Associations of carbohydrates and carbohydrate-rich foods with incidence of type 2 diabetes

Dietary carbohydrates have long been expected to be associated with risk of type 2 diabetes; however, the associations for many carbohydrates and carbohydrate-rich foods remain inconclusive. This study analysed associations between intakes of six types of carbohydrates and thirteen carbohydrate-rich foods with incident type 2 diabetes in 26 622 participants (61% women) in the Malmö Diet and Cancer Study in southern Sweden. Dietary intake was assessed at baseline (1991-1996) by using a modified diet history method. During mean follow-up of 18 years, 4046 cases were identified. Adjusting for potential confounders (including lifestyle, BMI, and dietary factors), comparing highest v. lowest quintile of intake, monosaccharides (hazard ratio (HR) 0·88; 95% CI 0·79, 0·98; P trend = 0·02) and fruits (HR 0·91; 95% CI 0·82, 1·01; P trend = 0·03) were inversely associated with incident type 2 diabetes, while disaccharides (HR 1·17; 95% CI 1·04, 1·30; P trend = 0·002) and sweets (HR 1·09; 95% CI 1·00, 1·19; P trend = 0·02) were positively associated. After stratification by sex, marmalade/honey/jam (HR 0·82; 95% CI 0·72, 0·94; P trend < 0·001) and vegetables (HR 0·85; 95% CI 0·73, 0·98; P trend = 0·06) were inversely associated with incident type 2 diabetes in men, and chocolate (HR 1·26; 95% CI 1·09, 1·46; P trend < 0·001) was positively associated in women. In conclusion, we identified inverse associations for intake of monosaccharides and fruits with type 2 diabetes risk, and positive associations for disaccharides and sweets. Additional sex-specific associations were also identified. Future studies are needed to explore these associations further.