Objective: Type 2 diabetes mellitus (T2D) prevalence continues to increase, and age of incidence continues to decrease. More information is needed to target interventions to the ages where they can be most effective. The objective of this study was to explore the degree to which the association between diet and T2D incidence changes through adulthood.
Methods: Participants were a large number (N = 2818) of community living adults in Canberra and Queanbeyan, Australia across three cohorts; young (20-24 followed to 32-36), mid-life (40-44 followed to 52-56) and late-life (60-64 followed to 72-76). Self-report dietary pattern scores at baseline and diabetes incidence across 12 years follow-up were measured, alongside confounders of caloric intake, sex, smoking status, years of education, hypertension, BMI and physical activity.
Results: Cox proportional hazards indicated that neither Western nor Prudent dietary pattern scores were significantly associated with T2D incidence when confounders were included in the model. Unadjusted estimates suggested a positive association between Western dietary pattern scores and subsequent diabetes incidence (HR = 1.40, 95% CI [1.18, 1.64]). Compared with the mid-life cohort, a higher Western dietary pattern score posed a lower risk for incident T2D in the young cohort (unadjusted HR = 0.46, 95% CI [0.22, 0.96]), who also had significantly lower BMI and higher physical activity. No such significant effects were found for the late-life cohort.
Conclusions: Our findings indicate that mid-life may be a period of heightened vulnerability to the effects of an unhealthy diet on diabetes risk, but this effect is attenuated when risk factors related to diet, such as adiposity, are taken into account.