Program for healthier school cafeterias in Rio Grande do Sul, Brazil: Protocol for a community-based randomized trial

Abstract

Background: School cafeterias can promote poor eating habits, as these retail outlets have a variety of foods considered to be nonnutritive and unhealthy. However, despite the need for effective preventive strategies, there is still disagreement on the best approach due to the lack of evidence on interventions to prevent and treat obesity in the school settings. Objective: We aim to verify the efficacy of an educational intervention program to improve the hygienic conditions and the composition of the menu offered in school cafeterias in the state of Rio Grande do Sul, Brazil. Methods: We will conduct a randomized, parallel, two-arm, community-based controlled study. Elementary and high schools, both public and private, in the State of Rio Grande do Sul, Brazil, that have a cafeteria will be eligible. Schools will be recruited and randomly assigned to the intervention (n=27) or control (n=27) group. The intervention group will receive an educational intervention program based on the guidelines issued by the Ministry of Health of Brazil, consisting of a 160-hour distance-learning qualification course, for 10 weeks, and using the Moodle platform and WhatsApp app. The intervention targets the owners and people in charge of the cafeterias, food handlers, principals, vice principals, teachers, pedagogical coordinators, dietitians, representatives of students' parents, and students over 16 years old. Meanwhile, the control group will receive only a printed copy of the book containing the guidelines used. The efficacy of the intervention will be determined by the hygienic conditions of the cafeteria and the composition of the menu offered, also considering the levels of processing of food sold. All outcomes will be analyzed as intention-to-treat and per-protocol. We will use covariance analysis or a generalized linear model for continuous data and ordinal logistic regression for ordinal categorical data. The level of statistical significance considered will be P