Aliment Pharmacol Ther. 2021 Mar 11. doi: 10.1111/apt.16322. Online ahead of print.

Christopher J Black ^1 2, Yan Yiannakou ³, Elspeth Guthrie ⁴, Robert West ⁴, Lesley A Houghton ¹, Alexander C Ford ^1 2

Author information

• ¹Leeds Institute of Medical Research at St. James's, University of Leeds, Leeds, UK.
• ²Leeds Gastroenterology Institute, St. James's University Hospital, Leeds, UK.
• ³County Durham and Darlington NHS Foundation Trust, Durham, UK.
• ⁴Leeds Institute of Health Research, University of Leeds, Leeds, UK.

Abstract

Background: Conventionally, irritable bowel syndrome (IBS) is subgrouped using predominant stool form, yet it is a complex disorder, with multiple biopsychosocial contributors. We previously derived and validated a latent class model subgrouping people with IBS into seven clusters based on gastrointestinal and extraintestinal symptoms and psychological profile.

Aims: To conduct longitudinal follow-up examining the natural history and prognostic value of these clusters.

Methods: Participants completed a 12-month follow-up questionnaire. We applied our model to these data, comparing cluster membership between the two time points in those still meeting Rome IV criteria at follow-up, including stratifying the analysis by predominant stool pattern, and level of psychological burden, at baseline. We examined whether baseline cluster predicted the course of IBS, and whether starting new treatment was associated with changing cluster.

Results: Eight hundred and eleven participants met Rome IV criteria for IBS at baseline, of whom 452 (55.7%) responded, and 319 (70.6%) still met Rome IV criteria for IBS at follow-up. Of these, 172 (53.9%) remained in the same IBS cluster as at baseline and 147 changed cluster. Cluster membership stratified according to psychological comorbidity was more stable; 84% of those in a cluster with high psychological burden at baseline remained in such a cluster at follow-up. People in clusters with high psychological burden at baseline had more severe symptoms (P < 0.001), received a higher mean number of subsequent treatments (P < 0.001), and were more likely to consult a doctor than people in clusters with low psychological burden (P < 0.001). There was no significant association between starting a new treatment and changing cluster at follow-up.

Conclusions: Longitudinal follow-up demonstrated little transition between clusters with respect to psychological burden, and these appeared to predict disease course. Directing treatment according to cluster, including earlier use of psychological therapies, and exploring how this approach influences outcomes in IBS, should be examined.