Review Article

The Relationship Between Obesity, Bariatric Surgery, and Inflammatory Bowel Disease

Fiorella Cañete; Míriam Mañosa; Ariadna Clos; Eduard Cabré; Eugeni Domènech


Aliment Pharmacol Ther. 2018;48(8):807-816. 

In This Article

Development of de Novo IBD After Bariatric Surgery

Clinical Features of Patients With de Novo IBD After Bariatric Surgery

Current knowledge about the role of bariatric surgery in the development of de novo IBD is scarce. To date, 17 cases of ulcerative colitis (UC), 60 cases of CD and three cases of unclassified colonic IBD after gastrointestinal bariatric surgery have been reported in the literature, most of them with endoscopic and histological findings consistent with IBD.[21–30]All cases but one were reported from 2011 to date, and more than half come from the largest and recently published case series.[28] The available pertinent information regarding 17 UC patients, 54 CD patients and two patients with IBD unclassified (a total of 73 patients) is summarised in Table 1 (detailed information in Table S1a). No clinical data are available for one of the cases with unclassified colonic IBD[28] and six cases reported in abstract form from an Austrian series;[30] for this reason, they are not included in the table. The largest case series published from a study from the Mayo Clinic and the Washington University School of Medicine, included 44 incident cases after bariatric surgery.[28] This study included a prospective database of 3709 patients undergoing bariatric surgery at the Mayo Clinic between 1996 and 2006, with a median follow-up time of 5.1 (IQR, 1.9–9.8) years, and a total follow-up of 22 457 person-years. Six of the total 44 de novo IBD cases reported in this study were operated on and diagnosed with IBD at the Mayo Clinic, resulting in an estimated incidence rate of de novo IBD after bariatric surgery of 26.7 per 100 000 person-years (4.5 for UC and 22.3 for CD). In fact, the standardised incidence ratio of CD adjusted for gender and age resulted in over 3.3 in patients older than 30 year old. This incidence rate should be considered with caution due to the possible referral bias with case findings at two tertiary IBD centres, which may have led to an overestimation of the true incidence of de novo IBD after bariatric surgery. Moreover, the incidence of de novo IBD was higher for CD than for UC (in contrast to what occurs in the general population), suggesting a role of bariatric surgery (or obesity) in CD development. Similarly, Ungaro et al. recently published the results of a case-control study using a US national database from Source Healthcare Analytics LLC. After adjusting for confounders, a past history of bariatric surgery was associated with and increased risk of new-onset IBD (OR 1.93, 95% CI 1.34–2.79).[29]

In relation to the baseline characteristics of the total 73 completely reported patients, and as expected among the population undergoing bariatric surgery, there was a marked predominance of female gender (82%), median age was 45 years, and median BMI before surgery was 47 kg/m2. No case of family history of IBD was reported. Only in the largest case series[28] was the history of perianal disease collected and it occurred in 32% of patients who were further diagnosed with CD. Regarding bariatric surgery procedures, Roux-en-Y gastric bypass was most frequently used (80% of cases), followed by gastric banding. Bariatric surgery was clinically effective in achieving significant weight loss and a decrease in BMI was achieved after 6–12 months in almost all patients. The potentially IBD-related symptoms that prompted clinical suspicions (diarrhoea, abdominal pain and weight loss not explained by the bariatric procedure) occurred within 1 month to 16 years after the procedure. Phenotypically, the most common disease location in CD was ileocolonic (44%), and ileal (31%). In patients diagnosed with UC, 50% had extensive colitis, 25% left-sided colitis, and 25% proctitis. Unfortunately, data regarding disease behaviour in CD patients was only reported in the largest case series, which included 32 patients with CD, showing 26% and 10% of penetrating and stricturing disease pattern respectively.[28] Unfortunately, smoking habit, the most important environmental factor associated with IBD susceptibility, was not provided in most of the cases. Regarding the management of IBD, biological therapy was started in 23 patients (32%) and thiopurines in 18 patients (25%), most of them in CD cases (90%). Eight patients (11%) required surgery (one proctocolectomy with ileostomy in a patient with UC, and two ileocecal resections, four proctocolectomies and one stricturoplasty among CD patients).

Which Could the Pathogenic Relationship Between Bariatric Surgery and IBD be?

The reported cases raise the question of a possible association between bariatric surgery and the development of IBD, and particularly of CD. However, there is controversy about this relationship. Ahn et al, postulated that the anatomic changes secondary to bariatric surgery might promote intestinal dysbiosis, which is the trigger in genetically predisposed patients to exaggerated intestinal inflammation, resulting in the development of CD.[22] In this sense, recent evidence has shown that a cluster of abundant Bacteroides species in the human gut was depleted in the microbiome of obese individuals; however, Bacteroides thetaiotaomicron (a member of Bacteroides species) exhibited a considerable increase in obese individuals following bariatric surgery and was associated with body weight reduction.[31] Korelitz et al, suggested that CD is present previous to bariatric surgery in a preclinical phase and is then nurtured or provoked to become symptomatic by the mechanical or metabolic events entailed by disrupting surgery.[27] They also suggested that early onset and late onset of de novo IBD following bariatric surgery may respond to different mechanisms. In early onset cases, loss of gastric defenses may expose the intestine to different toxins and chemicals, while in late onset cases, major anatomical alterations caused by surgery may lead to bacterial overgrowth and changes in the microbiome. Other proposed mechanisms for the development of CD include the dumping of gastric contents into the small intestine and the key role of the high content of inflammatory cytokines including tumour necrosis factor (TNF–α) in fat tissue when released during rapid weight loss.[5]

Given the small number of cases reported to date, the significance of the development of IBD after bariatric surgery remains unclear. The hypothetical pathogenic mechanisms are only speculative theories and it cannot be ruled out that the association of IBD with bariatric surgery is simply a question of chance.

Is Obesity a Risk Factor for the Development of IBD?

Another possibility is that obesity itself is the real risk factor for IBD and bariatric surgery is only the trigger for the phenotypical expression of the disease. In fact, this proposal is supported by several cohort studies that suggest a possible association between obesity and the development of IBD. A recently large prospective female cohort from the USA (The Nurses' Health Study), showed that BMI >30 kg/m2 was associated with an increased risk of CD but not UC, though the authors were not able to elucidate the biological mechanisms by which excess adiposity may increase the risk of CD.[32] A Danish cohort study of 75 000 women also demonstrated an increased risk of CD (but not for UC) among obese women.[33] Moreover, these data are in line with the predominance of CD over UC in the reported IBD cases occurring after bariatric surgery. Hemminki et al[34] observed, in a population-based Swedish hospital discharge register, that among hospitalised patients with a primary diagnosis of obesity, the relative incidence of CD was highest when obesity was diagnosed before 30 years of age, and decreased with increasing age (40–49 years of age) at obesity diagnosis. Also, Mendall et al[35] observed that obesity at diagnosis of IBD was significantly more common in subjects with CD than in UC, concluding that obesity may play a role in the pathogenesis of CD.

In contrast to these findings, in 2013 a European prospective cohort study including 300 724 adults in 10 countries (EPIC study), did not find any association between obesity and the development of IBD.[36] The demographic features of the above mentioned studies (which included almost exclusively young women) as opposed to those of the EPIC study (which included both men and women, and a larger proportion of elderly population) seem to point to the role of gender and age in obesity as a risk factor for IBD development. However, most of the available epidemiological data on the prevalence of obesity in IBD is cross-sectional and only three prospective cohort studies have been reported to date (Nurses' Health Study, Danish National Birth Cohort and EPIC study). Therefore, these studies might not be representative of the general IBD population and that limits our ability to infer causality in the association between obesity and IBD.

Potential Pathogenic Role of Adipose Tissue in IBD

Obesity has been postulated as a pro-inflammatory condition on the basis that adipose tissue is responsible for the secretion of pro and anti-inflammatory adipokines and the conversion of pre-adipocytes to macrophages.[5] Subcutaneous and visceral adipose tissue compartments display distinct metabolic and immunological profiles. In this sense, mesenteric visceral adipose tissue has a predominance of pro-inflammatory macrophages that secrete several inflammatory cytokines, including TNF–α and interleukin–1, recognised as being the metabolically active fraction of adipose tissue.[37] Patients with CD have higher visceral fat volumes than those without CD and this visceral fat might be more predictive of the risk of developing CD than overall obesity as determined by BMI. Moreover, patients with CD show a unique locally restrictive form of visceral adipose tissue called "creeping fat", a locally restricted hyperplasia of the mesenteric fat adjacent to the inflamed segments of the intestine.[38] Creeping fat is thought to be more immunologically active than other visceral adipose tissue, and the extent of creeping fat correlates closely with the extent of histological inflammation and the degree of lymphocyte or macrophage infiltration. Furthermore, it has been reported that the expression of leptin, adiponectin and other adipokines, such as resistin, is increased in the creeping fat of CD patients, thereby supporting the role of the interaction of adipokines and cytokines in promoting mucosal inflammation.[39]

A primary increased production of pro-inflammatory mediators in mesenteric fat due to a genetic predisposition should also be considered as a possible contributor in the development of CD. In addition, massive cytokine production in the inflamed colon and translocating bacteria that further induce the production of pro-inflammatory mediators in the adjacent adipose tissue could result in a vicious circle, in which inflammatory conditions in the intestine and the mesenteric fat reinforce each other.[39]

Potential Pathogenic Role of Obesity-related Dysbiosis in IBD

Normal gut flora contains approximately 100 trillion microorganisms and their genes may be up to 150 times greater in number than total human gene content. The majority of microbes residing in the gut have a profound influence on human physiology and nutrition, contribute to extracting energy from food and are crucial for human life.[40] Over 90% of the bacterial sequences in the human faecal microbiota consist of gram-negative Bacteroidetes and gram-positive Firmicutes, with Proteobacteria, Actinobacteria, and Fusobacteria being the other common minor components.[41] Dysbiosis is considered to be an alteration in the homeostasis of the microbiota, characterised by a reduction in bacterial biodiversity, lower bacterial population with anti-inflammatory properties and/or an increase in the proportion of bacteria with pro-inflammatory properties.[42] Obesity exhibits a decreased diversity of microorganisms and abnormal gut bacterial translocation, which can lead to adipocyte and pre-adipocyte activation, with ensuing alterations in pro-inflammatory cytokine expression and immune homeostasis.[3] This bacterial translocation can occur secondary to a high fat diet and also due to a loss of bacteria that may maintain intestinal barrier function, like Bifidobacterium.[42] Turnbaugh et al,[43] compared the gut microbial communities of 23 dizygotic and 31 monozygotic female twin pairs, and observed that obesity was associated with a significant decrease in microbial diversity as well as a lower proportion of Bacteroidetes and a higher proportion of Actinobacteria. A Danish study that compared 169 obese and 123 non-obese individuals showed that patients with lower bacterial richness (decreased bacterial diversity) had more marked overall adiposity, insulin resistance, and dyslipemia and a more pronounced inflammatory phenotype when compared with high bacterial richness individuals.[44] The authors found that anti-inflammatory species such as Faecalibacterium prausnitzii were more likely to be found in high bacterial diversity subjects, postulating that individuals with low bacterial diversity harbour an inflammation-associated microbiota. In fact, patients with IBD (mainly CD patients), have lower proportions of Firmicutes, and particularly of F. prausnitzii. These lower proportions were consistently associated with an increased risk of CD.[45]