A single bout of exercise lowers state anxiety in young people, with heterogeneity that undercuts the number

Design

Systematic review and random-effects meta-analysis with GRADE assessment, search to March 2026

Population

739 participants aged 6 to 18 across 12 studies

Primary outcome

State anxiety immediately after a single bout of exercise

Effect

Pooled SMD -0.77 (95 per cent CI -1.37 to -0.16, P = 0.01) with I-squared 94.9 per cent; subgroup estimates -0.19 to -0.62 apart from an outlying -2.97 in adolescents

Twelve studies with 739 participants aged 6 to 18 were pooled to test whether one session of exercise reduces state anxiety immediately afterwards. It did: standardised mean difference -0.77 (95 per cent CI -1.37 to -0.16, P = 0.01), graded as moderate certainty.

The heterogeneity, though, was 94.9 per cent, which means the studies were measuring effects of wildly different sizes. That shows in the subgroups. Sensitivity-adjusted analyses produced effects of -0.19 for aerobic exercise, -0.25 for multimodal, -0.28 against passive controls and in non-stress-induced conditions, -0.30 for group-based interventions and -0.62 in high-autonomy settings - all small - alongside an estimate of -2.97 (-3.39 to -2.55) in adolescents aged 13 to 18, which is an implausibly large effect for a single exercise session and sits oddly against every other figure in the paper.

Read the small consistent estimates rather than the headline. A single session of group-based aerobic or mixed exercise, chosen rather than imposed, produces a small immediate reduction in anxiety in young people. That is worth knowing when a school or a parent asks, and it is a reasonable thing to suggest alongside treatment. It is not a treatment for an anxiety disorder, the effect is measured minutes to hours later with no follow-up, and the authors themselves label the subgroup findings exploratory.

• Suggest a single bout of chosen, group-based activity for acute anxiety, expecting a small effect.
• Do not present this as treatment for a diagnosed anxiety disorder; state anxiety was the outcome.
• Read the -0.19 to -0.62 subgroup estimates rather than the -0.77 headline or the -2.97 outlier.
• Autonomy mattered more than modality here - the child choosing the activity is part of the intervention.
• Note the absence of any follow-up; nothing is known about duration of effect.

The statistics, in plain English

An I-squared of 94.9 per cent means almost all the variation between studies is real difference rather than chance, which makes any single pooled figure close to meaningless as a summary. The adolescent subgroup estimate of -2.97 with a narrow interval is the kind of result that usually signals a small number of unusual studies rather than a genuine developmental effect, and it is inconsistent with every other subgroup in the same analysis. GRADE moderate certainty applies to the existence of an effect, not to its size.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.