Fresh Fruit Consumption, Physical Activity and Five-year Risk of Mortality among Patients with Type 2 Diabetes: A Prospective Follow-up Study

Abstract Background and Aims We explored the associations among fruit consumption, physical activity and their dose-response relationship with all-cause and cardiovascular disease (CVD) mortality in type 2 diabetic patients. Methods and Results We prospectively followed 20,340 community-dwelling type 2 diabetic patients aged 21-94 years. Information on diets and physical activity was collected using standardized questionnaires. All-cause and CVD mortality were assessed. Hazard ratios (HRs) for all-cause mortality were estimated with Cox regression models, and HRs for CVD mortality were derived from a competing risk model. Restricted cubic spline regression was used to analyse dose-response relationships. We identified 1,362 deaths during 79,844 person-years. Compared to non-consumption, fruit consumption >42.9 g/d was inversely associated with all-cause mortality (HR 0.76; 95% CI 0.64-0.88), CVD mortality (HR 0.69, 0.51-0.94) and stroke mortality (HR 0.57, 0.36-0.89), but not with heart disease mortality (HR 0.93, 0.56-1.52). The HRs comparing the top vs bottom physical activity quartiles were 0.44 (0.37-0.53) for all-cause mortality, 0.46 (0.33-0.64) for CVD mortality, 0.46 (0.29-0.74) for stroke mortality and 0.51 (0.29-0.88) for heart disease mortality. Lower fruit consumption combined with a lower physical activity level was associated with a greater mortality risk. A nonlinear threshold of 80 g fruit/day was identified, all-cause mortality risk was reduced by approximately 24% at this value. A physical activity threshold of 8 metabolic equivalents (MET) h/day was also identified, after which the risk of mortality did not decrease. Conclusions Fruit consumption and physical activity may reduce all-cause, CVD and stroke mortality in type 2 diabetic patients.