An increased use of antibiotics after the Second World War coincided with the increased incidence of IBD and has led some investigators to speculate that antibiotic use may cause IBD [33]. A possible mechanism includes “dysbiosis,” which is the disruption of gut microflora that leads to an imbalance between protective and pathogenic bacteria [34]. Early evidence supporting antibiotics culpability showed an approximately threefold increase in antibiotic usage in IBD patients, but these data were retrospective and may have been influenced by recall and indication bias (i.e., patients receiving antibiotics for insidious symptoms not yet diagnosed as IBD) [16, 29, 35]. A case-control, population-based study in Manitoba using documented antibiotic exposure has shown a threefold increase in antibiotic use in childhood IBD patients compared to controls [36]. However, given the observational methodology of these studies, causation could not be established, and though some studies identified specific offending antibiotics (penicillin and extended-spectrum penicillin) [37], small sample sizes limit generalization. These results may indicate reverse causality and that infection more than the antibiotic use may actually cause IBD (see “Infections” section below). In addition, the prevalence of antibiotic use varies significantly between countries and does not appear to correlate well with the incidence of IBD [22]. Although there is no definitive causation between early antibiotic use and IBD, liberal use of antibiotics in childhood should be avoided with special attention to doxycycline for acne [38], penicillin, or extended penicillin [37].

A possible mechanism leading to IBD as discussed above is dysbiosis, which leads to an imbalance of protective and pathogenic bacteria. Probiotics and prebiotics are thought to reestablish equilibrium. Probiotics are live, nonpathogenic microbial food ingredients, usually of the genus Bifidobacterium or Lactobacillus, that alter the enteric flora and have been associated with beneficial effects. Some noninvasive coliforms and nonbacterial organisms such as Saccharomyces boulardii are also categorized as probiotics [39]. Prebiotics are selectively fermented short-chain fatty acids including fructo-oligosaccharides and galacto-oligosaccharides [40], and together prebiotics and probiotics are termed synbiotics. Given their ability to prevent the overgrowth of potentially pathogenic organisms and stimulate the intestinal immune defense system [41], synbiotics are being increasingly used as an adjuvant or alternative therapy for IBD [42]. In two controlled studies, a nonpathogenic strain of Escherichia coli was as effective as a 5-ASA preparation in maintaining remission in patients with UC [43, 44]. Probiotic combinations have the strongest evidence in the treatment of chronic pouchitis, reducing the relapse rates when compared to placebo [45, 46], and in the primary prevention of pouchitis postsurgery [47], though not all studies were favorable [48]. The use of probiotics and prebiotics in IBD is discussed elsewhere in this book.

Since the 1970s, several case reports as well as case-control and cohort studies have described an increased risk of IBD in women who use oral contraceptives (OCPs) [49–51]. Although cohort studies including more than 80,000 women reported increases in IBD risk ranging from 40 % to threefold, the results were not statistically significant, especially after adjusting for cigarette smoking [50–53]. Other case-control studies have also suggested an association between OCP use and IBD, especially CD [54, 55]. The risk appears to be higher among longtime users [54, 56, 57] and among users of high-dose estrogen preparations [54].

In a meta-analysis of two cohort studies and seven case-control studies from 1995, the pooled OR for UC among OCP users was 1.29, but did not reach statistical significance (95 % CI: 0.9–1.8) [58]. A more recent meta-analysis from 1983 to 2007 that included a total of 75,815 patients (14 studies, with 36,797 exposed to OCP and 39,018 nonexposed women) showed that the pooled relative risk (RR) for UC in women currently taking the OCP was 1.53 (CI 1.21–1.94, P = 0.001) and 1.28 (CI 1.06–1.54, P = 0.011), adjusted for smoking. With cessation of smoking, the RR for UC did not change, but was no longer statistically significant [59].

In summary, available evidence supports a weak association between OCP use and IBD that increases with length of exposure, but with cessation of OCP, the risk reverts to that on the nonexposed population. The thrombogenic potential of OCPs, leading to multifocal gastrointestinal infarctions mediated by chronic mesenteric vasculitis, similar to those of smoking, is the proposed mechanism underlying the effect of OCPs [53, 60]. No recommendations can be made regarding the use of OCPs and the risk of developing IBD.

NSAIDs are one of the most commonly used medications worldwide and have been implicated as a cause of flares secondary to an inhibitory effect on prostaglandins and through uncoupling mitochondrial oxidative phosphorylation [61]. The association between NSAIDs and IBD flares has not been clearly established, though several studies have examined this relationship [62–68]. In the United States, more than 70 million NSAID prescriptions and 30 billion over-the-counter preparations are sold every year [69]. Although NSAID use has typically been associated with the development of gastroduodenal injury, evidence implicating these agents in inducing and exacerbating damage in the distal gastrointestinal tract is also mounting. Colonic injury ranging from colitis resembling inflammatory bowel disease to colonic perforation and bleeding has been described [65, 70, 71].

More than 80 % of patients with IBD interviewed in one study reported use of NSAIDs within the previous month, and approximately one-third of these patients thought that there was an association between their IBD symptoms and NSAID use. In contrast, only 2 % of the IBS population used as a control group reported worsening symptoms following NSAID use [64, 66]. Most studies are case series or case reports and have reported an association between NSAIDs and IBD, though given the methodological shortcomings, causality cannot be established [62, 63, 67, 68]. For example, in new cases of IBD, there was a significant self-reported exposure to NSAIDs/salicylates within 3 months prior to presentation when compared to sex-matched community controls (OR = 9.1, 95 % CI:4.5–21.9) [68]. Some studies did not find a relationship between NSAIDs and IBD activity though these had significant limitations [64–66].

The exact mechanism by which NSAIDs can lead to exacerbations of IBD is not fully understood, though some speculate that small bowel mitochondrial dysfunction and colonic effects of prostaglandin are central. The key enzyme in the inhibition of colonic prostaglandin (PG) synthesis is cyclooxygenase (COX), which exists in two isoforms, COX-1, the constitutive enzyme involved in maintaining mucosal integrity in the GI tract, and COX-2, an inducible enzyme that is expressed at sites of inflammation [72]. COX-2 expression is significantly increased in the colonic mucosa of patients with active IBD when compared to inactive disease or healthy controls [73]. COX-2 appears to have a beneficial effect in healing experimental colitis, and in theory, COX-2 inhibition might impair colitis healing [72]. Alternatively, NSAIDs uncouple mitochondrial oxidative phosphorylation and reduce ATP levels, which can lead to increased permeability because of dysfunction at the mucosal tight junctions [74].

Because of the evidence suggesting that COX-2-specific inhibitors are less toxic to the gastrointestinal tract than traditional NSAIDs, patients and physicians have hoped that selective inhibition of COX-2 would result in anti-inflammatory and analgesic effects without exacerbating IBD. However, cases of IBD flares associated with the use of COX-2 inhibitors have been reported in the literature [75–77]. In a series of 33 patients with IBD who were prescribed with celecoxib or rofecoxib, 39 % experienced exacerbation of their disease [78]. A multicenter, randomized, double-blinded, placebo-controlled trial that enrolled 222 subjects with UC in remission found that rates of UC exacerbation were similar between groups who were taking 200 mg of celecoxib or placebo [67]. The general expert consensus is that the use of COX-2-specific inhibitors in patients with IBD should be viewed with the same caution as the use of traditional NSAIDs [79, 80].

There is no simple solution for patients who require NSAIDs and have significant IBD activity. When patients are using NSAIDs to control the pain from IBD-related arthritis, the intestinal disease should be treated aggressively hoping that the severity of the arthritis will decrease as the intestinal inflammatory activity resolves. Non-NSAID analgesics can be prescribed in the interim to control joint pain. Non-NSAID analgesics and local measures can be used for the treatment of trauma-related pain and inflammation in patients with IBD. If these fail, a short course of NSAIDs or COX-2-selective inhibitors may be prescribed with close monitoring of symptoms and side effects.

Smoking is the best-described environmental factor affecting IBD. The overall effects of smoking on UC are summarized in Table 20.2. Other modifiable factors such as diet, exercise, and stress are less supported and are summarized in Table 20.3.

IBD appears to result from the interaction of three essential cofactors: host susceptibility, enteric microflora, and mucosal immunity. Therefore, it has been proposed that intestinal bacteria may play a role in triggering and perpetuating chronic bowel inflammation. In susceptible individuals, a breakdown in the regulatory constraints of the mucosal immune response to enteric bacteria may result in the development of IBD [39]. Non-enteric systemic infections have also been proposed as causing a flare of IBD by releasing cytokines. The role of infections in the development and during the course of IBD is summarized in Table 20.4.

Appendectomy has been consistently found to be protective against the development of UC [29, 179, 204–207]. A meta-analysis of 17 case-control studies including more than 3,600 cases and over 4,600 controls showed that appendectomy was associated with a 69 % reduction in the subsequent risk of UC [53, 208]. The results of cohort studies have been less consistent, with two large series producing conflicting results. A Swedish inpatient registry of 212,963 patients with more than 5 million person-years of follow-up showed that patients who underwent appendectomy for appendicitis and mesenteric lymphadenitis had a 25 % reduction of the risk of developing UC. The protective effect was only seen if the appendectomy was performed before the age of 20 years. Appendectomy for noninflammatory conditions such as nonspecific abdominal pain did not appear to confer protection against UC [209]. In another large cohort from Denmark, 154,000 patients who had undergone appendectomy were followed for over 1 million person-years. Although the cohort was found to be 13 % less likely to be diagnosed with UC than previously documented national averages, the difference was not statistically significant [209]. Despite these somewhat conflicting results, most evidence from case-control and cohort studies suggest that appendectomy is a protective factor against UC [53].

The influence of appendicitis and appendectomy on UC is not limited to the onset of the disease. Appendectomy also appears to influence the clinical course of UC. When compared to patients with UC and an intact appendix, patients who have undergone appendectomy and develop UC are diagnosed at an older age [210, 211], develop less recurrent symptoms [210], require colectomy less frequently [206, 212], and require less immunosuppressive therapy to control the disease [206]. The effect of appendectomy on the clinical course of patients with known UC is limited to case reports and small case series and results are conflicting [53].

The mechanism by which appendectomy protects against UC is unknown. The appendix is part of the mucosa-associated lymphoid tissue system and is involved in B-lymphocyte-mediated immune responses and extrathymic T lymphocytes. A T-cell receptor alpha chain knockout mouse model of colitis showed that inflammation was suppressed in animals that underwent appendectomies [213]. Because of its role as a reservoir for enteric bacteria, removal of the appendix may influence the mucosal immune system and the antigenic exposure in the bowel lumen [201].

The role of certain environmental and lifestyle factors on the onset, severity, and course of IBD is significant. Patient education and, when possible, modifications of these risk factors should be an integral part of the care provided to patients with IBD. Smoking is the best studied of the disease modifiers in IBD, and smoking cessation should be encouraged for all IBD patients. Although achieving long-term smoking cessation is difficult, IBD patients, including those with UC, should be encouraged to quit smoking. The benefits of smoking cessation outweigh the risk of aggravating UC, and providers caring for these patients should be prepared to adjust the medical regimen to mitigate the adverse effects of nicotine discontinuation. While appendectomy appears to be protective against UC, the chronic use of both traditional NSAIDs and selective COX-2 inhibitors appears to exert a negative effect on the onset and course of IBD. Breastfeeding may offer some protection from IBD and given its other beneficial attributes should be encouraged. The influence of other factors such as diet, childhood infections, socioeconomic factors, psychological stress, and oral contraceptives is less clear, and specific recommendations cannot be generalized at this time for our patients.