No Association of Vitamin D Pathway Genetic Variants with Cancer Risks in a Population-Based Cohort of German Older Adults

Multiple studies have related vitamin D and its different metabolites to the occurrence of several types of cancer (1). The active form of vitamin D and the receptor to which it binds, the vitamin D receptor (VDR), regulate biologically plausible mechanisms linked with the development of cancer (2). Multiple mostly relatively small studies have assessed associations of SNPs in the VDR with incidence of colorectal, breast, prostate, lung, and other cancer sites with equivocal results (3). RFLPs in the VDR were usually investigated, but some of these SNPs are subject to methodologic limitations (4) and are independent of 25-hydroxyvitamin D (25(OH)D) concentrations, the best biomarker of vitamin D status. More recently, two new VDR SNPs were identified (5), which appeared to be associated with cancer incidence, but this finding has not been validated in a larger setting. Additional SNPs associated with 25(OH)D and located in genes encoding for key enzymes involved in vitamin D hydroxylation (CYP2R1), transport (GC), degradation (CYP24A1), and cholesterol synthesis (DHCR7) were identified in genome-wide association studies (6). These SNPs may better reflect long-term vitamin D status than a single 25(OH)D measurement because they are independent of lifestyle or nutritional changes. Thus, we aimed to investigate whether vitamin D pathway genetic variants alone, or together with 25(OH)D concentrations, may be related to total and site-specific cancer risk.

We analyzed data from the ESTHER study, an ongoing population-based cohort study from Saarland, Germany. Details regarding ethical approval, baseline data collection, blood samples, 25(OH)D immunoassay, SNP selection and genotyping, and cancer ascertainment are available from previous publications (7, 8). For these analyses, participants from the baseline survey with missing genotype data at all given loci and any history of cancer were excluded (n = 2,064). The following SNPs were assessed: rs3755967 (C/T); rs11603330 (A/C); rs12794714 (G/A); rs17216707 (T/C); rs2239179 (T/C); and rs7968585 (T/C). All SNPs were in Hardy–Weinberg equilibrium, and no significant linkage disequilibrium between SNPs was observed. A genotype score was constructed summing the number of minor alleles associated with lower 25(OH)D concentrations of SNPs in DHCR7, GC, and CYP2R1 genes as done previously (7).

Cox proportional hazards models were used to calculate HRs and 95% confidence intervals (CI) for the association of vitamin D SNPs and genotype score with total, colorectal, breast, prostate, and lung cancer incidence. To assess whether the association of an SNP with cancer incidence was mediated by vitamin D status, serum 25(OH)D concentration was also included in the model as a covariate. Other potential confounders were also added as covariates. Statistical significance was defined by a two-sided P value lower than 0.05. To correct for multiple testing, the Bonferroni method was used, with a correction factor equal to the number of SNPs. Statistical power was determined using the “powerSurvEpi” package. All analyses were conducted in R version 3.2.3 (R Core Team, 2015).

The association of vitamin D pathway genetic variants with cancer incidence endpoints is shown in Table 1. The HR estimates shown are unadjusted for confounders. Including 25(OH)D concentration and other potential confounders in the models did not alter the risk estimates. Overall, no statistically significant associations with cancer risks were observed for most of the variants. The genotype score including the variants rs3755967 (GC), rs11603330 (DHCR7), and rs12794714 (CYP2R1) was statistically significantly associated with higher lung cancer incidence (HR, 1.20; 95% CI, 1.03–1.39). This estimate was no longer statistically significant after correction for multiple testing. For VDR genetic variants rs2239179 and rs7968585, higher colorectal, breast, prostate, and lung cancer risks with increasing number of risk alleles were suggested, but none of the associations reached statistical significance.

In this population-based cohort study, overall we did not observe an association of vitamin D pathway SNPs with cancer incidence. However, a genotype score including SNPs in genes involved in metabolism (DHCR7 and CYP2R1) and transport (GC) of vitamin D was associated with higher lung cancer risk, but this association was no longer significant if corrected for multiple testing.

Our results do not support the findings from an earlier review of RFLP studies that suggested associations of vitamin D SNPs particularly with colorectal, breast, and prostate cancer (3). Also, our results do not confirm the findings of a previous study suggesting an association of VDR SNPs with total cancer incidence (5). Our study had sufficient power (>90%) to detect the effect estimates on total cancer incidence that were reported in that smaller study (5), possibly attributable to chance. Our study excludes the possibility of a moderate or strong effect of vitamin D genetic variants on cancer incidence. However, our power to detect small effect sizes (HR < 1.30) was limited. Therefore, larger epidemiologic studies or meta-analyses of multiple studies are needed to verify a potential modest influence of genetic variation in the association of vitamin D with the risk of cancer.

T.J. Wang reports receiving commercial research support from Diasorin. No potential conflicts of interest were disclosed by the other authors.

Conception and design: J.M. Ordóñez-Mena, H. Brenner

Development of methodology: J.M. Ordóñez-Mena

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.M. Ordóñez-Mena, B. Schöttker, K.U. Saum, B. Holleczek, B. Burwinkel, T.J. Wang, H. Brenner

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J.M. Ordóñez-Mena, B. Schöttker, B. Burwinkel, H. Brenner

Writing, review, and/or revision of the manuscript: J.M. Ordóñez-Mena, B. Schöttker, K.U. Saum, B. Holleczek, B. Burwinkel, T.J. Wang, H. Brenner

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.M. Ordóñez-Mena, K.U. Saum, B. Burwinkel

Study supervision: H. Brenner

The ESTHER study was funded by the Baden-Württemberg state Ministry of Science, Research and Arts (Stuttgart, Germany), the Federal Ministry of Education and Research (Berlin, Germany), and the Federal Ministry of Family Affairs, Senior Citizens, Women and Youth (Berlin, Germany). The work of José M. Ordóñez-Mena was supported by a scholarship from the Klaus Tschira Foundation (Klaus Tschira Stiftung gemeinnützige GmbH) within the framework of a PhD program in the Network Aging Research (Netzwerk Alternsforschung).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.