Changes in out-of-home food purchasing following England’s calorie labelling regulations: a population-level controlled interrupted time series

Background Large out-of-home (OOH) food businesses in England have been required by law to display calorie information on menus since 6^th April 2022. This study investigates whether the implementation of this policy was associated with changes in calories purchased OOH.

Methods Controlled interrupted time series analysis was used to estimate changes in calories purchased from all OOH outlets in England (intervention group). Secondary outcomes included purchases from large chain outlets, non-chain outlets and five sub-types of purchases (meals, lower-calorie coffee, higher-calorie coffee, sandwiches, and fish and chip meals). The control series consisted of purchases from non-chain outlets in Scotland and Wales. Itemised OOH food and non-alcoholic drink purchase records, matched to calorie (kcal) content at one point in time, covered 13 weeks pre- and 34 weeks post-intervention (3^rd January to 27^th November 2022) and were aggregated to population level average weekly kcal purchase estimates. Linear regression, adjusted for season and inflation, was used to model level and trend changes compared to the counterfactual of no mandatory policy. The counterfactual was constructed from the intervention series’ pre-intervention level and the control series’ post-intervention level and trend. Subgroup analyses explored effects by age, sex, socioeconomic status (SES), weight status, and weekday/weekend purchases.

Results Compared to the counterfactual, we found no evidence of a change in overall calories purchased OOH associated with mandatory calorie labelling. There was also no robust evidence of changes in calories purchased OOH for secondary outcomes and by subgroups compared to the counterfactual of no mandatory calorie labelling. Small changes observed in these analyses were sensitive to analytical choices.

Discussion Our findings suggest that mandatory calorie labelling in England did not reduce overall calories purchased from OOH venues. Findings should be interpreted taking into account that the data used did not capture possible changes to menus. Our findings support existing evidence that information provision alone is unlikely to secure significant changes in food purchasing behaviour at population level.