Gluten is a generic term applied to the storage proteins found in wheat, barley and rye. Gluten contains large amounts of gliadin, which is resistant to degradation by gastric acids, pancreatic enzymes and the brush border enzymes, hindering digestion(Reference Kupfer and Jabri1).
Some individuals develop a reaction to gluten, showing symptoms after consuming this protein. One of the diseases related to gluten consumption is non-celiac gluten sensitivity (NCGS). The diagnosis of this little-studied condition is exclusively clinical, identified by exclusion of other gluten-related disorders, as celiac disease, wheat allergy and gluten ataxia, and improvement in symptoms after exclusion of gluten from the diet, given that, to date, no specific biochemical or immunological markers have been discovered(Reference DeGeorge, Frye and Stein2). Manifestations of NCGS include gastrointestinal symptoms, such as abdominal pain, constipation, indigestion, diarrhoea and flatulence, as well as extraintestinal symptoms, such as migraine, mental confusion, fatigue, depression, anxiety and body aches(Reference Sapone, Bai and Ciacci3). According to the literature, when individuals diagnosed with NCGS are placed on a gluten-free diet, they show improvement in symptoms(Reference Fasano, Sapone and Zevallos4,Reference Volta, Giorgio and Caio5) . NCGS affects mainly adult women and appears to be more prevalent than other gluten-related disorders. The first cases were reported about 30 years ago; however, only recently have the disease and its mechanism of action gained interest in the scientific community, after numerous reports of improvement in symptoms with the exclusion of gluten from the diet(Reference Czaja-Bulsa6).
Fermentable oligosaccharides, disaccharides, monosaccharides and polyols (FODMAP) are naturally present or added to foods. These carbohydrates undergo fermentation by intestinal bacteria and cause changes in water absorption. FODMAP pass through the small intestine without being digested and reach the intestinal colon where they are fermented by microbiota, producing SCFA. These metabolites provide several benefits to healthy individuals, including protection of the colon epithelium and immune system, participation in appetite regulation and beneficial alteration of lipid metabolism(Reference Morrison and Preston7). However, in addition to SCFA, fermentation of FODMAP leads to methane and hydrogen production, causing bloating and flatulence in some individuals(Reference Biesiekierski, Newnham and Shepherd8).
The presence of FODMAP in the intestinal lumen alters osmolarity, as the amount of solute becomes greater than that within enterocytes. Such a change increases the amount of water in the lumen, which, in healthy individuals, improves intestinal transit and stool consistency. In more sensitive individuals, however, this condition can cause diarrhoea or bloating(Reference Muir, Varney and Ajamian9). Thus, low FODMAP diets have been proposed as dietary treatment for diseases that cause intestinal hypersensitivity, such as irritable bowel syndrome(Reference Solar, Santos and Yamashita10), to reduce symptom severity and improve the quality of life of patients.
Recently, researchers have investigated the effects of FODMAP on the occurrence of NCGS manifestations(Reference Priyanka, Gayam and Kupec11). FODMAP, especially fructans, are present in foods that also contain gluten, such as bread, pasta and breakfast cereals(Reference Biesiekierski, Rosella and Rose12). Although NCGS can be triggered by the consumption of gluten-containing foods, its aetiology is not yet well elucidated. There is still no consensus on the main cause of NCGS symptoms(Reference Catassi, Elli and Bonaz13). This review aimed to examine scientific literature on the relationship between FODMAP intake and NCGS symptoms.
Methods
This systematic review followed the guidelines of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) and was registered in the PROSPERO database (CRD42020199164). Article selection was based on the PICO (population, intervention, comparison, outcome) framework, as follows:
P: adults of both sexes aged 18–60 years and diagnosed with NCGS;
I: FODMAP-restricted diet;
C: unrestricted diet, rich in FODMAP and
O: improvement of gastrointestinal and extraintestinal signs and symptoms as well as quality of life.
Study eligibility criteria
The following eligibility criteria were used: randomised clinical trials, published between 2010 and 2020 in Portuguese, English or Spanish, without geographical restriction. Exclusion criteria included studies with pregnant women or individuals with celiac disease, wheat allergy, gluten ataxia, dermatitis herpetiformis, irritable bowel syndrome, short bowel syndrome, Crohn’s disease, ulcerative colitis or inflammatory bowel disease.
Search strategy
For the systematic search of the literature, uniterms were defined according to the objectives of this review using Medical Subject Headings of the USA the National Library of Medicine (MeSH), Health Sciences Descriptors (DeCS) and keywords of articles identified in a previous search. The search was conducted in Scopus, PubMed, SciELO, Cochrane Library, CINAHL, Embase and VHL (LILACS) databases in April 2020 by the three authors.
Uniterms were separated into two groups, the first related to diagnosis (NCGS) and the second to treatment (FODMAP). Uniterms in the same group were separated by the Boolean operator ‘OR’, whereas groups were separated by the Boolean operator ‘AND’. The simplified search strategy was as follows (“Non-celiac gluten sensitivity” OR “Gluten sensitivity” OR “Gluten sensitive” OR “Gluten-related disorders” OR “Gluten intolerance” OR “Spectrum of gluten-related disorders” OR “Sensibilidad no celiaca al gluten” OR “Non-celiac wheat sensitivity OR “non-celiac”) AND (“Low fodmap diet” OR “FODMAPs” OR “FODMAP” OR “Fermentable oligo-, di- and mono-saccharides and polyols” OR “monossacarídeos” OR “monosaccharides” OR “monosacáridos” OR “dissacarídeos” OR “disaccharides” OR “disacáridos” OR “oligossacarídeos” OR “oligosaccharides” OR “oligosacáridos” OR “polioles” OR “polióis” OR “polyols” OR “frutossacarídeos” OR “fructooligosaccharides” OR “fructooligosacáridos” OR “frutose” OR “fructose” OR “fructosa”). The search strategy used in each platform can be found in Table S1 (online Supplementary material). To ensure that all relevant studies were selected, each author conducted a separate search.
Article selection
The selection of eligible articles was carried out independently and in pairs, by two researchers (RAK and LBAFD), using RAYYAN software(Reference Ouzzani, Hammady and Fedorowicz14). The first selection was made by reading the titles. Then, the abstract of the chosen articles was read, and articles were further selected considering the PICO framework. At both stages, any discrepancies were resolved by a third researcher (ABN). Following this stage, the remaining articles were read in full and the data extracted.
Data extraction
As a strategy to control measurement bias, data extraction was performed independently by two researchers (LBAFD and RAK). Extracted data included authorship, year, location, intervention, follow-up time, variables, outcome and conclusion. Variables were extracted from articles but are not reported in the extraction table. Any discrepancies were resolved by a third researcher (ABN). The risk of bias was evaluated using the Cochrane tool (Risk of Bias 2·0), specific for randomised trials.
Results
Literature search
Initially, 1133 articles were retrieved, 203 of which were duplicates, including identical articles published in different languages. After duplicate exclusion, the titles of the remaining 930 articles were read, and 842 articles were excluded on the basis of the following criteria: studies on different topics, systematic reviews, editorials, letters and interventions with children and/or adolescents. In the subsequent stage of selection, the abstracts of the remaining eighty-eight articles were read. Of these, ten articles were selected for full reading, seven of which were excluded because they were unfinished manuscripts (only the abstracts were published). Finally, three articles that met all inclusion criteria were included in this review. The selected papers are original articles published between 2013 and 2018 on low-FODMAP dietary interventions in adult individuals with NCGS. Fig. 1 shows a flowchart of the article selection process.
Fig. 1. Flowchart of the study selection process.
Study sample and instruments
The selected studies were randomised clinical trials conducted in Germany, Australia or Norway. In total, 125 individuals were analysed, 115 of which were self-diagnosed with NCGS and 10 formed the control group; 106 participants were women and 19 men and the mean age was 42·5 years. In all studies, diagnosis of celiac disease was excluded by examinations performed at baseline. The study by Biesiekiesrki et al.(Reference Biesiekierski, Peters and Newnham15) comprised two distinct interventions; thus, for this review, the article will be treated as two distinct studies. The selected articles analysed several variables; however, only those that were within the scope of the review (outcome of gastrointestinal symptoms and quality of life) were assessed.
The studies used different instruments to evaluate outcomes related to gastrointestinal symptoms, lacking standardisation. Biesiekiesrki et al.(Reference Biesiekierski, Peters and Newnham15) used a visual analogue scale (VAS), Dieterich et al.(Reference Dieterich, Schuppan and Schink16) used the Gastrointestinal Symptom Rating Scale for Irritable Bowel Syndrome, and Skodje et al.(Reference Skodje, Sarna and Minelle17) used both. VAS is used for measuring symptom intensity: a 10-cm horizontal line is anchored by the numbers 0 (no pain) and 10 (unbearable pain), and results are scaled by measuring the distance in millimetres. The Gastrointestinal Symptom Rating Scale for Irritable Bowel Syndrome is a 13-item questionnaire with scores ranging from 1 to 7, where 7 represents more severe symptoms. Instruments used to assess quality of life and psychological symptoms also differed between studies. Biesiekierski et al. (Reference Biesiekierski, Peters and Newnham15) applied the Daily Fatigue Impact Scale and the Short Form-36 (SF-36), Skodje et al.(Reference Skodje, Sarna and Minelle17) used the Giessen Subjective Complaints List and the VAS to assess quality of life specific to symptoms, and Dieterich et al.(Reference Dieterich, Schuppan and Schink16) used the Psychological General Well-Being Index (PGWBI). The SF-36 is a scale in which the higher the result, the better the quality of life, whereas Daily Fatigue Impact Scale and TGSC analyse the severity of outcomes.
Study quality and risk of bias
The GRADE approach revealed some concerns about the risk of bias of studies by Dieterich et al.(Reference Dieterich, Schuppan and Schink16) and Skodje et al.(Reference Skodje, Sarna and Minelle17) Both interventions of the study of Biesiekierski et al. (Reference Biesiekierski, Peters and Newnham15) were classified as having a low risk of bias. Biases were mainly observed in the measurement of outcomes, selection of reported results and deviations from the intended intervention. The study of Dieterich et al. (Reference Dieterich, Schuppan and Schink16) was not a double-blind clinical trial because, according to the authors, participants would easily be able to identify which diet they had been assigned to by consulting the internet. In this case, participants’ knowledge of their treatment group might have interfered with the outcome. Skodje et al.(Reference Skodje, Sarna and Minelle17) did not report the results of gastrointestinal symptoms assessed by the VAS scale; results were stratified by intervention period only. Graphical representations of the risk of bias for each study and analysis of the domains are presented in Fig. 2.
Fig. 2. Risk of bias of selected studies, as assessed using the Cochrane Collaboration’s tool.
Data extraction and analysis
Table 1 provides a summary of the characteristics, results and conclusions of the selected articles.
Table 1. Data extraction and analysis
Gastrointestinal symptoms
The studies used different instruments and protocols for evaluation of variables of interest. By using VAS, Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15) (both interventions) and Skodje et al.(Reference Skodje, Sarna and Minelle17) found that FODMAP and, more specifically, fructans may cause gastrointestinal symptoms in NCGS patients. Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15) reported improvements in abdominal pain, bloating, fatigue, flatulence and stool consistency (P < 0·0001) with dietary restriction of FODMAP. Both studies identified that bloating improved with the exclusion and worsened with the inclusion of FODMAP(Reference Biesiekierski, Peters and Newnham15,Reference Skodje, Sarna and Minelle17) .
Using Gastrointestinal Symptom Rating Scale for Irritable Bowel Syndrome, Dieterich et al.(Reference Dieterich, Schuppan and Schink16) and Skodje et al.(Reference Skodje, Sarna and Minelle17) associated low-FODMAP diets with less severe gastrointestinal symptoms, particularly in relation to abdominal pain. For other gastrointestinal symptoms, such as reflux, indigestion and diarrhoea, fructan consumption was associated with symptom worsening. In these studies, gluten-free diets were compared with low-FODMAP diets, the gluten-free diet produced superior improvements in gastrointestinal symptoms, including bloating, diarrhoea and abdominal pain.
One of the factors limiting study comparison, in addition to differences in assessment scales, was the difference between dietary interventions. Biesiekierski et al. (Reference Biesiekierski, Peters and Newnham15) and Dieterich et al.(Reference Dieterich, Schuppan and Schink16) used low-FODMAP diets, whereas Skodje et al. (Reference Skodje, Sarna and Minelle17) used muesli bars containing fructans (a type of FODMAP). Furthermore, in the first intervention of Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15), it was not possible to compare between gluten-free, gluten-containing, low-FODMAP and placebo diets. Overall, the clinical trials revealed an association between increased gastrointestinal symptoms and FODMAP consumption in patients with NCGS.
Quality of life
The three studies measured psychological symptoms and quality of life by using different instruments and scales, making it difficult to compare them. In general, studies were unanimous in identifying that dietary intake of FODMAP worsened psychological symptoms at baseline. Low-FODMAP diets were associated with fewer psychological symptoms in Dieterich et al.(Reference Dieterich, Schuppan and Schink16) and Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15) compared with the gluten-free diet. These findings can be attributed to the severity of symptoms reported by participants, which directly affect quality of life. Such results were corroborated by those of Skodje et al.(Reference Skodje, Sarna and Minelle17), who found that the fructan diet worsened weakness and vitality in comparison with the gluten diet and enhanced the intensity of symptoms related to quality of life.
Discussion
The low-FODMAP diet is not yet well established. A protocol was developed in 1999 by Gibson and Shepard at the Monash University, Melbourne, and proposed, in 2005, for individuals with Crohn’s disease and, posteriorly, irritable bowel syndrome by the same researchers. However, the diet is based on the typical foods and food culture of Australia, not providing knowledge on the amount of FODMAP in the diets of other populations. The study of Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15) was developed in Australia, and those of Dieterich et al.(Reference Dieterich, Schuppan and Schink16) and Skodje et al.(Reference Skodje, Sarna and Minelle17) were conducted in Europe; all were likely influenced by local food cultures. None of the studies specified the foods that composed the low-FODMAP diets.
Skodje et al.(Reference Skodje, Sarna and Minelle17) did not use a low-FODMAP diet; rather, the FODMAP treatment consisted of food bars containing fructans. Fructan is but one of the existing FODMAP, generating the possibility that other factors interfered with symptoms, given that diet was not controlled. In the study of Biesiekierski et al.(Reference Biesiekierski, Peters and Newnham15), the entire diet was supplied to patients. In the other studies, researchers provided dietary guidelines to participants for adherence to a low-FODMAP diet, thereby generating the possibility of bias, as other factors possibly influencing the onset of symptoms could not be controlled. Furthermore, this method makes it difficult to monitor adherence to diets.
According to Muir et al.(Reference Muir, Varney and Ajamian9), the cut-off values for foods to be considered rich in FODMAP are the following: <0·15 g of fructose per serving; <0·3 g of oligosaccharides per serving in grain and cereal products; <0·2 g of oligosaccharides per serving in fruits, vegetables and all other products or <0·2 g of sorbitol/mannitol per serving. These values refer to a clinical trial with patients with irritable bowel syndrome; thus, the tolerable amount of these substances for individuals with NCGS has not yet been described. Furthermore, there may be individual variability in the tolerance to FODMAP, as well as specific foods that may increase sensitivity, although there have been no reports of such occurrence in current studies on FODMAP(Reference Whelan, Martin and Staudacher18).
Foods that contain gluten are included in the list of foods that contain FODMAP, because wheat contains a huge number of oligosaccharides, e.g. fructans. This fact could justify why a gluten-free diet (or a wheat-free diet) improves gastrointestinal symptoms in individuals with NCGS, as it simultaneously removes foods rich in FODMAP and fibres, thereby reducing intestinal bacterial fermentation. In the evaluated studies, the gluten-free diet led to greater improvement in individual gastrointestinal symptoms than the low-FODMAP diet, demonstrating that gluten seems to have a greater influence on individuals with NCGS than FODMAP. It should be noted that low-FODMAP diets may generate psychosocial problems, resulting from the high restriction of food items, hindering socialisation and causing changes to the intestinal microbiota resulting from reduced fibre intake(Reference De Giorgio, Volta and Gibson19).
One of the limitations of this review was the low number of studies on the topic; this is attributed to the fact that both topics have been studied for a short time. Furthermore, it was not possible to perform a meta-analysis of the review given the scarcity of data and the use of scales that cannot be compared. The four studies used different intervention periods and segments, hindering conclusive results regarding symptom improvement with low-FODMAP diets. Further studies are needed to analyse the effect of low-FODMAP diets in improving gastrointestinal symptoms and quality of life in individuals with NCGS and to standardise instruments used for outcome evaluation between interventions.
Conclusion
Analysis of the selected studies revealed that FODMAP can increase gastrointestinal symptoms in patients with NCGS and worsen quality of life and psychological symptoms. However, improvements in gastrointestinal symptoms were observed when these individuals adhered to gluten-free diets, without the need for FODMAP restriction. More studies are needed with scales and methods that can be compared with obtain resolutions with greater clinical applications for the study population.
Acknowledgements
The authors thank Dr. Ana Carolina Junqueira Vasques and Dr. Daniela Barbieri Hauschild for critical reading and contributions to the manuscript.
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
L. B. A. F. D and R. A. K were responsible for the study design, carrying out the study, data analysis, findings interpretations and writing — original draft preparation, review and editing. A. B. N. was responsible for the research question, study design, data analysis, findings interpretations and writing — review and editing. All authors read and approved the final manuscript.
There are no conflicts of interest.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S0007114522002884
