In the present study, we found interaction between the IL10 polymorphisms and intake of dietary fibre. High intake of fibre was associated with lowered risk of CRC among the IL10 rs3024505 homozygous wildtype allele carriers, but not among variant allele carriers (P-value for interaction = 0.01). Furthermore, in tertile analyses, risk of CRC was reduced among C-592A variant allele carriers by high fibre intake compared to wildtype allele carriers who had no risk reduction by fibre intake (P-value for interaction = 0.02). No statistically interactions between IL10 polymorphisms and dietary meat, cereals, or fish intake or between IL10 rs3024505 polymorphism and smoking or NSAID use were found in relation to CRC.
The protective effect of dietary fibre on risk of CRC is well documented
. To our knowledge, this is the first study which suggests that the protective effect of dietary fibre may involve an interaction with anti-inflammatory regulation. Fibre are the indigestible portion of plant foods whereof the soluble fibre are fermented by bacteria in the digestive tract and the insoluble fibre have bulking action. Of particular interest, dietary intake of starch have been found to impact the composition of the gut microbiota
. Commensal bacteria such as Clostridia-related species and Bacteroides fragilis have been shown to affect host IL-10 responses and result in enhanced resistance to intestinal infection in mouse models
[55–58]. Similarly, increased IL-10 level in blood was found after 4 weeks intake of dietary fibre in overweight humans
. Our results support the notion of biological interaction between fibre intake and host IL-10 synthesis. Furthermore, our results suggest this has beneficial effects in individuals carrying the C-592A variant allele and/or two alleles of rs3024505.
A case–control study of 495 gastric cancer (GC) cases and 495 controls found no interaction between smoking and IL10 genotypes
. Both carriers of the C-592A homozygous wildtype genotype and carriers of the variant genotypes had high risk of GC among smokers (odds ratio (OR) 2.88 (95% CI: 1.67-4.95) and 2.06 (95% CI: 1.20-3.55), respectively) compared to the non-smoking reference group (p for interaction = 0.541)
. No other studies on IL10 polymorphisms and dietary fibre, meat, cereals, or fish intake, smoking status or NSAID use in relation to CRC or any other cancer were found.
The present study found no association between IL10 rs3024505 polymorphism per se and risk of CRC. This polymorphism is a marker polymorphism with no known function which has been associated with IBD but which has not been studied in relation to CRC
[41, 42]. Associations between the low activity-associated IL10 C-592A and IL10 -1082 (rs1800896) variant genotypes and high risk of CRC have been found by some case–control studies
[37–39] but not all
. Thus, carriers of the C-592A homozygous variant genotype were at high risk of CRC compared to the homozygous wildtype carriers in the studies by Čačev et al. and Vogel et al. (odds ratio (OR) of 4.07 (95% CI: 1.28-12.96) and OR 1.25 (95% CI: 0.72-2.18), respectively)
[37, 38]. IL10 has been studied in relation to other cancers, including gastrointestinal cancers
[61–66], however, the results have been ambiguous. IL10 has not been associated with CRC in GWAS
[67, 68]. Our study suggests that intake of fibers may play a role in individuals carrying the C-592A variant allele and in individuals carrying the rs3024505 homozygous wildtype genotype. In the combination analysis of the two polymorphisms, additive effects were found, suggesting that both polymorphisms contribute to the found effects. Our results suggest that IL10 polymorphisms may be most important in populations with low fibre intake.
This study used a nested case-cohort design which is efficient and has the major advantage that data and samples were collected before diagnosis. A main strength of our study is a well-characterized study population. Information on diet and lifestyle factors was collected at enrolment for all participants which minimizes the risk of differential misclassification of cases and comparison group. However, data were only collected once, and may thus not be representative for the lifestyle during follow-up. This is, however, not expected to result in differential misclassification. Furthermore, information on food intake was based on a semi-quantitatively food frequency questionnaire
[43, 49], which was evaluated and found usable
. In addition, the results were adjusted for known confounding factors affecting the risk of CRC in this cohort including dietary factors, body mass index (BMI), alcohol, smoking status and NSAID use. Cases were generally older than the comparison group at inclusion into the Diet, Cancer and Health cohort. This is because the comparison group was chosen as a random sample of the study cohort, whereas cases were defined by their age at diagnosis. However, care was taken to ensure that all risk estimates were age adjusted and this means that we generally had longer follow up on the members of the comparison group than of the cases, because the CRC cancer cases generally belonged to the older age fraction of the recruited persons. Fibre intake was lower among the cases compared to controls. To prevent confounding, all risk estimates were adjusted for fibre intake. A main limitation of the study is the limited sample size.
A power calculation showed that we had 82% chance of detecting an OR of 1.5
. Therefore, since no gene dose effects were observed, heterozygous and homozygous variant genotype carriers were combined for the analyses of interactions to obtain sufficient statistical power. Nevertheless, in the light of the obtained P-values and the number of statistical tests performed, we cannot exclude that our positive findings may be due to chance. Moreover, we cannot exclude that associations and/or interactions may not have been identified due to insufficient power.
The present result is in agreement with previous studies proving the concept that gene-environmental interactions may be utilized for identification of underlying biological pathways in carcinogenesis
[31, 32]. The mentioned studies suggested that the intestinal transporter ABCB1 and nuclear factor kappa B (NFκB) are involved in meat carcinogens.