=0.85 0.11
Constitutive skin pigmentation (Hispanics only)
    Dark 213 17.7 (7.9) 104 18.4 (8.7) 109 17.1 (7.0)
    Intermediate 156 18.6 (7.3) 0.07 (0.05) 0.13 74 19.8 (7.4) 0.10 (0.07) 0.15 82 17.4 (7.1) 0.04 (0.06) 0.54
    Light 68 20.3 (9.1) 0.14 (0.06) 0.02 36 21.4 (9.8) 0.16 (0.09) 0.07 32 18.9 (8.1) 0.11 (0.09) 0.20
ptrd=0.01 ptrd=0.04 ptrd=0.21 0.71
Open in a new tab
1

Plasma 25(OH)D levels on the original scale were analyzed using a generalized linear model with natural log link and a gamma distribution, adjusting for season and age at blood draw, ethnicity, in addition to mutual adjustment for covariates listed in the table; interaction between ethnicity and constitutive skin pigmentation was also adjusted.

2

With further adjustment for case-control status.

3

Additional adjustment for stage and time since diagnosis did not alter the results.

4

Interaction between the exposure of interest and case-control status.

5

Within a 20 km radius of the residence at diagnosis (cases) or selection into study (controls).

Overall, the significant non-genetic factors (BMI, total vitamin D intake, outdoor activity, and constitutive skin pigmentation) accounted for 15% of the variability in circulating 25(OH)D levels (22% in controls and 13% in cases), after adjusting for season and age at blood draw. Each of the three significant SNPs accounted for approximately an additional 1% of the variability

Statistical interactions between SNPs and non-genetic predictors of circulating 25(OH)D levels

Statistically significant interactions were observed between GC SNP rs2282679 and BMI (p for interaction=0.02, Table 8). Specifically, 25(OH)D level was inversely associated with the number of rs2282679_C alleles only in women who had a BMI <30 kg/m2. Significant interaction was also found between NADSYN1/DHCR7 SNP rs12785878 and outdoor activity (p for interaction=0.005). The rs12785878_G allele was associated with increased 25(OH)D levels in women who spent less than 2 hours per week in outdoor activities, but reduced 25-OH)D levels in those who spent more time in outdoor activities. No statistically significant interaction was observed between any of the SNPs and total vitamin D intake (data not shown).

Table 8.

Interactions Between Genotypes and Non-genetic Factors in Cases and Controls from the San Francisco Bay Area Breast Cancer Study

Genotypes N 25(OH)D (ng/mL) Mean (SD) Adjusted beta1 Adjusted P value1
P interaction 2
rs2282679
Body mass index <30 kg/m2 AA 348 22.3 (9.4)
AC 246 20.8 (8.1) -0.29 (0.08) 0.0006
CC 34 20.3 (8.3) -0.45 (0.18) 0.01
ptrd<10-4
Body mass index ≥30 kg/m2 AA 199 17.5 (7.7)
AC 130 17.3 (7.4) -0.02 (0.10) 0.83
CC 26 16.0 (6.1) -0.11 (0.19) 0.55
ptrd=0.59 0.02
rs12785878
Outdoor activities <2 hours/week3 TT 102 18.6 (7.6)
TG 140 18.5 (8.8) 0.16 (0.13) 0.22
GG 51 19.7 (9.1) 0.39 (0.17) 0.02
ptrd=0.03
Outdoor activities ≥2 hours/week3 TT 198 23.0 (8.7)
TG 225 21.7 (8.8) -0.05 (0.10) 0.57
GG 77 18.2 (8.5) -0.31 (0.14) 0.03
Ptrd=0.05 0.005
Open in a new tab
1

Using standardized 25(OH)D levels as the dependent variable; multiple regression models adjusted for age at blood draw (continuous), body mass index (BMI, at blood draw if available; if not, BMI at interview for controls and BMI in the reference year for cases, continuous), season of blood draw (4 categories), and genetic ancestry (continuous).

2

Interaction between the genotypes and non-genetic factors.

3

Outdoor activities within the 6 months prior to blood collection.

Discussion

In this study of 1,605 Hispanic and 354 NHW women, several common SNPs identified by GWAS as predictors of 25(OH)D levels in populations of European descent were also significantly associated with 25(OH)D levels in Hispanics. Specifically, the minor alleles of rs7041 and rs2282679 in the GC gene were associated with lower levels of 25(OH)D, whereas the minor allele of the CYP2R1 rs2060793 was associated with higher 25(OH)D levels. Associations did not differ significantly between Hispanics and NHWs. In Hispanic controls, circulating 25(OH)D levels were significantly associated with the haplotypes of SNPs rs12785878 and rs1790349 in the NADSYN1/DHCR7 region. Our study confirms associations between 25(OH)D levels and known non-genetic predictors, including BMI, total vitamin D intake, outdoor activity, and constituent skin pigmentation. Statistically significant interactions were observed between the GC rs2282679 SNP and BMI, and between NADSYN1/DHCR7 SNP rs12785878 and outdoor activity.

GC is the major binding protein of vitamin D metabolites in the circulation and responsible for their transportation to target tissue [35]. In the GC gene, two non-synonymous SNPs, rs7041 (Glu416Asp) and rs4588 (Thr420Lys, which we did not examine) have been associated with circulating 25(OH)D levels [13]. In experimental studies, these two SNPs have been linked with altered binding affinity [36] and metabolic turnover of the GC protein [37]. However, in two GWAS [8,9], a non-coding SNP, rs228679, was an even stronger predictor of 25(OH)D levels, compared to SNP rs7041. It has been speculated that the association of rs228679 with 25(OH)D may be due to its tight LD with rs4588 (r2=1.0 in HapMap-CEU panel and 0.7 in HapMap-MEX panel) [16,18]. The exact mechanism linking changes in GC affinity and circulating levels to a change in circulating 25(OH)D levels is not yet established, although an influence on 25(OH)D clearance from the circulation, reabsorption in renal tubules, and/or further metabolism into 1,25-(OH)2 have been suggested [8,13,38].

Our finding of a statistically significant interaction between the GC SNP rs2282679 and BMI is intriguing. Vitamin D is fat soluble and studies in animals and humans have shown that adipose tissue is the major storage site of 25(OH)D [39,40]. A strong inverse association between BMI or body fat content and circulating 25(OH)D levels has been repeatedly observed [41-43], consistent with our findings. It is biologically plausible that variants in the GC gene that affect the efficiency of transporting vitamin D to adipose tissue interact with the overall body fat content to influence circulating vitamin D levels. The same interaction with BMI was not observed for the GC SNP rs7401. Different effects of the two SNPs (or the true functional variants they are in strong LD with) on the GC protein could be a potential explanation.

Although a previous small study suggested that vitamin D supplementation may result in differential changes in serum 25(OH)D levels according to the rs4588 genotypes [44], we observed no significant interactions between rs2282679 (in higher LD with rs4588) and self-reported vitamin D intake.

Overall, the role of the GC protein as the major transporter of vitamin D in circulation and the repeatedly observed associations between common GC SNPs and 25(OH)D levels across diverse populations, which include NHW ([8,9,14,15] and references in [13]), African-Americans [19], Asians [16,18], and Hispanics (in the present study and in [19]), provide compelling evidence for the contribution of the GC gene to individual variability of circulating 25(OH)D levels.

Compared to the GC gene, CYP2R1, has been less well studied as a potential contributor to variation of 25(OH)D levels [13,15,45-47]. It encodes 25-hydroxylase which catalyzes the conversion of vitamin D to 25(OH)D. Experimental data on the functional impact of the common polymorphisms in this gene are still lacking. Consistent with our results, SNP rs206793, located around 2 kilobases up-stream of the transcription start site, was associated with 25(OH)D levels in the GWAS by Ahn et al. [9] and in a candidate gene study of primarily NHW participants [15]. However, in the Chinese study by Lu et al. [18], this SNP only showed a marginally significant association in subgroup analysis. Several SNPs in or near CYP2R1 that are in various degrees of LD with rs206793 have also been associated with 25(OH)D levels in some other studies [14,46,48]. Taken together, these findings suggest that multiple yet-to-identified “causal” variants may exist in the CYP2R1 gene or the chromosomal region.

DHCR7 catalyzes the UV radiation-induced synthesis of vitamin D3 (cholecalciferol) from 7-dehydroxy cholesterol in the skin after sun exposure. SNPs in NADSYN1/DHCR7 region were first associated with 25(OH)D levels in the two GWAS [8,9] and later replicated in Chinese by Lu et al. [18]. In our study, SNPs rs12785878 (an intergenic SNP in the region) and rs1790349 (an intronic SNP in DHCR7) were not significantly associated with 25(OH)D levels. However, a statistically significant association was observed for the rs12785878T_rs1790349G haplotype in Hispanic controls, supporting a role of these SNPs as markers of functional variants in this region. Although the exact reason is unclear, the lack of replication in NHW women could be potentially explained by different LD structure and a smaller sample size in this group. Interestingly, in the controls and cases from the SFBCS, we observed a significant interaction between the rs12785878 SNP and time spent in outdoor activities, a measure of sunlight exposure which is required for cutaneous vitamin D synthesis. Given the role of DHCR7 in modulating the availability of 7-dehydrocholesterol (provitamin-D3), in the skin, it is reasonable to postulate that functional variants in DHCR7 may have different effects on vitamin D synthesis in people with different sunlight exposure levels. However, because the rs12785878 genotype frequencies in the Mexican control group showed a significant departure from the HWE, our results for this SNP should be interpreted with caution.

The main strengths of this study include the large sample size of Hispanic women, and the availability of measures of known environmental determinants of 25(OH)D levels, which allowed us to examine important gene-environment interactions. One limitation of our study is that circulating 25(OH)D was measured with different assays between the MBCS and SFBCS, and between cases and controls in the SFBCS. To minimize the variability introduced by the different measurement methods, we standardized 25(OH)D levels according to group means and standard deviations and used this variable in the regression analysis. A caveat of this approach is that we could not evaluate the effects of factors such as ethnicity, skin pigmentation and disease status on 25(OH)D levels and their interactions with the SNPs in the overall study population. Nevertheless, similar conclusions regarding the SNPs and 25(OH)D associations could be drawn from the pooled analysis on standardized 25(OH)D levels and from the stratified analysis on the non-standardized 25(OH)D levels.

Our study provides further support for the contributions of common polymorphisms in the vitamin D metabolic pathway to variability in circulating 25(OH)D levels and risk of vitamin D deficiency among individuals. However, in contrast to known non-genetic predictors, these common variants accounted for less than 4% of the inter-individual variability in 25(OH)D levels. Additional variability may be explained by rare variants in these genomic regions, by variants in genes that regulate the vitamin D pathway, and by interactions between genetic and non-genetic factors.

Acknowledgements

This work was supported by the Breast Cancer Health Disparities Study funded by grant CA14002 from the National Cancer Institute to Dr. Slattery, and by grant 161B-0039 to Dr. Wang from the California Breast Cancer Research Program. The San Francisco Bay Area Breast Cancer Study was supported by grants CA63446 and CA77305 from the National Cancer Institute, grant DAMD17-96-1-6071 from the U.S. Department of Defense and grant 7PB-0068 from the California Breast Cancer Research Program. The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201000036C awarded to the Cancer Prevention Institute of California; and the Centers for Disease Control and Prevention’s National Program of Cancer Registries, under agreement #1U58 DP000807-01 awarded to the Public Health Institute. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute or endorsement by the State of California Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors. The Mexico Breast Cancer Study was funded by Consejo Nacional de Ciencia y Tecnología (CONACyT) (SALUD-2002-C01-7462). We would also like to acknowledge the contributions of Dr. Myles Cockburn for providing the solar UV data; Jennifer Herrick and Sandra Edwards for data harmonization oversight and management; Erica Wolff and Michael Hoffman for laboratory support; Carolyn Ortega for data management for the Mexico Breast Cancer Study; and Jocelyn Koo for data management for the San Francisco Bay Area Breast Cancer Study.

Abbreviations

AIMs

ancestral informative markers

BMI

body mass index

CV

coefficients of variation

FDR

false discovery rate

HWE

Hardy-Weinberg equilibrium

GWAS

genome-wide association studies

LD

linkage disequilibrium

MAF

minor allele frequency

MBCS

Mexican Breast Cancer Study

NHW

non-Hispanic White

SD

standard deviation

SFBCS

San Francisco Bay Area Breast Cancer Study

SNP

single nucleotide polymorphisms

25(OH)D

25-hydroxyvitamin D

References

  • 1.Holick MF. Vitamin D deficiency in 2010: health benefits of vitamin D and sunlight: a D-bate. Nat Rev Endocrinol. 2011;7:73–75. doi: 10.1038/nrendo.2010.234. [DOI] [PubMed] [Google Scholar]
  • 2.Byers T. Anticancer vitamins du Jour-The ABCED’s so far. Am J Epidemiol. 2010;172:1–3. doi: 10.1093/aje/kwq112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Rosen CJ, Adams JS, Bikle DD, Black DM, Demay MB, Manson JE, Murad MH, Kovacs CS. The nonskeletal effects of vitamin d: an endocrine society scientific statement. Endocr Rev. 2012;33:456–492. doi: 10.1210/er.2012-1000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Holick MF, Chen TC. Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr. 2008;87:1080S–1086S. doi: 10.1093/ajcn/87.4.1080S. [DOI] [PubMed] [Google Scholar]
  • 5.Shea MK, Benjamin EJ, Dupuis J, Massaro JM, Jacques PF, D’Agostino RB Sr, Ordovas JM, O’Donnell CJ, Dawson-Hughes B, Vasan RS, Booth SL. Genetic and non-genetic correlates of vitamins K and D. Eur J Clin Nutr. 2009;63:458–464. doi: 10.1038/sj.ejcn.1602959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hunter D, De Lange M, Snieder H, MacGregor AJ, Swaminathan R, Thakker RV, Spector TD. Genetic contribution to bone metabolism, calcium excretion, and vitamin D and parathyroid hormone regulation. J Bone Miner Res. 2001;16:371–378. doi: 10.1359/jbmr.2001.16.2.371. [DOI] [PubMed] [Google Scholar]
  • 7.Snellman G, Melhus H, Gedeborg R, Olofsson S, Wolk A, Pedersen NL, Michaelsson K. Seasonal genetic influence on serum 25-hydroxyvitamin D levels: a twin study. PLoS One. 2009;4:e7747. doi: 10.1371/journal.pone.0007747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wang TJ, Zhang F, Richards JB, Kestenbaum B, van Meurs JB, Berry D, Kiel DP, Streeten EA, Ohlsson C, Koller DL, Peltonen L, Cooper JD, O’Reilly PF, Houston DK, Glazer NL, Vandenput L, Peacock M, Shi J, Rivadeneira F, McCarthy MI, Anneli P, de Boer IH, Mangino M, Kato B, Smyth DJ, Booth SL, Jacques PF, Burke GL, Goodarzi M, Cheung CL, Wolf M, Rice K, Goltzman D, Hidiroglou N, Ladouceur M, Wareham NJ, Hocking LJ, Hart D, Arden NK, Cooper C, Malik S, Fraser WD, Hartikainen AL, Zhai G, Macdonald HM, Forouhi NG, Loos RJ, Reid DM, Hakim A, Dennison E, Liu Y, Power C, Stevens HE, Jaana L, Vasan RS, Soranzo N, Bojunga J, Psaty BM, Lorentzon M, Foroud T, Harris TB, Hofman A, Jansson JO, Cauley JA, Uitterlinden AG, Gibson Q, Jarvelin MR, Karasik D, Siscovick DS, Econs MJ, Kritchevsky SB, Florez JC, Todd JA, Dupuis J, Hypponen E, Spector TD. Common genetic determinants of vitamin D insufficiency: a genome-wide association study. Lancet. 2010;376:180–188. doi: 10.1016/S0140-6736(10)60588-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ahn J, Yu K, Stolzenberg-Solomon R, Simon KC, McCullough ML, Gallicchio L, Jacobs EJ, Ascherio A, Helzlsouer K, Jacobs KB, Li Q, Weinstein SJ, Purdue M, Virtamo J, Horst R, Wheeler W, Chanock S, Hunter DJ, Hayes RB, Kraft P, Albanes D. Genome-wide association study of circulating vitamin D levels. Hum Mol Genet. 2010;19:2739–2745. doi: 10.1093/hmg/ddq155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Engelman CD, Meyers KJ, Ziegler JT, Taylor KD, Palmer ND, Haffner SM, Fingerlin TE, Wagenknecht LE, Rotter JI, Bowden DW, Langefeld CD, Norris JM. Genome-wide association study of vitamin D concentrations in Hispanic Americans: the IRAS family study. J Steroid Biochem Mol Biol. 2010;122:186–192. doi: 10.1016/j.jsbmb.2010.06.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Thomas DC, Haile RW, Duggan D. Recent developments in genomewide association scans: a workshop summary and review. Am J Hum Genet. 2005;77:337–345. doi: 10.1086/432962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.McCarthy MI, Abecasis GR, Cardon LR, Goldstein DB, Little J, Ioannidis JP, Hirschhorn JN. Genome-wide association studies for complex traits: consensus, uncertainty and challenges. Nat Rev Genet. 2008;9:356–369. doi: 10.1038/nrg2344. [DOI] [PubMed] [Google Scholar]
  • 13.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. doi: 10.1016/j.jsbmb.2010.03.073. [DOI] [PubMed] [Google Scholar]
  • 14.Bu FX, Armas L, Lappe J, Zhou Y, Gao G, Wang HW, Recker R, Zhao LJ. Comprehensive association analysis of nine candidate genes with serum 25-hydroxy vitamin D levels among healthy Caucasian subjects. Hum Genet. 2010;128:549–556. doi: 10.1007/s00439-010-0881-9. [DOI] [PubMed] [Google Scholar]
  • 15.Shui IM, Mucci LA, Kraft P, Tamimi RM, Lindstrom S, Penney KL, Nimptsch K, Hollis BW, Dupre N, Platz EA, Stampfer MJ, Giovannucci E. Vitamin D-related genetic variation, plasma vitamin D, and risk of lethal prostate cancer: a prospective nested case-control study. J Natl Cancer Inst. 2012;104:690–699. doi: 10.1093/jnci/djs189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Gozdzik A, Zhu J, Wong BY, Fu L, Cole DE, Parra EJ. Association of vitamin D binding protein (VDBP) polymorphisms and serum 25(OH)D concentrations in a sample of young Canadian adults of different ancestry. J Steroid Biochem Mol Biol. 2011;127:405–412. doi: 10.1016/j.jsbmb.2011.05.009. [DOI] [PubMed] [Google Scholar]
  • 17.Signorello LB, Shi J, Cai Q, Zheng W, Williams SM, Long J, Cohen SS, Li G, Hollis BW, Smith JR, Blot WJ. Common variation in vitamin D pathway genes predicts circulating 25-hydroxyvitamin D Levels among African Americans. PLoS One. 2011;6:e28623. doi: 10.1371/journal.pone.0028623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lu L, Sheng H, Li H, Gan W, Liu C, Zhu J, Loos RJ, Lin X. Associations between common variants in GC and DHCR7/NADSYN1 and vitamin D concentration in Chinese Hans. Hum Genet. 2012;131:505–512. doi: 10.1007/s00439-011-1099-1. [DOI] [PubMed] [Google Scholar]
  • 19.Engelman CD, Fingerlin TE, Langefeld CD, Hicks PJ, Rich SS, Wagenknecht LE, Bowden DW, Norris JM. Genetic and environmental determinants of 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D levels in Hispanic and African Americans. J Clin Endocrinol Metab. 2008;93:3381–3388. doi: 10.1210/jc.2007-2702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Yetley EA. Assessing the vitamin D status of the US population. Am J Clin Nutr. 2008;88:558S–564S. doi: 10.1093/ajcn/88.2.558S. [DOI] [PubMed] [Google Scholar]
  • 21.Slattery M, John E, Torres-Mejia G, Lundgreen A, Herrick J, Baumgartner K, Hines L, Stern M, Wolff R. Genetic variation in genes involved in hormones, inflammation, and energetic factors and breast cancer risk in an admixed population. Carcinogenesis. 2012 Aug;33:1512–21. doi: 10.1093/carcin/bgs163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.John EM, Phipps AI, Davis A, Koo J. Migration history, acculturation, and breast cancer risk in Hispanic women. Cancer Epidemiol Biomarkers Prev. 2005;14:2905–2913. doi: 10.1158/1055-9965.EPI-05-0483. [DOI] [PubMed] [Google Scholar]
  • 23.John EM, Schwartz GG, Koo J, Wang W, Ingles SA. Sun exposure, vitamin D receptor gene polymorphisms, and breast cancer risk in a multiethnic population. Am J Epidemiol. 2007;166:1409–1419. doi: 10.1093/aje/kwm259. [DOI] [PubMed] [Google Scholar]
  • 24.Angeles-Llerenas A, Ortega-Olvera C, Perez-Rodriguez E, Esparza-Cano JP, Lazcano-Ponce E, Romieu I, Torres-Mejia G. Moderate physical activity and breast cancer risk: the effect of menopausal status. Cancer Causes Control. 2010;21:577–586. doi: 10.1007/s10552-009-9487-8. [DOI] [PubMed] [Google Scholar]
  • 25.Fejerman L, Romieu I, John EM, Lazcano-Ponce E, Huntsman S, Beckman KB, Perez-Stable EJ, Gonzalez Burchard E, Ziv E, Torres-Mejia G. European ancestry is positively associated with breast cancer risk in Mexican women. Cancer Epidemiol Biomarkers Prev. 2010;19:1074–1082. doi: 10.1158/1055-9965.EPI-09-1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wang J, John EM, Horn-Ross PL, Ingles SA. Dietary fat, cooking fat, and breast cancer risk in a multiethnic population. Nutr Cancer. 2008;60:492–504. doi: 10.1080/01635580801956485. [DOI] [PubMed] [Google Scholar]
  • 27.Horn-Ross PL, John EM, Lee M, Stewart SL, Koo J, Sakoda LC, Shiau AC, Goldstein J, Davis P, Perez-Stable EJ. Phytoestrogen consumption and breast cancer risk in a multiethnic population: the Bay Area Breast Cancer Study. Am J Epidemiol. 2001;154:434–441. doi: 10.1093/aje/154.5.434. [DOI] [PubMed] [Google Scholar]
  • 28.Tatalovich Z, Wilson JP, Mack T, Yan Y, Cockburn M. The objective assessment of lifetime cumulative ultraviolet exposure for determining melanoma risk. J Photochem Photobiol B. 2006;85:198–204. doi: 10.1016/j.jphotobiol.2006.08.002. [DOI] [PubMed] [Google Scholar]
  • 29.Cauley JA, Lacroix AZ, Wu L, Horwitz M, Danielson ME, Bauer DC, Lee JS, Jackson RD, Robbins JA, Wu C, Stanczyk FZ, LeBoff MS, Wactawski-Wende J, Sarto G, Ockene J, Cummings SR. Serum 25-hydroxyvitamin D concentrations and risk for hip fractures. Ann Intern Med. 2008;149:242–250. doi: 10.7326/0003-4819-149-4-200808190-00005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Genetics. 2005 Oct;171:783–90. doi: 10.1534/genetics.104.036699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Falush D, Stephens M, Pritchard JK. Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics. 2003;164:1567–1587. doi: 10.1093/genetics/164.4.1567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–959. doi: 10.1093/genetics/155.2.945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Stram DO, Leigh Pearce C, Bretsky P, Freedman M, Hirschhorn JN, Altshuler D, Kolonel LN, Henderson BE, Thomas DC. Modeling and E-M estimation of haplotype-specific relative risks from genotype data for a case-control study of unrelated individuals. Hum Hered. 2003;55:179–190. doi: 10.1159/000073202. [DOI] [PubMed] [Google Scholar]
  • 34.Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357:266–281. doi: 10.1056/NEJMra070553. [DOI] [PubMed] [Google Scholar]
  • 35.Speeckaert M, Huang G, Delanghe JR, Taes YE. Biological and clinical aspects of the vitamin D binding protein (Gc-globulin) and its polymorphism. Clin Chim Acta. 2006;372:33–42. doi: 10.1016/j.cca.2006.03.011. [DOI] [PubMed] [Google Scholar]
  • 36.Arnaud J, Constans J. Affinity differences for vitamin D metabolites associated with the genetic isoforms of the human serum carrier protein (DBP) Hum Genet. 1993;92:183–188. doi: 10.1007/BF00219689. [DOI] [PubMed] [Google Scholar]
  • 37.Lauridsen AL, Vestergaard P, Nexo E. Mean serum concentration of vitamin D-binding protein (Gc globulin) is related to the Gc phenotype in women. Clin Chem. 2001;47:753–756. [PubMed] [Google Scholar]
  • 38.Weinstein SJ, Stolzenberg-Solomon RZ, Kopp W, Rager H, Virtamo J, Albanes D. Impact of Circulating Vitamin D Binding Protein Levels on the Association between 25-Hydroxyvitamin D and Pancreatic Cancer Risk: A Nested Case-Control Study. Cancer Res. 2012;72:1190–1198. doi: 10.1158/0008-5472.CAN-11-2950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Rosenstreich SJ, Rich C, Volwiler W. Deposition in and release of vitamin D3 from body fat: evidence for a storage site in the rat. J Clin Invest. 1971;50:679–687. doi: 10.1172/JCI106538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Mawer EB, Backhouse J, Holman CA, Lumb GA, Stanbury SW. The distribution and storage of vitamin D and its metabolites in human tissues. Clin Sci. 1972;43:413–431. doi: 10.1042/cs0430413. [DOI] [PubMed] [Google Scholar]
  • 41.Arunabh S, Pollack S, Yeh J, Aloia JF. Body fat content and 25-hydroxyvitamin D levels in healthy women. J Clin Endocrinol Metab. 2003;88:157–161. doi: 10.1210/jc.2002-020978. [DOI] [PubMed] [Google Scholar]
  • 42.Young KA, Engelman CD, Langefeld CD, Hairston KG, Haffner SM, Bryer-Ash M, Norris JM. Association of plasma vitamin D levels with adiposity in Hispanic and African Americans. J Clin Endocrinol Metab. 2009;94:3306–3313. doi: 10.1210/jc.2009-0079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr. 2000;72:690–693. doi: 10.1093/ajcn/72.3.690. [DOI] [PubMed] [Google Scholar]
  • 44.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. doi: 10.1016/j.clinbiochem.2009.03.008. [DOI] [PubMed] [Google Scholar]
  • 45.Wjst M, Altmuller J, Faus-Kessler T, Braig C, Bahnweg M, Andre E. Asthma families show transmission disequilibrium of gene variants in the vitamin D metabolism and signalling pathway. Respir Res. 2006;7:60. doi: 10.1186/1465-9921-7-60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ramos-Lopez E, Bruck P, Jansen T, Herwig J, Badenhoop K. CYP2R1 (vitamin D 25-hydroxylase) gene is associated with susceptibility to type 1 diabetes and vitamin D levels in Germans. Diabetes Metab Res Rev. 2007;23:631–636. doi: 10.1002/dmrr.719. [DOI] [PubMed] [Google Scholar]
  • 47.Ramos-Lopez E, Kahles H, Weber S, Kukic A, Penna-Martinez M, Badenhoop K, Louwen F. Gestational diabetes mellitus and vitamin D deficiency: genetic contribution of CYP27B1 and CYP2R1 polymorphisms. Diabetes Obes Metab. 2008;10:683–685. doi: 10.1111/j.1463-1326.2008.00879.x. [DOI] [PubMed] [Google Scholar]
  • 48.Wjst M, Altmuller J, Braig C, Bahnweg M, Andre E. A genome-wide linkage scan for 25-OH-D(3) and 1,25-(OH)2-D3 serum levels in asthma families. J Steroid Biochem Mol Biol. 2007;103:799–802. doi: 10.1016/j.jsbmb.2006.12.053. [DOI] [PubMed] [Google Scholar]

Articles from International Journal of Molecular Epidemiology and Genetics are provided here courtesy of e-Century Publishing Corporation

RESOURCES