Address for Correspondence Mohammad Bashashati, Department of Physiology and Pharmacology, Health Sciences Centre, University of Calgary, 3330 Hospital Drive NW, Calgary, AB T2N 4N1, Canada. Tel: +1 403 220 6031; fax: +1 403 283 2700; e-mails: firstname.lastname@example.org; email@example.com
Background Low-grade inflammation has been increasingly implicated in the pathophysiology of irritable bowel syndrome (IBS). Imbalances of pro- and anti-inflammatory cytokines and polymorphisms in cytokine genes have been reported in IBS; however, these findings have not been consistently observed. This may be due to small sample sizes and differences in ethnicities. Therefore, we performed a meta-analysis on the studies that investigated cytokine gene polymorphisms in IBS patients compared to healthy controls.
Methods A PubMed and EMBASE search was performed, and cytokine gene polymorphisms, which had been investigated in at least two case-control studies, were evaluated. Pooled odds ratios (OR) for the genotypes were calculated using random- or fixed-effects models.
Key Results Five studies that investigated interleukin-10 (IL-10; −1082 G/A), transforming growth factor-β1 (TGF-β1; +869 T/C and +915 G/C) and tumor necrosis factor (TNF; −308 G/A) polymorphisms in IBS patients and controls were included. High producer IL-10 (−1082 G/G; OR: 0.64 [95% CI: 0.48–0.87]) was significantly associated with a decreased risk of IBS. The intermediate producer TGF-β1 (+915 G/C) genotype showed a tendency toward decreasing the risk of IBS. No associations were found between TNF (−308 G/A) genotypes and IBS in the whole meta-analysis although an analysis of Asian studies revealed an association between TNF (−308 G/A and G/G) genotypes and IBS (OR: 0.50 [95% CI: 0.29–0.85]), and 1.82 [95% CI: 1.08–3.07], respectively).
Conclusions & Inferences This meta-analysis indicates a role for IL-10 polymorphisms in IBS in general and TNF in Asian populations. Whether or not gene polymorphisms are associated with alterations in cytokine levels leading to functional effects at the level of the gut needs further investigation.
Irritable bowel syndrome (IBS) is a functional gastrointestinal (GI) disorder without significant histopathological abnormalities in the GI tract.1 The pathophysiology of IBS remains to be fully established.2 It has been shown that GI inflammation or infection may predispose individuals to developing IBS.3 Some patients with IBS have a significant history of GI infections; however, others do not recall any form of significant GI infection.4 Studies in post-infectious-IBS (PI-IBS), and those with non-PI-IBS and animal models mimicking features of IBS, reveal subtle changes in pro- or anti-inflammatory cytokines in the blood or GI mucosa; although gross or microscopic inflammation is absent.2,5–8 Pro-inflammatory cytokines are now well recognized as having the potential to alter GI functions; motility, epithelial permeability, and visceral sensation.5,9–12 Cytokines are well-accepted etiologic agents in the development of GI inflammation,13 but their role in the development of IBS or after enteritis in PI-IBS is less well understood.5 It has been shown that cytokine gene polymorphisms may over- or under-produce the encoded cytokines.14 To establish whether or not there is an involvement of cytokines in the development of IBS, recent studies have examined the possible role of cytokine gene polymorphisms in patients with IBS.15–20 In these studies, pro- and anti-inflammatory cytokine gene polymorphisms and their associations with IBS were examined. To date, they have not revealed a clear picture of cytokine gene polymorphisms in IBS; some studies show associations, but others do not. The reasons for the reported discrepancies between these studies potentially include the statistical power of the studies, differences in ethnic composition or differences in the classification of IBS. The discrepancy among the studies on cytokine gene polymorphisms is not only seen in IBS but is also observed in studies of organic GI diseases such as inflammatory bowel diseases (IBD), notably Crohn’s disease.21 In IBD no correlations with cytokine single nucleotide polymorphisms (SNPs) were found. Therefore, despite their pathophysiological involvement with a GI disease, cytokine gene polymorphisms may not readily explain its etiology. Nevertheless, it is important to understand gene associations with diseases as these may be used diagnostically, even if they are not etiologic factors.
Given the importance of understanding potential etiologic agents in IBS and to overcome some of the problems of individual studies, we performed a meta-analysis on the role of cytokine gene polymorphisms in IBS. Our results suggest that cytokines gene polymorphisms are significantly associated with IBS and consequently these gene polymorphisms may be involved in the pathophysiology of this widespread functional GI disorder.
Materials and Methods
Search strategy and study selection
In February 2012, PubMed and EMBASE (OVID) electronic database search was done by two independent reviewers by using the following search terms: (i) ‘irritable bowel syndrome,’ (ii) IBS, (iii) cytokine, (iv) interleukin, (v) TNF-α, (vi) TGF-β, (vii) TGF-β1, (viii) ‘transforming growth factor-β,’ (ix) ‘tumor necrosis factor-α,’ (x) ‘tumor necrosis factor,’ (xi) ‘tumor necrosis factor-alpha,’ (xii) ‘transforming growth factor-beta,’ (xiii) ‘iii or iV or V or Vi or Vii or Viii or iX or X or Xi or Xii,’ (xiv) ‘i and xiii,’ (xv) ‘ii and xiii,’ and (xvi) ‘xiv or xv.’ After removal of duplicates, the abstracts were screened for the relevance to the topic by two independent reviewers. After this step, the full-text of all articles, which assessed cytokine levels or gene expression in human blood or GI mucosa or evaluated cytokine gene polymorphisms in IBS were retrieved. The full text of the articles was evaluated to find case-control studies on cytokine gene polymorphisms in IBS patients vs healthy controls. Selection criteria for the case-control studies were: (i) the diagnosis of IBS based on physicians’ opinion and specific diagnostic criteria (i.e. ROME or Manning), (ii) sufficient evaluation for the absence of any GI diseases in the control groups, (iii) reported information for the calculation of odds ratio (OR) in cases vs controls, and (iv) publication in English. The flow diagram of the study selection is shown in Fig. 1.
In total, five studies were selected for this meta-analysis.15–20,22 The following cytokine gene polymorphisms were evaluated: IL-2 (−330 G/T, +166 G/T), IL-4 (−1098 G/T, −590 C/T, and −33 C/T), IL6 (−174 C/G and +nt565 A/G), IL-10 (−1082 A/G, −819 T/C, −592 A/C), IL-12 (−1188 C/A), IFN-γ (UTR +5644 A/T), TGF-β1 (+869 T/C and +915 G/C), and TNF (−308 G/A and −238 A/G). Gene polymorphisms, which were assessed in at least two studies from different continents, were selected and the relevant data were extracted. Based on these criteria, IL-10 (−1082 A/G), TGF-β1 (+869 T/C and +915 G/C) and TNF (−308 G/A) were studied further. Based on the full text of the articles and raw data whenever it was available, data related to the frequency of genotypes in the case or control groups were retrieved separately by two reviewers. Whether or not significant correlations were found, we examined whether there were any correlations based on regional or ethnic differences in the groups. We also studied the subtypes of IBS if these were available. These were described as: constipation-predominant IBS (C-IBS), diarrhea-predominant IBS (D-IBS) or alternating-type (mixed) IBS (M-IBS).
Genotype frequencies were converted into individual 2 × 2 tables and reported as OR with 95% confidence interval (95% CI). Moreover, pooled OR was calculated for the genotypes using random- or fixed-effects models according to the I2 statistics, where I2 > 50% indicated inconsistency and heterogeneity and necessitated using the random-effect model.23 Review Manager version 5.1 (The Nordic Cochrane Centre, Copenhagen, Denmark; The Cochrane Collaboration, 2011) was used to analyze the data and the final results were presented as forest plots.
Overall, 529 IBS patients and 860 healthy controls were included in the analyses. In the individual studies that were examined there was considerable variation in the associations between genotypes and IBS. However, in our meta-analysis, statistically significant associations were discovered. High producer IL-10 (−1082 G/G) genotype was associated with a decreased risk of IBS; whereas the low producer (A/A), and the intermediate producer (G/A) genotypes were not significantly associated with IBS (Fig. 2). When we stratified the IL-10 data by either ethnicity or region, we observed a significant association between IL-10 (−1082 G/G) and IBS in European population (OR: 0.67 [95% CI: 0.49, 0.91]; P-value = 0.01)17,18 and a tendency but not a significant association between IL-10 (−1082 G/G) and IBS in the Asian population (OR: 0.33 [95% CI: 0.08, 1.31]; P-value = 0.11).16,19,20
Transforming growth factor-β1 (TGF-β1; +869 T/C) genotypes were not significantly associated with IBS (Fig. 3); however, the intermediate producer TGF-β1 (+915 G/C) genotype showed a tendency toward decreasing the risk of IBS (Fig. 4).
Statistical analysis indicated that TNF (−308 G/A) is not a risk factor for IBS in general, as there was no association between its polymorphisms and IBS (Fig. 5). However, in the Asian studies, the G/A and G/G genotypes of TNF (−308 G/A) polymorphisms were associated with a decreased and increased susceptibility to IBS, respectively (Fig. 6).
Stratifying the data based on IBS subtypes and performing meta-analysis did not show a significant association between the cytokine gene polymorphisms and IBS; although, IL-10 (−1082 G/A) polymorphisms had a borderline association with C-IBS (Figs S1 and S2).
Irritable bowel syndrome is a complex condition manifesting with abdominal pain and altered bowel habits. Many studies have suggested that there is a local, subclinical inflammatory component to IBS.2,5,24 Studying cytokines as markers of subclinical inflammation is important in IBS patients; because the absence of gross inflammation makes GI histological evaluation inconclusive. In the current meta-analysis we have shown that there are associations of cytokine gene polymorphisms and IBS.
Cytokine gene polymorphisms are important as they may be associated with changes in cytokine profiles. For IL-10 (−1082 G/A), the presence of allele A is associated with low production of IL-10; however, allele G is considered a high producer allele. For TGF-β1 (+869 T/C) and TGF-β1 (+915 G/C), alleles T and G are the high producer alleles, respectively. The combinations of high and low producer alleles give high, intermediate, and low producer genotypes for IL-10 and TGF-β1. For TNF (−308 G/A), G is the low producer and A is the high producer allele. Tumor necrosis factor (−308 G/G) is considered a low producer; however, both TNF (−308 G/A) and TNF (−308 A/A) are considered high producer genotypes.14 As a result, any change in the genotype may potentially reflect the systemic and the intestinal cytokines levels.
Genetic studies conducted to date have not revealed a clear picture of the possible association of cytokine gene polymorphisms and IBS. In part this may be due to the relatively low sample sizes and study power. We therefore performed a meta-analysis of cytokine gene polymorphisms in IBS. Herein, we show that IL-10 polymorphisms in general and the pro-inflammatory cytokine (TNF) gene polymorphism in Asian populations are associated with IBS.
Interleukin 10 (IL-10; −1082 G/G), a high producer IL-10 genotype, negatively correlates with the development of IBS. In other words, being positive for IL-10 (−1082 G/G) decreases the risk of IBS. Different studies evaluated the systemic levels of IL-10 in IBS patients vs controls. Some studies showed decreased, and others showed no change in the serum IL-10 levels.5,8,25–31 In addition, studies on the intestinal mucosa showed decreased IL-10 expression in female patients with IBS.29,32 Definitive conclusions on whether serum or intestinal IL-10 is decreased in IBS cannot be drawn at present. Interestingly, inconsistencies in IL-10 levels are not only true for IBS but are also seen in patients with IBD.33–36 Small sample sizes and ethnic differences, difference in cytokines measurement methods, the phases of the disease or the anti-inflammatory medications in IBD patients may alter IL-10 levels. As a result, up to now, the possible functional changes to altered IL-10 levels in the gut are not known. Although the immunoregulatory role of IL-10 on the production of pro-inflammatory cytokines has been shown,37,38 whether it directly affects GI function and visceral sensation or whether its effect on the GI function is indirect by a possible anti-inflammatory role is not completely understood.
Our meta-analysis detected a correlation between the common TNF (−308 G/A) polymorphism and IBS in the Asian population, but not in the overall meta-analysis. Low producer TNF (−308 G/G) genotype was correlated with an increased risk of IBS and TNF (−308 G/A) genotype was protective in the Asian population. To our knowledge, from 10 studies, which have measured systemic and/or intestinal levels of TNF, only two studies showed an increase in its levels in IBS.5 In one of these studies the increased level was seen in the group of patients with extra-intestinal co-morbidities, limiting the association with IBS.27 In the other study, patients with post-infectious IBS tended to have the highest level of TNF.39 Interestingly, TNF (−308 G/A) gene polymorphisms were associated with ulcerative colitis, but not Crohn’s disease, again in the Asian population, but not in European cohorts.21,40–42 This finding in the Asian cohorts is in agreement with our meta-analysis. Thus, this observation is an interesting finding as it points to differences in susceptibility and possibly pathophysiology in different populations and needs further attention.
It needs to be noted that we cannot make a clear comparison between IBS and IBD regarding the TNF (−308 G/A) polymorphism as with the studies on IBS patients, the studies on IBD do not come to a clear conclusion on the roles of this gene polymorphism in the development of the disease. Moreover, we cannot rule out the associations between gene polymorphisms in other loci of the TNF gene and IBS. The similarities in TNF (−308 G/A) polymorphism in the Asian population with IBS and IBD potentially suggest an association between these diseases that may depend on ethnicity. However, we know that IBD is less prevalent and less severe in Asian countries. On the other hand, the incidence and prevalence of IBD is rapidly increasing across Asia. This might be related to the Westernization of diet, improved hygiene or other environmental issues.43 Therefore, genetic differences may in part, but not completely, explain the differences in prevalence between Asian and Western countries. Another point, which should be taken into account for IBS is the hygiene issue in developing countries, compared with Western countries. Acute gastroenteritis is common in the developing countries of Asia, but the prevalence of IBS appears to be lower in those regions. The discrepancy between Western and Asian countries may be due to acquired factors, such as the tolerance of hosts to repeated GI infections.44 High producer TNF (−308 G/A) is associated with the lower risk of IBS in Asian studies; however, a recent study on gene polymorphisms in Canadian PI-IBS patients showed no genetic variation in TNF coding genes of IBS patients vs controls.45 Moreover, in a study by van Der Veek et al.18 the high producer TNF (−308 G/A) was more prevalent in European IBS patients, compared to the controls. They showed that the combination of high producer TNF and low producer IL-10 genotypes were more prevalent in IBS patients. This study illustrates the importance of considering the ratio of cytokines in IBS, possibly linked to gene polymorphisms. Additional studies are required to establish whether these findings are observed in other populations.
We also studied TGF-β1 gene polymorphisms, but failed to find any significant associations. The role of TGF-β1 in IBS is not well understood.5 Macsharry et al.32 showed a decrease in the intestinal mucosal level of TGF-β1 in IBS. To our knowledge, a direct role of TGF-β1 in controlling GI motility and sensation has not been shown. TGF-β1 knockout mice present with systemic inflammation including colitis.46 Therefore, alterations in TGF-β1 genotypes may indirectly affect GI function through activation of inflammatory pathways. TGF-β1 (+915 G/C) causes the substitution of arginine with proline in codon 25 and decreases the levels of TGF-β. Transforming growth factor-β1 is an anti-inflammatory cytokine, but whether it is protective is under debate as it may act as an immunosuppressive or may also induce fibrosis.47,48 Studies with larger sample sizes are needed to allow a better conclusion on the effect of TGF-β1 gene polymorphisms in the development of IBS.
Although significant correlations between cytokine gene polymorphisms and IBS were seen in our study, it should be taken into account that published studies with a higher weight in the analysis may skew the results in this meta-analysis. To overcome this limitation, further case-control studies with larger sample sizes are required. Moreover, as the polymorphisms may occur on different parts of a gene, studying SNPs reveals some, but not all possible polymorphisms; this may alter genomic function and signaling. Therefore, more accurate methods like genome wide association studies may more clearly address the role of cytokine genes in IBS.49 In addition, as environmental factors like GI infections or psychological stress are likely involved in the etiology of IBS, epigenetic studies conducted in the future may in part answer questions related to its complex pathophysiology.50 Because limited investigations have been performed on IBS subtypes, and although the categorization of PI-IBS is problematic in developing countries, because of repeated GI infections, this topic requires considerable further investigation. Moreover, gender differences and ethnicities should also be considered in future studies.
In summary, carriers of the high producer anti-inflammatory cytokine IL-10 gene polymorphisms are less likely to have IBS. Tumor necrosis factor (−308 G/A) gene polymorphisms are associated with IBS in Asian populations. Studies on systemic or intestinal cytokine levels in IBS patients with cytokine gene polymorphisms will help to understand their possible functional role in the development of IBS. The significance of this work lies not only with a better understanding of etiology but also with regard to diagnosis and predictor of response in these disorders. Stratifying the patients based on cytokine gene polymorphisms or cytokine profile may predict the response to anti-inflammatory treatments in IBS.51,52
We thank Ms. Lorraine Toews who helped us with the electronic database search. KAS is an Alberta Heritage Foundation for Medical Research Medical Scientist and the Crohn’s and Colitis Foundation of Canada Chair in Inflammatory Bowel Disease Research at the University of Calgary.
This work was supported by the Canadian Institutes of Health Research (to KAS).
No competing interests declared.
MB, NR, and ASH did the literature review; MB and HB performed the statistical analysis; MB, NR, KAS, and MS drafted the paper; MB, NR, NED, KAS, and MS performed the critical revision of the manuscript. All authors have approved the final draft manuscript submitted.