Search

Individual-level data of health outcomes and their contributing factors are rarely available at a neighbourhood level, hindering local efforts to address complex issues such as dietary behaviours and their associated inequalities. Spatial microsimulation can be used to approximate individual-level data for small areas (eg, lower or middle super output areas), creating novel synthetic data to support local decision making. We aimed to estimate small-area fruit and vegetable intake to explore geographical inequalities of diet in England. In this spatial microsimulation, cross-sectional study, we performed the spatial microsimulation using individual-level data from the 2014–19 UK National Dietary and Nutritional Survey (NDNS) and aggregated data from the 2011 UK Census. Iterative proportional fitting was done in R (version 1.4.1106) to estimate daily servings of fruits and vegetables for each lower super output area (n=32 844) in England. Internal validation techniques such as measures of total absolute error were evaluated and external validation including comparison to variables of interest (eg, correlation with obesity) undertaken. 3633 adults aged 16 years or older (2092 women and 1541 men) were included in the analysis. We estimated that 26% of adults in England meet the daily recommendations of fruit and vegetable intake, with 7% consuming fewer than one serving per day. This result mirrored NDNS findings as expected. Fruit and vegetable consumption was unevenly distributed within and across regions in England (16–46% meeting the five-a-day recommendations). There were higher rates of fruit and vegetable consumption than those among populations in the least deprived neighbourhoods with 32% meeting daily recommendations versus 20% in the most deprived. Preliminary model validation suggests estimates were robust for internal validation. External validations were mixed. Wide geographical inequalities in fruit and vegetable consumption exist across England, varying greatly by level of deprivation. This study presents a novel framework for small area estimation of diet indicators and could support better nuanced local decision-making. Economic and Social Research Council, North West Social Science Doctoral Training Partnership (reference 2107539). [ABSTRACT FROM AUTHOR]