Introduction Fermentable carbohydrates can trigger symptoms of irritable bowel syndrome (IBS) but the underlying mechanisms are unclear. Our aim was to test whether fructose or inulin ingestion would cause more symptoms than glucose in IBS patients and to determine which magnetic resonance imaging (MRI) biomarkers could explain the origin of their symptoms.
Method A 3-period, 3-treatment, randomised, double-blind crossover study. 29 IBS patients meeting Rome III criteria for IBS, who also reported bloating, completed 3 study days at least 1 week apart: median age 34 (IQR 25–48.5), 22 female, 7 male. Procedures: symptom score, breath hydrogen (H2) measurement and MRI scan at baseline (fasted) and 0, 60, 120, 180, 240 and 300 min post-prandially. Symptoms and H2were also measured at 30 and 90 min. Symptoms of flatulence, bloating, pain and diarrhoea were recorded on a simple 0–3 integer scale and by visual analogue score (VAS). Treatments: ingestion of 40 g of glucose (control), fructose or inulin in 500 ml water.
Primary endpoint: a rise from baseline of ≥3 in a composite symptom score of the integer scales (CSS – max total 12). Secondary endpoints: composite symptom intensity (VAS), H2and MRI measures of small bowel water content (SBWC), colonic volume and gas.
Results Table 1[mean (SD) or median (IQR)]. More patients had a rise in CSS ≥3 after inulin than glucose (p < 0.04). Symptoms peaked 50 min after fructose and 240 min after inulin. H2rose after both but more so after inulin with a later peak (90 vs. 300 min). While fructose increased SBWC, inulin increased colonic gas and total volume more than fructose or glucose. There were no significant differences between those who reported a rise in symptoms and those who did not. In those with symptoms after inulin there was a correlation between peak rises in symptom intensity and colonic gas (p < 0.05, r2= 0.33). H2did not correlate with symptoms (r2= 0.08). For fructose the best correlate of symptoms was SBWC (p = 0.05, r2= 0.26).
Conclusion Inulin induced more symptoms than glucose probably by distending the colon while any fructose effect appears to be mediated by small bowel distension. The similarity of findings between patients with and without a symptom increase suggests that visceral hypersensitivity is the key to symptoms.
Disclosure of interest G. Major: None Declared, S. Pritchard: None Declared, K. Murray: None Declared, C. Hoad: None Declared, L. Marciani: None Declared, P. Gowland: None Declared, R. Spiller Grant/ Research Support from: Lesaffre, Ironwood, Norgine, Falk Pharma, Consultant for: Almirall, Astellas, Yuhan, Danone.