Abstract 19: Time-Varying Exposure to Food Retailers and Cardiovascular Disease Incidence and Mortality in the Netherlands: A Nationwide Prospective Cohort Study

Background: The food environment is suggested as a risk factor of cardiovascular disease (CVD), but current evidence is inconsistent and does not account for changes in the food environment. We aimed to explore if time-varying exposure to the food environment was associated with hospitalisation and mortality due to total and specific types of CVD in The Netherlands. We hypothesized that favourable time-varying trends in food environment exposure would be associated with decreased risk of CVD, stroke, heart failure (HF), and coronary heart disease (CHD) morbidity and mortality among the Dutch general population. Methods: A prospective cohort study was performed among all Dutch residents aged 35+ years who did not change residence in 2002-2018 without CVD in 2004, identified through registry data from Statistics Netherlands (n=4,641,435; 51.3% female, 91.2% Dutch). Time-varying Food Environment Healthiness Index (FEHI) scores (range: -5 to 5) and time-varying Kernel density of six specific food retailers (food delivery outlets, fast food outlets, restaurants, local food shops, supermarkets and convenience stores) around the home location were based on bi-annual updates between 2004-2018. Hospitalisation and mortality due to overall CVD, stroke, HF, and CHD occurring between 2004 and 2020 were used as outcomes, based on hospital and death registries. Models were adjusted for age, sex, ethnic origin, household composition, income, marital status and neighbourhood urbanisation level and effect modification by age, sex, urbanization and income was studied. Results: Each unit increase in the FEHI, representing a 10% improvement, was associated with a lower hospitalisation and mortality of CVD (hospitalization Hazard Ratio (HR h )=0.90, 0.89-0.91, mortality Hazard Ratio (HR m )=0.85, 0.82-0.89), CHD (HR h =0.88, 0.85-0.91, HR m =0.80, 0.75-0.86), stroke (HR h =0.89, 0.84-0.93, HR m =0.89, 0.82-0.98), and HF (HR h =0.90, 0.84-0.96, HR m =0.84, 0.76-0.92). Increased density of local food shops, fast food outlets, supermarkets and convenience stores, and decreased density of food delivery outlets and restaurants were associated with a higher risk of CVD, CHD, stroke and HF incidence and mortality, albeit with small effect sizes. Stratified analyses suggest slightly stronger associations in highly urbanized areas, younger adults and those with higher incomes. Conclusions: Favourable trends in time-varying food retailer exposure were associated with a risk reduction in CVD hospitalisation and mortality, in particular in urbanized areas and for younger adults and those with higher incomes.