Article type
Abstract
Background: Worldwide, school meals bridge socioeconomic disparities by promoting better nutrition and learning outcomes for all children. Our goal was to update the 2007 Kristjansson School Meals review. The sheer volume, diversity, and complexity of these interventions and studies present challenges. School Meals studies have multiple interacting components, vary in design and implementation, and occur in diverse contexts. Our Cochrane systematic review tackled these complexities to assess the impact of school meals across educational, cognitive, and nutritional outcomes in varying socioeconomic conditions and contexts.
Methodology: Our team was made up of systematic review experts, statisticians, nutritionists, a psychologist, equity experts, an information specialist, and students. We found more than 45,000 articles, so we required a large team to help with hand-searching, screening, reading 505 full-text articles, extracting data, and performing ROB calculations. We included 45 studies with 90 papers. We encountered several challenges when performing the analyses: the studies were in very diverse settings (30 countries), the meals given differed widely between settings, more than 25 outcomes were presented, and diverse measures and statistical methods were used for each outcome.
Results: We meta-analyzed data from HIC and LMIC separately and RCTs and NRSIs separately. We were able to conduct 12 main meta-analyses on our primary outcomes. We also conducted 12 equity-focused subgroup analyses; some by sex and others by socioeconomic disadvantage. Academic achievement and cognitive outcomes used different measures; therefore, we used SMDs from an online calculator. EXCEL calculations were done to derive the necessary numbers. To understand why and how some meal programs worked or not, a team member performed a realist review. We also performed subgroup analyses by the level of energy given.
School meal interventions probably increased math but not reading scores. Girls demonstrated meaningfully greater change than boys, whereas the level of disadvantage seemed to be unrelated to change. School meals probably had a small but significant impact on growth in intelligence scores. Disadvantaged children gained more in intelligence than less disadvantaged children. The realist review found factors in the context and implementation that affected school meal programs. This information is invaluable to policymakers and planners.
Methodology: Our team was made up of systematic review experts, statisticians, nutritionists, a psychologist, equity experts, an information specialist, and students. We found more than 45,000 articles, so we required a large team to help with hand-searching, screening, reading 505 full-text articles, extracting data, and performing ROB calculations. We included 45 studies with 90 papers. We encountered several challenges when performing the analyses: the studies were in very diverse settings (30 countries), the meals given differed widely between settings, more than 25 outcomes were presented, and diverse measures and statistical methods were used for each outcome.
Results: We meta-analyzed data from HIC and LMIC separately and RCTs and NRSIs separately. We were able to conduct 12 main meta-analyses on our primary outcomes. We also conducted 12 equity-focused subgroup analyses; some by sex and others by socioeconomic disadvantage. Academic achievement and cognitive outcomes used different measures; therefore, we used SMDs from an online calculator. EXCEL calculations were done to derive the necessary numbers. To understand why and how some meal programs worked or not, a team member performed a realist review. We also performed subgroup analyses by the level of energy given.
School meal interventions probably increased math but not reading scores. Girls demonstrated meaningfully greater change than boys, whereas the level of disadvantage seemed to be unrelated to change. School meals probably had a small but significant impact on growth in intelligence scores. Disadvantaged children gained more in intelligence than less disadvantaged children. The realist review found factors in the context and implementation that affected school meal programs. This information is invaluable to policymakers and planners.