Effectiveness of a multicomponent intervention on adherence and acceptability to school meals: a multi-arm, school-based randomized trial in Sumidouro - RJ

The National School Feeding Program plays a key role in promoting healthy eating for millions of schoolchildren throughout Brazil. However, national surveys indicate low adherence and acceptability to school meals in the public school system. Interventions based on choices architecture and standardization of gastronomic techniques have been implemented aiming to improve this scenario and encourage the development of healthy eating habits. Thus, the objective of this study was to evaluate the effectiveness of a multicomponent intervention in adherence and acceptability to school meals. A randomized trial was conducted with students from 4th to 9th grade from 3 public schools of Sumidouro, a city located in Rio de Janeiro, in 2019. The participating schools were randomly allocated into three groups: one school in the control group (no intervention), another school in the intervention group 1 (environmental modifications and addition of a self-service system) and other one in the intervention group 2 (environmental modifications, addition of a self-service system and inclusion of new dishes in the menu). The school from the intervention group 1 received environmental interventions based on the principles of choices architecture, which consisted of: creation of posters and displays with motivational messages and images to promote healthy eating habits and encourage the consumption of school meals; implementation of tablecloth with colorful images of fresh food on the cafeteria tables; allocation of fruits in a prominent position, in a fruit stand; placement of a banner with the daily menu at the cafeteria’s entrance; and application of footprints on the floor that, in a playful way, directed students to the cafeteria. Moreover, a self-service system was implemented through the addition of a thermal food distribution counter in the cafeteria. The school from intervention group 2 received the same interventions as the school from intervention group 1, plus the standardization of gastronomic techniques for changes in the dishes that were part of the menus in 2019, and also the creation of new ones, based on the perceptions about the reasons for non-adherence to school meals reported in focus groups carried out with lunch ladies and schoolchildren. Adherence to school meals (≥3 times a week) was assessed through a question inserted in a questionnaire applied at three different moments (baseline, 1st follow-up [after 5 months of intervention] and 2nd follow-up [after 8 months of intervention]). Longitudinal analyzes were performed through the application of generalized linear models that take into account repeated measures and missing participant outcome data. Piecewise analyzes were also performed, allowing the identification of parameters in interval 1 (baseline - 1st follow-up) and in interval 2 (1st follow-up - 2nd follow-up). An increase in adherence to school meals was observed in the intervention school 1 vs. control school (RR=1.20; p-value=0.0166) and greater increase in intervention school 2 vs. control school (RR=1.27; p-value=0.0041). It was also observed an increase in adherence in the intervention school 1 vs. control school in interval 1 and interval 2 (RR=1.18; p-value=0.0146 and RR=1.14; p-value=0.0424, respectively), and greater increase in adherence in intervention school 2 vs. control school in interval 1 (RR=1.22; p-value=0.0095) and in interval 2 (RR=1.20; p-value=0.0070). There was no statistically significant difference in the variation of adherence between intervention 2 and 1 schools. The acceptability of school meals was evaluated by applying an acceptability test referring to the menu served in schools before the interventions and, to the other menus, after the beginning of the interventions, through the use of facial and verbal hedonic scale. For the longitudinal analyses, generalized linear models were used, taking into account repeated measures and missing outcome data from the participants. There was a significant increase in the acceptability index in the groups of preparations based on meat and sweets among participants in the intervention school 1 (RR=1.36; p-value=0.0008 and RR=3.99; p- value=<.0001, respectively) and intervention school 2 (RR=1.34; p-value=0.0029 and RR=2.22; p-value=<.0001, respectively), when compared to the control school over time. The increase in fruit acceptability was greater among participants in the intervention school 2, when compared to those in the intervention school 1 (RR=1.60; p-value=0.0052). Reduction in the acceptability of bean-based preparations was observed between intervention 2 school and intervention 1 and control schools (RR=0.60; p-value=0.0003 and RR=0.57; p-value=0.0013, respectively). The results suggest that multicomponent interventions based on low cost and complexity environmental modifications, and also on changes in the sensory aspects of meals offered at school through the use of gastronomic techniques, can be effective in increasing adherence and acceptability to school meals, assisting in the creation of public policies aimed at promoting healthy eating.