Associations between diet and disease activity in ulcerative colitis patients using a novel method of data analysis

Abstract

Background: The relapsing nature and varying geographical prevalence of ulcerative colitis (UC) implicates environmental factors such as diet in its aetiology. Methods: In order to determine which foods might be related to disease activity in UC a new method of dietary analysis was developed and applied. Eighty-one UC patients were recruited at all stages of the disease process. Following completion of a 7 d diet diary, clinical assessment including a sigmoidoscopic examination (scale 0 (normal mucosa) to 6 (very active disease)) was conducted. Food weights for each person were adjusted (divided) by the person's calorific intake for the week. Each food consumed was given a food sigmoidoscopy score (FSS) calculated by summing the products of the (adjusted) weight of food consumed and sigmoidoscopy score for each patient and occurrence of food and dividing by the total (adjusted) weight of the food consumed by all 81 patients. Thus, foods eaten in large quantities by patients with very active disease have high FSSs and vice versa. Foods consumed by <10 people or weighing <1 kg for the whole group were excluded, leaving 75 foods. Results: High FSS foods were characterized by high levels of the anti-thiamin additive sulfite (Mann-Whitney, p < 0.001), i.e. bitter, white wine, burgers, soft drinks from concentrates, sausages, lager and red wine. Caffeine also has anti-thiamin properties and decaffeinated coffee was associated with a better clinical state than the caffeine containing version. Beneficial foods (average intake per week) included pork (210 g), breakfast cereals (200 g), lettuce (110 g), apples and pears (390 g), milk (1250 ml), melon (350 g), bananas (350 g), bacon (120 g), beef and beef products (500 g), tomatoes (240 g), soup (700 g), citrus fruits (300 g), fish (290 g), yogurt (410 g), cheese (110 g), potatoes (710 g) and legumes (120 g). Conclusions: The dietary analysis method described provides a new tool for establishing relationships between diet and disease and indicates a potentially therapeutic diet for UC. © 2005 Magee et al; licensee BioMed Central Ltd.