Background Children in the USA are experiencing obesity and overweight at epidemic rates. Schools have started to make policy decisions based on a popularly presumed connection between nutrition and academic achievement. This study aimed to determine whether such a relationship exists, and if so, its nature.
Methods Data from the Food Consumption Questionnaire administered to approximately 12 000 fifth graders as part of the nationally representative Early Childhood Longitudinal Study-Kindergarten Cohort was analysed using regression analysis. The analysis used propensity score matching to examine the relationship between students' reported fast-food consumption and their test scores in reading and mathematics. Propensity score matching attempts to control for a host of background characteristics that might be correlated with both fast-food consumption and test scores.
Results Standard ordinary least squares regression showed that test scores decreased as reported fast-food consumption increased. In the propensity-score-matched analysis, which controlled for 25 student background characteristics, higher-than-average fast-food consumption (‘four to six times in the last 7 days’ or more) was associated with significantly lower test scores in both reading (−11.15 points or 0.48 SD) and math (−11.13 points or 0.52 SD), even when teacher experience, school poverty level and school urbanicity were also included in the model.
Conclusions Although this study is preliminary, its results indicate that the relationship between poor nutrition and test scores may in fact be quite negative, strengthening the impetus for schools to consider policies that support students' healthy eating. In order to strengthen these findings and investigate possible mechanisms through which poor nutrition might affect test scores, there is a need for further research.