Mixture model analysis identifies irritable bowel syndrome subgroups characterised by specific profiles of gastrointestinal, extraintestinal somatic and psychological symptoms

Background: Current subgrouping of Irritable Bowel Syndrome (IBS) is exclusively based on stool consistency without considering other relevant gastrointestinal (GI), extraintestinal somatic or psychological features. Aim: To identify subgroups based on a comprehensive set of IBS-related parameters. Methods: Mixture model analysis was used, with the following input variables: 13 single-item scores from the IBS-specific Gastrointestinal Symptom Rating Scale, average stool consistency and frequency from a 7-day Bristol Stool Form diary, 12 single-item extraintestinal symptom scores from the Patient Health Questionnaire-12, and anxiety and depression subscale scores from the Hospital Anxiety and Depression scale. The resulting latent subgroups were compared regarding symptom profiles using analysis of variance followed by pair-wise comparisons. Results: One hundred and seventy-two IBS patients (Rome III; 69% female; mean age 33.7 [range 18-60] years) were included. The optimal subgrouping showed six latent groups, characterised by: (I) constipation with low comorbidities, (II) constipation with high comorbidities, (III) diarrhoea with low comorbidities, (IV) diarrhoea and pain with high comorbidities, (V) mixed GI symptoms with high comorbidities, (VI) a mix of symptoms with overall mild severity. The subgroups showed differences in the distribution of Rome III-subtypes, IBS severity, presence of anxiety and depression, and gender, but not regarding age, IBS duration or reported post-infectious onset of IBS. Conclusions: This model-based subgrouping of IBS partly supports the distinction of subgroups based on bowel habits, but additionally distinguishes subgroups with or without co-morbid extraintestinal somatic and psychological symptoms. The resulting groups show specific profiles of symptom combinations.