Key messages

• School meal programs in low- and middle-income countries lead to a small increase in math test scores, but may have little to no effect on reading test scores. School meals lead to a small increase in enrollment, but may have little to no effect on attendance. School meals probably result in small gains in height and weight for age. We are unsure whether they decrease overweight and obesity.

• Research should be a key part of planning and running school meal programs. To make it easier to compare results from different studies, researchers, parents, and policymakers should work together to agree on a list of the most important goals for children's health, development, and success in school.

Why did we want to review studies on school meals?

Worldwide, many children do not get enough nutritious food to keep them healthy and ready to learn in school. This affects their ability to enroll, attend school regularly, and learn there. School meals aim to reduce hunger and improve children's learning, focus, and overall health.

School meals are one of the world's largest social safety nets. We sought to understand their effectiveness for children in disadvantaged circumstances across the globe. Most previous reviews of the evidence are limited in geographical scope, and many use narrative analyses.

What did we want to find out?

We asked the following two health equity (fairness) questions.

• Do school meals improve the physical health of socioeconomically disadvantaged children worldwide?

• Do school meals benefit children who are disadvantaged by sex and low family income more than they benefit less disadvantaged children?

What did we do?

We searched for studies that evaluated the effects of school meals for school students aged five to 19 years. We included randomized studies, which allocate students or schools to different study groups at random. However, because randomized studies are not always possible, we included non-randomized studies that compared school meal groups with groups that received no school meals. These non-randomized studies needed to measure outcomes at least twice; once at the beginning of the study and once at the end. We rated our confidence in the findings separately for each outcome.

What did we find?

We found 13 randomized studies and 27 non-randomized studies, which collectively involved more than 91,995 students aged five to 19 years. One study included 59,613 students, while others had student populations ranging from 40 to 6038 students. Across all studies, there was a similar number of boys and girls, although three studies included only girls. Most studies took place in low- and middle-income countries (34 studies), while six were from high-income countries.

In low- and middle-income countries, school meals lead to a small increase in math test scores, but may have little to no effect on reading test scores. School meals lead to a small increase in enrollment, but may have little to no effect on attendance. School meals probably result in small gains in height and weight for age. We are unsure whether they decrease overweight and obesity.

In high-income countries, one study presented very uncertain evidence that school meals may increase attendance slightly.

What are the limitations of the evidence?

Because we found very few eligible studies conducted in high-income countries, we do not recommend applying our findings to this setting.

We are confident in the results for math test scores and enrollment. We are moderately confident in the results for height and weight for age, because we had some concerns about the study methods. We have little confidence in the results for reading test scores and attendance, because we had some concerns about the study methods, and the results differed substantially among the trials. We are not confident in the results for overweight/obesity, because we have some concerns about the study methods, the two studies measured the outcome in different ways, and one of the studies was very small.

Several studies faced barriers due to their context, such as conflict, droughts, bureaucratic delays, and difficulties in distributing funds to caterers, as well as eagerness to initiate school meal programs in control schools.

We need more evidence on school meals and their impact on disadvantaged school students—especially those in secondary school—in countries of all income levels. We also need more studies that assess the difference between disadvantaged groups and more advantaged children in high-income countries. We suggest that researchers, school feeding experts, teachers, and parents collaborate to develop a core set of key outcomes, and that these outcomes be put in an open-access repository.

How up-to-date is this evidence?

This review updates and expands our 2007 review of school meal programs. The evidence is current to November 2024.

Funding

Dubai Cares funded this review through the World Food Program School Feeding Program. The London School of Hygiene and Tropical Medicine also funded this review. The funders had no influence on the review.