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

Introduction: Large out-of-home (OOH) food businesses in England have been required by law to display calorie information on menus since 6 April 2022. This study investigated whether the implementation of this policy was associated with changes in calories purchased OOH by consumers. 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 chains, non-chain outlets and five subtypes 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 to avoid spillover labelling in chains across Great Britain. We aggregated self-reported itemised OOH food and non-alcoholic drink purchases from a rolling consumer panel of ~7500 individuals spanning 13 weeks pre intervention and 34 weeks post intervention to population-level average weekly per-person calorie (kcal) purchase estimates. Linear regression, adjusted for season and inflation, modelled level and trend changes compared with the counterfactual of no mandatory policy. Subgroup analyses explored effects by age, sex, occupational socioeconomic status, weight status and weekday/weekend purchases. Results: Compared with the counterfactual, we found no evidence of a change in overall calories purchased OOH associated with mandatory calorie labelling (level change -95.6 kcal, 95% CI -471.2 to 280.0; trend change 5.1 kcal, 95% CI -5.5 to 15.8). There was also no robust evidence of changes in calories purchased OOH for secondary outcomes and by subgroups compared with the counterfactual of no mandatory calorie labelling. Small changes observed in these analyses were sensitive to analytical choices. Conclusions: This study supports existing evidence that calorie labelling alone is unlikely to secure significant changes in food purchasing behaviour at population level. Possible changes to menus were not included in the study and warrant further investigation.