Prospective associations of overall dietary diversity and within-food group diversity with type 2 diabetes risk: The CAVAS (Cardiovascular Disease Association Study)

Abstract

Type 2 diabetes (T2D) is a prevalent non-communicable disease with significant health and economic consequences worldwide. Lifestyle interventions, including dietary modifications, have proven effective in preventing T2D and other chronic conditions. The recommendation to "eat a variety of foods" is widely accepted to reduce the risk of major chronic diseases, including T2D. National policies and dietary guidelines emphasize the consumption a diet that includes a variety of foods to prevent chronic diseases such as diabetes. To date, studies have focused on overall dietary diversity and T2D. Although diversity within a food group may be associated with intake of a greater variety of nutrients, research on within-group diversity is limited. To better operationalize dietary diversity, it is necessary to establish standardized method for measuring dietary diversity. The study aimed to investigate the relationship between overall/within-group/within-subgroup Minimum Dietary Diversity Score (MDDS) and the risk of T2D. The study had three specific objectives, which were to test the hypotheses that: 1) Cumulative average overall MDDS would be associated with lower T2D incidence, 2) Cumulative average within-group MDDS would be associated with lower T2D incidence, and 3) Cumulative average within-subgroup MDDS would be associated with lower T2D incidence. A total of 16,666 participants who did not have T2D were enrolled. Cumulative average overall/within-group/within-subgroup MDDS were estimated using a repeated food frequency questionnaire. Incidence rate ratios (IRRs) with 95% confidence intervals (CIs) were estimated using modified Poisson models with a robust error estimator. Nonlinear dose-response association were assessed by restricted cubic spline model. Over 96,655 person-years of follow-up, 945 participants developed T2D. Significant linear dose response inverse associations were observed between cumulative average overall MDDS and T2D in both men and women (IRR = 0.56, 95% CI = 0.40–0.79, in the highest quartile compared with the lowest quartile, P-trend = 0.0003 for men; IRR = 0.73, 95% CI = 0.54–0.98, P-trend = 0.0353 for women). The cumulative average within-group MDDS of fruits and milk·dairy products in men, and vegetables and milk·dairy products in women were inversely associated. There were significantly nonlinear dose-response association in grains·its products-MDDS and milk·dairy products-MDDS in both men and women, and meat·seafood·eggs·legumes·nuts-MDDS for men. Furthermore, the cumulative average within-subgroup MDDS of grains, rice cake, breads, nuts and mushrooms in both men and women has been inversely associated with the risk of T2D. Our study suggests that greater cumulative average overall MDDS may decrease the risk of T2D. Cumulative average within-group MDDS of vegetables, fruits, and milk·dairy products, and cumulative within-subgroup MDDS of grains, rice cake, breads, nuts and mushrooms were inversely associated with T2D risk.