FODMAPs

What is it?

FODMAPs are fermentable carbohydrates — oligosaccharides, disaccharides, monosaccharides and polyols — that some people absorb poorly. Gibson and Shepherd (2010) give the mechanism the whole axis rests on: eating them delivers rapidly fermentable substrate AND water to the distal small intestine and proximal colon, which distends the lumen. That is why the symptoms are bloating, pain and altered stools rather than anything systemic. Their review is about the approach and carries no food value, so it sets no level for anything.

The protocol, and what it is for

The Monash three-step protocol is to avoid all FODMAPs for two to six weeks, then reintroduce one subgroup at a time over about three days each to find which ones matter, then settle into a long-term diet that restricts only those. The elimination phase is a diagnostic tool and not a destination. Monash's own serve thresholds — oat milk low to 104 ml and high at 250, coconut milk low at three-quarters of a cup with sorbitol as the limiting sugar — are thresholds per serve and not composition figures: for grams per 100 g the papers answer, not the app.

Grams per portion, which is the form that helps

Biesiekierski (2011) measured fructan in 55 grains, cereals, breads, pulses and biscuits and reported it per portion as eaten. Couscous was the highest at 1.12 g, wheat-free muesli 0.96, a muesli fruit bar 0.81, dark rye bread 0.60, oats 0.11 and spelt bread 0.07. Only the eight foods its abstract names carry a figure, and it covers grain products and pulses — nothing about fruit, vegetables or dairy.

The bread result the authors themselves call surprising

Whelan (2011) measured fructan by AOAC 999.03 in breads sold in the UK. The whole range is 0.61 to 1.94 g per 100 g, RYE BREAD is the richest at 1.94, and GLUTEN-FREE BREAD carries 1.00 — about the same as the others. A bread chosen to avoid one problem does not avoid this one. That study measures fructan only and only in bread, so it cannot set a FODMAP level on its own, and brand turned out to matter more than type.

Vegetables and fruit, and what the numbers are and are not

Muir (2007) measured total fructans in 60 vegetables and 43 fruits and found the highest in garlic, artichoke, shallots, leek bulb and onions — a range of 1.2 to 17.4 g/100 g given ACROSS those five rather than for any one of them, so it supports membership and not a ranking inside the group. Muir (2009) quantified FODMAPs in 45 vegetables and 41 fruits and names the foods in exactly one finding: apple, pear, mango, clingstone peach and watermelon all contained fructose in excess of glucose. Ma (2014) separated fructose, sorbitol, glucose and sucrose in four fruits — peach, apple, watermelon and cherry — by HPLC-ELSD, with fructose from 5.79 to 104.01 mg/g and sorbitol from 8.70 to 19.13; it is a method paper on four fruits, and a hundred records once cited it for their own sugar profile.

Two rankings that have to be read as rankings

Hogarth (2000) separated three fructooligosaccharide oligomers by ion chromatography and found that of the vegetables investigated, ARTICHOKES contained by far the most FOS, followed by onion and banana. Campbell (1997) ordered foods by BOTANICAL FAMILY instead: samples of the Compositae were highest in total FOS, followed by the Allium species. A family is not a food — giving one member its family's rank is not something either study supports — and neither publishes a figure per 100 g.

One name, two foods

Perovic (2021) sets out the chicory disjunction: the leaf carries around 0.1 per cent fructans of fresh weight against 15 to 20 per cent in the ROOT. Chicory leaf and chicory root are not the same ingredient on this axis, and a single name covering both is the trap. Jaime (2001) adds what storage does to an onion: in four of five varieties fructans are the main carbohydrate, and storage raises free fructose markedly through fructan hydrolysis. Onion only — no leek, no garlic, no shallot.

What fermentation and ripeness remove

Rehms and Barz (1995) screened 46 tempe-forming Rhizopus strains and found R. oryzae and R. stolonifer consuming raffinose, stachyose and melibiose completely. Jiang (2020) compared 54 edamame genotypes against the mature soybean of the same line and found the MATURE soybean highest in stachyose and total sugars. Together they explain why edamame, soy and tempeh from the same bean are not the same food here — though the first measures the fungi rather than a finished tempeh, and the second gives a direction rather than an absolute figure.

Where a whole diet's fructans come from

Moshfegh (1999) used the USDA food-intake survey across more than 15,000 Americans and found an average of 2.6 g of inulin and 2.5 g of oligofructose a day. No composition table can say that, and no composition table is what it is: it publishes no per-food content, so it sets no level. Its shares describe the American diet of the mid-1990s.

Polyols behave differently from the rest

Lenhart (2017) screened 1,823 articles and kept 79 on what polyols do to the gut. Malabsorption is dose-dependent in healthy people, and it INCREASES WHEN POLYOLS ARE EATEN IN COMBINATION — a pear and a sugar-free mint are not two separate small doses, they compound. Yao (2014) put a number on "poorly absorbed": challenged with 10 g, only 33 per cent of healthy people absorbed sorbitol completely, and 40 per cent of those with IBS. Lenhart reviews effects and not composition, so there is no polyol content for any food in it and no threshold in grams; Yao's composition tables are in the paper rather than its abstract.

Where the sugar figures come from, and what they cover

Tuck (2018) measured FODMAPs in common plant-based and processed foods — canned coconut milk at 0.24 g per serve classified low, tempeh 0.26, soy cheese 0.03 — and contains no soy milk at all. Prichard (2016) measured twenty foods eaten by ethnic minority groups in the UK, of which five came out significant sources, contents per 100 g; it says nothing about any food outside its list. USDA FoodData Central supplies sugars per VARIETY where a category cannot — Deglet Noor dates at 19.6 g fructose against 19.9 glucose, Medjool at 32.0 and 33.7, grapes at 8.13 and 7.20 — and publishes no FODMAP figure as such.

Source: Gibson P.R. & Shepherd S.J. (2010), for the mechanism the axis rests on — a review of the approach with no food value in it; Monash University FODMAP composition tables and app, for the three-step protocol and for serve thresholds, which are not composition figures; Biesiekierski J.R. et al. (2011), for fructan per portion as eaten in 55 grain products and pulses, of which the abstract names eight; Whelan K. et al. (2011), for fructan in UK breads, 0.61-1.94 g/100 g, rye highest and gluten-free at 1.00; Muir J.G. et al. (2007), for total fructans in 60 vegetables and 43 fruits, the 1.2-17.4 g/100 g range given across five vegetables rather than per food; Muir J.G. et al. (2009), for the five fruits with fructose in excess of glucose; Ma C. et al. (2014), for fructose, sorbitol, glucose and sucrose in four fruits — a method paper; Hogarth A.J. et al. (2000), for the artichoke-onion-banana FOS ranking; Campbell J.M. et al. (1997), for the ranking by botanical family, which is not a ranking of foods; Perovic J. et al. (2021), for the chicory leaf-versus-root disjunction; Jaime L. et al. (2001), for what storage does to onion fructans; Rehms H. & Barz W. (1995), for Rhizopus consuming the raffinose family — it measures the fungi and not a finished tempeh; Jiang G.-L. et al. (2020), for mature soybean against edamame, a direction and not an absolute figure; Moshfegh A. et al. (1999), for inulin and oligofructose in a whole diet — an intake survey with no per-food content; Lenhart A. & Chey W.D., Adv Nutr 8 (2017) 587-596, for polyol malabsorption being dose-dependent and compounding in combination; Yao C.K. et al. (2014), for the measured absorption of a 10 g sorbitol challenge; Tuck C. et al. (2018), for FODMAPs per serve in plant-based and processed foods — it contains no soy milk; Prichard R. et al. (2016), for twenty foods eaten by ethnic minority groups in the UK; USDA FoodData Central, for sugars per variety and never per brand

Frequently asked questions