Abstract

--- - Background - In 2016, Chile implemented a multiphase set of policies that mandated warning labels, restricted food marketing to children, and banned school sales of foods and beverages high in nutrients of concern ("high-in" foods). Chile's law, particularly the warning label component, set the precedent for a rapid global proliferation of similar policies. While our initial evaluation showed policy-linked decreases in purchases of high-in, a longer-term evaluation is needed, particularly as later phases of Chile's law included stricter nutrient thresholds and introduced a daytime ban on advertising of high-in foods for all audiences. The objective is to evaluate changes in purchases of energy, sugar, sodium, and saturated fat purchased after Phase 2 implementation of the Chilean policies. - Methods and findings - This interrupted time series study used longitudinal data on monthly food and beverage purchases from 2,844 Chilean households (138,391 household-months) from July 1, 2013 until June 25, 2019. Nutrition facts panel data from food and beverage packages were linked at the product level and reviewed by nutritionists. Products were considered "high-in" if they contained added sugar, sodium, or saturated fat and exceeded nutrient or calorie thresholds. Using correlated random-effects models and an interrupted time series design, we estimated the nutrient content of food and beverage purchases associated with Phase 1 and Phase 2 compared to a counterfactual scenario based on trends during a 36-month pre-policy timeframe. Compared to the counterfactual, we observed significant decreases in high-in purchases of foods and beverages during Phase 2, including a relative 36.8% reduction in sugar (-30.4 calories/capita/day, 95% CI -34.5, -26.3), a 23.0% relative reduction in energy (-51.6 calories/capita/day, 95% CI -60.7, -42.6), a 21.9% relative reduction in sodium (-85.8 mg/capita/day, 95% CI -105.0, -66.7), and a 15.7% relative reduction in saturated fat (-6.4 calories/capita/day, 95% CI -8.4, -4.3), while purchases of not-high-in foods and drinks increased. Reductions in sugar and energy purchases were driven by beverage purchases, whereas reductions in sodium and saturated fat were driven by foods. Compared to the counterfactual, changes in both high-in purchases and not high-in purchases observed in Phase 2 tended to be larger than changes observed in Phase 1. The pattern of changes in purchases was similar for households of lower versus higher socioeconomic status. A limitation of this study is that some results were sensitive to the use of shorter pre-policy time frames. - Conclusions - Compared to a counterfactual based on a 36-month pre-policy timeframe, Chilean policies on food labeling, marketing, and school food sales led to declines in nutrients of concern during Phase 2 of implementation, particularly from foods and drinks high in nutrients of concern. These declines were sustained or even increased over phases of policy implementation. - Author summary - Why was this study done? - Chile implemented a multiphase set of policies to halt increases in obesity prevalence and noncommunicable disease risk. - These policies mandated front-of-package warning labels on foods and beverages, restricted food marketing to children, and banned the sale of foods and beverages high in nutrients of concern (e.g., sugar, saturated fat) in schools. - These policies were implemented in 3 phases, with increasingly strict nutritional thresholds in each phase and the introduction of a daytime marketing ban in Phase 2. We aimed to examine the pre-post association of these policies with food purchases during the Phase 2 implementation. - What did the researchers do and find? - Using longitudinal data on household food purchases, we estimated how purchases high in nutrients of concern were associated with the Phase 1 and Phase 2 implementations compared to a counterfactual scenario, or hypothetical food purchases if the policies had not been enacted, based on trends during a 36-month pre-policy timeframe. - After the Phase 2 implementation, compared to the counterfactual, purchases of foods high in nutrients of concern decreased, ranging from approximately a 16% relative reduction for saturated fat purchases to a 37% relative reduction in sugar purchases, while purchases of foods not high in nutrients of concern increased. Compared to counterfactual, there were larger drops in purchases of nutrients of concern from high-in foods and beverages in Phase 2 compared to Phase 1; there were also larger increases in purchases from not high-in foods and beverages. - There were few differences in post-policy food and beverage purchases by socioeconomic strata. - Overall, there was a reduction in purchases of nutrients of concern during Phase 2 of implementation. - What do these findings mean? - The Phase 2 implementation led to declines in purchases of nutrients of concerns, especially from high-in foods and drinks. - Purchasing changes of high-in and not-high in foods were more pronounced in Phase 2 than in Phase 1, suggesting that post-policy purchasing changes were maintained or even strengthened over time. - Reductions in nutrients of concern were similar across socioeconomic strata, suggesting that this policy did not disproportionately advantage or disadvantage any group. - The main limitations of this study are that the data included are only a portion of what Chilean households purchase, and total dietary intake data are needed to fully evaluate the policies.