Review Article

Biological Mechanisms for Symptom Causation by Individual FODMAP Subgroups

The Case for a More Personalised Approach to Dietary Restriction

Xiao Jing Wang; Michael Camilleri; Stephen Vanner; Caroline Tuck


Aliment Pharmacol Ther. 2019;50(5):517-529. 

In This Article

Plausibility and Mechanisms of Individual FODMAP Subgroups in Symptom Generation

The FODMAP acronym encompasses specifically those carbohydrates shown to be poorly absorbed and rapidly fermented (Table 4). In studies reporting improved symptoms following the low-FODMAP diet, abdominal pain, gas and/or bloating are most often the symptom types with response.[2,3,6] While in other studies the primary endpoint of "adequate symptom relief" was not different from the comparator diet, there were still reductions in abdominal pain.[7] The primary mechanisms proposed to support the reductions in abdominal pain, bloating and gas, are the reduction in substrates available for colonic fermentation and hence reduced gas production. The altered gas production has been shown in studies using breath testing. For example, a low-FODMAP diet has reduced breath hydrogen production in healthy controls and in patients with IBS,[6] however, other studies have found breath testing is not a reliable means to detect clinical response.[1] Additionally, MRI studies have shown increased colonic gas associated with consumption of inulin (a fructan used as a prebiotic) both in healthy volunteers and patients with IBS.[8] While the gas volume was similar in healthy and IBS, impaired transit and tolerance to gas has been shown in IBS patients providing a mechanism behind pain and bloating experienced in this patient group.[9] It has also been suggested that the individual metabolic output of the microbiota including amount and types of gases produced, may be important mediators in the symptom response to the low-FODMAP diet.[8] While hydrogen and methane are often reported due to ease of measurement in breath testing, carbon dioxide production is likely two- to four-fold higher within the colonic lumen[10] and hence may be important for symptom induction although this has not been specifically measured. Furthermore, nongaseous metabolites could also contribute to pain signalling mechanisms. For example, changes in urinary histamine, a metabolite that sensitises nociceptive intestinal neurons has been shown to correlate with levels of FODMAP intake in subsets of patients.[11] Visceral hypersensitivity induced by such pathways could explain why IBS patients might develop symptoms in response to a given volume of gas whereas healthy controls remain asymptomatic with similar volumes.[8]

The second key mechanism of action of a low-FODMAP diet is reduced osmotic effect caused by unabsorbed carbohydrates in the lumen resulting in increased intestinal water and resulting distention. Using ileostomates, increased water content of ileal effluent was noted with a high-FODMAP diet.[12] Reduced osmotic effects of the low-FODMAP diet may have provided the improvement in satisfaction with bowel habits seen in a mostly diarrhoea-predominant IBS cohort.[3]

This discussion in part highlights that individual FODMAP subgroups have different mechanisms of symptom generation, a concept that is addressed in more depth below.

Disaccharide— Lactose

Approximately 65% of the human population has a reduced ability to digest lactose after infancy. Lactose malabsorption in adulthood is most prevalent in people of East Asian descent, affecting more than 90% of adults in some communities.[13] Breath testing for lactose malabsorption is sensitive, and can provide a useful picture of malabsorption status to assist in management planning. However, it is only when lactose malabsorption results in symptom generation, that is, lactose intolerance, that any alterations to diet should be considered.[14] Symptoms reported to be associated with lactose intolerance include diarrhoea, abdominal pain, bloating and gas.[15] Colonic fermentation of lactose has been shown to increase carbon dioxide and hydrogen, but only a small fraction of the gases are rectally excreted. Intra- and inter-individual variations in fermentation and visceral sensitivity on a given day may alter the perceived response.[15]

Implications for Clinical Management

In addition to a patient's ethnicity, a key clinical consideration is the dose consumed. When lactose intake is limited to the equivalent of 240 ml of milk (12–15 g lactose) or less per day, symptoms are likely to be negligible and the use of lactose-digestive aids unnecessary.[16,17] Persons with milk intolerance learn to recognise the amount they can tolerate though some may be erroneous in the attribution of their gastrointestinal symptoms.[16] In addition to dose, how the lactose is consumed may alter symptom response. That is, consumption of lactose with other foods likely slows gastric emptying and small intestinal transit, allowing more time for the disaccharide to be hydrolysed and absorbed.[15] This may result in reduced symptom generation from an equivalent dose of lactose consumed alone. Alternative strategies such as lactase supplementation either taken orally or pre-incubated with milk and yogurts can also improve tolerance, and is not necessary with intake of hard cheeses or creams due to their naturally low levels of lactose.[14] Yogurts also contain bacteria that provide lactase.[18]

Monosaccharide—Excess Fructose

Monosaccharides, including fructose, are transported by carrier-mediated mechanisms across the enterocyte brush border, and up to 50% of this transport is dependent on a sodium ion gradient maintained by a sodium-potassium ATPase. There are five functional, mammalian, facilitated hexose carriers (GLUTs). In the gut, glucose is absorbed across the apical membrane primarily by sodium-dependent active transport mechanisms via SGLT1 and then across the basolateral membrane by the facilitative transporter GLUT-2.[19] Fructose can be absorbed via GLUT-2 in conjunction with glucose, but when luminal glucose concentrations are low, fructose in excess of glucose ("excess fructose") can only be absorbed slowly via GLUT-5, which is primarily a fructose carrier.[20] This slow absorption was shown experimentally when 40 g fructose was administered without glucose, leading to the excess fructose being osmotically active[21] based on MRI measurement of small bowel water content. However, it is unclear whether the results of fructose administration alone, either in the MRI experiments, or in a standard fructose-hydrogen breath test (in which typically 35 g fructose is administered to a fasting patient) realistically reflects the conditions under which most dietary fructose is ingested. Thus, when fructose is ingested as part of a solid food (which retards gastric and small bowel transit compared to transit of an aqueous solution[22]), it does not reach the colon for at least 2 hours; in addition, the concomitant ingestion of glucose in the food enhances fructose absorption (via GLUT-2). The actual amount of fructose delivered to the colon for bacterial fermentation is proposed to be critical for symptom generation but due to difficulties in quantification, the amount reaching the colon is unknown.

In contrast to uncontrolled studies,[23,24] other studies[25–28] (summarised in Table 5) did not show a difference in fructose malabsorption rates between patients with functional GI disorders (FGID) and controls, and one of two studies documented higher symptom scores in patients with evidence of fructose malabsorption compared to patients without fructose malabsorption.[29] The maximum tolerated dose of fructose in one study of IBS patients believed to be fructose intolerant suggested 66% tolerated a 50 g fructose solution, while 21% tolerated 28 g and 13% tolerated only 14 g.[30] Patients reported higher overall symptoms and specifically, more bloating associated with consumption of a fructose solution compared to glucose as control.[30] This may imply that increased osmotic effect and/or gas production is resulting in symptom induction. There has been a disparity reported between fructose malabsorption and symptom induction, suggesting that the malabsorption of fructose may not be the main mechanisms by which fructose induces symptoms.

The applicability of breath testing for fructose malabsorption is fraught with pitfalls, such as use of doses (25–35 g) higher than consumed through a normal diet,[14] poor test re-test reproducibility,[31,32] poor correlation of symptoms with outcome on breath test,[31] and lack of predictive value of test result with outcomes on a low fructose diet.[33] In daily living, consumption of fructose, such as that found in fruit, is generally in the presence of glucose and fibre (eg cellulose), which changes absorption characteristics, slows gastric and small bowel transit time, and significantly decreases rates of fructose malabsorption.[34] Even studies using high-fructose corn syrup (HFCS which consists of 43% or 55% fructose, 45% glucose and up to 10% other saccharides) compared to pure fructose given in breath testing showed significantly lower rates of malabsorption of fructose with HFCS and no association with symptoms after HFCS as opposed to fructose alone.[26]

There are several important principles that need further analysis in considering possible effects of fructose on the development of symptoms as a result of osmotic effects or fermentation. First, there is improved absorption of excess fructose in the presence of equimolar amounts of glucose;[35] this is facilitated by fructose absorption via GLUT-2. Second, although glucose absorption is carrier mediated, and fructose absorption (alone) is linear (concentration dependent, passive absorption), the latter was only about 20% slower than glucose absorption when the two sugars were perfused separately at the same concentrations (1.0, 2.5 and 5.0 g/100 mL) in the human small intestine.[36,37] Third, when saccharides are ingested in aqueous solution without a meal, they empty exponentially from the stomach, and rapidly traverse the small intestine to reach the colon within 30 minutes, as illustrated by 33% of patients with chronic diarrhoea given glucose.[38] In contrast, the mean small bowel transit T1/2 in healthy volunteers administered a mixed meal consisting of cheese, crackers and water (200 calories) was approximately 180 minutes.[39] Therefore, fructose ingested as part of a meal (often containing glucose) has ample time to be absorbed in spite of the diverse absorption kinetics (velocity and affinity) of GLUT-2 and GLUT-5. However, these findings may be dependent on the relative amount of excess fructose in the meal. Overall, the low-FODMAP diet as a whole reduces 14-h effluent output in patients with ileostomy by 20% compared with a high-FODMAP diet[12] suggesting a reduced osmotic load with the low-FODMAP diet. Whether this is attributed to a single FODMAP subgroup (eg excess fructose), or the combination of FODMAPs was not established.

Implications for Clinical Management

First, it is not currently recommended to add glucose as a clinical treatment for possible fructose-related symptoms due to lack of symptom improvement seen when glucose is added to either solutions or foods in an attempt to prevent fructose induced functional gastrointestinal symptoms.[27] Second, it is still unproven that the same dose of fructose administered in real food (see Table 6 for fructose contents of commonly ingested foods) actually causes symptoms compared to 35 g fructose administered in water in the fasted state during a fructose-hydrogen breath test. For example, to reach the 35 g fructose intake, one would need to eat two pears, two medium apples and drink a 16 ounce glass of apple juice, and one would also have to assume that fructose in the solid food empties from the stomach in the same exponential manner as a sugar drink, which is clearly not the case. Lastly, an enzymatic replacement strategy used to convert fructose to glucose using xylose isomerase has shown some promise of improved symptom control[32] but more studies are needed before it is used in routine clinical practice.

In summary, it is still unclear to which degree dietary fructose causes gastrointestinal symptoms, and the threshold dose required for symptoms from fermentation of sugar in the colon, and whether this is the mechanism of symptom generation. For sensitive individuals, management may include reducing intake of foods such as apples, pears and fruit juices known to have high excess fructose; however, the level of reduced fructose intake to result in beneficial effects is unclear and may need to be individualised. This point is reflected in the general recommendation that the low-FODMAP diet can indeed be liberalised after the initial significant reductions in FODMAP intake.

Polyols—Mannitol and Sorbitol

Polyols (sugar alcohols: maltitol, xylitol, mannitol, sorbitol) are incompletely absorbed, osmotically active in the small intestine and can also be fermented in the colon, leading to symptoms of bloating, gas and abdominal pain.[40] Malabsorption of sorbitol has been suggested to occur in 67% of healthy controls and 60% of patients with IBS.[41] On the contrary, mannitol malabsorption was shown to be similar to sorbitol in healthy controls at 57% but much lower in IBS patients at 20% compared to healthy controls. In general, there is poor correlation between polyol malabsorption and gastrointestinal symptoms, with sorbitol and mannitol both increasing symptoms independent of the presence or absence of malabsorption.[41] It is conceivable that the dose of sorbitol and mannitol reaching the small intestine and colon will influence symptom response, although this has not been studied. While breath hydrogen studies used 5–20 g of sorbitol or mannitol to evaluate malabsorption, dietary intake in over 2000 respondents of the UK National Diet and Nutrition Survey showed average intake of polyols over a single meal to be 1.9 g, with 95th percentile of ingested polyols at 5.6 g. Average daily intake of polyols was 3.5 g with the 95th percentile 10.4 g. Notably, items classified as medication tablets and hard candy as well as some meat/fish-based products did contain up to 12 g of polyols.[42] Therefore, it is unclear whether malabsorption testing using 5–20 g of polyols is clinically relevant given average daily consumption of these polyols of 3.5 g.

A SRMA on the effect of polyols on gastrointestinal physiology and IBS showed that, while polyol malabsorption is dose-dependent and compounded by ingestion of polyols in combination, there is considerable intra- and inter-individual variability.[43] The discordant response in IBS patients between malabsorption status (based on breath hydrogen testing) and intolerance (based on symptom response) negates the use for polyol breath testing to guide clinical management.[41] While malabsorption is likely to be partly responsible for the dose-dependent symptom provocation, polyol ingestion may also lead to intestinal dysmotility and can, like other FODMAPs,[44–46] alter the microbiome and metabolome, which may be additional factors in symptom induction and require further elucidation.[43]

Implications for Clinical Management

Small studies have investigated the possibility of adding glucose[47] or amino acids[48] to sorbitol to improve tolerance, but there is insufficient evidence to warrant their use. Restriction of high polyol containing food is current best practice, but further research needs to be conducted to understand the impact of polyols on overall gastrointestinal physiology. Restrictions can be targeted to patients consuming larger quantities of artificial sugars in hard candies which may contain up to 5.7 g sorbitol, chewing gums and mints, which have high polyol content, as well as certain fruits which contain either one or multiple FODMAP subgroups such as both sorbitol and mannitol or sorbitol and excess fructose, for example, peaches contain 1.3 g sorbitol and 0.7 g mannitol, while pears contain 3.8 g sorbitol and 6.2 g excess fructose per piece of fruit.[41,49] As with fructose, the degree of restriction might need to be individualised in the absence of robust data from which to make recommendations.

Oligosaccharides—Fructans and Galacto-oligosaccharides

This is the subgroup of FODMAPs on which there should be little controversy. Humans lack the enzymes to digest fructose-fructose bonds in fructans or the galactose-galactose bonds in galacto-oligosaccharides (GOS) and the arrival of fructans and GOS in the colon leads to bacterial metabolism, fermentation and symptoms. As oligosaccharides are poorly absorbed in everyone, they have potential to be symptom inducing in patients through the actions of metabolites generated by fermentation in patients with existing visceral hypersensitivity and/or through the aforementioned mechanisms. Symptoms associated with fructans have been reported to be overall worsening of global abdominal symptoms, as well as pain, bloating and gas.[30] While there is likely a dose response with consumption of both fructans and GOS, this has not been well studied. Due to differences in chemical structure, degree of polymerisation and fermentation patterns, it is possible that individual patients will have different symptomatic responses to fructans vs GOS. While prevalence of sensitivity to fructans in IBS patients is largely unknown, sensitivity to GOS occurred in 68% of patients with IBS in one study investigating the effects of 5–8 g/d GOS for 3 days through provided food.[50] Specific symptom types generated from a 3-day high GOS diet compared to a low-FODMAP run-in period included abdominal pain, bloating, nausea, gas and fatigue.[50]

Implications on Clinical Management

Current strategies targeting fructan and GOS are to restrict intake followed by strategic re-challenge as part of the low-FODMAP diet approach. The use of enzyme therapy has been shown to be effective in those who are GOS sensitive, but will only target foods containing GOS.[50] There is no available enzyme or other strategy to improve tolerance to fructans. Restriction of oligosaccharides should be targeted to the individuals' usual diet, with those consuming regular servings of rye (containing 0.6 g fructan per serving), wheat (0.36 g fructan per serving),[51] onion (0.34 g fructan per serving) and garlic (0.52 g fructan per serving)[52] more likely to benefit from fructan restriction, compared to GOS restriction in those consuming legumes (red kidney beans contain 0.49 g GOS per serve)[53] and nuts.

Additive Effects and Carbohydrate Interactions

Simultaneous consumption of FODMAP subgroups, such as excess fructose and sorbitol, has resulted in increased symptoms and malabsorption compared to consumption of either sugar alone.[28] Likewise, compared to fructose alone, a solution containing both fructose and fructans resulted in a significantly greater symptom response.[30] These findings suggest interactions between FODMAP subgroups, either affecting absorption capacity, altering the fermentation process or the microbiota profile and their metabolic profiles. FODMAP subgroups are preferentially fermented by specific bacteria, for example, Bifidobacteria increase with GOS consumption,[46] but this fermentation may be altered in the presence of higher quantities of fructan. Indeed it is the additive effect of the FODMAP subgroups that has been proposed to be the key driver for symptomatic response of the low-FODMAP diet compared to more modest symptom reductions in previous studies assessing each FODMAP subgroup individually. This area requires more study as results could help to simplify the application of the low-FODMAP diet.

Challenges in Interpretation of Symptom Response Related to Mechanistic Understanding

While it is proposed that more than one mechanism of action underlies the effect of FODMAPs, the mechanisms and symptoms of each subgroup are overlapping. For example, reduced diarrhea may occur as a result of a lactose free diet, but ongoing bloating may continue due to intake of fructose and fructans. This lack of consistency with individual symptom improvement, as well as frequent co-existence of multiple sources of dietary triggers within foods, has also hampered efforts to demonstrate efficacy of dietary therapies. With the combination of the FODMAP subgroups and greater overall symptom response, efficacy has been easier to establish. However, for the clinician faced with interpreting symptom response and providing management guidance, the overlapping nature makes pinpointing specific foods as triggers prior to dietary modifications difficult. In addition, patients frequently report extra-intestinal symptoms, but other than fatigue, this is rarely monitored in dietary research studies and warrants further investigations. Furthermore, it is widely recognised that patient response to changes in food intake cannot always be traced to known plausible mechanisms and can reflect a wide variety of factors ranging from patient biases and expectations, to yet to be elucidated mechanisms.