Low-FODMAP Diet for Gastrointestinal Symptoms in Endometriosis: A Systematic Review

This systematic review was prospectively registered with the International Prospective Register of Systematic Reviews (PROSPERO; CRD420261388786). The review was designed and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [ 37 ]. The PRISMA 2020 checklist is provided in Supplementary Material . Clinical studies were eligible if they investigated a low-FODMAP diet or structured FODMAP restriction, either as a primary intervention or as part of a stepwise dietetic pathway, in adolescent or adult women with surgically confirmed, imaging-based, clinically diagnosed, or participant-reported prior endometriosis diagnosis accepted by the source study protocol, and reported at least one outcome relevant to GI symptoms, pain, QoL, adherence, nutritional adequacy, acceptability, or adverse effects. Endometriosis ascertainment included laparoscopic confirmation, transvaginal sonographic or magnetic resonance imaging (MRI)-based diagnosis, physical examination or clinical diagnosis by a treating physician, or participant-reported prior diagnosis when this was accepted by the original study protocol, in line with diagnostic pathways recognized in current guidelines [ 10 , 11 , 12 ]. Eligible designs included randomized and non-randomized controlled trials, prospective and retrospective cohort studies, and individual case reports with structured intervention detail. Survey-based dietary-practice studies that did not isolate a structured low-FODMAP intervention with extractable endometriosis-specific outcome data were excluded from the evidence tables. Editorials, narrative reviews, letters without primary data, guidelines, protocols, conference abstracts without sufficient methodological detail, and animal or in vitro reports were excluded. No language restrictions were applied. A systematic literature search was conducted in PubMed/MEDLINE, the EBSCOhost platform (including Academic Search Premier, APA PsycArticles, APA PsycInfo, CINAHL, and MEDLINE), and the Bielefeld Academic Search Engine (BASE), from database inception to 30 April 2026. The PubMed search combined MeSH and free-text terms for endometriosis with free-text terms for FODMAPs and related fermentable-carbohydrate concepts: (“Endometriosis”[Mesh] OR endometriosis[tiab] OR endometrioma*[tiab]) AND (FODMAP*[tiab] OR “low FODMAP”[tiab] OR “low-FODMAP”[tiab] OR “fermentable oligosaccharides”[tiab] OR “fermentable oligo-saccharides”[tiab] OR “fermentable carbohydrate”[tiab] OR “fermentable carbohydrates”[tiab] OR “short-chain carbohydrate”[tiab] OR “short-chain carbohydrates”[tiab] OR “short chain carbohydrate”[tiab] OR “short chain carbohydrates”[tiab] OR (oligosaccharides[tiab] AND disaccharides[tiab] AND monosaccharides[tiab] AND polyols[tiab])). The EBSCOhost search applied a parallel construction across MH (subject heading), TI (title), AB (abstract), and TX (all text) fields: (MH “Endometriosis+” OR TI (endometriosis OR endometrioma*) OR AB (endometriosis OR endometrioma*)) AND (TI (FODMAP* OR “low FODMAP” OR “low-FODMAP”) OR AB (FODMAP* OR “low FODMAP” OR “low-FODMAP”) OR TX (“fermentable oligosaccharides” OR “fermentable oligo-saccharides” OR “fermentable carbohydrate” OR “fermentable carbohydrates” OR “short-chain carbohydrate” OR “short-chain carbohydrates” OR “short chain carbohydrate” OR “short chain carbohydrates”) OR (TX oligosaccharides AND TX disaccharides AND TX monosaccharides AND TX polyols)). The BASE search used the all-fields query “Low-FODMAP” AND “Endometriosis” . Reference lists of all included reports and relevant systematic reviews were screened for additional records. The PubMed/MEDLINE search retrieved 21 records. The EBSCOhost search retrieved 51 records, of which 26 were duplicates of PubMed/MEDLINE records, and none represented an additional eligible study. The BASE search retrieved 44 records; after consolidation of 18 repeated BASE entries referring to the same reports and subsequent eligibility screening, BASE contributed one additional included report [ 38 ]. The selection process is summarized in the PRISMA 2020 flow diagram ( Figure 1 ). Two reviewers (R.W., S.K.) independently screened titles and abstracts and subsequently assessed full-text reports against the eligibility criteria. There were no disagreements. No automation tools were used. Five primary clinical reports met the inclusion criteria and were retained for descriptive synthesis [ 38 , 39 , 40 , 41 , 42 ]. Data were extracted directly from the full texts of the included studies into a structured evidence table by one reviewer and verified by a second. Extracted variables were organized into the following categories: (1) study metadata (authors, year, journal, setting, registration); (2) population characteristics (sample sizes at consent, start, and completion; age; menopausal status; diagnostic basis for endometriosis; Rome criteria and IBS comorbidity); (3) intervention and comparator details (follow-up duration; low-FODMAP protocol specifics such as restriction duration, target intake, food provision versus advice, lactose handling, and masking; dietitian involvement; educational materials; reintroduction and personalization protocols); (4) outcomes and instruments (primary and secondary definitions; GI and pain outcomes, including endometriosis-specific tools like EHP-30/23; GI-specific QoL via Gastrointestinal Quality of Life Index (GIQLI); psychological outcomes via DASS-21 and equivalents); and (5) adherence and safety (assessment methods and results, dropout reasons, acceptability, post-study continuation, and adverse events). Where reported, substrate-specific reintroduction tolerance was extracted using the original substrate identifiers and dose categories. Numerical values are reported in their original metric. Risk of bias was assessed using tools matched to the study design. The revised Cochrane risk-of-bias tool for randomized trials (RoB 2) [ 43 ], with the additional considerations for crossover trials, was applied to the EndoFOD trial [ 39 ]. The Risk Of Bias In Non-randomized Studies of Interventions tool (ROBINS-I) [ 44 ] was applied to the prospective cohort study by Keukens et al. [ 40 ], the prospective patient-preference pilot study by van Haaps et al. [ 41 ], and the retrospective audit by Moore et al. [ 42 ]. The Joanna Briggs Institute critical appraisal checklist for case reports [ 45 ] was applied to the case report by Jankovich and Watkins [ 38 ]. Risk-of-bias judgements were performed independently by two reviewers and reconciled by discussion with a third reviewer. Risk-of-bias-relevant limitations, methodological strengths, and the resulting overall judgement for synthesis are presented in Table A2 . A quantitative meta-analysis was not appropriate given substantial heterogeneity in study design, control condition, intervention duration, low-FODMAP protocol implementation, endometriosis ascertainment, IBS comorbidity criteria, follow-up, and outcome instruments. The evidence was synthesized narratively. Findings were organized by populations and clinical contexts, diet implementation and adherence, GI outcomes, including global response, bloating and abdominal distension, and stool pattern, pain and endometriosis-specific QoL outcomes, GI-specific QoL and psychological outcomes, acceptability and safety, and risk-of-bias-relevant limitations. The synthesis distinguishes restriction-phase data from reintroduction- and personalization-phase data. Only the randomized crossover feeding trial isolated the short-term restriction phase under controlled dietary conditions. The non-randomized prospective studies included reintroduction and personalization components but did not allow phase-specific causal attribution. Effect estimates from individual studies are reported in the original metric without recoding or post hoc imputation. Where studies separated low-FODMAP-arm-specific from pooled-diet effects, both estimates are reported, and the limits on causal attribution are stated explicitly.

Endometriosis is a chronic estrogen-dependent disease of gynecologic origin with systemic consequences [ 1 , 2 ]. The disease affects approximately one in ten women of reproductive age, often begins in adolescence, and produces a cumulative burden of pain, subfertility, fatigue, reduced work capacity, psychosocial distress, and prolonged healthcare contact [ 1 , 2 , 3 ]. Pathogenetically, endometriosis is multifactorial and can be described with the “seed-and-soil” metaphor [ 4 ]. The retrograde menstruation hypothesis proposed by Sampson approximately 100 years ago describes the route of dissemination for the “supercharged” altered eutopic endometrium, but the complexity of endometriosis pathogenesis includes hormonal and immune dysregulation, peritoneal inflammation, neurogenic inflammation, and peripheral and central sensitization that contribute to chronic pelvic and visceral pain [ 5 , 6 , 7 ]. The pain phenotype frequently extends beyond cyclic dysmenorrhea to non-cyclic pelvic pain, dyspareunia, and bowel- and bladder-related symptoms, in patterns that map only weakly onto lesion location and stage [ 6 , 7 , 8 ]. Diagnostic delay of several years to more than a decade is common and contributes to disease progression and accumulating disability [ 8 , 9 ]. Current guidelines have shifted from mandatory histological confirmation toward symptom- and imaging-based diagnosis [ 10 , 11 , 12 ]. The research on non-invasive biomarkers from saliva [ 13 ], serum [ 14 ], urine [ 15 ], or menstrual effluent [ 4 ] is expanding, but still far from clinical application. Standard therapy combines analgesia, hormonal suppression, and surgery [ 8 , 10 ]. Each modality has limitations: surgery is invasive, operator-dependent, and unable to address all components of disease-related morbidity; hormonal suppression controls endometrium-driven pain components but is contraceptive, carries cardiovascular considerations, and does not relieve all symptom domains [ 7 , 8 , 16 ]. Many patients continue to report bothersome symptoms despite optimized standard care, driving interest in adjunctive non-hormonal strategies for persistent symptoms [ 7 , 17 , 18 ]. GI symptoms are a frequent component of endometriosis, including abdominal pain, bloating, flatulence, constipation, diarrhea, nausea, painful or urgent defecation, incomplete evacuation, and cyclic fluctuation [ 19 , 20 ]. In a case–cohort study of 109 patients and 65 controls, women with endometriosis had higher severity of abdominal pain, constipation, bloating, flatulence, defecation urgency, incomplete evacuation, impaired psychological well-being, and daily life interference, with only a weak association to lesion localization apart from nausea and vomiting in those with bowel-associated lesions. Combined oral contraceptives or progestins did not reduce intestinal symptoms over follow-up [ 19 ]. The symptom burden involves mechanisms beyond mechanical bowel infiltration, including pelvic inflammation, altered bowel function, medication effects, visceral hypersensitivity, and comorbid disorders of gut–brain interaction [ 19 , 20 , 21 ]. Hansen et al. described a “visceral syndrome” cluster in women with endometriosis comprising non-menstrual abdominal pain, dysuria, dyschezia, altered bowel habit, irregular bleeding, nausea or vomiting, and fatigue [ 22 ]. As a result, the same patient often consults multiple specialists—gynecologists, gastroenterologists, urologists, and pain clinicians—although all the symptoms are caused by the same underlying condition [ 22 , 23 ]. Bloating and abdominal distension, often intensifying in the luteal phase and during menstruation and associated with increased intestinal wall sensitivity, have been referred to as “endo belly” [ 20 ]. Bloating, weight fluctuation, surgical scars, pain, and impaired body function contribute to body-image distress, food-related anxiety, and disordered-eating concerns in a subset of patients [ 24 ]. Endometriosis and irritable bowel syndrome (IBS) share abdominal pain, bloating, altered bowel habit, visceral hypersensitivity, symptom fluctuation, and quality-of-life (QoL) impairment [ 20 , 21 , 25 ]. Reduced intestinal stretch pain thresholds comparable to those reported in IBS have been documented in endometriosis [ 20 , 21 ]. Women with endometriosis have higher odds of IBS than controls (OR 2.97, 95% CI 2.17–4.06), with a pooled IBS prevalence of 23.4% and individual estimates ranging from 10.6% to 52% [ 25 ]. The overlap may reflect true comorbidity, diagnostic misclassification, shared inflammatory and neural pain pathways, or common mechanisms of visceral hypersensitivity [ 21 , 25 , 26 ], and raises the question whether interventions developed for IBS are relevant to selected patients with endometriosis-associated GI symptoms, a question that requires disease-specific evidence. The gut microbiome links endometriosis, IBS-like symptoms, and dietary responsiveness. Dysbiosis has been described in both conditions, with proposed effects on intestinal permeability, immune activation, inflammatory signalling, microbial metabolites, estrobolome-mediated estrogen metabolism, and gut–brain axis regulation [ 20 , 26 ]. Although the endometriosis-specific data are still largely associative, they provide a biological rationale for studying interventions that modify luminal substrates and fermentation, reduce osmotic load, and alter microbial products and food-related symptoms [ 26 , 27 ]. The low fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (low-FODMAP) diet restricts short-chain carbohydrates that are poorly absorbed in the small intestine and rapidly fermented by colonic microbiota [ 28 , 29 , 30 ]. In IBS, randomized trials and network meta-analytic synthesis establish the low-FODMAP diet as an effective dietary intervention for global symptoms and abdominal pain [ 27 , 29 , 31 , 32 ]. Implementation follows three phases—restriction, reintroduction, and personalization—with dietitian guidance to preserve nutritional adequacy and avoid prolonged restriction [ 30 , 33 ]. The reintroduction phase remains heterogeneous in real-world practice, with variability in challenge foods, dose escalation, timing, sequencing, and follow-up [ 33 ]. Examples of higher-FODMAP foods and common lower-FODMAP alternatives are provided in Table A1 ; individual tolerance depends on portion size, preparation, baseline diet, and structured reintroduction. Evidence from IBS provides a specific rationale for studying FODMAP restriction in endometriosis-associated GI symptoms. In diarrhea-predominant IBS, low-FODMAP feeding has been associated with improved colonic barrier structure and function, reduced mast-cell recruitment and activation, lower mast-cell mediator levels, and TLR4-dependent effects linked to fecal lipopolysaccharide [ 27 ]. Pre-intervention fecal supernatants induced barrier dysfunction in experimental models, whereas post-intervention samples did not, and lipopolysaccharide removal, TLR4 antagonism, mast-cell stabilization, or mast-cell deficiency each prevented barrier dysfunction [ 27 ]. Although direct evidence in endometriosis populations is currently lacking, these mechanisms may explain the symptom overlap in women whose clinical phenotype features bloating, abdominal pain, altered bowel habits, and visceral hypersensitivity. The nutrition literature on endometriosis remains heterogeneous, with low-quality evidence, variable exposures and outcomes, and limited randomized data. Diet may influence inflammation, oxidative stress, hormone metabolism, symptoms, and quality of life (QoL), while low-FODMAP treatment appears plausible mainly for patients with coexisting endometriosis and IBS-compatible symptoms [ 34 , 35 , 36 ]. In this population, restrictive interventions require evaluation against feasibility, adherence, nutritional adequacy, psychological safety, and long-term personalization [ 24 , 33 , 34 ]. This systematic review summarizes the evidence on whether a low-FODMAP diet or FODMAP restriction improves GI symptoms, bloating or abdominal distension, abdominal or bowel-related pain, IBS-like symptom burden, pelvic pain, dysmenorrhea, QoL, and patient-reported symptom burden in women with endometriosis. Secondary outcomes include stool pattern, constipation, diarrhea, dyspareunia, adherence, adverse effects, acceptability, and patient satisfaction. GI outcomes are separated from gynecologic pain outcomes where the primary studies permit, and controlled, uncontrolled, retrospective, and single-case evidence are interpreted according to their methodological strength.

Five primary clinical reports met the inclusion criteria and were retained for descriptive synthesis: one randomized controlled crossover feeding trial [ 39 ], one prospective single-centre cohort with a structured elimination and reintroduction pathway [ 40 ], one prospective patient-preference pilot study with a parallel no-diet control arm [ 41 ], one retrospective audit of a private community-based IBS clinic with a laparoscopically defined endometriosis subgroup [ 42 ], and one structured single-patient case report from a private dietetic practice [ 38 ]. Study designs, eligibility criteria, comparators, and outcome instruments are summarized in Table 1 , protocol implementation and adherence in Table 2 , GI outcomes in Table 3 , pain, QoL, acceptability and safety in Table 4 , and risk-of-bias-relevant limitations together with the overall judgement for synthesis in Table A2 . The included studies varied substantially in sample size and recruitment context ( Table 1 ). The largest cohort, by Moore et al., retrospectively analyzed 160 women meeting Rome III criteria for IBS at a single private clinic, 59 of whom had concurrent laparoscopically diagnosed endometriosis [ 42 ]. At the other end of the spectrum, Jankovich and Watkins provided a detailed single-case report of a 23-year-old woman with laparoscopically treated ovarian endometriosis and IBS [ 38 ]. Among the prospective studies, Varney et al. randomized 35 women with a prior diagnosis of endometriosis and a baseline overall GI VAS score above 30 mm; notably, Rome criteria were deliberately not required, and 25 women completed both diet periods [ 39 ]. Keukens et al. enrolled 47 premenopausal adults with debilitating bowel symptoms, though this study required a clinical diagnosis of endometriosis (via examination, imaging, or laparoscopy), and 58.3% of completers met Rome III criteria for IBS at baseline [ 40 ]. Van Haaps et al. took a different approach, using a patient-preference design with 62 per-protocol participants who self-selected into a low-FODMAP arm ( n = 22), an endometriosis-targeted diet arm ( n = 21), or usual care ( n = 19); eligibility required a pain score of at least 3 out of 10 despite ongoing medical treatment [ 41 ]. Implementation, dietetic input, and reintroduction structure varied widely across studies ( Table 2 ). The EndoFOD trial used a fully supplied feeding design with a low-FODMAP arm below 5 g/day FODMAPs and a control arm of approximately 20 g/day FODMAPs, both modelled on the Australian Dietary Guidelines and matched for energy, macronutrients, fibre, gluten exposure, and low lactose content. Daily adherence was 75% on low-FODMAP days versus 71% on control days, with dropouts of 2 during diet 1, 2 during washout, and 6 during diet 2 [ 39 ]. Keukens et al. used a 4-week strict elimination followed by an at least 10-week dietitian-guided reintroduction, with a mean non-adherence of 8% in completers and 50% choosing to continue the diet after study end; the 13 of 47 pre-start withdrawals and the 10 of 34 post-start dropouts together constrain the analysis to motivated completers [ 40 ]. Van Haaps et al. delivered three 1 h and three 30 min consultations during the 3-month guided phase, followed by 3 months of independent continuation; 81.8% of low-FODMAP participants and 35 of 43 across both diet arms wished to continue at least partially [ 41 ]. Moore et al. taught a single low-FODMAP regimen to all Rome III patients in the clinic, with adherence captured by direct questioning, reaching 55 of 59 (93%) in the endometriosis subgroup and 91 of 101 (90%) in the IBS without known endometriosis subgroup [ 42 ]. Jankovich and Watkins applied a stepwise 16-week pathway with two individual and two group meetings, including a 2-week first-line dietary and lifestyle phase, a 6-week strict elimination including lactose, and a structured 10-week reintroduction; the substrate-specific tolerance map at full clinical re-challenge showed full-dose tolerance for lactose, sorbitol and galacto-oligosaccharides, no tolerance for mannitol and garlic fructans, minimal tolerance for fructose and wheat fructans at half of the initial challenge dose, and tolerance of the initial dose for onion fructans [ 38 ]. GI outcomes generally favoured symptom improvement after low-FODMAP restriction or dietitian-guided dietary intervention, although the strength of evidence differed by design and comparator ( Table 3 ). In the EndoFOD trial, responders were defined by a decrease of more than 20 mm in the 100 mm overall GI visual analogue scale (VAS) from baseline to the end of the dietary intervention and/or an end-of-intervention score below 30 mm. In intention-to-treat analysis, 21 of 35 participants (60%) responded to the low-FODMAP diet versus 9 of 35 (26%) to the nutritionally matched control diet ( p = 0.008). Per-protocol responders were 18 of 25 (72%) versus 8 of 25 (32%) ( p = 0.01), and week-4 overall GI symptom scores were 35 mm versus 58 mm ( p < 0.001). Patient-Reported Outcomes Measurement Information System GI (PROMIS-GI) bloating T-scores improved on the low-FODMAP diet versus baseline and control, normal stool form was recorded on 71% of low-FODMAP days against 43% at baseline and 56% on control, and PROMIS-GI diarrhea improved versus control [ 39 ]. Keukens et al. reported a primary outcome of decreased constipation after reintroduction, with the Agachan Constipation Scoring System falling from a median of 7.0 (interquartile range 5) to 5.0 (interquartile range 4), giving a mean difference of 2.1 (95% confidence interval (CI) 0.4 to 3.7, p = 0.023). Binary bloating prevalence did not change significantly; the transient decrease after elimination was non-significant ( p = 0.125), and all respondents with post-reintroduction data again reported bloating. Separately, 84% of completers reported decreased bowel symptoms, and 53% of completers reported less bloating after the diet [ 40 ]. In van Haaps et al. [ 41 ], the diet groups combined showed less bloating than controls over 6 months (mean difference −0.84, 95% CI −1.68 to −0.004, p = 0.049), but the separated low-FODMAP versus control comparison for bloating was not significant (mean difference −0.69, 95% CI −1.66 to 0.27, p = 0.159). Within the low-FODMAP arm, bloating improved from baseline ( p < 0.001), and dysuria improved from baseline ( p = 0.015). Moore et al. recorded a greater than 50% abdominal symptom reduction in 43 of 59 (72%) women with endometriosis and IBS versus 49 of 101 (49%) patients with IBS without known endometriosis ( p = 0.001; odds ratio (OR) 3.11, 95% CI 1.5 to 6.2) [ 42 ]. In the case study [ 38 ], bloating decreased from 8 of 10 to less than or equal to 3 of 10 and abdominal pain from 7 of 10 to less than or equal to 3 of 10 after the 6-week elimination, together with a transition from Bristol type 1 to 2 stools every 2 to 3 days at baseline to type 3 stools every 1 to 2 days at final follow-up. Endometriosis-specific pain and QoL outcomes were assessed most extensively in the randomized feeding trial [ 39 ] and the prospective cohort study [ 40 ]. The controlled patient-preference evaluation [ 41 ] provided narrower comparative data, whereas the retrospective audit [ 42 ] and the case report [ 38 ] included limited or no validated assessment of these domains ( Table 4 ). In the EndoFOD trial, Endometriosis Health Profile-30 (EHP-30) pain improved on low-FODMAP versus control, with a median of 22.7 versus 40.9 ( p = 0.005). EHP-30 total scores were lower on low-FODMAP than on control, 45.3 versus 47.1 ( p = 0.028), and the control and powerlessness subdomain improved nominally ( p = 0.045); GIQLI overall health-related QoL improved on low-FODMAP versus control with a median of 88 versus 81 ( p = 0.003), with parallel gains in GIQLI symptoms and physical function. According to the Depression Anxiety Stress Scales-21 (DASS-21), depression and anxiety did not differ across diets in the full intention-to-treat analysis. Among participants with abnormal baseline scores, depression decreased on low-FODMAP and anxiety showed a trend; stress decreased versus baseline, but no diet-specific difference was observed in the abnormal-stress subgroup. Pelvic pain, dysmenorrhea and dyspareunia were not assessed as specific symptom endpoints. One serious adverse event of appendicitis occurred during the washout after the control diet and was considered unlikely to be related to the intervention; two participants developed gastroenteritis-like symptoms during the control diet, with one later withdrawing, citing life stress [ 39 ]. Keukens et al. reported an EHP-30 pain improvement from 47.8 +/− 20.1 to 29.2 +/− 17.3 ( p = 0.002), with significant gains across control and powerlessness (69.4 to 36.7, p < 0.001), emotional well-being (45.2 to 29.2, p = 0.001), social support (46.4 to 31.3, p = 0.017), self-image (51.2 to 40.5, p = 0.035), work life (35.0 to 21.7, p = 0.003), and sexual intercourse (61.6 to 45.7, p = 0.023). Of those who started the diet, 63% rated it as easy to maintain. Among completers, 65% reported less pain. Across all participants, 87% described the diet as a good addition to current therapies, and 90% said they would recommend it; no harmful effects were reported [ 40 ]. Van Haaps et al. observed within-arm low-FODMAP improvements in dysuria and bloating from baseline and a separated low-FODMAP versus control reduction in deep dyspareunia (mean difference −1.15, 95% confidence interval (CI) −2.2 to −0.10, p = 0.032), together with an EHP-30 medical-profession improvement (mean difference −17.14, p = 0.018). Dysmenorrhea, chronic pelvic pain, and GIQLI did not differ significantly between the separated low-FODMAP and control arms [ 41 ]. Moore et al. [ 42 ] did not capture endometriosis-specific QoL or post-diet pain endpoints, and Jankovich and Watkins [ 38 ] reported no validated QoL or gynecological pain endpoints, although improved symptom control and trigger identification were documented at one-month follow-up after final consultation. The five reports differ substantially in their methodological rigour ( Table A2 ). The EndoFOD trial represents the highest methodological quality, with prospective registration, nutritionally matched supplied diets, menstrual-cycle-controlled timing, validated instruments, and a prespecified responder definition, but it is limited by single blinding without formal assessment of blinding effectiveness, subjective outcomes, a 10 of 35 dropout rate, single-centre Melbourne recruitment via advocacy channels, absence of original diagnostic reports, deliberate non-requirement of Rome criteria, a control FODMAP dose that may exceed habitual intake, no mechanistic biomarkers, and no reintroduction or long-term follow-up [ 39 ]. The cohort study [ 40 ] is most informative for real-world feasibility, acceptability, and within-subject symptom change among motivated completers, but lacks a control group, randomization, and blinding, with high attrition before and after diet initiation and completer-based efficacy estimation. The controlled pilot study [ 41 ] adds a no-diet comparison, but patient-preference allocation introduces selection bias; the diet arms were often pooled, no sample size calculation was performed, no follow-up extends beyond 6 months, and low-FODMAP-specific causal interpretation is limited. The audit study [ 42 ] identifies a candidate subgroup in a private clinic, but is limited by restriction to follow-up returners, absence of a placebo or control diet, use of a non-validated 40-item questionnaire, and a comparator group without systematic exclusion of endometriosis by laparoscopy or imaging. The case report [ 38 ] illustrates clinical implementation, but has no control, uses a non-validated practice symptom score, and includes concurrent first-line dietary, lifestyle, laxative, probiotic, and stress modifications. Notably, the patient had previously self-reported lactose intolerance after an unstructured home challenge, whereas structured clinical re-challenge later demonstrated full-dose lactose tolerance. Across studies, GI outcomes generally improved after low-FODMAP restriction or dietitian-guided dietary intervention ( Table 3 ). The EndoFOD trial is particularly informative because it compared a supplied low-FODMAP diet directly with a nutritionally matched control diet (ca. 20 g/day FODMAPs), giving the control condition greater dietary specificity than standard dietary advice [ 39 ]. Endometriosis-specific QoL and pain outcomes were most extensively assessed in the randomized feeding trial [ 39 ] and the prospective cohort study [ 40 ]. The controlled patient-preference evaluation [ 41 ] added narrower comparative data for deep dyspareunia and the Endometriosis Health Profile-30 (EHP-30) medical-profession domain, while dysmenorrhea, chronic pelvic pain, and GIQLI were not significantly improved in the separated low-FODMAP versus control analysis ( Table 4 ). Evidence on reintroduction and personalization remains sparse and comes from non-randomized or case-report-level evidence. The only randomized controlled trial specifically tested the supplied restriction phase only ( Table 2 ). The EndoFOD trial can therefore be interpreted as the methodologically strongest source for short-term restriction-phase efficacy, Keukens et al. [ 40 ] and van Haaps et al. [ 41 ] as the principal sources of dietitian-guided real-world feasibility and patient-reported outcomes, Moore et al. [ 42 ] as a candidate-subgroup identifier, and the case report [ 38 ] as an illustration of structured reintroduction and tolerance testing ( Table A2 ).

Five clinical reports met the inclusion criteria and generally favoured improvement of GI symptoms after low-FODMAP restriction or dietitian-guided FODMAP reduction, with different levels of causal interpretability across designs. The EndoFOD trial provides the most direct low-FODMAP-specific evidence, as its 28-day supplied intervention was tested against a nutritionally matched control diet and showed higher responder rates and lower overall GI symptom scores [ 39 ]. The two prospective non-randomized studies extend the evidence toward dietitian-guided practice, acceptability, QoL, and symptom evolution beyond short restriction, although causal attribution is limited by uncontrolled or patient-preference allocation, attrition, and pooled diet-arm analyses in one study [ 40 , 41 ]. The retrospective audit and the case report add clinical plausibility and implementation detail, but their designs cannot separate diet effects from selection, expectation, clinical contact, or co-interventions [ 38 , 42 ]. Women with endometriosis report higher rates of abdominal pain, bloating, constipation, urgency, incomplete evacuation, nausea, fatigue, and impaired daily functioning than controls [ 19 , 22 ]. The term “endo belly” further describes cyclic or peri-menstrual abdominal distension as a clinically relevant symptom, cautioning against reducing all endometriosis-associated bloating to IBS [ 20 ]. FODMAP restriction is therefore best understood as a GI symptom-directed intervention in women with endometriosis, not as a treatment for endometriosis itself. In gastroenterology, the low-FODMAP diet in IBS is supported by a large evidence base. A recent network meta-analysis of dietary interventions in IBS included 28 randomized controlled trials with 2338 patients and confirmed that the low-FODMAP diet was superior to the habitual diet for global IBS symptoms, abdominal pain, and abdominal bloating or distension, although confidence in many network comparisons remained limited [ 46 ]. Dietary approaches in this field are not interchangeable as they target different endpoints, including global-symptom burden, abdominal pain, bloating, stool pattern, and inflammatory or quality-of-life measures, and are therefore difficult to compare on any single outcome [ 46 ]. In endometriosis specifically, the diet literature remains heterogeneous in exposures and outcomes, and global dietary patterns such as the Mediterranean diet have been proposed as a safer long-term basis, with FODMAP restriction reserved as a symptom-directed option for patients with IBS-like complaints [ 35 , 47 ]. A low-FODMAP diet represents therefore one GI symptom-directed option among several, chosen according to phenotype, feasibility, and nutritional and psychological safety. FODMAPs may increase small-intestinal water content and undergo colonic fermentation, thereby promoting luminal distension, pain, bloating, and stool disturbance in susceptible patients [ 27 , 29 ]. In endometriosis, this rationale is plausible because IBS-like complaints, visceral hypersensitivity, and chronic pelvic pain frequently overlap, but the Rome criteria have not been validated specifically for women with endometriosis [ 39 ]. The endometriosis nutrition literature also stresses that restrictive diets may expose patients to nutritional and psychological burden, whereas global dietary patterns may provide a safer long-term basis, and targeted low-FODMAP intervention may be reserved for selected patients with IBS-compatible digestive symptoms [ 47 ]. The included studies support a GI symptom-specific interpretation of FODMAP restriction and do not establish a disease-modifying effect in endometriosis. The controlled feeding design in [ 39 ] is well suited to assessing short-term fermentable-carbohydrate restriction in women with endometriosis and poorly controlled GI symptoms, but is hardly applicable to patients whose dominant complaints are dysmenorrhea, dyspareunia, infertility, fatigue, or non-GI pain [ 39 ]. In the real-world cohort by Keukens et al., improvements in constipation and several EHP-30 domains were observed among completers with debilitating bowel symptoms, although the absence of a control group leaves motivation, regression to the mean, dietitian contact, and symptom fluctuation unresolved [ 40 ]. The 6-month controlled pilot supports cautious interpretation because several main analyses pooled low-FODMAP and endometriosis-diet arms. Bloating improved within the low-FODMAP arm, while the separated low-FODMAP-versus-control comparison was not significant [ 41 ]. The most plausible clinical candidates are patients with prominent bloating, abdominal pain, altered stool form, constipation, diarrhea, urgency, or IBS-like symptom clusters. The audit by Moore et al. is consistent with this selection principle, as women with Rome III IBS and laparoscopically diagnosed endometriosis had higher bowel-symptom response rates after low-FODMAP instruction than women with IBS without known endometriosis [ 42 ]. This comparison remains limited by the lack of systematic screening for endometriosis in the comparator group and by the inclusion of follow-up returners only. The main biological rationale for low-FODMAP is a reduction in osmotic load, fermentation, gas production, luminal distension, and stimulation of visceral pain pathways. The included endometriosis studies did not measure breath hydrogen or methane, intestinal permeability, inflammatory biomarkers, microbiome composition, or metabolomic change. Among potentially transferable mechanisms are toll-like receptor 4-, lipopolysaccharide-, and mast-cell-dependent epithelial barrier pathways recently characterized in diarrhea-predominant IBS, in which low-FODMAP feeding reduced fecal lipopolysaccharide load and prevented mast-cell-driven barrier dysfunction in experimental models [ 27 ]. Zhao et al. therefore correctly noted that mechanistic claims in the EndoFOD context remain inferential without physiological endpoints and that future studies should include objective measures if they aim to link symptom response with fermentation, barrier function, microbiome change, or inflammation [ 48 ]. The broader IBS microbiome literature argues against prolonged unguided restriction. Low-FODMAP restriction can improve symptoms but may reduce Bifidobacterium and alter microbial composition during the restrictive phase [ 49 ]. This concern is relevant to endometriosis because many patients already experiment with gluten-free, dairy-free, “anti-inflammatory”, low-histamine, low-nickel, or other elimination diets, often without dietetic supervision [ 47 ]. This mirrors observations in celiac disease, where FODMAP restriction has been proposed to address persistent IBS-like symptoms, while gluten avoidance remains first-line therapy [ 50 ]. When implemented, a low-FODMAP diet should be structured, time-limited, professionally supervised, and followed by reintroduction when restriction results in a clinical improvement. The broader endometriosis-specific diet literature further complicates a blanket application of an IBS-derived low-FODMAP model. Onion and garlic are commonly restricted during the low-FODMAP phase because of their fructan content, but both belong to food groups with experimental or clinical endometriosis-specific data: quercetin inhibited endometriosis-cell proliferation through cyclin-D1- and microRNA-related pathways, high-dose Allium cepa reduced proliferative features and Ki67 expression in rat endometriotic implants, and garlic tablets reduced pelvic pain, back pain, dysmenorrhea, and dyspareunia in a randomized placebo-controlled trial [ 51 , 52 , 53 ]. S-allyl-L-cysteine, a major aged-garlic constituent, also inhibited lesion growth in a mouse model and modulated adhesion-, apoptosis-, and immune-inflammatory pathways, while aged-garlic extract has been proposed as a prophylactic candidate on mechanistic grounds [ 54 , 55 ]. These data argue against presenting low-FODMAP restriction as an intrinsically anti-endometriosis diet. If a low-FODMAP diet is applied, temporary restriction of onion and garlic remains reasonable during the restriction phase to reduce fructan-related symptom provocation. High-FODMAP foods with plausible endometriosis-relevant bioactivity should therefore be treated as context-dependent restriction targets: they may be temporarily restricted during a defined GI symptom trial but should be systematically reintroduced when response permits. Reintroduction and personalization remain insufficiently studied in the endometriosis-specific context. As the sole randomized trial examined only the 28-day restriction phase, it demonstrates short-term symptom control but does not address long-term dietary self-management [ 39 ]. The two prospective non-randomized studies incorporated reintroduction and personalization components, but their designs do not allow phase-specific causal attribution [ 40 , 41 ]. The case report illustrates the clinical role of structured re-challenge, as supervised lactose testing did not confirm prior self-reported dairy intolerance after an unstructured home challenge and helped identify tolerated and poorly tolerated FODMAP subgroups in the individual patient [ 38 ]. Safety and acceptability were reassuring but incompletely assessed. No diet-related serious safety concern was reported in the randomized trial, and no side effects were reported in the controlled patient-preference study [ 39 , 41 ]. Feasibility was more complex in the prospective cohort, where 13 of 47 participants withdrew before starting, mostly because of motivation or time constraints, and 10 of 34 discontinued after starting, although completers reported low average non-adherence and high satisfaction [ 40 ]. These data suggest good feasibility among motivated completers, with lower certainty for the broader eligible population. Longer follow-up is needed because the included studies do not adequately assess nutritional adequacy, food-related QoL, microbiome effects, or sustainability. In IBS cohorts, validated screening tools have detected possible eating disorders in approximately one-fifth to one-quarter of patients, with avoidant/restrictive food intake disorder presentations in up to one-third of neurogastroenterology referrals [ 56 , 57 ], and reported prevalence ranging more widely depending on the screening instrument applied [ 58 ]. In women with endometriosis, screening studies point to roughly one-third for possible eating disorders and close to half for concurrent IBS-compatible symptoms, alongside a near-threefold genetic liability to eating disorders independent of body mass index, chronic pain, and IBS [ 59 ]. Chronic pain with cyclic GI exacerbation, body-image distress related to bloating, weight fluctuation and surgical scars, and food-related anxiety from repeated symptom-trigger learning may further raise the risk of disordered eating in this population [ 24 , 56 , 59 ]. None of the included studies screened for eating-disorder symptoms before or during intervention or reported food-related QoL as an endpoint. When low-FODMAP intervention is offered, reasonable safeguards include pre-intervention screening with a validated instrument such as Sick, Control, One stone (14 lb/6.35 kg), Fat, Food eating-disorder screening questionnaire (SCOFF) or the nine-item avoidant/restrictive food intake disorder (ARFID) screen, attention to unintended weight loss, and avoidance of unsupervised indefinite restriction [ 56 , 57 , 58 , 59 ]. The EndoFOD trial strengthened the evidence base by testing FODMAP restriction under randomized, controlled feeding conditions. Its supplied meals, nutrient matching, menstrual-cycle timing, validated GI and QoL instruments, and prespecified responder definition strengthen internal validity [ 39 ]. The limitations discussed by Zhao et al. include small sample size, attrition, Melbourne-only recruitment related to food delivery, recruitment through advocacy networks, limited cultural and dietary diversity, exclusion of women without English proficiency or with irregular cycles, single blinding without formal assessment of blinding success, subjective outcomes, absence of mechanistic biomarkers, and possible carryover effects despite washout [ 48 ]. The non-randomized studies provide valuable data on feasibility, selection effects, and clinical applicability. Dietitian-guided pathways appeared acceptable for motivated completers, but attrition before or after diet initiation shows that the intervention can be demanding even when professional support is available [ 40 ]. The controlled patient-preference design extends observation to 6 months, although self-selection into the diet arms limits causal attribution because motivation and expectation were probably unevenly distributed between groups [ 41 ]. The retrospective IBS-clinic audit contributes a clinically relevant comparison between women with IBS and known endometriosis and women with IBS without known endometriosis, but its comparator cannot be treated as a systematically screened endometriosis-negative group [ 42 ]. Future trials should define whether low-FODMAP restriction reduces symptoms, which patients benefit, how long improvement persists, and under which dietary conditions the intervention remains feasible. Parallel-group designs may be preferable when durable dietary and microbial adaptations are expected. Comparators should include active dietary alternatives, such as Mediterranean, anti-inflammatory, gluten-free, or simplified FODMAP-focused approaches, alongside habitual-diet or no-diet controls, since head-to-head IBS evidence shows comparable global-symptom responder rates across some approaches with potentially lower implementation burden [ 46 ]. A recent RCT-focused review across GI disorders similarly emphasized heterogeneous protocols, variable outcome measures, the need for dietitian supervision, and caution against treating low-FODMAP restriction as broadly indicated [ 32 ]. Baseline phenotyping should separate bowel lesion status, IBS criteria, constipation- versus diarrhea-predominant symptoms, hormonal treatment, opioid use, prior surgery, menstrual suppression, bloating severity, and “endo belly” pattern. Future pooling would be more reliable if trials used a shared minimum outcome set, including a validated global GI response measure, an endometriosis-specific pain assessment such as the EHP-30 pain domain, a structured dyspareunia visual analogue scale, dietary adherence, nutritional adequacy, food-related QoL, adverse effects, and long-term continuation. Cost and access also require assessment, because gluten-free and low-FODMAP-compatible products are often more expensive than standard counterparts, while access to dietitian support varies across health systems [ 56 ]. Breath testing, stool microbiome profiling, fecal metabolomics, inflammatory markers, intestinal permeability measures, and menstrual-cycle-linked symptom diaries could clarify whether responders share identifiable physiological characteristics. Such data would help distinguish fermentable-carbohydrate sensitivity from visceral hypersensitivity, peripheral and central sensitization, bowel endometriosis, medication-related dysmotility, defecatory dysfunction, or other overlapping mechanisms. A negative response to a supervised restriction phase should lead to stopping the diet and reassessing the symptom pathway instead of intensifying avoidance. For current practice, a low-FODMAP diet may be cautiously considered when GI symptoms are prominent, particularly bloating, abdominal pain, altered stool pattern, constipation, diarrhea, urgency, or IBS-like complaints. It should be introduced as a trial of GI symptom management, ideally by a dietitian trained in FODMAP protocols and familiar with endometriosis-related dietary burden. The restrictive phase should be short, response should be assessed with predefined symptom targets, and continuation should depend on meaningful improvement. Reintroduction and personalization should follow to minimize unnecessary restriction and identify individual tolerance patterns. The diet should not be presented as a general “anti-endometriosis” diet or a substitute for gynecological care. Figure 2 summarizes the proposed clinical pathway: pre-intervention assessment, time-limited restriction with predefined response criteria, structured substrate-by-substrate reintroduction, long-term personalization, and continuous safeguards including cycle-aligned symptom recording, eating-disorder screening, and co-monitoring with gynecological care. Several aspects of this review warrant caution. The search was focused on low-FODMAP and related fermentable-carbohydrate terms, so it was not designed to evaluate all dietary interventions in endometriosis. Only five clinical reports met the eligibility criteria, and their designs, populations, comparators, duration, and outcome instruments differed substantially. The inclusion of a case report and a retrospective audit was justified by the small evidence base, but these designs were interpreted descriptively. Publication bias is possible because small dietary studies with null findings may be less likely to appear in indexed literature.

The available evidence indicates that a low-FODMAP diet may reduce GI complaints in selected endometriosis patients, including bloating, abdominal pain, altered stool pattern, constipation, or IBS-like symptoms. The limited evidence base provides the strongest support for short-term restriction under controlled conditions, and weaker supporting evidence from real-world dietitian-guided pathways. Current data do not justify presenting the low-FODMAP diet as a “treatment” method for endometriosis itself. In clinical practice, FODMAP restriction should therefore be considered a time-limited, dietitian-guided GI symptom intervention, followed by structured reintroduction if clinical improvement occurs. Future studies should clarify patient selection, durability after reintroduction, and the nutritional and psychosocial consequences of longer-term use.