Genetic Variation in Calcium-Sensing Receptor and Risk for Colon Cancer

Substantial evidence from epidemiologic (1-7) and experimental studies (8-12) support the role of dietary calcium as a protective factor in colon or colorectal cancer development (13). Similar findings were shown in a recent meta-analysis for milk and dairy products, the major sources of calcium in Western countries (14), and a pooled analysis evaluating milk and calcium consumption (15). In addition, three randomized trials found a reduction in colorectal adenoma recurrence with calcium supplementation (1,200-2,000 mg/d for 3 to 4 years; refs. 16-19), an effect which persisted up to 5 years after the end of treatment (20). However, results from the Women's Health Initiative, the only trial of calcium supplementation (1,000 mg/d for 7 years) with colorectal cancer as an endpoint, did not support the findings from observational studies (21) and trials of adenoma recurrence; but high calcium intake among participants at enrollment and a high proportion of calcium supplement use in the placebo group complicate the interpretation of this trial (22, 23).

Evidence from experimental studies suggests that calcium prevents colon cancer development through two potentially relevant mechanisms: indirectly by binding free fatty acids and bile acids, and directly by inducing apoptosis and differentiation through binding to the calcium-sensing receptor (CASR; ref. 24). In addition to the important role of vitamin D in maintaining calcium homeostasis, the discovery of vitamin D response elements in the promoter regions of CASR provides an intriguing link between calcium and vitamin D at the molecular level (Fig. 1; ref. 25). The primary focus thus far in the calcium–vitamin D pathway has been on polymorphisms in the vitamin D receptor (VDR); however, CASR, also expressed in the gastrointestinal tract, is an important candidate gene in this pathway that may prove helpful in elucidating mechanistic pathways through which calcium acts.

The few studies (26-28) that have investigated the association between CASR and colorectal tumors have focused primarily on 3 nonsynonymous single-nucleotide polymorphisms, potentially overlooking the additional genetic variation in this large gene (102 kb) that is not sufficiently captured by these 3 single-nucleotide polymorphisms; hence, the need to take advantage of the recent genetic characterization of CASR to explore functional effects associated with its genetic variation (29). In our previous study, we observed no association between any specific genotype of these three single-nucleotide polymorphisms and colorectal adenoma risk but a significant association with CASR diplotypes, suggesting that another polymorphism may be linked with these diplotypes (26). We comprehensively assessed the genetic variation of CASR in relation to colon cancer risk in a population-based case-control study of individuals from California, Minnesota, and Utah. We also examined whether CASR variants modified associations among colon cancer, calcium, and vitamin D intake, and sun exposure, a marker for endogenous vitamin D production. Because calcium and vitamin D are strongly interrelated at the physiologic level, we explored gene-gene interaction between CASR variants with previously genotyped variants in vitamin D–related genes, including VDR, CYP24A1 (the vitamin D deactivating enzyme 24-α hydroxylase), and CYP27B1 (the vitamin D activating enzyme 1-α hydroxylase).

Colon cancer cases and controls were recruited from the Kaiser Permanente Medical Care Program of Northern California, an eight-county area in Utah, and the metropolitan Twin Cities area of Minnesota. This multicenter population-based case-control study included 1,993 cases and 2,410 controls. Cases were 30 to 79 years old at time of diagnosis of first primary colon cancer (International Classification of Diseases for Oncology, 2nd edition, codes 18.0, 18.2-18.9) between October 1991 and September 1994. Cases with tumors in the rectosigmoid junction or rectum and cases with pathology report indicating familial adenomatous polyposis, Crohn disease, or ulcerative colitis were not eligible. Incident cases of colon cancer were identified using a rapid-reporting system, with most interviewed within 4 months of diagnosis. Of the cases that were asked to participate in the study, 76% cooperated, which represents 65% of the eligible cases (for further details, see ref. 30).

Controls were matched to cases by 5-year age groups and sex. Controls from Kaiser Permanente Medical Care Program were randomly selected from membership lists. In Utah, controls who were younger than 65 years were randomly selected from lists generated using random-digit dialing and driver license lists, and those 65 years and older were randomly selected from Health Care Financing Administration lists. In Minnesota, control participants were identified from driver's license or state identification lists. Of all controls asked to participate, 64% cooperated (30).

Tagging single-nucleotide polymorphism selection was based on existing resequencing data from SeattleSNPs (http://pga.gs.washington.edu/), which was limited to 23 HapMap CEPH individuals with European ancestry, given that White was the predominant racial group in our study population. The promoter region (2 kb upstream of exon 1), complete coding sequence (7 exons, including intron-exon boundaries), conserved intronic regions, and a portion of the untranslated region (UTR; 2 kb downstream of STOP) of CASR were resequenced. Linkage disequilibrium structure of CASR was characterized by 4 blocks, encompassing areas in introns 1, 3, and 4, and intron 5–exon 6–intron 6 (the small block in intron 4 is not described because we only successfully genotyped one single-nucleotide polymorphism within it). To include potentially functionally relevant single-nucleotide polymorphisms, we selected all nonsynonymous single-nucleotide polymorphisms. Thereafter, we identified tagging single-nucleotide polymorphisms using the htSNP program developed by Clayton (31) to capture the common variation in the gene (minimum r², 0.80; minor allele frequency, >5%; all nonsynonymous single-nucleotide polymorphisms were predefined as tagging single-nucleotide polymorphisms). Finally, we examined how well the selected tagging single-nucleotide polymorphisms were able to identify common haplotypes (>5%). To preserve a parsimonious set of tagging single-nucleotide polymorphisms, additional single-nucleotide polymorphisms required for capturing haplotypes were tested to determine whether they could replace previously selected single-nucleotide polymorphisms, yet still maintain a minimum r² of 0.80. A total of 22 single-nucleotide polymorphisms were selected for genotyping in CASR.

Of the 4,403 cases and controls with valid study data, 3,680 (83% of cases and 85% of controls) provided a blood sample. Genomic DNA was extracted from peripheral blood lymphocytes or immortalized cell lines and was available from 1,600 cases (80% of cases) and 1,949 controls (81% of controls). Staff was blinded to case-control status and duplicate quality control samples, which were interspersed among plates (147 duplicates from controls). All genotyping was done by Matrix Assisted Laser Desorption/Ionization- Time of Flight (MALDI-TOF) mass spectrometry on the Sequenom MassARRAY7K platform using the iPLEX Gold (low plex) reaction. In total, we received genotyping results for 17 CASR single-nucleotide polymorphisms.

The call rate was >98% for all 17 single-nucleotide polymorphisms. Blinded duplicate samples displayed >98% concordance for any single-nucleotide polymorphism. Using a goodness-of-fit test, the allele frequencies among Caucasian controls were consistent with Hardy-Weinberg equilibrium (P > 0.01) for all single-nucleotide polymorphisms. The 17 successfully genotyped tagging single-nucleotide polymorphisms covered the genetic variation of CASR with a minimum r² of 0.38 and a mean r² of 0.97.

Detailed in-person interviews were used to collect demographic, dietary, and lifestyle data from all eligible study participants. Study participants were asked about their lifestyle during the 2 years before the date of diagnosis (cases) or selection (controls). During the in-person interview, information was collected on dietary intake, physical activity, medical history and drug use, demographic factors, smoking, reproductive history (for women), and family history of cancer and colorectal polyps. Quality control methods have been described in detail elsewhere (32).

Dietary intake was ascertained using a diet-history questionnaire, designed and validated for the Coronary Artery Risk Development in Young Adults study (33, 34). Participants were asked to recall foods eaten, the frequency they were eaten, serving size, and use of fats during food preparation. The nutrient database was based on the University of Minnesota Nutrient Data System for Research. In a comparative validity study, assessment of dietary calcium intake was measured well (Pearson correlations of 0.79 and 0.70 for men and women, respectively; ref. 35). Multivitamin, calcium, and vitamin D supplement use were also ascertained. We defined supplemental vitamin D as 10 μg (400 IU)/d if a participant reported either regular multivitamin or vitamin D supplement use (thrice a week for at least 1 month). Total vitamin D combines dietary and supplemental vitamin D intake. Participants were asked if they used antiacid medications (Rolaids or Tums); however, calcium was restricted to dietary intake only because supplemental dosage was not provided and doses in multivitamins are often low and highly variable.

UV-weighted hours of sun exposure (UV index-h/wk) was based on the average hours per week spent outdoors in the daylight reported by subjects during each season (spring, summer, fall, winter) of the reference year, multiplied by the UV index for each season in the geographic area of the study center, and divided by four to average over the four seasons (36). Height and weight measurements were used to determine body mass index at reference year. Physical activity was assessed using a questionnaire adapted from the Coronary Artery Risk Development in Young Adults study to include more details on amount of time and frequency of activities over 20 years (37, 38). Regular nonsteroidal anti-inflammatory drug use was defined as at least three times a week for at least one month.

Unconditional logistic regression, adjusted for age, sex, race, and study center, was used to estimate odds ratios and 95% confidence intervals (95% CI) for the association between individual polymorphisms and colon cancer. Dietary factors included as potential confounders or effect modifiers were energy adjusted using the residual method (39). The distribution of intake among controls was used to determine cut points. Additional adjustment for education, income, body mass index, cigarette smoking, physical activity, long-term alcohol use, nonsteroidal anti-inflammatory drug use, family history of colorectal cancer, dietary fiber, folate, red meat, fat intake, and multivitamin use did not result in meaningful changes of risk estimates and were not included in the analyses. Global tests of association were conducted by simultaneously, including genotypes (heterozygote and homozygote rare allele) of all CASR single-nucleotide polymorphisms in a model and comparing it to a model that included none of the genotypes. Adjustment for multiple testing was achieved through this multilocus global test [degrees of freedom (df) = 2 × number of single-nucleotide polymorphisms in a gene] within the likelihood ratio χ² test. When single-nucleotide polymorphisms were highly correlated (r² > 0.80), only one of the single-nucleotide polymorphisms were included in the gene-based model. All analyses were conducted using SAS version 9.1 (SAS Institute).

Haploview was used to examine measures of linkage disequilibrium (D′ and r²) between single-nucleotide polymorphisms and to define haplotype-block structure based on Gabriel's definition (40). Haplotype frequencies were estimated and associations were evaluated using HaploStats (Version 1.3.1) in R (version 2.4.1), assuming an additive model unless otherwise indicated. The most common haplotype was used as the reference group, and rare haplotypes (frequency, <5%) were combined into a single category. Analyses for haplotypes were restricted to non-Hispanic Whites, the predominant racial group in our study, and haplotype frequencies and associations differ by racial group. All haplotype analyses were adjusted for age, sex, and study center. A global score statistic was used to evaluate the overall difference in haplotype frequencies between cases and controls.

Polytomous regression was used to estimate associations between genotypes and risk for proximal colon cancer (cecum, ascending colon, hepatic flexure, and transverse colon) and distal colon cancer (splenic flexure, descending colon, and sigmoid colon).

To evaluate whether the presence of a genotype or haplotype modified the association between calcium and vitamin D intake (as well as sun exposure) and colon cancer risk, and whether there was evidence of a gene-gene interaction among CASR, CYP24A1, CYP27B1, and VDR, we investigated interactions through the inclusion of cross-product terms in the regression models. We did an omnibus test for interaction between CASR with the variable of interest (e.g., calcium) by simultaneously including all cross-product terms for the CASR genotypes (coded as dummy variables for heterozygotes and variant homozygotes) with the variable of interest in a model (coded as continuous for dietary factors or UV-weighted sun exposure) and comparing that to a model including only the main effects for genotypes and the variable of interest. In omnibus tests for interaction between two genes, genotypes for both genes were reduced to a binary variable combining heterozygotes and variant homozygotes. If the omnibus test for interaction was statistically significant, we further investigated multiplicative interaction between individual single-nucleotide polymorphisms and the variable of interest, using the log likelihood ratio test to compare the fit of models with and without interaction terms.

Several publicly available Web-based tools were used to assess the potential functional significance of variants, including the University of California Santa Cruz Genome Browser to obtain conservation scores (41), SIFT (42), and PolyPhen (43) to evaluate the functional effect of nonsynonymous single-nucleotide polymorphisms; the Splice Site Prediction tool by Neural Network (44) to identify potential splice sites; ESEfinder (45) and RESCUE-ESE (46) to evaluate whether a single-nucleotide polymorphism was located in a potential exonic splice enhancer; and UTRScan to identify areas within the UTR that may be associated with mRNA stability and translation (47).

Most of the study population was non-Hispanic Whites, with a slightly higher proportion among controls (Table 1). Controls were more likely to be female, be college graduates, and have a higher income. Dietary intake of calcium and vitamin D were higher among controls than cases. Multivitamin use and use of calcium and vitamin D supplements were slightly higher in controls. An inverse association between the highest quartile of dietary calcium and colon cancer was observed (proximal: odds ratio, 0.72; 95% CI, 0.56-0.91; distal: odds ratio, 0.63; 95% CI, 0.49-0.80) but nonsignificant inverse associations for the highest quartiles of dietary vitamin D (proximal: odds ratio, 0.83; 95% CI, 0.66-1.05; distal: odds ratio, 0.80; 95% CI, 0.63-1.01) and sunshine exposure (proximal: odds ratio, 0.71; 95% CI, 0.55-0.92; distal: odds ratio, 0.84; 95% CI, 0.65-1.09). These results have been previously published (7).

Overall, we did not observe a statistically significant association between any of the CASR polymorphisms and colon cancer risk (Table 2). The global association P value for single-nucleotide polymorphisms in CASR and colon cancer was 0.52. Because the association with colon cancer differs for several risk factors by anatomic site, we examined associations for distal and proximal colon cancer. Four CASR variants displayed a statistically significant association with risk for proximal colon cancer (Table 2); the global P value was low but not significant (P = 0.10; df = 32). We observed a positive association between a CASR variant, IVS3+1048, and risk for proximal colon cancer (all ethnicities: odds ratio TT versus GG, 1.35; 95% CI, 1.01-1.81; non-Hispanic Whites: odds ratio, 1.38; 95% CI, 1.02-1.85). CASR variant IVS3-685 was associated with a decreased risk for proximal cancer (all ethnicities: odds ratio AA+AG versus GG, 0.84; 95% CI, 0.71-1.00; non-Hispanic Whites: odds ratio, 0.81; 95% CI, 0.68-0.97) for carriers of at least one copy of the rare allele. A statistically significant inverse association with proximal colon cancer risk was found for IVS5-90 (all ethnicities: odds ratio CC+CT versus TT, 0.83; 95% CI, 0.70-0.98; non-Hispanic Whites: odds ratio, 0.80; 95% CI, 0.67-0.95). Furthermore, we observed an inverse association for IVS6+16 and proximal colon cancer risk among all ethnicities (odds ratio CC versus TT, 0.43; 95% CI, 0.19-0.97) and among non-Hispanic Whites (odds ratio, 0.44; 95% CI, 0.20-1.00). The IVS3+1048 variant was in strong linkage disequilibrium with IVS3-685, IVS5-90, and IVS6+16 (D = 1.0) but weakly correlated with these three variants (r² = 0.11, r² = 0.12, and r² = 0.07, respectively). IVS3-685 was highly correlated with IVS5-90 (r² = 0.89) and moderately correlated with IVS6+16 (r² = 0.59). The two variants IVS5-90 and IVS6+16 were in strong linkage disequilibrium (D = 1.0) and moderately correlated (r² = 0.53) with each other. The associations observed for variants IVS3+1048, IVS3-685, IVS5-90, and IVS6+16 were slightly stronger after adjustment for each other and all other single-nucleotide polymorphisms in CASR (data not shown).

Haplotype analysis did not reveal any statistically significant associations between CASR haplotypes and risk for colon cancer overall; however, an association with proximal colon cancer was observed (Table 3). The haplotype carrying the rare allele of IVS3+1048 in block 2 (A-T) displayed a positive but statistically nonsignificant association with proximal colon cancer risk. This weaker association for haplotype compared with genotype is to be expected because the haplotype analysis is conducted under an additive model whereas the genotype analysis showed statistically significant associations under a recessive model for IVS3+1048. As with the genotype analyses for proximal colon cancer, we observed a statistically significant association for a haplotype in block 3 of CASR, containing IVS5-90 and IVS6+16 (P value for global test = 0.08; Table 3). We observed an inverse association between the C-C haplotype and proximal colon cancer risk, with an odds ratio of 0.80 (95% CI, 0.67-0.97).

We did not observe any statistically significant interaction between genotypes in CASR and UV-weighted hours of sun exposure (global P = 0.09; df = 32), dietary calcium (global P = 0.75; df = 32), or total vitamin D (global P = 0.59; df = 32), with overall colon cancer risk or by cancer subsite (data not shown) with one exception. An interaction between CASR genotypes and total vitamin D was observed (global P = 0.05; df = 31) with distal colon cancer risk, but further single-nucleotide polymorphism by single-nucleotide polymorphism investigation did not portray an obvious association. There was a suggestion of a decrease in risk for distal colon cancer associated with a change of 10 μg of total vitamin D among individuals with either none (odds ratio, 0.79; 95% CI, 0.65-0.97) or both copies (odds ratio, 0.46; 95% CI, 0.26-0.82) of the rare allele for IVS1-4705C>T; however, this relationship is difficult to explain from a biological perspective because no clear risk is associated with the carrying of a specific allele. Associations observed for CASR haplotypes and colon cancer risk did not differ significantly by sunlight exposure, dietary calcium, or total vitamin D and were similar by cancer subsite (data not shown). There was no evidence of an interaction between genotypes in CASR and genotypes in VDR (Bsm1 and Fok1) with colon cancer (global P = 0.39; df = 32), proximal cancer (global P = 0.88; df = 32), and distal cancer risk (global P = 0.37; df = 32). Similarly, genotypes in CASR did not display a statistically significant interaction with CYP24A1 or CYP27B1 genotypes and risk for colon cancer overall or site-specific colon cancer (data not shown).

We observed a statistically significant association between four genetic variants in CASR and proximal colon cancer risk. The global test of association between CASR and proximal colon cancer risk was low but not significant. We hypothesized that variants in CASR might modify the association between UV-weighted hours of sun exposure, calcium, and vitamin D intake or genetic variants in vitamin D related genes on colon cancer risk because it is a key gene in the calcium–vitamin D pathway; however, there was no meaningful evidence of such interactions between CASR genotypes or haplotypes with any of these factors, either overall or by anatomic site.

CASR is suggested to be a key component of the pathway through which calcium mediates its anticarcinogenic effect on colorectal cancer development (48-53). CASR detects extracellular calcium and has a crucial role in calcium homeostasis. CASR is expressed throughout the entire gastrointestinal tract (54-58); in the colon, it is expressed on both the basolateral and luminal surfaces of colon cells, suggesting that CASR senses changes in calcium concentrations in the lumen of the colon as well as circulating concentrations (51, 56, 58). A recent review (51) of the molecular mechanisms of CASR in colorectal cancer indicated that signaling pathways involved in cell growth control and differentiation are activated by calcium through CASR, including the promotion of E-cadherin expression, suppression of β-catenin–T cell factor (48) activation, and activation of the p38 mitogen-activated protein kinase cascade (59). Furthermore, studies have shown that the cell proliferative effect of low intestinal calcium concentrations is likely to be mediated by CASR via protein kinase C–signaling activation, leading to the up-regulation of c-myc expression (49, 50).

To our knowledge, this is the first study to investigate CASR variants and colon cancer using a comprehensive approach to capture common genetic variation; only the three nonsynonymous variants [Ex7+1224 (A986S), Ex7+1236 (R990G), Ex7+1299 (E1011Q)] we examined have been previously evaluated in relation to colorectal adenoma risk. Consistent with our results, that study found no association between these single-nucleotide polymorphisms but rather a diplotype with distal colorectal adenoma risk (26). In our study, IVS3+1048 was associated with an increased risk for proximal colon cancer, but it is located within an intronic area that is not highly conserved (conservation score, 0.01; ref. 41) and thus may not be responsible for the observed association. This variant is in linkage disequilibrium (r² > 0.5) with nine other CASR variants, including nonsynonymous variant Ex7+1224 (A986S; D = 1.0 and r² = 0.57) and a variant in the 3 UTR Ex7-1125delA (D = 1.0 and r² = 0.89). We observed a consistent but statistically nonsignificant increased risk for proximal colon cancer with Ex7+1224 (A986S); however, this variant is suggested to have little functionality (SIFT, tolerated; PolyPhen, benign) despite being highly conserved (conservation score, 1.0). A second linked variant (Ex7-1125delA) is a highly conserved (conservation score, 0.99) deletion in the 3′UTR but is also not predicted to be functionally important. A variant in the 3′UTR, nonetheless, could be important because elements located in this region are involved in the stability and expression of mRNA (60); however, no such element was detected by UTRScan in the area immediately surrounding this variant. Hence, evidence supporting a specific variant as being responsible for the association with IVS3+1048 is lacking.

Three linked intronic CASR variants, IVS3-685, IVS5-90, and IVS6+16, were associated with a decreased risk for proximal colon cancer in our study. However, all three variants are not evolutionarily conserved (conservation score, <0.1). IVS3-685, IVS5-90, and IVS6+16 are in strong linkage disequilibrium with 13 other CASR variants (r², 0.46-1.0). Only one of these variants, IVS4+2762, is correlated with all three tagging single-nucleotide polymorphisms (r², 0.53-1.0) and highly conserved (conservation score, 0.76), but the potential biological importance of this intronic single-nucleotide polymorphism is unknown. Although not residing in a highly conserved area, 2 variants in the 3 UTR were also strongly correlated (Ex7-107: r² = 0.90 and 0.80 with IVS5-90 and IVS3-685, respectively; Ex7-212: r² = 0.86 with IVS6+16). Accordingly, there are several potentially interesting candidates; but because all three single-nucleotide polymorphisms are in linkage disequilibrium, it is unclear whether one or all of these variants are responsible for the associations observed.

Growing evidence suggests that the effect of several risk factors may differ in the development of tumors within the proximal and distal colon; molecular and functional differences result in one site being more susceptible to specific exposures, such as diet (61-63). In particular, the prolonged presence in the proximal colon of bile acids produced by the liver and modified by the intestinal microflora could induce an increase in cell proliferation rates (63). Because calcium has the ability to bind and neutralize secondary bile acids and free fatty acids, the beneficial effect of calcium may be stronger in the proximal colon than the distal colon, where because of such binding as well as resorption, the bile acid concentrations are much lower. Results from our study suggest that calcium may also be mediating its effect in the proximal colon through CASR. Experimental studies have shown that most calcium absorption in the large intestine occurs in the proximal (cecum and ascending) colon (64-66). This is consistent with our findings of a statistically significant association between CASR variants and cancer in the proximal colon, where calcium may be more influential in reducing risk. However, results from epidemiologic studies have been mixed. Twelve observational studies have presented results on the effect of calcium by colon subsite (3-7, 67-73). Four of these studies (3-5, 67) found inverse associations for calcium to be stronger for proximal colon cancer than distal colon cancer, six studies (6, 7, 68, 69, 71, 72) observed stronger associations for distal colon cancer, and two did not find a statistically significant association for either site (70, 73). In our study, dietary calcium was associated with a reduced risk for colon cancer in both sites, but this was slightly stronger for distal cancer (7). Further studies assessing the role of CASR variants are required to verify whether this association is, indeed, site specific.

Our study has some strengths as well as limitations. Because this gene has been investigated only to a limited extent, this study provides unique information on the role of CASR and its association with colon cancer risk. By basing our tagging single-nucleotide polymorphism selection on resequencing data, we were able to conduct a more comprehensive analysis of common genetic variation of CASR and colon cancer risk than the traditional candidate single-nucleotide polymorphism approach. However, because CASR is a very large gene, the entire intronic region of CASR was not resequenced, making it probable that we may have missed some genetic variation (41% of gene sequenced). Nonetheless, the resequencing strategy by SeattleSNPs focused on regions of the gene that are likely to be most functionally relevant (promoter, coding regions, UTRs, and conserved intronic regions); thus, resequencing of additional areas is not likely to reveal additional putatively functional variants. Because of genotyping failure, our tagging single-nucleotide polymorphisms do not fully cover the entire genetic variation of CASR. However, our tagging single-nucleotide polymorphisms did well by tagging 123 detected CASR variants with a mean r² of 0.97. Of the 123 common variants detected through resequencing, only 7 were tagged at an r² between 0.38 to 0.71.

Other advantages of this study are the large number of cases and the detailed data collected on dietary habits and other environmental factors related to colon cancer, thus allowing us to control for important confounding factors. Because this is a case-control study, dietary data were collected retrospectively and may be subject to recall bias; however, data were obtained rapidly after enrollment and collected with detailed and validated procedures (33, 34). A limitation of our data is that only dietary calcium intake was assessed. Inclusion of supplemental calcium from multivitamins or antiacid medications, potentially important sources of calcium, would provide a better indicator of total calcium intake than diet alone. The average daily dietary calcium intake among this study population suggests a high calcium status (1,056 mg), about 200 mg higher than the U.S. national average (863 mg; ref. 74); it is possible that specific CASR variants may be functionally relevant only among a subset of individuals with nutrient deficiency, which may explain the lack of interaction between CASR variants and calcium intake. This study had 80% power to detect a multiplicative interaction odds ratio of 2.2, assuming a main-effects odds ratio of 1.4 for the CASR genotype and an odds ratio of 0.67 for calcium intake above the mean.

In summary, we observed a statistically significant association between CASR variants and proximal colon cancer risk but not distal colon cancer risk. Although this subset distinction is plausible, these results should be considered preliminary and additional studies are needed to confirm these results.

No potential conflicts of interest were disclosed.

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.