Changes in physical activity during the transition from primary to secondary school in Belgian children: what is the role of the school environment?

The promotion of physical activity (PA) among young people has become a public health priority because of the well-known health benefits [1, 2]. The belief that the behavior of young people is easier to control, to develop or to change into more healthy behavior than the behavior of adults [3], underscores the importance of primary prevention in young people. Furthermore, persistent PA at a young age considerably increases the probability of being active in adulthood [4] which in turn can contribute to the quality of adult life. Despite these benefits, a large proportion of school-aged youth does not achieve the public health recommendations of 60 minutes moderate-to-vigorous physical activity (MVPA) per day [5]. Moreover, PA levels tend to decline during adolescence [6, 7].

Research has shown that important life events are key periods that coincide with changes in PA patterns [8]. For children, the transition from primary to secondary school is an important life event. Many European countries including Belgium have distinct primary and secondary schools. In the Belgian educational context, children spend the first 9 years in a primary school (3–11 year olds) after which they normally move to a secondary school (mostly located elsewhere) where they spend 6 years (12–18 year olds). Different longitudinal studies have shown that the transition from primary to secondary school is characterized by changes in PA patterns has been shown by different longitudinal studies [8‐12]. However, the directions of these changes are not straightforward. Two longitudinal studies have shown that PA levels gradually decreased during childhood but were characterized by an obvious drop during the transition from primary to secondary school [8, 9]. Specifically, Niven and colleagues observed a decrease in total PA, PA during school day break and lunch time and PA after school. In contrast, Cooper and colleagues described that objectively determined MVPA slightly increased during this transition [10]. Jago and colleagues [11] investigated after-school and weekend MVPA and concluded that objectively determined after-school MVPA tended to decline during the transition from primary to secondary school, while objectively determined weekend MVPA tended to increase during this transition. Furthermore, Cardon and colleagues found an increase in time spent bicycling to school during the transition from primary to secondary school [12]. Based on these results, the changes in PA during the transition from primary to secondary school seem to be dependent on the context of PA. Furthermore, according to several authors the changes seem to be more apparent in boys than in girls [9, 11]. The results of these longitudinal studies clearly underscore the need for studies to further scrutinize the change in context-specific PA during the transition from primary to secondary school.

An important question to inform the development of PA interventions is why PA changes during this transition. In line with the perspectives of the ecological models, behavior change occurs as a result of the dynamic interaction between the individual and the environment in which he/she lives, learns, works and plays [13]. Young people spend a large proportion of their time at school. Virtually all children and adolescents attend school-based education. In general, children and adolescents spend between 4 and 8 hours per day for 12 years or more of their lives at school. Schools therefore have the opportunity to offer an environment with the potential to facilitate and to promote PA [14] in a large number of children and adolescents, across different ethnic and socio-economic groups. To foster regular PA, schools worldwide are recommended to strive for a ‘whole-school’ approach that maximizes the opportunities for school-aged youth to be active [15, 16]. The ‘whole-school’ approach that has been outlined in the Toronto Charter for Physical Activity [17] goes beyond providing physical education classes. Key components of a ‘whole-school’ approach are: providing resources and suitable physical environments to support structured as well as unstructured PA, supporting active travel to school and provision of opportunities to be active during the school day, during class, during breaks at lunch time and after school (Toronto charter). In line with the ‘whole-school’ approach, a conceptual framework for PA programs within school-community partnerships was developed for the schools in Flanders (Belgium) to offer schools guidelines to develop extracurricular PA programs [18]. The framework consists of five complementary components: sports and PA during lunch break, active school yards or playgrounds, active commuting to school, health education policy and sports and PA after school. Since a few years, Flemish primary and secondary schools are sensitized and supported to have a school policy that promotes PA. However, no specific strategies were developed by the Flemish government: the Flemish schools are free to choose which promotion programs they implement. Consequently, the extent of PA promotion differs; some schools focus on one of the five framework components, while other schools focus on all five framework components [19]. The differences in approaches between schools may contribute to the change in PA observed when moving schools.

To our knowledge, the only study, investigating the contribution of school environmental factors to the change in PA when moving schools was a qualitative study [20]. Through the use of narrative interviews, Knowles and colleagues (2011) investigated factors (PA opportunities, environment, sense of self when active and individual issues) related to the decrease in PA during the transition between primary and secondary school in 14 British girls. A main conclusion of this study was that the change in school environment plays a central role in the change in PA patterns that occur when moving schools [20]. Specifically, according to this study, a positive environment could be provided by ensuring a choice of activities in the physical education lessons, reducing the focus on competence and competition and recognizing the importance of social support by the involvement of friends in PA [20].

Based on the shortcomings in the present literature, the aim of this study was threefold.

A first study aim was to investigate how PA behavior changes when Belgian primary school children make the transition to secondary school. Changes in self-reported walking and cycling to and from school, extracurricular activity at school and total PA as well as changes in pedometer and accelerometer determined step counts and MVPA were examined. A second aim was to investigate if school environmental characteristics with regard to the five framework components (sports and PA during lunch break, active school yards or playgrounds, active commuting to school, health education policy and sports and PA after school) differed between Belgian primary and secondary schools. The third and last aim of this study was to investigate if changes in school environmental characteristics can predict changes in PA in Belgian schoolchildren. Since in previous studies the changes in PA during the transition from primary to secondary school were found to be more apparent in boys than in girls [9, 11], the moderating effects of gender was investigated with regard to the first and third research question. Furthermore, since change in PA over time has shown to be related to the start level of PA [21], the moderating effects of the level of PA at baseline was also investigated with regard to the first and third research question.

The results of this study confirm that the transition from primary to secondary school is characterized by a change in PA. In line with the literature [8, 11, 12] the changes in PA seem to be dependent on the context of PA. Comparable to the results of the study of Cardon and colleagues [12], the data presented here indicated an increase in the self-reported number of minute’s active transport to and from school after the transition from primary to secondary school. Furthermore, a decline was found in self-reported extracurricular PA and total PA. This is in line with the results of the study of Niven and colleagues [8].

Accelerometer determined weekday MVPA increased while for pedometer/accelerometer determined weekday step counts a large decline was observed among those who achieved the PA guidelines in primary school. Among those who did not achieve the guidelines in primary school weekday step counts remained nearly stable. This result underscores the fact that the start level of PA predicts the change in weekday step counts after the transition from primary to secondary school. To our knowledge, this was the first study to investigate changes in weekday PA determined by accelerometers or pedometers. More research is clearly needed to investigate this in more detail. A somewhat surprising result was that no differences between boys and girls were noted in the changes in PA. In two previous longitudinal studies the changes in PA were more apparent in boys than in girls [9, 11]. Despite the fact that in the present study no gender differences were found in the change in PA, the results of previous studies underscore the fact that future research still should be encouraged to investigate differences between boys and girls.

The implementation scores of sports and PA during lunch break, active schoolyards and playgrounds and health education policy were higher (increase > 0.50) in secondary schools than in primary schools. Although the change in the score of sports and PA after school was small, the score was significantly lower in secondary compared to primary schools (decrease < 0.50) and the score of active commuting to school showed no difference. In general, secondary schools seem more likely to foster strategies to promote PA during school hours than primary schools who seem more likely to foster strategies to promote PA after school. Primarily, secondary schools are overall larger and have therefore more available space and facilities than primary schools. This may explain the higher implementation score of active schoolyards and playgrounds. Secondly, the availability of space and facilities may in turn be conducive to organize sports and physical activity during lunch break. This may explain the higher implementation score on this component. Third, larger schools typically have more teachers. Consequently, teachers following refresher courses on sports and PA during school hours can be replaced more easily. This ultimately may encourage following refresher courses on sports and PA. Following this reasoning, secondary schools may have a higher score on the item concerning training for teachers that contributes to the implementation score of health education policy. Further, a higher implementation score on the component sports and PA during lunch break and the component active schoolyards or playgrounds may affect the subjective norm with regard to the importance of sport and PA for the school. The subjective norm with regard to the importance of sport and PA for the school is one of the items that determined the component health education policy. And as last, the implementation score of health education policy is also determined by how pupils are involved in the decision making about sports and physical activity. The cognitive abilities of secondary school children may be a plausible explanation for a higher involvement in the decision making process. A possible reason for the higher score of sports and PA after school in primary schools can be that in primary schools after school PA programs are seen as a solution for children who are not allowed to go home after school without supervision. Sports and PA programs after school are then used as a form of after-school care.

An earlier study investigating the implementation of the different components of the PA promotion framework in Flemish primary and secondary schools did not find a difference in the implementation score of the different components of the framework to promote extracurricular PA [19]. However, since the implementation score in the study of Cardon and colleagues is an overall score, including the different implementation scores on the different components of the PA promotion framework and in the present study separate scores were conducted for the five components comparison of the study results is difficult.

To our knowledge this was the first longitudinal study that investigated if changes in the school environment can predict changes in self-reported and pedometer/accelerometer determined PA in different contexts. It was promising to find that changes in three components of a framework that was developed to promote extracurricular PA programs were associated with changes in PA.

An increase in the implementation score of health education policy was found to predict an increase in self-reported extracurricular PA independent of gender or the level of PA at baseline. This result underscores the need to further endeavor for a sport and PA minded school-atmosphere. Efforts to change the overall school spirit, the involvement of students in the decision making process about sports and PA and the support teachers receive to have training on sports and PA may be promising.

A positive association was found between the implementation score of active schoolyards and playgrounds with pedometer/accelerometer determined weekday step counts. Furthermore, among boys, total PA decreased less when the implementation score of active schoolyards and playgrounds increased from primary to secondary school while among girls the decrease in total PA was larger when the implementation score of active schoolyards and playgrounds increased from primary to secondary school. Based on this result it seems that the availability of facilities and equipment seems to be important for boys’ total PA but not for girls. This finding is in accordance to the conclusions of several observational studies concluding that PA facilities in school settings are predominantly used by boys while girls less claim the activity facilities [45, 46]. A possible explanation can be that the facilities available are more interesting for activities preferred by boys than by girls. Furthermore, for girls recess and lunch break are considered to be an opportunity to socialize with friends [47].

Active commuting to school was the third framework component that was of importance. The association found between the change in the implementation score of active commuting to school and both self-reported total PA and accelerometer determined MVPA was dependent on the level of PA at baseline. For both total PA and MVPA, an increase in the implementation score of active commuting to school was found to have a more positive effect on children who did not achieve the guidelines in primary school compared with those who did achieve the guidelines in primary school. Children who were not sufficiently active in primary school seem more impressible to the promotion of active commuting to school, the availability of facilities for active commuting and safety issues related to active commuting, than children who did achieve the guidelines in primary school. It is possible that children who were not sufficiently active in primary school are not interested in sports activities but are more receptive for physical activities that are not sports activities. Surprisingly, this component does not directly predicts the change in active commuting, A possible explanation for these findings can be that schools that promote active commuting to school, provide facilities for active commuting, are situated in a safe neighborhood and consequently score higher on the implementation score of active commuting, score also higher on other factors. For example, it is possible that schools with higher scores on the implementation score of active commuting score also higher on the implementation score of active schoolyards and playgrounds This may have had an influence influences on the results.

Two components were no predictors of changes in PA: sports and PA after school and sports during lunch break. This finding is somewhat surprising. Earlier studies showed that factors like access to activity facilities [48], access of equipment to be physically active [49‐52], provision of after-school PA [53] and recess PA programs [54] have been associated with participation in extracurricular PA. These factors were used in this study to determine the implementation scores of sports and PA after school and sports during lunch break. However, there is limited data available on adolescents. It is possible that among primary school children but not among older, secondary school children, these components are of importance and can contribute to PA. Consequently, there is a need for additional research to investigate this in more detail.

This study provides more insight into the changes in PA during the transition from primary to secondary school, the differences in school environmental characteristics with regard to the five components of the Flemish framework to promote extracurricular PA, and the effect of changes in these school environmental characteristics on changes in self-reported and pedometer/accelerometer determined PA. Based on the results of this study it was possible to conclude that the changes found in PA were dependent of the context of PA. Furthermore, secondary schools seem more likely to foster in-school strategies (sports and PA during lunch break, active schoolyards and playgrounds and health education policy) than primary schools while primary schools are more likely to foster sports and PA after school than secondary schools. As last, the implementation of three components of the framework developed to promote extracurricular PA were found to be important to predict changes in PA during the transition from primary to secondary school: active commuting to school, active schoolyards and playgrounds and health education policy. As these components can contribute to higher levels of PA, efforts are needed to extent the implementation of these components in all primary and secondary schools. Moreover, components need to be implemented in school settings as early as possible to encounter the changes in PA in a positive manner during the transition from primary to secondary school. However, the contribution of other social and physical environmental factors to the changes in PA need to be further explored.