The Evidence on Teaching

Physical Activity, Fitness, Cognitive Function, and Academic Achievement in Children: A Systematic Review (American College of Sports Medicine position stand)

Donnelly JE, Hillman CH, Castelli D, Etnier JL, Lee S, Tomporowski P, Lambourne K, Szabo-Reed AN · 2016

grade Creviewindependentfailed
Sample
64 studies on cognition/learning/brain plus 73 studies on academic achievement/attention, published January 1990 - September 2014
Population
Children aged 5-13
Design
Systematic review commissioned as an American College of Sports Medicine position stand, grouping evidence by design (cross-sectional, longitudinal, acute, intervention) and assigning ACSM evidence-category grades. Graded C — a narrative-plus-vote-count synthesis of predominantly observational work, with no pooled effect sizes and no publication-bias assessment. The authors themselves flag that many included studies were high risk of bias for lack of randomization, inadequate power, no blinding and unaccounted confounders.
Key findings
The sports-medicine establishment's own position stand splits the verdict exactly along the line this archive draws. Physical activity gets evidence category B for cognition and brain function — driven substantially by cross-sectional and acute laboratory studies. Physical activity and physical education get evidence category C for academic achievement, with the explicit judgement that they have a "NEUTRAL effect on academic achievement". The review notes that favourable cross-sectional and longitudinal results are not reproduced by controlled experiments. It also records that only ONE study on PA and learning met inclusion criteria at all, and that one was cross-sectional. The genuinely defensible claim it makes is the negative one: increasing PA does not HARM achievement, so activity time is not bought at an academic cost.
Genetic confound
Not addressed. The bulk of the underlying corpus is cross-sectional and longitudinal observational work in which fit, active children are compared with unfit, inactive ones — the textbook passive gene-environment correlation setup, given fitness heritability of roughly 55-72%.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Cognition, brain structure and functionACSM evidence categorycategory B — positive influence claimedstandardizedmixed designs; largely cross-sectional and acute lab tasksunclearunclearnear-transfer
Academic achievementACSM evidence categorycategory C — "neutral effect"standardizedmixed designsunclearuncleardomain-skill
Learningnumber of qualifying studies1 study, cross-sectionalunknowncross-sectionalnonenot-applicabledomain-skill
Harm from displacing instructional timequalitativeno indication that increases in PA negatively affect cognition or academic achievementmixedacross corpusunclearnot-applicabledomain-skill

Cited by