Dietary diversity and type 2 diabetes risk: evidence from a prospective Korean cohort study

Abstract Although consuming a diverse diet is generally considered beneficial for chronic disease prevention, its relationship with type 2 diabetes (T2D) remains unclear. This study investigated the associations of overall and within-group dietary diversity scores (DDS) with T2D incidence in the Korean Genome and Epidemiology Study–CVD Association Study. A total of 16 666 participants aged 40 years and older without T2D at baseline were followed for 96 609 person-years, during which 953 incident T2D cases were identified. Cumulative average overall DDS and within-group DDS were derived from repeated FFQ administered at each visit. Modified Poisson regression with a robust error variance estimator was used to estimate incidence rate ratios (IRR) and 95 % CI. A significant inverse linear trend was observed between overall DDS and T2D risk (IRR = 0·56, 95 % CI = 0·40–0·79 for men; IRR = 0·73, 95 % CI = 0·54–0·98 for women). Among the five within-group DDS, those for ‘Milk and Dairy products’ in both sexes, ‘Fruits’ in men and ‘Vegetables/Seaweeds/Mushrooms’ in women were inversely associated with T2D (Q4 v . Q1). No significant sex-related interactions were observed (all p for interaction > 0·05). Higher overall DDS and greater within-group diversity in ‘Milk and Dairy products,’ ‘Fruits’ and ‘Vegetables/Seaweeds/Mushrooms’ were associated with lower T2D risk. These findings suggest that maintaining overall dietary diversity – particularly emphasising non-energy-dense food groups such as fruits, vegetables, and dairy – may help prevent T2D.