Does cutting down on your food consumption lead to a net improvement in nutritional intake? A panel data approach using data from the UK Biobank

Abstract

Abstract Background Food diets are complex and a policy targeting one item of a person’s diet does not affect their nutritional intake in a solely additive or subtractive manner. Policies tackling unhealthy diets are more likely to be adopted by governments if there is robust evidence to support them. To evaluate dietary policies, it is important to understand the correlations and interdependencies between food groups, as these can lead to unintended negative consequences. We aimed to see whether reductions in consumption of a particular group is related to a net improvement in nutritional intake, after taking into account patterns of consumption and substitution across food groups. Methods Detailed dietary data was collected using a 24-h online dietary assessment from the UK Biobank and Oxford Web Q (n = 185,611). We used panel data fixed effects methods to estimate changes in energy, saturated fat, total sugar, and fibre following a 100gram reduction across 44 food groups. We compare these estimates against the average nutritional value of that food group from the UK National Diet and Nutrition Survey. Results We find evidence of variation in whether a food is compensated between the main confectionery products. Crisps, savoury snacks, and sugar confectionery are less likely to be compensated, whereas chocolate confectionery, biscuits, and buns/cakes/pastries and pies are compensated. The result is particularly striking for chocolate confectionery which shows that while chocolate confectionery often has a high energy content, eating less chocolate confectionery is not associated with an equal reduction in energy. Instead, we find individuals switch or compensate for their reduction in chocolate confectionery consumption with other high energy food items. Conclusions We find that sugar confectionery and crisps and savoury snacks are less likely to result in substitution than chocolate confectionery. This would suggest that food policies aiming to reduce the consumption of these food groups are more likely to result in overall lower consumption of unhealthy foods.