Background Complications of diverticular disease are increasingly common, possibly linked to increasing obesity. Visceral fat could contribute to the development of symptomatic diverticular disease through its pro-inflammatory effects. Objective The study had 2 aims. A) to develop a semi-automated algorithm to measure abdominal adipose tissue from 2-echo magnetic resonance imaging (MRI) data; B) to use this to determine if visceral fat was associated with bowel symptoms and inflammatory markers in patients with symptomatic and asymptomatic diverticular disease. Design An observational study measuring visceral fat using MRI together with serum adiponectin, leptin, stool calprotectin and patient-reported somatisation and bowel habit. Setting Medical and imaging research centres of a university hospital. Participants MRI scans were performed on 55 patients after an overnight fast measuring abdominal subcutaneous and visceral adipose tissue volumes together with small bowel water content (SBWC). Blood and stool samples were collected and patients kept a 2 week stool diary and completed a somatisation questionnaire. Main outcome measures Difference in the volume of visceral fat between symptomatic and asymptomatic patients. Results There were no significant differences in visceral ( p = 0.98) or subcutaneous adipose ( p = 0.60) tissue between symptomatic and asymptomatic patients. However measured fat volumes were associated with serum adipokines. Adiponectin showed an inverse correlation with visceral adipose tissue (VAT) (Spearman ρ = -0.5, p = 0.0003), which correlated negatively with SBWC (ρ = -0.3, p = 0.05). Leptin correlated positively with subcutaneous adipose tissue (ρ = 0.8, p < 0.0001). Overweight patients (BMI > 25 kgm -2 ) showed a moderate correlation between calprotectin and VAT (ρ = 0.3, p = 0.05). Somatization scores were significantly higher in symptomatic patients ( p < 0.0003). Conclusions Increasing visceral fat is associated with lower serum adiponectin and increased faecal calprotectin suggesting a pro-inflammatory effect which may predispose to the development of complications of diverticulosis.

No abstract