In the first phase of the project we used virtual methods to study the impact that different types of dietary intervention (structural vs. information-based) have on consumer behaviour. Using online platforms, we have conducted a series of experiments in which we exposed participants to different interventions and then examined their hypothetical food choices in response to these interventions. In particular, we have tested whether calorie labelling of food menus (information-based intervention) promotes healthier food choices and whether this is more or less likely in people from higher vs. lower social disadvantage backgrounds. Similarly, we tested whether increasing the availability of healthier menu options (structural-based intervention) promotes healthier food choices and whether this is more or less likely in people from higher vs. lower social disadvantage backgrounds. The results of these experiments were very consistent; the structural based interventions tested resulted in participants choosing healthier food options and fewer calories overall. Importantly these results were consistent across people from more vs. less social disadvantage, suggesting that in line with the project hypotheses people from both advantaged and disadvantaged backgrounds benefit from structural interventions. Conversely, we found limited evidence that information-based interventions tested had a beneficial effect on behaviour of participants irrespective of level of social disadvantage. These initial findings confirmed our hypotheses that structural based interventions have the most potential to be ‘socially equitable’ in their benefits and that information-based interventions may be less effective.
In the next phase of the project we conducted a series of laboratory and real-world trials. These studies confirmed that structural interventions impact on dietary behaviour and they do so socially equitably. For example, we have found no evidence that changing the food environment structures (e.g. portion sizes) has a stronger or weaker effect on diet for people from differing social circumstances. In our real-world trials we also found that a type of information-based intervention (warning labels on foods) changed consumer behaviour and that people from high levels of social disadvantage found this intervention to be particularly effective.
The final phase of the project was the epidemiological modelling work. This work involved modelling the impacts on population level diet, obesity and non-communicable disease of different types of food intervention policies. Counter to our original hypotheses before the grant, the policies we modelled which were information-based did not widen social inequalities in health or diet based on the modelling results. Counter to our hypotheses, we also did not find that a structural policy (taxing unhealthy food) had a bigger impact than an information-based policy (better food labelling) when modelling across European countries (e.g. Belgium, Germany). Nonetheless, these results have still been very important as they have outlined evidence for adoption of these policies in the UK and other European countries. During the project we were active in disseminating findings to other scientists, government policy makers and the general public.
