Effectiveness and feasibility of lowering playground density during recess to promote physical activity and decrease sedentary time at primary school

As physical activity during childhood is associated with numerous health benefits [1–3], social well-being [4, 5] and academic achievement [6], it is recommended for children to engage in at least 60 minutes of moderate-to vigorous-intensity physical activity (MVPA) per day [7, 8]. Independent of the amount of physical activity, high levels of sedentary time (sitting and lying activities that do not increase energy expenditure substantially above the resting level [9]) are associated with obesity and metabolic risks [10, 11]. Therefore, the amount of sedentary time should be limited [12]. Despite numerous health benefits associated with being sufficiently active and limiting sedentary time, many children do not meet the physical activity guidelines (e.g. only 4.6% of the girls and 16.8% of the boys in Europe meet MVPA guidelines with 1.7% of the Belgian girls and 14.0% of the Belgian boys meeting the guidelines) and spend too much time being sedentary [13].

Physical activity in children declines as children age [14–16], and therefore, it is important to promote physical activity during childhood. As most children spend a large proportion of their waking hours at school, elementary school is an ideal setting to promote physical activity. Moreover, children from different ethnic and socio-economic groups can be reached in elementary schools [17]. During a school day, recess and physical education offer opportunities to contribute to the recommended 60 minutes of MVPA per day [18]. Moreover, recess takes place on a daily basis and accounts for nearly a quarter of the average school day in Flanders. Nonetheless, children’s physical activity levels during recess are low, especially in girls [19]. Ridgers et al. hypothesized that children should engage in MVPA for a minimum of 40% of recess; however, only 14.9% of the boys and 4.3% of the girls met this recommendation in the UK [19]. These findings show that increasing children’s MVPA during recess is an essential and realistic objective and strategic interventions are needed to increase physical activity levels of children during recess [19, 20].

Recently, two reviews studied the interventions that have been trialed to promote physical activity in recess settings [21, 22]. Different intervention strategies were identified. Most of the interventions added equipment or materials to the regular playground offerings and also playground markings were frequently used. Furthermore teacher involvement, physical structures (e.g. soccer goal posts, basketball hoops), color coded playground areas and zones, court rotation, organized activities and active video games in order to increase children’s physical activity levels during recess were investigated. Most of these interventions were successful in increasing children’s physical activity levels during recess [21, 22]. While in a recent focus group study, lack of access to space was recognized as an important barrier for active play during recess according to 10- to 13-year old children [23], the effect of more play space for children during recess on their physical activity has not been studied yet.

In a cross-sectional study, available play space per child was a significant negative predictor of observed sedentary time and a positive predictor of observed vigorous activity [24] and Harten et al. found in an explorative study that boys (mean age = 9.3 years ± 0.5) in large-space areas were more objectively active compared to boys in small-space areas [25]. In a subsample of the latter study, the effect of increasing playground density by using markers to decrease play space on children’s estimated energy expenditure, based on activity counts measured by activity monitors, was also investigated. By decreasing playground space, it was found that boys were less active in smaller play areas (13.0 - 15.8 m² per child) compared to medium (71.6 - 86.8 m²) and large (130.4 – 157.9 m²) play areas [25].

It is not obvious to put more play space at children’s disposal during recess, as many schools do not have enough space to increase the playground area. A possible solution for this problem is to decrease playground density by splitting up recesses and consequently decreasing the number of children sharing the playground. Not only children’s play space will increase, but children will also have better access to sports materials (e.g. soccer and basketball hoops, playground markings, loose materials,..) as fewer children are having recess at the same time. Moreover, such an intervention is free of cost and does not require any additional training of supervisors.

In preschool children, lowering playground density, by splitting up recesses, had a positive effect on preschoolers’ objectively measured physical activity levels and was effective in lowering sedentary time [26]. However no studies were found investigating the effectiveness of lowering playground density during recess in elementary schools.

Therefore, the primary aim of this pilot study was to test the effectiveness of lowering playground density in order to increase physical activity levels and to decrease sedentary time in 9–12 year old-children. Harten et al. concluded that play space was an important variable in determining the free play activity of boys, but not of girls [25]. Therefore it is expected that this intervention might affect boys and girls differently and the moderating effect of sex in the relation between play space and children’s physical activity was tested.

More active children are more autonomously motivated [27] and are more likely to be active, independently of the available play space. Consequently it is hypothesized that less active children will benefit more from this intervention compared to more active children. Therefore, moderating effects of children’s baseline physical activity during recess were tested. Furthermore Fremeaux et al. stated that more activity at one time (e.g. during the intervention recess) will be compensated for by less activity at another time (e.g. during another recess) [28] due to the effect of intrinsic biological control on physical activity [29]. This implies that the decrease of playground density might be successful during recess, but cannot be detected in a measure of physical activity during the whole school day [28]. Therefore, the intervention effects on children’s physical activity and sedentary behavior during the entire school day were also tested. By doing so, the activitystat hypothesis, was tested [28]. The term ‘activitystat’, refers to the central control of PA according to a set point for energy expenditure. As the actual implementers of the intervention were the teachers, and not the researchers it is important to investigate if the intervention was attainable to conduct in order to increase sustainability. Therefore, a third aim of this study was to test the feasibility of lowering playground density during recess in primary school. Additionally, cross-sectional baseline associations between play space and physical activity were investigated.

Lowering playground density resulted in significant changes of sedentary time, LPA and MVPA during recess. Children engaged in less sedentary time, less light-intensity physical activity, and more moderate- to vigorous-intensity physical activity during split recesses. These changes were comparable to the changes in a similar intervention that was conducted in preschoolers [26]. However, these positive changes were rather small compared to other recess interventions aiming to increase physical activity [21] (e.g. Verstraete et al. found an increase of 13% in MVPA by providing game equipment during recess while MVPA decreased in the control group [20]). This can be attributable to the short duration of the recesses in which the playground density was lowered (15–20 minutes) and the short period of the intervention (one week) compared to other studies were interventions were conducted during 4 weeks [40] or 3 months [41]. It might be possible that children had insufficient time to adapt their play behavior to the increased available space. Another possible explanation for the small effects is the fact that the available area per child was already very reasonable and quite high; therefore it might be possible that children already had enough space to move around as they wished. Or alternatively the increase in play space may have been not large enough to alter children’s PA or sedentary time during recess. In a study by Harten et al. children were more active in medium (71.6 - 86.8 m² per child) and large (130.4 – 157.9 m² per child) play areas compared to small play areas (13.0 – 15.8 m² per child); these areas were much larger than the intervention space in the present study [25].

Larger effects could possibly be found if a multicomponent intervention was used instead of only focusing on the increased available space as many intervention studies confirmed the positive effects of game equipment [20, 42] and playground markings [40, 43, 44] on children’s physical activity levels during recess at elementary school. This intervention also might be more effective in schools with smaller playgrounds (e.g. inner-city schools): participating schools in this study already had large playgrounds compared to schools in other studies [24]. Also, implementing this intervention during lunch break could increase the effectiveness as lunch recess lasts longer than other recesses (± 1 hour) [20].

Similar to other studies we found that boys engaged in less sedentary time and more MVPA during recess compared to girls [24, 45]. These differences remained significant during split recesses. Blatchford et al. showed that boys in primary school were more involved in ball games (e.g. throwing and catching, soccer, basketball and other derived games), whereas girls were more engaged in sedentary play, conversation, and rope skipping [46]; which explains the differences in physical activity levels during recesses in boys and girls. On the contrary, Mota et al. found in a small study sample that girls were more engaged in MVPA during recess compared to boys [47]; the majority of boys spent much of their playtime in sedentary play such as swapping game cards or simply talking; whereas girls continued participation in traditional, apparently more active, playground games.

The intervention effects during recess were similar in boys and girls. This was contrary to what was expected based on the findings of Harten et al. [25] who found that play space was a more important variable in determining the free play activity of boys compared to girls. The difference between the current study and the study of Harten et al. can be explained by the fact that Harten al. investigated the effect of change in play area by using markers to define out-of-bound areas, whereas in the current study, the area of the playground remained the same during the intervention, but only the number of children sharing the playground decreased. By decreasing the number of children sharing the playground, children have better access to the available play equipment (e.g. jumping ropes, balls,..), as this equipment need to be shared with less children. This may have led to the fact that intervention effects were similar in boys and girls as both had easier access to their favorable play equipment. However during the entire school day, lowering playground density was effective in increasing boys’ MVPA but not girls’ MVPA, so boys did not compensate more MVPA during recess at another moment during the school day. Girls who engaged in more MVPA during recess, compensated for this increase by decreasing their MVPA during another moment during the school day (e.g. being less active during physical education class). The finding that physical activity interventions were more effective in boys is consistent with the results of an environmental intervention study in the US, aiming at increasing physical activity at schools [48]. Therefore, tailored physical activity interventions are necessary; aiming at increasing girls physical activity during the school day.

The intervention was more effective in the least active children, compared to the most active children during recess. Children that were least active during usual recess engaged in less sedentary time and more MVPA during the split recess, whereas these effects were not significant in the most active children. Apparently the most active children want to be active during recess, regardless of the available space, probably because they are more autonomously motivated towards PA [27]; whereas less active children probably experience more opportunities to be physically active due to the increase in play space when recess was split. By increasing the play space during recess, it is easier to participate in active games and make use of game equipment and playground markings during recess which results in higher physical activity levels during recess in least active children. The positive intervention effect that was found during recess in least active children was not found during the entire school day. It seems that the least active children compensate less sedentary time and more MVPA during recess with more sedentary time and less MVPA at another moment during the school day [28]. Consequently, strategies are needed to maintain the increase of MVPA in least active children during the entire day, by implementing multi-component interventions to increase physical activity during the school day. During the entire school day, most active children engaged in less sedentary time, more LPA and more MVPA during the intervention condition, whereas no significant differences were found during recess. These effects are probably not due to the playground intervention, as this intervention was not effective during recess in most active children but can be due to other things (e.g. more activity during lunch recess on the intervention days, physical education during intervention days,…). Therefore, intervention effects during the entire school day need cautious interpretation.

Although this intervention was free of costs, did not require any adaptations of the school facilities and did not require teacher trainings, the school principals reported some difficulties when implementing the intervention. The need for more supervisors, the incomplete lesson times before and after recess (as lesson times were split up because recess started earlier or later as usual), nuisance from recess for other classes and the absence of friends of children from other classes that did not have recess the same time were some of the difficulties reported in the questionnaire. Also the low response rate of the schools (30.0%) could be explained by the low perceived feasibility of the intervention by the school principals as those principals did not find it feasible to implement the intervention. Due to these reported problems, none of the participating principals fully agreed that it is possible to use this intervention further in the future. However, one school principal would implement the intervention in the future on rainy days. As elementary schools in Flanders often have lack of place inside for recess when it rains, lowering the playground density can be a valuable solution to solve this problem. Van Cauwenberghe et al. conducted a comparable study in preschoolers [26]. In preschools, principals reported a higher feasibility and this is possibly due to the fact that preschools have less strict timetables. A possible solution for some of these difficulties is the implementation of the intervention before and after lunch time instead of implementing this intervention during the shorter recess before noon or during the afternoon. In Belgium, children have lunch when recess during the noon starts (during 20–30 minutes); after lunch time children have noon recess outside. Half of the children could start the noon recess with lunch time, while the other half of the children could have recess outside before lunch time. By doing so, lessons will not be disturbed by the nuisance of the recess, lesson times will remain complete and positive intervention effects can be maintained or even enlarged as recess during noon usually lasts longer than recess before noon or in the afternoon. However, it might be possible that some children who go home during lunch recess (16.8% of Flemish school children, unpublished data) miss the positive intervention effects of decreasing playground density. Also the need for more supervisors cannot be solved by implementing the intervention during lunch time.

Despite these difficulties, all principals agreed that children enjoyed the intervention and positive intervention effects were found, so it might be worth to further investigate the effects of lowering playground density during lunch time.

The small number of children and schools included in the study due to the low perceived feasibility; the short intervention period were limitations of this study. As accelerometers were handed out on Monday morning before the first recess and collected on Friday afternoon after the last recess, Mondays and Fridays were measured incomplete which made our data unsuitable to investigate the intervention effects during the entire day (including data before and after school time). This also forms a limitation of the study. Baseline characteristics concerning play equipment and type of play space differed between the schools, however as a within-subject design was used, the influence of these different baseline characteristics was minimal. However it is possible that a larger increase of PA was measured in the schools with more playground equipment, as this equipment became more available for the children during intervention. The use of accelerometers as objective measurement tools for physical activity and sedentary time, the high external validity due to the within subjects design and the measurements during the whole school day were strengths of the study.

The present study showed that lowering playground density can be an effective intervention for decreasing children’s sedentary time and increasing their physical activity levels during recess. Although positive intervention effects were small, decreasing playground density could be very valuable and effective in a larger multi-component intervention in schools to increase physical activity and to decrease sedentary time. Further research should focus on the effect of this intervention during lunch time in schools with denser playgrounds, as this can enlarge the intervention effects and forms a solution for a lot of the reported difficulties by the principals.