A family-school-community collaborative PA intervention was associated with improvements in exercise cognition and perceived benefits in primary school students, while incremental between-group gains in PA were modest and not statistically robust after accounting for class clustering.
Key Findings
Results
The family-school-community collaborative PA intervention was associated with significantly higher exercise cognition compared to the control group.
Adjusted class-clustered ANCOVA showed β = 3.94, 95% CI: 1.29–6.60; p = 0.003; Hedges' g = 0.26
515 fifth-grade students (10–11 years) from a public primary school in Guangdong, China participated
Intervention group: n = 255; control group: n = 260
Exercise cognition was measured using the EBBS-CN instrument
The study was a quasi-experimental, one-academic-year design
Results
The intervention was associated with significantly higher perceived exercise benefits compared to the control group.
Adjusted class-clustered ANCOVA showed β = 3.46, 95% CI: 1.03–5.89; p = 0.008; Hedges' g = 0.24
Perceived exercise benefits were a subscale of the EBBS-CN instrument
The effect size was small (Hedges' g = 0.24)
Results
The adjusted between-group effect of the intervention on physical activity was not statistically significant.
Adjusted class-clustered ANCOVA showed β = 0.11, 95% CI: -0.15–0.36; p = 0.418; Hedges' g = 0.18
PA was assessed using the PAQ-CN instrument
PA increased from pre- to post-intervention in both groups
The effect was described as 'modest and not statistically robust after accounting for class clustering'
Results
The intervention had no statistically significant effect on perceived exercise barriers.
Adjusted class-clustered ANCOVA showed β = 0.55, 95% CI: -0.60–1.70; p = 0.354; Hedges' g = 0.08
Perceived exercise barriers were a subscale of the EBBS-CN instrument
The effect size was very small (Hedges' g = 0.08)
Results
Post-intervention physical activity was positively associated with exercise cognition and perceived exercise benefits, but not with perceived exercise barriers.
Exploratory class-clustered regression models with baseline PA adjustment showed p ≤ 0.001 for both exercise cognition and perceived exercise benefits
The association with perceived exercise barriers was not statistically significant (p = 0.453)
These were exploratory associations, not pre-specified primary outcomes
Methods
The study employed a quasi-experimental design with class-level clustering in a single public primary school in Guangdong, China.
515 fifth-grade students aged 10–11 years were allocated to intervention (n = 255) or control (n = 260) groups
The intervention lasted one academic year
The control group received usual physical education
Intervention effects were estimated using class-clustered ANCOVA adjusting for baseline outcome values and sex
The single-school design was noted as a limitation
Kong L, Chen X, Gong B, Zhang X, Guo M. (2026). Effectiveness of a family-school-community collaborative physical activity intervention.. Frontiers in public health. https://doi.org/10.3389/fpubh.2026.1767961