Abstract
Background: In many countries, children and youth are not meeting global physical activity guidelines and while schools are regarded as appropriate settings for increasing physical activity, it remains unclear what specific interventions are most effective. A growing societal culture of ‘risk aversion’ is restricting children’s play and the degree of risk and challenge that children are allowed to encounter, even though exposing children to manageable risk helps develop confidence and ability to assess real danger. Very few studies appear to have evaluated the potential of risky play to increase physical activity within the school environment.
Objectives: To investigate if increasing opportunities for children to experience risk and challenge through changing equipment and reducing rules i) was acceptable to schools, and ii) increased children’s physical activity, reduced their body mass index (BMI) and altered their interactions with one another during play.
Methods: A two-year (one-year intervention, one-year follow-up) cluster randomised controlled trial (PLAY Study) was undertaken involving 902 children (aged 5-12) recruited from 16 primary schools in the Otago and Auckland regions of New Zealand. Intervention schools were provided with start-up funds alongside tailored recommendations detailing changes that could be made to improve the play experience for children at their school. Recommendations included limiting rules, incorporating loose parts and allowing scooter use during break time. Control schools were asked to not change any aspects of play or physical activity over the two years. Exit interviews were held with all 16 schools regarding the changes they had made and the acceptability of the intervention. Height, weight, waist circumference and physical activity (accelerometry) were measured in children at baseline, one and two years. Children, parents and teachers completed questionnaires at all three time points investigating how children interacted with each other at school.
Results: Respondents from all intervention schools revealed benefits from participating in the project, citing improvements in children’s behaviour, physical activity and confidence. Play space evaluations showed that intervention schools increased the opportunities for risk and challenge (score difference 1.00 (95% CI: 0.11, 1.89) and wheeled play (1.25 (0.18, 2.32)). However, this did not translate into intervention differences in physical activity. Differences (95% CI) in counts per minute during lunch time were 35 (-51, 120) at one year and -44 (-164, 75) at two years. A similar lack of effect was observed for moderate-to-vigorous physical activity (0.4 minutes (-1.1, 2.0)) and BMI z-score (0.02 units (-0.04, 0.07)) at two years. Assessment of children’s interactions showed that, when compared with the control group, intervention children at two years reported higher odds of being happy at school (OR 1.64 (1.20, 2.25)). However, at this time point teachers from intervention schools reported more bullying (score difference 1.04 (95% CI: 0.49, 1.60)) than control schools.
Conclusions: All schools were positive about changing their play environment to one that included more risk and challenge. Despite schools believing that children were more physically active, no significant differences were observed in objectively measured physical activity. This suggests that children may have changed rather than added to their break time activities. Alternatively, the use of accelerometers to measure physical activity may have been inadequate because children at play are not necessarily expending high amounts of energy. This highlights the need for other methods to assess children’s playtime activities. The findings elicited from questionnaires regarding ‘bullying’ were intriguing and, although statistically significant, the differences observed were relatively small. Further work is required to more accurately assess the complexities of children’s interactions with one another.