The problem this thesis addresses is unhealthier diets consumed in high income countries that contribute to rising obesity rates. This problem is most evident amongst those in lower socio-economic positions (SEPs) who benefit less from traditional information-based interventions, possibly due to lower cognitive resource for self-control. Conversely, altering environmental cues appears to change behaviour with less conscious deliberation. This thesis aims to investigate whether interventions targeting the “proximity effect” i.e. the farther away the food, the lower the chance it is selected, consistently affects food selection in general populations in ways that could be used to improve diets without relying on cognitive resource. Three laboratory-based experiments were conducted.

Study 1 investigated the proximity effect in a general population sample, including those from lower SEPs. Fewer participants selected an unhealthier food when it was far (53.8%) compared with near (63.3%), an effect which increased in size when participants who moved the bowl were excluded (39.3% vs 63.9%).

Study 2 extended Study 1 using a statistically more powerful design in which participants were also randomised to a cognitive load intervention, assessing whether the proximity effect was similar regardless of differences in manipulated cognitive resource. Fewer participants selected an unhealthier food when it was far (57.7%) compared with near (70.7%) an effect not moderated by cognitive resource.

Study 3 built on Studies 1 and 2, (in which only one unhealthier food was available), to assess the proximity effect when both healthier and unhealthier foods were available. The selection of healthier food was not significantly affected by its proximity (OR=0.61, 95% CI[0.28,1.32]) and that of competing unhealthier foods (OR=1.54, 95% CI[0.52,4.55]). By contrast, the selection of unhealthier food was influenced by its proximity (OR=0.39, 95% CI[0.18,0.82]) and that of competing healthier foods, although not significantly (OR=2.83, 95% CI[0.98,8.33]). These differing effects require further testing and replication to determine their reliability.

The thesis concludes by discussing how the findings of these studies can inform development of interventions capitalising on the “proximity effect” to improve diets for all SEPs, the research needed to address key remaining uncertainties, and the implications of such interventions for addressing the rise in obesity.