The new Rome V criteria for irritable bowel syndrome (IBS) capture more people than Rome IV, and identify a group with milder symptoms, but the authors estimate the criteria still leave out 30% to 50% of people who say they have IBS, according to a UK registry study.
Published online in Clinical Gastroenterology and Hepatology, the study compared the Rome III, IV, and V criteria for diagnosing IBS. Rome V expands the definition to include abdominal discomfort as well as pain and lowers the required frequency of symptoms from once a week to three days a month. It also specifies that abdominal pain cannot be continuous.
“This analysis should not be interpreted as demonstrating that one set of criteria is definitively better than another,” first author Kyle Staller, MD[, MPH — paper lists MD only; confirm], of the Center for Neurointestinal Health, Division of Gastroenterology, Massachusetts General Hospital and Harvard Medical School, Boston, told GI & Hepatology News. “Rather, it shows that diagnostic criteria are not merely semantic: they influence which patients receive a label, how patients understand their illness, and which populations are represented in research.”
Dr. Staller and colleagues analyzed survey data from adults enrolled in ContactME-IBS, a UK registry of 4,280 people who report having IBS and have volunteered to take part in research. Participants were recruited by email in July 2021 and completed online questionnaires covering digestive symptoms, mental health, quality of life, impact on work and daily activities, and IBS-related health care costs. All were invited to complete the same questionnaires 12 months later. Among the 1,278 people recruited, mean age was 47.2 years (range, 18 to 89 years), 85% were female, and 96.6% were white. A total of 148 participants (11.6%) said their IBS symptoms began after an acute enteric infection.
Because Rome V was not published until 2026, the researchers could not administer a Rome V questionnaire. Instead, they applied the Rome III, Rome IV, and Rome V scoring algorithms retroactively to symptom data that had been collected with the Rome III and Rome IV questionnaires, which contain most of the items making up the Rome V criteria.
The researchers ran two analyses that differed in how they applied the Rome V exclusion for continuous abdominal pain. Both used a yes/no item from the Rome IV questionnaire asking whether abdominal pain had been continuous or almost continuous, defined as never going away during waking hours.
In the main analysis, participants were excluded from the Rome V definition only if they answered yes and also reported pain every day of the week — the approach recommended by the developers of the Rome V diagnostic questionnaires. In a sensitivity analysis, anyone answering yes was excluded regardless of how often pain occurred. The authors ran the second analysis because the Rome V validation study found that some people reported continuous pain even when pain was not present daily.
Among 1,275 participants with complete data, 995 (78%) met Rome III criteria for IBS, 752 (59%) met Rome IV criteria, and 893 (70%) met Rome V criteria. Nearly all who met Rome V also met Rome III (98.9%), while 72.1% also met Rome IV. Agreement between Rome V and Rome III was substantial (kappa = 0.75), while agreement between Rome V and Rome IV was only fair (kappa = 0.40). Kappa measures how closely two sets of criteria classify the same individuals, adjusted for agreement that would occur by chance; higher values indicate closer agreement, with 1.0 representing perfect agreement.
Under the stricter sensitivity definition, 598 participants (46.9%) met Rome V criteria. Agreement dropped to fair against Rome III (kappa = 0.38) and slight against Rome IV (kappa = 0.17). Among participants who met Rome III but not Rome V criteria, 93.8% in the main analysis and 98.5% in the sensitivity analysis reported continuous abdominal pain.
The groups differed substantially in symptom severity. Compared with the 108 participants who met Rome IV but not Rome V criteria, those who met Rome V criteria had lower mean scores on the IBS Severity Scoring System, a 0 to 500 scale on which higher scores indicate worse symptoms and scores of 300 or above denote severe disease (252.6 vs. 385.0; P < .001). They were also less likely to fall into the severe category (32.5% vs. 85.2%), to report that IBS limited activities at least half the time (67% vs. 90.7%), to experience urgency most days (22.8% vs. 56.5%), or to report fecal incontinence at least once a week (13% vs. 28.7%). All comparisons were significant at P < .001.
Participants who met Rome V criteria also reported fewer psychological symptoms. On the Hospital Anxiety and Depression Scale, which scores anxiety and depression separately from 0 to 21 with scores of 11 or higher considered abnormal, 42.2% of the Rome V group had abnormal anxiety scores, compared with 62% of those who met Rome IV but not Rome V criteria. Abnormal depression scores were reported in 17.1% vs. 46.3%. Mean scores followed the same pattern for anxiety (9.8 vs. 12.6) and depression (6.6 vs. 10.0), with all comparisons significant at P < .001.
The Rome V group reported better disease-specific quality of life on the IBS-QOL, scored from 0 to 100 with higher scores indicating better quality of life (54.7 vs. 32.0), and was less likely to report IBS-related absenteeism (22.7% vs. 50.9%; P < .001 for both).
Self-reported IBS-related direct health care costs in the 12 months before recruitment were also lower among participants meeting Rome V criteria, at £415 vs. £1,046 — roughly $556 vs. $1,400 at the conversion rate the authors applied (P < .001). In the sensitivity analysis, the gap was £342 vs. £759.
Only 604 of the 1,275 participants (47.3%) were successfully followed up at 12 months, and 603 provided complete symptom data. Among them, 78.2% of those who met Rome V criteria at baseline still met them a year later, compared with 73.6% for Rome IV and 83.8% for Rome III.
Agreement between baseline and follow-up status, which accounts for participants who moved into a diagnosis as well as those who moved out of it, was lower for Rome V (kappa = 0.31) than for Rome IV (kappa = 0.41). Among participants who did not meet Rome V criteria at baseline, 45.9% met them at 12 months, compared with 32.3% for Rome IV. Under the stricter sensitivity definition, 62.8% of the Rome V group still met criteria at 12 months (kappa = 0.29). The authors concluded that Rome V-defined IBS was less stable than Rome IV-defined IBS over follow-up.
The researchers said the findings could affect how IBS is identified in clinical practice and how results from studies using different Rome criteria are applied to patients.
The study had several limitations. Participants were recruited online, which may have introduced selection bias, and most were women and white. Fewer than half of participants completed 12-month follow-up. Researchers also used questionnaire items designed for earlier Rome criteria and could not rule out other GI diseases in every participant. The authors additionally noted that Rome IV asks about pain that is “continuous or almost continuous,” while Rome V specifies pain that is continuous throughout the day, and that this difference in wording may have altered which participants qualified under Rome V.
Tillotts Pharma UK Ltd. provided unrestricted research funding and had no role in study concept, design, analysis, or reporting. Dr. Staller reported research support from Ardelyx, an advisory role with Mindset Health and consulting for AbbVie, Ardelyx, Atmo, Laborie, Salix, and Takea. The other authors reported no conflicts of interest.
Expert Insight
GI & Hepatology News asked Dr. Staller to discuss what the findings mean for clinical practice.
How do you anticipate the adoption of Rome V criteria will change diagnostic evaluation and treatment decisions for patients who meet Rome IV criteria but no longer qualify for Rome V IBS?
I would not interpret failure to meet Rome V criteria as a reason to reflexively remove the IBS diagnosis or begin an entirely new, exhaustive diagnostic evaluation. The first step is to understand why the patient no longer meets criteria. In our analysis, the principal reason was continuous abdominal pain, and the Rome IV-positive/Rome V-negative group actually had greater symptom severity, psychological comorbidity, functional impairment, and health care costs — not a trivial or clinically unimportant symptom burden.
Clinically, these patients should be re-phenotyped rather than simply relabeled. A targeted evaluation for alternative organic disease remains appropriate when suggested by the history, bowel pattern, or alarm features, but the change in criteria alone should not prompt indiscriminate testing. Treatment should continue to address the dominant bowel symptoms and mechanisms. At the same time, when continuous pain is prominent, clinicians should consider whether centrally mediated pain or overlapping disorders of gut-brain interaction warrant greater emphasis on neuromodulation, brain-gut behavioral treatment, and multidisciplinary care. Most importantly, patients should not hear that they “no longer have IBS” and therefore no longer have a legitimate or treatable disorder.
Given that Rome V identifies a less symptomatic cohort, do you believe prior randomized trials conducted using Rome III/IV criteria remain generalizable to patients diagnosed under Rome V?
I believe the existing trial literature remains highly relevant, but its generalizability should not be assumed to be complete. There is substantial overlap between the populations: in the main analysis, almost all participants meeting Rome V criteria also met Rome III criteria, and approximately 72% also met Rome IV criteria. We therefore should not discard decades of therapeutic evidence simply because the diagnostic framework has changed.
Nevertheless, Rome IV trials were often enriched for patients with greater symptom severity and psychological comorbidity. Treatment effect sizes, placebo responses, adherence, and the perceived value of an intervention may differ in the less severely affected population identified by Rome V. Conversely, the greater severity of Rome IV-defined IBS may have made some dietary, pharmacologic, or brain-gut behavioral interventions appear less effective than they might in a broader community population.
The most useful approach will be to examine whether an individual patient resembles the trial population in terms of bowel subtype, pain severity, psychological burden, and prior treatment exposure — not merely whether the same version of the Rome criteria was used. Future trials should ideally report cross-classification by Rome III, IV, and V and separately characterize participants with continuous pain.
Should clinicians routinely evaluate patients with continuous abdominal pain who fail Rome V criteria for centrally mediated abdominal pain syndrome or other disorders of gut-brain interaction rather than IBS?
Yes, centrally mediated abdominal pain syndrome and other disorders of gut-brain interaction (DGBI) should be considered systematically, but continuous pain should not be treated as synonymous with centrally mediated abdominal pain syndrome. The exclusion was incorporated into Rome V specifically to distinguish IBS from centrally mediated abdominal pain syndrome, so it appropriately prompts clinicians to reconsider the formulation.
That reassessment should include whether the pain remains meaningfully related to defecation or stool changes, whether another bowel DGBI, such as functional diarrhea, chronic constipation, or functional abdominal bloating, better describes the presentation, and whether there is evidence of overlapping central pain amplification. The better clinical question is not necessarily “IBS or centrally mediated abdominal pain syndrome?” but rather, “What combination of bowel dysfunction and pain mechanisms is driving this patient's illness?”