When Food Avoidance Meets Compatible Diagnostic Criteria: Avoidant/Restrictive Food Intake Disorder Features in Irritable Bowel Syndrome

• IBS (irritable bowel syndrome): A common, long-lasting gut condition. It causes belly pain and changes in bowel habits (diarrhea, constipation, or both) without visible damage to the gut.

• DGBI (disorders of gut–brain interaction): The current name for "functional" gut disorders such as IBS. Symptoms come from how the gut and brain talk to each other, not from structural disease.

• FD (functional dyspepsia): Ongoing upper-stomach discomfort, such as feeling full too early, fullness after meals, or upper belly pain, with no identifiable cause.

• ARFID (avoidant/restrictive food intake disorder): An eating disorder in which a person eats too little or too narrow a range of foods. It is not driven by concern about weight or body shape. It comes in three forms: avoiding foods because of their sensory features ("picky"), fear of bad consequences of eating ("fear"), or low interest in food ("appetite").

• ARFID-like symptoms: Scoring above the cut-off on a screening questionnaire. This is not a diagnosis.

• DSM-5: The American Psychiatric Association's diagnostic manual for mental disorders.

• "DSM-5-compatible" ARFID: The authors' approximation of a diagnosis. It requires a positive screen, no other eating disorder, and evidence of real harm from the restrictive eating. It was not confirmed by a specialist interview.

• NIAS (Nine Item ARFID Screen): A short self-report screening questionnaire for ARFID. NIAS-Fr is the French version.

• Rome IV criteria: The international standard criteria used to diagnose IBS, FD, and other DGBI.

• Medical / psychosocial impairment: Physical harm (weight loss, nutrient deficiency, needing nutritional supplements or tube feeding) or harm to daily, social, and emotional life.

• Tertiary care center: A specialized hospital that sees referred, usually more severe, cases.

• Odds ratio (OR): How much the odds of an outcome change with a given factor. An OR above 1 means higher odds.

• Logistic regression: A statistical method that tests which factors are independently linked to a yes/no outcome while holding the other factors constant.

• Nagelkerke R²: A rough measure of how much of the outcome a statistical model explains. 0.20 means about 20%.

1. How common are ARFID-like symptoms, based on a positive screening questionnaire, among adults with IBS?

2. How common is ARFID when DSM-5-compatible criteria are applied, meaning a positive screen plus documented medical or psychosocial impairment and no other eating disorder?

3. Which clinical and psychological factors are independently linked to (a) ARFID-like symptoms and (b) DSM-5-compatible ARFID?

4. Do patients who only screen positive differ from those who meet the fuller DSM-5-compatible criteria?

The authors expected that screening alone would overestimate ARFID. They also expected that adding the impairment criteria would pick out a more clinically meaningful group.

Sample: 326 adults with IBS (Rome IV), seen at Rouen University Hospital, France, between January 2024 and February 2025. 80.4% were women, with an average age of about 43. On average, patients had moderate IBS severity, mild anxiety and depression, mild insomnia, and medium somatic symptom burden. 60.4% also had functional dyspepsia, and 4.9% had a diagnosed eating disorder other than ARFID, who were excluded from later analyses.

Prevalence

• 47.5% had ARFID-like symptoms (positive screen).

• 33.4% met DSM-5-compatible criteria for ARFID.

• "Fear" of aversive consequences was the most common presentation in both groups (about 72–73%), followed by low appetite and then picky eating.

How impairment was established in the DSM-5-compatible group (n = 109): Almost everyone (92.6%) qualified through psychosocial impairment, measured by a quality-of-life subscale on food avoidance. Far fewer had medical impairment: 13.8% had a nutritional deficiency, 5.5% significant weight loss, and 6.4% dependence on nutritional support.

Comparison with IBS patients without an eating disorder: Patients with DSM-5-compatible ARFID had:

• More functional dyspepsia (69.7% vs. 53.2%)

• About twice the rate of following a specific diet (41.3% vs. 21.9%)

• Clearly lower quality of life

• More severe IBS, more anxiety, depression and somatic symptoms, worse sleep, and more insomnia

ARFID was equally common across IBS subtypes (diarrhea, constipation, mixed). Age, sex, and BMI did not differ.

Factors independently linked to each definition

• ARFID-like symptoms: depressive symptoms (OR 1.10), somatic symptom severity (OR 1.06), and functional dyspepsia (OR 1.73). The model explained about 14% of the variation.

• DSM-5-compatible ARFID: depressive symptoms (OR 1.12) and IBS symptom severity (OR 1.004 per scale point). Functional dyspepsia was no longer significant. The model explained about 20%. The paper reports this as 0.203 in the abstract and table but 0.208 in the results text.

Screen-positive only vs. full criteria: Among patients who screened positive, those meeting full criteria had worse quality of life and more anxiety, depression, somatic symptoms, and IBS severity. They were also more likely to follow a specific diet (41.3% vs. 18.9%). In the adjusted model, only IBS symptom severity separated the two groups, and the effect per point was small (OR 1.008).

Screening alone is not enough. Screening captured more patients than the fuller criteria, and the two definitions were linked to different factors. So "screens positive" and "meets DSM-5-compatible criteria" don't describe the same population. Meeting the fuller criteria seems to go with more severe gut symptoms.

Overlap with functional dyspepsia. FD predicted screening positive but not meeting the full criteria. The authors suggest this is because FD symptoms, especially early fullness and post-meal distress, look like the questionnaire's "low appetite" items. Requiring evidence of real impairment may help separate true ARFID from this symptom overlap. The authors caution that the confidence interval was wide, so a meaningful FD effect can't be ruled out.

Depression. Psychological comorbidity is common in both IBS and ARFID. Having both conditions may make these difficulties worse. The models explained only about 20% of the outcome, so other factors likely matter. Candidates include hypervigilance to gut sensations, visceral hypersensitivity, major life stress, and other medical or psychological conditions.

Diets. Patients following a specific diet were more likely to meet the criteria. This matches earlier work linking self-imposed exclusion diets to ARFID symptoms. The study cannot tell whether the diet came before the ARFID or resulted from it.

Somatic symptom burden. Somatic symptoms were linked to ARFID-like symptoms, and in the authors' earlier work, to carbohydrate intolerance. The authors stress that these links are not causal.

Strengths

• A well-characterized IBS cohort diagnosed with Rome IV criteria

• Validated questionnaires

• Regression variables chosen in advance, with safeguards against overfitting

• Medical impairment assessed objectively from hospital records, not self-report

Limitations acknowledged by the authors

• The NIAS has not been validated in patients with gut symptoms.

• No ARFID diagnosis was confirmed by a specialist clinician.

• Using the IBS-QOL food-avoidance subscale (cut-off ≤ 40) to measure psychosocial impairment has not been validated. It may partly measure IBS burden itself and inflate prevalence. Stricter cut-offs gave mostly similar group differences. Counting medical impairment alone, only 5.8% of patients would qualify.

• Weight and shape concerns could not be checked systematically, so some cases may not truly be ARFID.

• Patients came from a single tertiary center, so the results may not apply to primary care.

• The DSM-5 gives no operational thresholds for impairment, and no validated ARFID tools exist for adult gut patients.

Recommendations

• Clinicians should know that IBS and ARFID often overlap. They should distinguish adaptive dietary changes made to manage symptoms from maladaptive restrictive eating.

• Patients who follow specific or exclusion diets deserve particular attention.

• All DSM-5 criteria, not just a screening score, should be considered for a personalized approach.

• Research should develop reliable, standardized ways to apply DSM-5 ARFID criteria in DGBI populations. It should also clarify mechanisms and the direction of these relationships.