NO CHANGE IN CIRCULATING CYTOKINES IN ACTIVE CROHN’S DISEASE OR HEALTHY CONTROLS DURING HIGH AND LOW EMULSIFIER DIETS IN BLINDED FEEDING STUDIES

Abstract Background & Aim Emulsifiers have been pathogenically implicated in Crohn’s disease (CD). Circulating cytokines may reflect intestinal barrier dysfunction and/or inflammatory activity. We aimed to examine the effect of closely controlled high-emulsifier (HED) and low-emulsifier (LED) diets on a panel of cytokines in patients with active ileal CD and healthy controls. METHODS In a double-blinded, controlled feeding study, patients with symptomatic and sonographically active CD with ileal involvement on stable medical therapy were randomised to four weeks of either a HED or LED, both modelled on healthy eating guidelines. Diets were matched for all nutritional components, other than the presence of emulsifiers, thickeners, and ultra-processed food content (median proportion HED 68 [IQR 63-82]% vs LED 27 [22-34]%). A panel of cytokines (interferon-γ, TNFα, interleukins 6, 8, 10, 12p40) was measured in plasma (ultrasensitive electrochemiluminescence-based multiplex assay) and compared within and between diets, and for binary clinical outcomes: sonographic response (reduction in IBUS-SAS) or 5% reduction in waist circumference. Cytokine concentrations were compared to those in healthy controls who received both diets for 4 weeks each in a cross-over trial. RESULTS 24 CD patients (median age 35 years IQR 27-45; 50% female) were randomised to HED or LED (2 HED, 3 LED withdrew). 18 healthy participants (median age 34 years IQR 23-46; 75% female) completed both diets. Dietary adherence was excellent. IBUS-SAS reduced similarly (p = 0.28) on HED (mean 51 [95% CI 35-68] to 33 [15-51]; p = 0.014) and on LED (57 [38-76] to 44 [29-59]; p = 0.010). Only IL-6 concentrations were higher in CD patients compared with healthy controls at baseline and during the diets (Figure). There were no differences in the other cytokines between or within HED or LED, nor between CD patients and healthy controls (all p > 0.10). Cytokine levels did not differ between diets in sonographic responders or those with a 5% reduction in waist circumference (all p > 0.05). CONCLUSION Circulating cytokines did not change in healthy subjects or patients with CD, despite a reduction in intestinal sonographic inflammation on both HED and LED. The cytokine assay confirmed the expected IL-6 elevation in active CD, supporting its sensitivity to detect clinically relevant inflammation. Marked variation in dietary emulsifier intake and ultra-processed food content had no specific effect on barrier function or inflammation as reflected by systemic cytokine profiles.