Source of dietary fibre and diverticular disease incidence: a prospective study of UK women

Abstract

Background: Previous prospective studies have found the incidence of intestinal diverticular disease decreased with increasing intakes of dietary fibre, but associations by the fibre source are less well characterised. We assessed these associations in a large UK prospective study of middle-aged women.

Methods and findings: During 6 (SD 1) years follow-up of 690 075 women without known diverticular disease who had not changed their diet in the last 5 years, 17 325 were admitted to hospital or died with diverticular disease. Dietary fibre intake was assessed using a validated 40-item food questionnaire and remeasured 1 year later in 4265 randomly-selected women. Mean total dietary fibre intake at baseline was 13.8 (SD 5.0) g/day, of which 42% came from cereals, 22% from fruits, 19% from vegetables (not potatoes) and 15% from potatoes. The relative risk (95% CI) for diverticular disease per 5 g/day fibre intake was 0.86 (0.84 to 0.88). There was significant heterogeneity by the four main sources of fibre (p<0.0001), with relative risks, adjusted for each of the other sources of dietary fibre of 0.84 (0.81 to 0.88) per 5 g/day for cereal, 0.81 (0.77 to 0.86) per 5 g/day for fruit, 1.03 (0.93 to 1.14) per 5 g/day for vegetable and 1.04 (1.02 to 1.07) per 1 g/day for potato fibre.

Conclusions: A higher intake of dietary fibre is associated with a reduced risk of diverticular disease. The associations with diverticular disease appear to vary by fibre source, and the reasons for this variation are unclear.

Keywords: diet; dietary fibre; diverticular disease; prospective.

Figure 1

Relative risk of diverticular disease by fifths of dietary fibre intake. Relative risks are stratified by region and adjusted for age, socioeconomic status, smoking, alcohol, body mass index, height, current use of hormone therapy for menopause, total energy intake, and type of meat consumed and are plotted against the mean remeasured dietary fibre intake in each fifth.

Figure 2

Relative risk (95% CI) of diverticular disease per 5 g/day dietary fibre intake by chosen characteristics. Stratified by region and adjusted for age, socioeconomic status, smoking, alcohol, body mass index, height, current use of hormone therapy for menopause, total energy intake and type of meat consumed (where appropriate). Heterogeneity of trends in relative risk between different subgroups was assessed using a χ² test. Complicated diverticular disease was defined as diverticula with abscess, bleeding or perforation (ICD-10 code: K570, K572 or K578). Current comorbidity was defined using the Charlson index.