This study demonstrates that symptoms compatible with a Rome IV bowel DGBI are highly prevalent, affecting over 50% of people diagnosed with NAFLD. Most of these patients fulfilled Rome IV IBS criteria, the most severe bowel DGBI. The prevalence of Rome IV IBS in our cohort of patients with NAFLD in the liver assessment clinic was 35.2%; nearly ten times the published prevalence of IBS in global and UK populations, based on an estimated Rome IV IBS prevalence of 4.3% in the UK 8.
Importantly, our findings corroborate previous studies reporting a prevalence of IBS in patients with NAFLD of 29.2% and 29.4% respectively14,15. Our findings are also consistent with recent data suggesting that patients with NAFLD are significantly more likely to develop IBS than those without NAFLD23. In contrast to epidemiological data in general populations where IBS is estimated to affect approximately 1.5 times more females than males24, our data in a NAFLD population did not show a significant difference in the prevalence of Rome IV IBS between males and females with NAFLD. This disparity most likely reflects the demographics of the NAFLD population studied, rather than sex-specific differences in the pathophysiology of IBS in patients with NAFLD.
Interestingly, a further twenty percent of patients in our study who did not meet the Rome IV criteria for IBS, met the criteria for other, less severe, Rome IV bowel DGBI which were also associated with significant impairment of QoL.
Similar to previous studies of patients with NAFLD and chronic diarrhoea25, there was no association between the presence of Rome IV IBS and liver stiffness measurements in our study. Moreover, there were no associations between IBS symptom severity, liver stiffness measurements, blood-based fibrosis assessments or steatosis measurements.
Despite the high prevalence of IBS in our study, only 16% of patients with Rome IV IBS had a prior diagnosis of IBS. This suggests that there is a large proportion of patients with undiagnosed IBS attending liver clinics with NAFLD, often with severe symptoms. The importance of this is that 86% of those with NAFLD who met the Rome IV criteria for IBS had moderate-to-severe symptoms. This degree of IBS symptom severity seen in the liver clinic is above the UK national average8 and is similar to that seen in tertiary IBS clinics26. As far as the authors are aware, none of the previous literature describing the incidence of IBS in NAFLD used any validated symptom severity scores for IBS. Indeed, our data do suggest that Rome IV IBS is associated with considerable morbidity with a significantly worse QoL, impaired ability to perform activities of daily living, and an increased incidence of severe anxiety and depression, compared to patients with NAFLD without IBS. This replicates findings from Jones-Pauley et al.14 and it is well documented that IBS is associated with increased levels of anxiety and depression11. Troublesome symptoms such as diarrhoea and incontinence can be distressing for patients and difficult to manage medically leading to frustration and a reduced QoL9,27. This study will therefore be important in raising awareness of the burden of overlapping Rome IV bowel DGBI and NAFLD in patients presenting to liver clinics.
Recent evidence-based clinical guidelines on bowel DGBI place an emphasis on making a positive diagnosis of IBS in the absence of alarm features, and also provide guidance on which patients would benefit from targeted investigations to exclude IBS mimics including coeliac disease, microscopic colitis, and bile salt malabsorption28. The findings of our study suggest that patients with NAFLD may not always be forthcoming about their bowel symptoms. The findings are therefore likely to be of importance to both luminal gastroenterologists and hepatologists looking after people with these conditions. This study suggests that screening for bowel DGBI in patients with NAFLD attending liver clinics has the potential to allow a positive diagnosis, adjust any medications (e.g. metformin) which may contribute to symptoms, order relevant tests to exclude mimics, and then proceed with holistic symptomatic treatments including dietary, medical and behavioural approaches to improve bowel symptoms.
As far as the authors are aware, this is the first study investigating the overlap of the entire spectrum of Rome IV bowel DGBI to include liver fibrosis assessment using liver stiffness measurements to determine liver disease severity, as well as a validated instrument to assess IBS symptom severity. However, there are several limitations to our study. Firstly, patients in ID LIVER hepatology assessment clinics did not undergo luminal gastroenterology evaluations to exclude IBS mimics other than coeliac disease. Reassuringly, however, all patients included had a negative coeliac serology as part of their autoimmune screen, and a structured past medical and surgical history was taken to exclude red flags and none of those included had any known organic gastrointestinal diagnoses. However, in the sub-set of patients with diarrhoea predominant symptoms (22/79 (27.8%) patients) with a Rome IV bowel disorder of gut brain interaction (functional diarrhoea or diarrhoea predominant IBS), it is recognised that around a third of patients would benefit from investigations to exclude bile salt malabsorption28, and up to 8% may have microscopic colitis29. Secondly, we cannot exclude responder bias. Only 146/340 (43%) of patients screened with NAFLD agreed to complete the bowel questionnaires. Those with significant bowel symptoms may have been more motivated to complete the bowel questionnaires than those without symptoms. One reason for the lower recruitment rate, and a potential source of accrual bias, is that some patients were unable to complete questionnaires due to language barriers (the questionnaires were only available in English). Thirdly, NAFLD was frequently found incidentally as part of investigations for other abdominal symptoms and hence referred into clinic. It is therefore possible that more symptomatic patients fulfilling Rome IV IBS were included in our study compared to the situation if patients were identified through population screening. However, the lack of pre-existing organic gastrointestinal diagnoses in our population, and the fact the only a minority of those with Rome IV IBS had a prior IBS diagnosis, suggests that most patients had not had their significant bowel symptoms recognised by a healthcare provider which is an important finding given the association with a worse QoL and higher reported psychological distress. Finally, to minimise questionnaire fatigue and improve study completion amongst participants of the ID LIVER project, a dedicated validated instrument for anxiety and depression such as the hospital anxiety and depression scale was not included. Whilst it was notable that anxiety or depression was reported more frequently in those with Rome IV IBS on the EQ-5D-5L questionnaire, future studies should investigate this further using specific anxiety and depression questionnaires.
To further investigate the co-existence of IBS in NAFLD, further large-scale cohort studies would be needed with targeted investigations, where appropriate, to exclude IBS mimics.