Common single-nucleotide polymorphism (SNP) variants around the melanocortin 4 receptor (MC4R) gene have recently been associated with obesity risk and insulin resistance. Obesity is a known risk factor for colorectal cancer (CRC) and we hypothesized that there might be a common inherited genetic component.
Methods and Results:
Four of the variants reported earlier were genotyped and tested for association with body mass index (BMI), waist circumference (WC), dietary energy intake (DEI) and CRC. Using a case–control genetic association study, we replicated the association with BMI (P=0.0001, additive genetic effect=0.37 kg/m2) and WC (P=0.005, additive genetic effect=0.70 cm) using over 3800 individuals. However, there was no association between these variants and CRC risk. Rare (highly penetrant) variants within the MC4R gene have been shown to influence eating behaviour and hyperphagia. We hypothesized that the newly identified common variants might also influence hyperphagia. Using DEI data recorded from a validated food frequency questionnaire, we found no significant genetic association between MC4R SNPs and DEI.
As the MC4R locus explains only 0.28% of the BMI and 0.14% of the WC phenotypic variance in the Scottish population, most of the genetic contribution to obesity remains to be identified.
During the last decade, the prevalence of obesity has steadily increased worldwide. Although the reasons for such a trend are complex it is likely that it is fuelled by easy access to energy-dense food and a sedentary lifestyle that leads to a permanent state of positive energy balance and concomitant accumulation of body fat. Obesity increases the risk of developing diseases such as type 2 diabetes, cardiovascular disease and cancer. In Scotland, the prevalence of obesity (that is, a body mass index (BMI)⩾30 kg/m2) in the adult population has increased by 46% between 1995 and 2003.1 Even more alarming is that this observation is not restricted to the adult population, with similar trends reported among Scottish children.2 This increase in the incidence of early-onset obesity suggests that the disease burden associated with obesity will rise sharply in the coming decades.
In large epidemiological studies, obesity is usually assessed through the body mass index (BMI=weight/height2), which measures overall adiposity, and waist circumference (WC), which measures central adiposity. Both BMI and WC are highly heritable and genetically correlated traits. Heritability estimates for BMI and WC range between 0.4 and 0.7, and are consistent among different ages and ethnic groups.3, 4 The genetic correlation, which measures the degree to which genes are shared among the two traits, ranges between 0.6 and 0.89,3, 4 indicating that genes affecting both traits do so in the same direction.
Until very recently, most of the known inherited contribution to obesity was restricted to rare variants with large effects.5, 6 Common variants with individually small effects are starting to emerge with the advent of large-scale genome-wide association studies.7, 8, 9, 10 Given the large genetic correlation between BMI and WC, most of these newly identified common variants have pleiotropic effects, that is influence both traits.7, 8
A wealth of epidemiological and biological evidence indicates a range of mechanistic links between obesity and cancer risk.11, 12 Among the most commonly hypothesized mechanisms, which vary with cancer site,11, 12 is the effect of circulating levels of insulin and insulin-like growth factor 1, both of which inhibit apoptosis and promote cell proliferation in vitro.11 This hypothesis is indirectly supported by the association between type 2 diabetes, characterized by insulin resistance and concomitant hyperinsulinaemia, and some types of cancer (for example, colorectal cancer (CRC)13) and the fact that free insulin-like growth factor 1 and insulin levels are elevated in overweight people.11, 12 Also, obesity increases the bioavailability of oestradiol in men and postmenopausal woman and testosterone in women, both of which promote cellular proliferation and inhibit apoptosis in some tissues (for example, the breast epithelium and the endometrium).11, 12 However, these mechanisms and their relative importance are not fully understood. It is also unlikely that they are the only mechanistic links between obesity and cancer and, hence, it is important to gain a further understanding of the biological basis common to both obesity and cancer. We hypothesize that part of the common biological basis is a shared inherited genetic component with susceptibility genes/loci exerting pleiotropic effects on cancer and obesity.
We recently conducted a two-phase genome-wide association study for CRC.14 In phase 1, we genotyped 1012 early-onset (<55 years of age), population-based CRC patients and 1012 controls matched by age, gender and area of residence using the Illumina HumanHap 300 and 240S arrays. In phase 2, the 15 008 single-nucleotide polymorphisms (SNPs) that showed the strongest level of statistical support in phase 1 were genotyped in 2057 Scottish colorectal cases (<80 years of age) and 2111 matched controls using the Illumina iSelect custom array. Two of these top-ranking SNPs from phase 1 (rs477181 and rs502933) were near the melanocortin 4 receptor (MC4R) gene and had been reported earlier to be associated with WC, BMI and insulin resistance.7 Intrigued by the overlap and the link between obesity, insulin resistance and colorectal cancer,11, 13 we investigated the four SNPs reported by Chambers et al.7 in our phase 1 data and the two SNPs initially associated with CRC in our combined phase 1 and 2 data.
Table 1 shows the total number of individuals with self-reported BMI, WC and dietary energy intake (DEI), which we derived from a semiquantitative food frequency questionnaire (Scottish Collaborative Group FFQ, version 6. 41), as well as the mean and standard deviation for each trait. This questionnaire was specifically developed and validated for use in a wide range of studies of diet and health in Scotland.15, 16 Subjects’ Scottish descent was assessed by the origin of their four grandparents. An extensive quality control of the data was performed as described elsewhere.14
Relevant covariates (that is, gender and age) were identified in preliminary analyses and included in the regression analysis to test for genotype–phenotype associations. BMI, WC and DEI were analysed using least squares, assuming an additive genetic model. The fit of the model containing the covariates was compared with that of the model containing the covariates and the SNP genotype using an F-test.17 The association between marker genotype and CRC risk was tested using binary logistic regression. The goodness-of-fit of nested models was tested using an analysis of deviance and compared with a χ2 distribution. All analyses were performed using the R software (http://www.r-project.org/).
The study was subject to all relevant approvals from the Multicentre and Local Research Ethics Committees as well as from National Health Service research and Development Management Committees in every participating hospital. Each study participant gave informed signed consent.
We tested the association between CRC risk and the ‘intermediate phenotypes’ (that is, DEI, BMI and WC) and found a significant association between CRC risk and DEI (P=8.6 × 10−6) but not with BMI (P=0.5) or WC (P=0.68). Each 1000 kcal per day increase in DEI increased CRC risk by 16% (95% CI: 1.09–1.24).
All SNPs were associated (P⩽0.01) with BMI and WC in the phase 1 data (Table 2). They also showed a high level of consistency in terms of the strength of the statistical support and the size of the genetic effect (expected due to the high level of linkage disequilibrium among them). The combined analysis of phase 1 and 2 data yielded highly statistically significant results for both BMI (P=0.0001) and WC (P=0.005). It is worth noting that the estimated size of the effect was almost halved in the combined data set compared with the phase 1 data set for both BMI and WC, due to the larger sample set providing more precise estimates of the effect. Our estimates of the effects for phases 1 and 2 are highly similar to those reported earlier.7 The borderline statistical association with CRC in the phase 1 data was not replicated in the phase 2 data (combined analysis: P=0.32), suggesting that either our original finding was false positive or we lack statistical power to detect the very small effect size (4% increased CRC risk in the combined data set).
Rare, highly penetrant mutations in the MC4R gene have been reported to be associated with hyperphagia and binge eating.5, 18 To test whether these common variants had a similar effect on eating behaviour, we tested whether the variants were associated with DEI. We observed a statistically significant association between rs12970134 and DEI (P=0.03). However, after adjusting for BMI, the association was no longer significant (P=0.10), which suggests that the association was just reflecting differences in caloric intake associated with a larger body size. Similarly, the analysis of the combined data set for rs477181 and rs502933 showed no association with DEI (P=0.54). Thus, we found no evidence that common variants at the MC4R locus are associated with a hyperphagic phenotype.
Obesity is a complex trait influenced by genetic and environmental factors as well as their complex interactions. Advances in high-throughput genotyping technologies have allowed the identification of the first common genetic variant associated with obesity risk. Nonetheless, progress has been slow and most of the genetic variation contributing to obesity remains to be identified.7, 8, 9, 10, 19 To do so, large sample sizes will be required, especially if one takes into account that the loci identified to date have required thousands of individuals and that these identified loci are likely to contain the variants with the largest effects. Obesity mediates the risk of developing other complex diseases such as CRC. It seems very unlikely that these diseases are linked only at the phenotypic level and so we hypothesize that there are pleiotropic genetic loci still to be identified. The extent to which the inherited risk to CRC and obesity is shared is largely unknown because of the difficulties in collecting the large family sizes required to estimate their genetic correlation.
The four SNPs reported here lay within the 467 kb region that spans between MC4R and PMAIP1, a putative hypoxia-inducible factor-α-regulated proapoptotic gene8 that encodes the protein Noxa. Noxa is upregulated by p53 as a response to DNA damage and hence PMAIP1 is a good candidate for CRC risk. Despite that, we did not find the locus to be associated with CRC risk. Identifying the causative variant or variants for obesity will require a considerable additional genotyping and resequencing effort.
The lack of association between BMI and WC with CRC in our cohort is not entirely surprising given the modest increase in risk for both ‘intermediate traits’. Renehan et al.20 performed a meta-analysis comprising 67 361 incident CRC cases and found that a 5 kg/m2 increase in BMI increased colon cancer risk by 24 and 8% in men and women, respectively. Risk of rectal cancer was increased in men by ∼9% but not in women. Similarly, Pischon et al.21 found that WC was significantly associated with colon cancer risk (P<0.008). The association between WC and rectal cancer risk was not significant. Moore et al.22 found similar results in the Framingham Study, in which a WC larger than 99.1 cm for women and larger than 101.6 cm for men was associated with about a twofold increase in risk of developing colon cancer.
One limitation of our study was the availability of SNPs genotyped in the whole cohort. SNPs genotyped in phase 2 were selected on the basis of CRC risk and not obesity risk. It is plausible that different SNPs within the region might impart CRC or obesity risk, and this will need to be systematically studied in future studies requiring large sample sizes comprising CRC patients and controls with BMI, WC and DEI measures and probably denser SNP genotyping and even deep resequencing to identify variants that are too rare to be included in ‘off-the-shelf SNP arrays’ but that might have moderate-to-large effects23 (for example, odds ratio between 2 and 5).
This study shows that self-reported trait values can yield proxy phenotype measurements, which are appropriate for gene discovery studies. Our results are not quantitatively different from those reported when using clinically measured traits.7 Other self-reported traits such as DEI may also play an important role in understanding the complex mechanisms underlying the onset of obesity and other diseases such as cancer.
Grant I, Fischbacher C, Whyte B . Obesity in Scotland: an epidemiology briefing. ScotPHO reports and papers 2007, http://www.scotpho.org.uk/home/Publications/scotphoreports/pub_obesityinscotland.asp.
Chinn S, Rona RJ . Prevalence and trends in overweight and obesity in three cross sectional studies of British children, 1974–94. BMJ 2001; 322: 24–26.
Wardle J, Carnell S, Haworth CM, Plomin R . Evidence for a strong genetic influence on childhood adiposity despite the force of the obesogenic environment. Am J Clin Nutr 2008; 87: 398–404.
Tang W, Hong Y, Province MA, Rich SS, Hopkins PN, Arnett DK et al. Familial Clustering for Features of the Metabolic Syndrome: The National Heart, Lung, and Blood Institute (NHLBI) Family Heart Study. Diabetes Care 2006; 29: 631–636.
Farooqi IS, Keogh JM, Yeo GS, Lank EJ, Cheetham T, O’Rahilly S . Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J Med 2003; 348: 1085–1095.
Farooqi IS, Wangensteen T, Collins S, Kimber W, Matarese G, Keogh JM et al. Clinical and molecular genetic spectrum of congenital deficiency of the leptin receptor. N Engl J Med 2007; 356: 237–247.
Chambers JC, Elliott P, Zabaneh D, Zhang W, Li Y, Froguel P et al. Common genetic variation near MC4R is associated with waist circumference and insulin resistance. Nat Genet 2008; 40: 716–718.
Loos RJ, Lindgren CM, Li S, Wheeler E, Zhao JH, Prokopenko I et al. Common variants near MC4R are associated with fat mass, weight and risk of obesity. Nat Genet 2008; 40: 768–775.
Dina C, Meyre D, Gallina S, Durand E, Korner A, Jacobson P et al. Variation in FTO contributes to childhood obesity and severe adult obesity. Nat Genet 2007; 39: 724–726.
Frayling TM, Timpson NJ, Weedon MN, Zeggini E, Freathy RM, Lindgren CM et al. A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science 2007; 316: 889–894.
Calle EE, Kaaks R . Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev Cancer 2004; 4: 579–591.
Renehan AG, Roberts DL, Dive C . Obesity and cancer: pathophysiological and biological mechanisms. Arch Physiol Biochem 2008; 114: 71–83.
Seow A, Yuan JM, Koh WP, Lee HP, Yu MC . Diabetes mellitus and risk of colorectal cancer in the Singapore Chinese Health Study. J Natl Cancer Inst 2006; 98: 135–138.
Tenesa A, Farrington SM, Prendergast JG, Porteous ME, Walker M, Haq N et al. Genome-wide association scan identifies a colorectal cancer susceptibility locus on 11q23 and replicates risk loci at 8q24 and 18q21. Nat Genet 2008; 40: 631–637.
Theodoratou E, Kyle J, Cetnarskyj R, Farrington SM, Tenesa A, Barnetson R et al. Dietary flavonoids and the risk of colorectal cancer. Cancer Epidemiol Biomarkers Prev 2007; 16: 684–693.
Masson LF, McNeill G, Tomany JO, Simpson JA, Peace HS, Wei L et al. Statistical approaches for assessing the relative validity of a food-frequency questionnaire: use of correlation coefficients and the kappa statistic. Public Health Nutr 2003; 6: 313–321.
Tenesa A, Visscher PM, Carothers AD, Knott SA . Mapping quantitative trait loci using linkage disequilibrium: marker- versus trait-based methods. Behav Genet 2005; 35: 219–228.
Branson R, Potoczna N, Kral JG, Lentes KU, Hoehe MR, Horber FF . Binge eating as a major phenotype of melanocortin 4 receptor gene mutations. N Engl J Med 2003; 348: 1096–1103.
Benzinou M, Creemers JW, Choquet H, Lobbens S, Dina C, Durand E et al. Common nonsynonymous variants in PCSK1 confer risk of obesity. Nat Genet 2008; 40: 943–945.
Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M . Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. The Lancet 2008; 371: 569–578.
Pischon T, Lahmann PH, Boeing H, Friedenreich C, Norat T, Tjonneland A et al. Body size and risk of colon and rectal cancer in the European Prospective Investigation Into Cancer and Nutrition (EPIC). J Natl Cancer Inst 2006; 98: 920–931.
Moore LL, Bradlee ML, Singer MR, Splansky GL, Proctor MH, Ellison RC et al. BMI and waist circumference as predictors of lifetime colon cancer risk in Framingham Study adults. Int J Obes Relat Metab Disord 2004; 28: 559–567.
Bodmer W, Bonilla C . Common and rare variants in multifactorial susceptibility to common diseases. Nat Genet 2008; 40: 695–701.
We are grateful to all participants in these studies and to nursing and administrative staff on the COGS and SOCCS studies. We also thank departments in central Scottish NHS, including Cancer Registry, Scottish Cancer Intelligence Unit of ISD and the Family Practitioner Committee. This study was funded by grants from Cancer Research UK (C348/A3758 and -A8896, C48/A6361); Medical Research Council (G0000657-53203); Scottish Executive Chief Scientist's Office (K/OPR/2/2/D333, CZB/4/449); and Centre Grant from CORE as part of the Digestive Cancer Campaign (http://www.corecharity.org.uk).
About this article
Association between investigator-measured body-mass index and colorectal adenoma: a systematic review and meta-analysis of 168,201 subjects
European Journal of Epidemiology (2018)
Effect of genetic variants and traits related to glucose metabolism and their interaction with obesity on breast and colorectal cancer risk among postmenopausal women
BMC Cancer (2017)
Future Oncology (2016)
Common variants in the obesity-associated genes FTO and MC4R are not associated with risk of colorectal cancer
Cancer Epidemiology (2016)
Nature Reviews Cancer (2015)