The Evidence on Teaching

Effects of physical activity interventions on cognitive and academic performance in children and adolescents: a novel combination of a systematic review and recommendations from an expert panel

Singh AS, Saliasi E, van den Berg V, Uijtdewilligen L, de Groot RHM, Jolles J, Andersen LB, Bailey R, Chang YK, Diamond A, Ericsson I, Etnier JL, Fedewa AL, Hillman CH, McMorris T, Pesce C, Puhse U, Tomporowski PD, Chinapaw MJM · 2019

grade Creviewindependentnot-applicable
Sample
58 intervention studies screened; only 11 (19%) rated high methodological quality; conclusions drawn from those 11
Population
Children and adolescents in physical-activity intervention studies
Design
PRISMA systematic review with explicit methodological-quality rating, plus an international expert panel (19 authors, including most of the field's principal investigators) that formulated research priorities. PROSPERO CRD42017082505. Graded C — it is a vote-count synthesis over a mixed-quality corpus, not a bias-corrected pooled estimate, and the expert panel is opinion, not evidence.
Key findings
The field's own consensus review says the evidence is inconclusive, which is the most important thing in it. Only 11 of 58 studies (19%) were rated high quality. Among those, benefits appeared in 10 of 21 (48%) cognitive constructs and 15 of 25 (60%) academic analyses — coin-flip vote counts. The headline positive claim, "strong evidence for beneficial effects of PA on maths performance", rests on 6 of 7 outcomes across just FIVE studies, which is a small-corpus vote count with no bias correction and no effect size attached. The panel itself named causality as the most pressing unresolved question, and the review conclusion is "inconclusive evidence for the beneficial effects of PA interventions on cognitive and overall academic performance."
Genetic confound
Restricted to intervention studies, so genetic confounding of the treatment contrast is not the issue; the issue is that 81% of the corpus was too weak to use.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Corpus qualityshare rated high methodological quality11 of 58 studies (19%)standardizedacross corpusnonenot-applicabledomain-skill
Cognitive performance (high-quality studies)vote count of significant beneficial effects10 of 21 constructs (48%)standardizedend of interventionbusiness-as-usualend-of-treatmentfar-transfer
Academic performance (high-quality studies)vote count of significant beneficial effects15 of 25 analyses (60%)mixedend of interventionbusiness-as-usualend-of-treatmentdomain-skill
Mathematics performancevote count6 of 7 outcomes beneficial, across 5 studies — the basis of the "strong evidence for maths" claimmixedend of interventionbusiness-as-usualend-of-treatmentdomain-skill

Cited by