Neighbourhood greenspace exposure throughout childhood and quality of decision-making in adolescence: Links across buffer sizes

Research on the role of neighbourhood greenspace in children's and adolescents' cognitive and academic functioning has emphasised the importance of granular measures of neighbourhood greenspace, but studies using such measures are few and typically cross-sectional. Our study attempted to fill this gap. Using the UK's Millennium Cohort Study, we examined neighbourhood greenspace across several circular buffers in Great Britain, and then investigated the role of such exposure throughout childhood on adolescents' cognitive functioning - measured with the Number Analogies Test (numerical reasoning), the Word Activity Test (vocabulary knowledge) and the Cambridge Gambling Task (CGT) (reward-based decision-making) - and academic achievement, measured with qualifications attained in secondary school. Neighbourhood greenspace exposure across the school years was assessed with a cumulative measure based on the percentage of greenspace in 400m, 800m, 1200m and 1600m around the family's residential postcode. Across infancy and the preschool years, it was measured as the mean relative greenery of the ward of residence at that period. Measures of urbanicity and air pollution, excluded from the main analysis given their interrelation and the strong relation of each with greenspace, were added separately as confounders in supplementary models. We found a positive association between percentage of neighbourhood greenspace in 800m, 1200m and 1600m around the residential postcode and the CGT's quality of decision-making (indexing the tendency to choose the more probable option under explicit probability conditions). Ward-level greenspace cover across the very early years was also associated with quality of decision-making. We did not find evidence for greenspace associations with any other adolescent outcome examined. Future research should explore greenspace 'effects' on other measures of quality of decision-making and, should a link generalise across measures, explore the mechanisms that may explain it.