The impact of Chile's multipronged food labelling and advertising law on early childhood excess weight: a cohort difference-in-differences study

Background The 2016 Chilean Food Labelling and Advertising Law (FLAL), featuring black octagonal front-of-package warning labels and marketing and school restrictions, was among the first sets of multiple healthy food policies globally. In spite of its relevance, no study has causally linked the FLAL implementation to health outcomes. We aimed to estimate the plausible causal effect of the implementation of phase 1 of the FLAL on the relevance of excess weight among young children. Methods In this cohort difference-in-differences approach, we assessed the effect of phase 1 of the FLAL on children's BMI when initital thresholds for critical nutrients were subsequently increased and implemented. Our analysis used national administrative data from the Chilean National Board of School Aid and Scholarships covering more than 300 000 school children aged 4–6 years across public and publicly subsidised schools nationwide from 2012 to 2017. We primarily used the Nutritional Map (Mapa Nutricional), available beginning in 2012, complemented with the Vulnerability Survey (Encuesta de Vulnerabilidad), available beginning in 2019. We defined cohorts by the year students entered prekindergarten, and compared unexposed cohorts (2012 and 2013, control group) with cohorts exposed to phase 1 (2014 and 2015, treatment group) during prekindergarten, kindergarten, and first grade. The primary outcome was a binary indicator of excess weight (ie, overweight or obesity) analysed using logit models and reported as marginal effects. We conducted subgroup analyses by children's school type, school area, maternal education and age at childbirth, birthweight, and gender. Findings The final analytical sample included 321 597 students (of cohorts 2012–15) covering the years 2012 to 2017 across prekindergarten, kindergarten, and first grade. Exposure to phase 1 of the FLAL led to significant reductions in the probability of a child having excess weight. Children exposed during both kindergarten and first grade (ie, 18 months of exposure) had the largest effects. Girls had a 2·85% lower probability of excess weight (95% CI –0·0407 to –0·0163), while boys had a 2·40% lower probability (–0·0358 to –0·0122). We also observed significant effects after 6 months of exposure (ie, first grade only): 1·91% lower probability (−0·0315 to −0·0068) for girls and 2·24% (−0·0345 to −0·0104) lower probability for boys. Interpretation Phase 1 of Chile's comprehensive FLAL plausibly caused a measurable decrease in the prevalence of excess weight among young school children. The results provide crucial, evidence-based support for policy makers worldwide who are considering food environment policies as a scalable, impactful strategy to combat the childhood obesity epidemic. Funding Bloomberg Philanthropies Grant.