The FODMAP Diet for Irritable Bowel Syndrome: Food Fad or Roadmap to a New Treatment Paradigm?

Article Outline

• Comment
• Copyright

Shepherd SJ, Parker FC, Muir JG, et al. (Department of Gastroenterology and Monash University Department of Medicine, Box Hill Hospital, Victoria, Australia). Dietary triggers of abdominal symptoms in patients with irritable bowel syndrome: randomized placebo-controlled evidence. Clin Gastroenterol Hepatol 2008;6:765–771.

The role of restricting dietary fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (FODMAPs) in patients with irritable bowel syndrome (IBS) was recently evaluated by Shepherd et al in a randomized, double-blinded, quadruple-arm, crossover, placebo-controlled, rechallenge trial. The investigators studied 25 Australian patients with IBS as defined by the Rome II criteria and fructose malabsorption (FM) diagnosed by a positive fructose hydrogen breath test (HBT) following a 35-g fructose load. Notably, all 3 IBS subtypes—constipation-predominant, diarrhea-predominant, and mixed bowel habits—were included. Eligible subjects had received instruction in a low FODMAP diet ≥3 months before recruitment, and experienced marked and sustained global improvement in gastrointestinal (GI) symptoms on the low FODMAP diet. Celiac disease, inflammatory bowel disease, other serious comorbidities, or medications influencing GI symptoms were all exclusion criteria.

All patients were provided with a low FODMAP diet for the duration of the study. The diet was a 4-week rotation of foods low in FODMAPs calculated for the specific energy requirements of each patient and in compliance with the Australian Recommended Daily Intakes for macronutrients and micronutrients. Excluded foods (those high in FODMAPs) were fruits containing fructose in excess of glucose (apples, pears, watermelon), vegetables containing fructan (onions, leeks, asparagus, artichokes), wheat-based products, foods containing sorbitol, foods containing raffinose (legumes, lentils, cabbage, brussel sprouts), and foods containing lactose (if the patient had lactase deficiency as diagnosed by a HBT after a 50-g lactose load). Patients were also provided with a list of suitable foods when dining out. Patient adherence to diet was determined by diary entries. Patients' specific symptoms (overall abdominal symptoms, wind, bloating, abdominal pain, tiredness, nausea) were assessed by a severity score on the 100-mm visual analog score (VAS), whereas more general assessments were obtained by daily diary entries comprising the global symptom question, “Were your symptoms adequately controlled in this phase?” This latter question was answered at the end of each dose phase and weekly during washout periods.

The subjects were provided with a low FODMAP diet for ≥10 days at the start of the study, and subsequently underwent 3 phases of the study in which they ingested premixed drinks containing 1 of 4 test substances (fructans, fructose alone, fructan and fructose mix, and glucose) at varying doses (low, medium, or high). The subjects consumed low-dose drinks with 3 meals per day for 3 days, followed by medium-dose drinks with meals for 3 days, and finally high-dose drinks with meals for the remainder of 2 weeks. Patients were allowed to withdraw from a phase early if they experienced intolerable symptoms. A washout period of ≥10 days was used between test phases with patients continuing a low FODMAP diet throughout. A subsequent test substance challenge was not permitted until a patient returned to his or her baseline symptom level for ≥7 days. The primary end point was the answer to the question, “Were your symptoms adequately controlled in this phase?”; secondary end points were VAS scores for individual symptoms at the highest test substance dose taken as well as at each specific dose.

With regard to the primary end point question—“Were your symptoms adequately controlled in this phase?”—a similar proportion of patients receiving fructose, fructans, or a mix of the 2 answered positively. However, a significantly greater proportion of patients receiving glucose answered this question positively as compared with patients receiving fructose, fructans, or a mix of the 2.

As compared with the glucose phase, patients in the fructose, fructans, or fructose and fructans mix phase reported significantly higher VAS scores for overall abdominal symptoms, abdominal pain, wind, and bloating. In addition, the reported intensity of each of these symptoms increased in relation to increasing doses of fructose, fructans, or a mix of the 2. There was no similar dose-dependent symptom intensity variation for glucose. There were no statistically significant differences in VAS scores for nausea or tiredness across all groups. As compared with fructose alone, a combination of fructose and fructans was associated with significantly higher symptom severity. A similar difference was not identified when comparing fructose to fructans, or fructans to a fructose and fructan mix.

Back to Article Outline

Comment 

IBS remains a disorder defined by the presence of characteristic symptoms, including some combination of abdominal pain and altered bowel habits. The clinical presentation of patients with IBS is highly variable. Not surprisingly, then, a unifying physiologic abnormality or biomarker has thus far not been identified for IBS. In fact, recent studies suggesting that celiac disease, microscopic colitis, and small intestinal bacterial overgrowth can masquerade as IBS should teach us that this condition likely represents a number of different diseases that happen to present with similar symptoms. As our understanding of the physiologic abnormalities that underlie the different disease subgroups that constitute IBS improves, so too will diagnostic testing strategies and, ultimately, our ability to choose the most appropriate therapy for an individual patient. However, until such a time, we are left with symptoms as the primary guide to the management of IBS patients.

One of the more consistent clinical features of IBS is an association between the development of symptoms and the ingestion of food. Nearly two thirds of patients with IBS associate symptoms with eating a meal, and this finding is particularly common in female patients who have underlying anxiety (Digestion 2001;63:108–115). Despite this very practical clinical observation, surprisingly little attention has been paid to the role of specific foods in the genesis of IBS symptoms.

IBS patients frequently report multiple food “allergies.” However, only a small subset of IBS patients have true food allergies based on serum immunoglobulin E testing (Eur J Clin Nutr 2006;60:667–672). The most common true food allergies include peanuts, tree nuts, fish, shellfish, milk, eggs, soy, and wheat (www.cfsan.fda.gov/∼dms/wh-alrgy.html). On the other hand, the prevalence of less well-characterized food hypersensitivities and intolerances remain unclear, but is likely to be higher than true food allergies (Ann Allergy 1989;62:94–99; Am J Gastroenterol 2005;100:1550–1557). The potentially important role of food in IBS symptoms is underscored by studies reporting benefits for exclusion diets (J Am Coll Nutr 2006;25:514–522; Gut 2004;53:1459–1464) as well as oral cromolyn sulfate, a mast cell inhibitor used to treat atopic conditions and food allergy (Am J Gastroenterol 1992;87:55–57).

Overall, IBS patients tend to report more symptoms when exposed to foods rich in lipids or poorly absorbed carbohydrates (Digestion 2001;63:108–115). It is well established that high-fat foods affect gut motility and transit. For example, high-fat meals tend to slow gastric emptying and can trigger an exaggerated gastrocolonic reflex in some IBS patients. Furthermore, Simren et al recently demonstrated that duodenal lipid infusion leads to colonic hypersensitivity in response to balloon distention and an altered viscerosomatic referral pattern. These changes occurred independent of predominant bowel habit, psychological factors, or gender (Clin Gastroenterol Hepatol 2007;5:201–208). As such, there is now evidence to suggest that fatty foods exert effects on motility and visceral sensation. It is not difficult to envision that such changes could underlie the development of symptoms in a subset of IBS patients.

The role of carbohydrate malabsorption/maldigestion in IBS patients has been investigated yielding mixed results. Older work tended to focus on lactose maldigestion in IBS patients. There is some evidence to suggest that lactose maldigestion is more prevalent in IBS patients than non-IBS controls, although this remains controversial (Am J Gastroenterol 2009;104[Suppl 1]:S1–S35). Whether or not lactose intolerance is more prevalent in IBS patients may not be the question of greatest clinical relevance. To clarify, the clinical consequences of lactose intolerance may be more profound in IBS patients than in healthy individuals. It seems reasonable to hypothesize that any circumstance that leads to a large load of unabsorbed, fermentable substrates in the colon could exacerbate IBS, a condition hallmarked by abnormalities in motility and visceral sensation. Further support for this hypothesis can be derived from recent studies that have identified inherent differences in the gut microflora and thus fermentation in IBS patients versus controls (J Clin Gastroenterol 2006;40:264–269). Unfortunately, rates of symptomatic improvement after lactose restriction in IBS patients have been inconsistent, ranging between 29% and 44% (Eur J Gastroenterol Hepatol 2001;13:219–225).

So, if the answer is not just lactose, what are other potential candidate foods that warrant consideration? Fermentable oligosaccharides, disaccharides, monosaccharides, and polyols or FODMAPs are by definition highly fermentable and poorly absorbed foods. Restricting dietary FODMAP intake could benefit IBS patients through a number of mechanisms. FODMAPs induce an osmotic effect in the colon by which stool fluid content increases (Scand J Gastroenterol 1992;27:819–828). In addition, there is gas production resulting from bacterial fermentation of FODMAPs in the intestines and colon (J Nutr 1998;128:11–19). In animal models, rats fed fructooligosaccharides developed injury of the colonic epithelium and increased intestinal permeability. The rats fed FODMAPs also developed more severe colitis when infected with Salmonella species (Gut 2003;52:1572–1578). This finding may be particularly relevant given evidence that 7%–30% of patients with acute gastroenteritis subsequently develop IBS (J Gastroenterol Hepatol 2005;20:381–386). Interestingly, FODMAPs have been found to possess prebiotic effects, selectively inducing proliferation of certain bacterial strains such as Bifidobacterium species (Gastroenterology 1995;108:975–982). Ingestion of FODMAPs may also exert extraintestinal effects relevant to IBS. For example, both lactose intolerance and FM have been associated with mild depression, and elimination of dietary fructose has been shown to improve depression symptoms in some patients (Scand J Gastroenterol 2000;36:1048–1052).

A key constituent of FODMAPs is fructose. Fructose is a common part of the Western diet and can be consumed as a free monosaccharide, a constituent of the disaccharide sucrose, or in a polymerized form referred to as fructans. The human intestine does not have a specific enzyme for digestion or transport of fructose; absorption of fructose primarily relies on facilitation by glucose transporters (GLUT 5 and GLUT 2), which can be overwhelmed after the ingestion of large amounts of fructose (Gastroenterology 1988;95:694–700). Fructose absorption is much more efficient in the presence of glucose; in fact, FM is much greater when taken alone as compared with in combination with glucose. In aggregate, the literature suggests that the prevalence of FM is similar between patients with functional disorders like IBS and healthy volunteers. There is some literature to suggest that FM is associated with the development of GI symptoms in a subset of individuals (Aliment Pharmacol Ther 2007;25:349–363; Neurogastroenterol Motil 2008;20:505–511; J Clin Gastroenterol 2008;42:233–238). Interestingly, in the study by Shepherd et al, nearly 30% of IBS patients were unable to tolerate a large load of fructose or fructans, although no such problem occurred in non-IBS patients with or without FM. This suggests that the clinical ramifications of fructose or fructan malabsorption may be different in IBS patients than otherwise healthy persons.

Other intriguing components of the FODMAP diet are fructans, which include wheat, rye, and barley among others. It is conceivable that the FODMAP hypothesis offers an explanation for the increased prevalence of anti-gliadin antibodies reported in IBS patients with no evidence of celiac sprue on small bowel biopsies compared with controls (Gastroenterology 2007;132:A-147[#986]) or a recent report of significant clinical improvement following a gluten-free diet in IBS patients with positive celiac antibodies but normal small bowel biopsies (Clin Gastroenterol Hepatol 2007;5:844–850).

There is an increasingly persuasive body of literature to confirm what many IBS patients have been telling us all along—that food is related to their symptoms. Unfortunately, the medical community has largely ignored this clinical information. In fact, it is interesting to note that the National Commission on Digestive Diseases priorities for funding in functional bowel disorders makes little mention of food as an investigative priority (Neurogastroenterol Motil 2008;20:1189–1203). Work by Shepherd and others will hopefully whet the appetite of basic and clinical investigators with an interest in functional disorders like IBS. A great deal of work is clearly necessary to better understand the complexities of the effects of specific foods on GI motor, sensory and immune function, as well as consequences to and from the host microflora. It is hoped that such research will pave the way to more practical and holistic dietary approaches to the management of IBS—something our patients tell us they want. Let's hope we listen this time around.