Background: About one in ten adults are living with diabetes worldwide. Intake of carbohydrates and carbohydrate-rich foods are often identified as modifiable risk factors for incident type 2 diabetes. However, strong correlation between food variables can make it difficult to identify true associations. The purpose of this study was to identify clusters of carbohydrate-rich foods and analyse their associations with type 2 diabetes incidence in the Malmö Diet and Cancer Study cohort in southern Sweden.
Methods: Dietary intake of 26 622 participants was assessed using a validated three-part diet history method: a 7-day food diary, a 168-item food frequency questionnaire, and a 60-minute interview. K-means clustering analysis identified five clusters from 21 food variables. The Cox proportional hazard regression model was applied to calculate hazard ratios (HR) and 95% confidence intervals (CI) of the association between clusters and incident type 2 diabetes.
Results: The cluster analysis resulted in five clusters; high vegetables/low added sugar, high sugar-sweetened beverages, high juice, high fruit, and high refined carbohydrates/low fruit & vegetables (reference). During mean follow-up of 18 years, 4046 type 2 diabetes cases were identified. After adjustment for potential confounding (including lifestyle, body mass index, and diet), a high fruit cluster (HR 0.86; 95% CI 0.78, 0.94) was inversely associated with type 2 diabetes compared to the reference cluster. No other significant associations were identified.
Conclusions: A dietary pattern defined by a high intake of fruits was associated with a lower incidence of type 2 diabetes. The findings provide additional evidence of a potential protective effect from fruit intake in reducing type 2 diabetes risk. Future studies are needed to explore this association further.
Keywords: Diet; Epidemiology; K-means clustering; Malmö Diet and Cancer Study; Nutrition; Type 2 Diabetes.
© 2023. The Author(s).