Quantitative results

Parts of this chapter have been reproduced from Jago et al.54 © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

This chapter provides the quantitative results from the study. The recruitment data are presented first, followed by data provision and study results. Where findings specifically relate to progression criteria, these are clearly shown.

School recruitment

The study required participation of eight primary schools from South Gloucestershire and four from North Somerset, making a total of 12 study schools. Schools were approached based on the percentage of their pupils claiming free school meals and their intake size. Eleven schools in South Gloucestershire were invited to participate. Two schools did not respond to the invitations, one school declined to take part after the initial meeting and eight signed study agreements to take part. The agreements outlined the study role and what was expected of the school in terms of allocating space and staff time for the clubs and also the expectations that the school could have of the project team. In all schools, members of the project team met with the head teacher, provided an overview of the study and answered any questions during a face-to-face meeting. Sixteen schools in North Somerset were invited to participate. Eight schools did not respond to the invitations, four declined to take part after receiving the study information and four signed study agreements to take part. It is notable that the recruitment rate of South Gloucestershire schools was considerably higher than that of North Somerset schools. We believe that this reflects our long-established working relationship with South Gloucestershire Council and support from council staff in approaching schools. Table 5 presents the recruitment rates.

Progression criterion 1 of the study was to recruit at least 25% of the schools approached. Forty-four per cent of all schools that were approached agreed to take part, so this criterion was met. An aim of the recruitment process was to recruit half of the schools in each local authority from above the median free school meals percentage of that local authority. The median free school meals percentage is 8.5% in South Gloucestershire and 9.3% in North Somerset. Six and four schools in South Gloucestershire and North Somerset, respectively, were above this level, indicating that these schools were in more disadvantaged areas.

Participant (pupil) recruitment

Opt-out: pre baseline

All Year 4 and 5 pupils had been asked to take part in the opt-out phase when they were in Years 3 and 4. Across all 12 schools, 1139 pupils were eligible to participate, and 1125 pupils participated (98.77%). Table 6 presents the number of eligible pupils, the number of pupils recruited (those who did not opt out), the number of pupils who opted out (those who did not want to take part) and the recruitment percentage for each school in the opt-out phase. The percentage of pupils recruited was similar when they were asked to wear an accelerometer for 7 days and complete the PAQ-C survey (99.01%) and when they were asked to complete only the survey (98.19%).

Baseline

All eligible Year 4 and 5 pupils were asked to take part in the main study when they were in Years 3 and 4. Across the 12 schools, of the 1139 pupils eligible to take part, 459 pupils expressed an interest in doing so by returning study consent forms (41.39%). Owing to health and safety, a maximum of 30 children per school were selected to participate (the maximum was set at 32 in one school because this was the total number of interested pupils and the school approved the inclusion of two extra pupils rather than turning down two pupils). Of the 459 pupils who returned consent forms, 48.66% were girls. In one school (school 31), only 36.36% of pupils who returned consent forms were girls. Table 7 presents the number of eligible pupils, the number and percentage of pupils who returned consent forms, the number and percentage of girls who returned consent forms and the number of pupils in each school who were selected to take part in the study.

Progression criterion 2 of the study was that at least 25% of eligible pupils express an interest in the study by returning consent forms, which was achieved (41.39%). Progression criterion 3 was that at least 40% of participants expressing an interest were girls, which was also achieved (48.66%).

An aim of the study (objective 2c, which studied an enhanced recruitment approach) was to compare two recruitment methods: briefing against briefing plus taster (see Chapter 2, Study population and recruitment section). A comparison of the two methods is shown in Table 8. The number and percentage of pupils recruited was similar using both recruitment methods. The mean PAQ-C score shows an indication of physical activity level. The score is on a scale of one to five, one being low and five being high. Standard procedures were followed to complete the scoring.30 There was no difference in mean PAQ-C score between the briefing plus taster and the briefing-only recruitment methods.

Another aim of the study (objective 2c) was to compare physical activity levels between pupils who expressed an interest in the study and pupils who did not. In particular, the study aimed to appeal to and recruit less active children. Table 9 compares the mean PAQ-C score and MVPA, assessed objectively with accelerometers, between pupils who did and pupils who did not consent to take part in the study. Minimal differences were seen between those who did and those who did not consent, which indicated that those who consented to the study were unlikely to have higher physical activity levels than those who did not consent. The data in the table also show that the participants were, on the whole, active and that the consented participants engaged in slightly more minutes of MVPA than non-consented participants (66.8 vs. 63.2 minutes of MVPA per day), but this difference was small.

Reasons for non-consent

During the participant recruitment process, brief letters were sent home to parents to collect data on why children did not want to take part in the study. Table 10 presents the number of consent forms that were not returned (i.e. the number of pupils who did not express an interest), and the number and percentage of reasons for non-consent forms returned, for each school. The most common reasons for non-consent were ‘my child does enough activities already’ (38 responses) and ‘my child is not interested in Action 3:30’ (23 responses). Thirteen responses reported reasons related to the letters sent home about the study: parents misreading the information sheet or not receiving it, or not returning the consent form on time.

School status

A total of 334 pupils were selected to take part in the study. Following randomisation, two schools (school 24 and school 28) that had been randomised to the intervention arm decided that they were unable to provide the intervention. For the purposes of this feasibility study, to obtain an indication of the impact of the intervention in those schools able to deliver it, these two schools were dropped from the primary analysis. Therefore, 165 pupils were randomly allocated to the control arm and 170 were allocated to the intervention arm. An interview was conducted with the key contact in school 28 to explore the school’s reasons for withdrawal, and those findings are presented in Chapter 6. School 28 agreed to continue to take part in the study (albeit without delivering the intervention) and therefore participated in follow-up measurements and post-study interviews. School 24 became unresponsive following its decision to withdraw from the intervention. Therefore, an interview could not be scheduled, and the school was withdrawn from the study and any further data collection.

Figure 1 presents the study Consolidated Standards of Reporting Trials (CONSORT)55 flow diagram and details of recruitment and retention throughout the study.

FIGURE 1 The CONSORT flow diagram. Reproduced from Jago et al.54 © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). (Figure not reproduced here; see the original article.)

Data provision

Data provision rates for all measures and accelerometer return are shown in Table 11, and these were similar at baseline and follow-up.

Baseline data

Baseline data are described in Tables 12 and 13. Data are presented as percentages in Table 12 and as means and SDs in Table 13. Overall, the trial arm were well balanced. Parent ethnicity varied slightly between the arms, with the control arm having a higher proportion of white British parents. Accelerometer-assessed MVPA and sedentary time were similar across the arms, as were psychosocial variables.

Table 14 presents a comparison of accelerometer-assessed MVPA at baseline between the trial arms for all participants and for boys and girls separately. Data are presented as means and SDs and percentages. Mean MVPA was similar between the trial arms for both boys and girls. In both arms, mean MVPA was lower for girls than for boys. In the control arm, girls did 10.89 minutes less MVPA than boys, and in the intervention arm, girls did 12.35 minutes less MVPA than boys. The total proportion of both boys and girls meeting 60 minutes of MVPA per weekday was similar between the arms. In both arms, approximately 19% fewer girls than boys met the 60-minute guideline.

Re-enrolment

Considering previous research, the study was designed to re-recruit pupils approximately half-way through the intervention if spaces became available. The number of spaces available was calculated by determining the average number of pupils attending each club on a weekly basis (from club attendance registers provided by the school) and subtracting that from the maximum capacity (n = 30 pupils). Table 15 presents the number of spaces available in each school and the number of pupils who expressed an interest in participating.

Teaching assistant attendance

Teaching assistants from all four intervention schools either were nominated by their school or volunteered to attend training and run the intervention in their schools. Two TAs from schools 25, 27 and 33 were recruited. Three TAs from school 32 volunteered. Eight TAs attended all 5 days of training and one TA from school 32 missed 1 day of the training.

Teaching assistant self-efficacy

Figures 2 and 3 show the results of the self-efficacy and perceived autonomy-support questionnaires completed by TAs before and after training and at three time points during the intervention. Measures of personal endorsement and intention to use an autonomy-supportive teaching style were completed pre and post training only (see Figure 3). Results from the teaching efficacy measure (see Figure 2) suggest that TAs became more confident across all teaching environment contexts in their ability to provide highly active after-school sessions under challenging circumstances following training and grew slightly more confident across all teaching contexts over the course of the intervention (process evaluation visit 3).

FIGURE 2 Self-perceived TA teaching efficacy. PEV, process evaluation visit. Reproduced from Tibbitts et al.56 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. (Figure not reproduced here; see the original article.)

FIGURE 3 Teaching assistant-perceived autonomy-supportive teaching style. a, Adapted version of the Sport Climate questionnaire,45 scale 1–7; b, Endorsement questionnaire,47 measured only pre and post training, scale 1–7. PEV, process evaluation visit. Reproduced from Tibbitts et al.56 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. (Figure not reproduced here; see the original article.)

The results of the personal endorsement item (see Figure 3) showed a small increase in mean score between pre (6.0 ± 0.31) and post (6.4 ± 0.57) training. Similarly, there was a small mean score increase in the intention to use item (see Figure 3) pre (6.7 ± 0.45) and post (6.8 ± 0.28) training. Results from the autonomy support measure (see Figure 3) show that TAs perceived that they adopted a more autonomy-supportive teaching style post training than pre training, and that they perceived this to have been maintained throughout the intervention. For example, TAs agreed that they provided their pupils with options and tried to understand their pupils’ perspectives before suggesting alternative ways of doing things. These results suggest that intervention delivery was consistent with underpinning SDT (objective 3c). Observational visits to schools at three time points during the intervention also affirmed this, as two project staff observed and agreed that an autonomy-supportive teaching style had been adopted by TAs in all four schools.

Session delivery (objective 3a)

Teaching assistants in all four intervention schools completed the log books (see Appendix 3), which were collected to assess intervention fidelity. As shown in Table 16, all 30 sessions were delivered in all four intervention schools. TAs described most of their sessions as partially delivered (69.2%), one-quarter of sessions as fully delivered, and a small number of sessions as not delivered at all (5.8%). Comments in log books and interviews with TAs (see Chapter 6) revealed that TAs felt that sessions were partially delivered because there was not enough time to complete all of the activities in the session plan. Furthermore, comments and TA interviews revealed that the sessions classified as not at all delivered were those where space had been limited and TAs chose to revisit previous session plans, which were more easily adaptable to the space, rather than deliver the current session plan.

Child-reported autonomy support and self-perceived exertion and enjoyment

Observations made by the project staff during three time points in each school aligned with results from the log books. All 12 sessions observed were thought to be either fully or partially delivered, the reason for partial delivery being the lack of time to fit in all of the activities. Overall, the intervention fidelity was good, and TAs used the session plans as intended.

Table 17 presents child-reported autonomy support scores at three process evaluation time points during the intervention, with scores derived from the short form of the Sport Climate Questionnaire.45 Results are presented after multiple imputation and did not differ from the results produced before multiple imputation (see Appendix 4). The results show that children perceived their Action 3:30 leaders (the TAs) to be autonomy-supportive (as indicated by a high score) throughout the intervention, with little difference between time points. This may suggest that TAs were able to adopt an autonomy-supportive teaching style from the start of the intervention and maintain this throughout. This aligns well with the results shown in Figure 3, as TAs also perceived their teaching style to be autonomy-supportive throughout the intervention. Observational visits also support this finding, as project staff agreed that TAs successfully implemented the autonomy-supportive teaching techniques that had been taught during training. Overall, these findings suggest that the intervention was effective at underpinning SDT and that intervention fidelity was good.

Figure 4 presents child-reported mean perceived exertion and enjoyment at three time points during the intervention. It should be noted that the session number varied between schools at each process evaluation visit (process evaluation visit time point) and therefore the mean results reflect a variety of sessions and activities. The results of the self-perceived exertion measure suggest that children found the sessions towards the end of the intervention less tiring than those at the start. It is important to note that findings from the focus groups with a selection of pupils in each school suggested that they did perceive sessions to be active and tiring throughout the intervention period (see Chapter 6).

FIGURE 4 Child self-perceived exertion and enjoyment. a, Score from 1 to 5; b, score from 1 to 10. PEV, process evaluation visit. (Figure not reproduced here; see the original article.)

The results of the enjoyment measure show that children did enjoy all of the sessions observed, with little variation between time points (as indicated by mean scores of > 4). These findings suggest that the intervention was effective at promoting enjoyment and are reflected in the focus groups with pupils and interviews with TAs. During observational visits, project staff also agreed that there was minimal disruptive behaviour.

Session attendance (objective 3b)

Table 18 presents the number, mean and percentage of intervention (club) attendance in each school and overall. N represents the total number of pupils who were enrolled in the study and the club, including re-enrolment pupils. The results are presented for all pupils enrolled into the study (including re-enrolment pupils and those who did not attend a single session) and for just those pupils who participated in at least one session, including re-enrolment pupils. Mean attendance refers to the number of sessions attended by pupils (not the number of pupils). Mean attendance is presented as total pupils and for boys and girls separately. The number and percentage of pupils attending > 50% of sessions are also presented overall and by sex.

Overall, the results suggest that attendance in all four schools was good. Table 18 shows that among those who took part in at least one session, a mean of 19 sessions out of 30 were attended in total, with school 32 having the highest mean attendance (21 sessions). Mean attendance was similar between boys and girls and between schools, which suggests that the intervention was successful in engaging all pupils. To meet progression criterion 4, > 50% of pupils were required to attend at least 50% of sessions. This was achieved when analysing data from all pupils in the intervention arm and only those who took part in at least one club session. Across all schools, an average of 73.5% of pupils attended > 50% of sessions, with school 25 attaining the highest percentage (88.2%). The results suggest that the reach of the intervention was good for those who enrolled in the study and started the club.

Primary outcomes (accelerometer-derived weekday moderate to vigorous physical activity)

Table 19 presents the results comparing accelerometer-derived weekday MVPA between the control arm and the intervention arm at follow-up for the overall sample and by sex. MVPA at follow-up was very similar between the arms and, when adjusted for local authority, recruitment strategy and baseline MVPA, the lack of difference between the arms was reflected in the adjusted difference and 95% CI. Among boys there was a slightly higher MVPA in the intervention arm than in the control arm; however, the adjusted difference and CI shows no notable difference. Among the group of girls, MVPA did not appear to differ between the arms and this was also reflected in the adjusted difference and 95% CI. None of these differences and 95% CIs met the 10-minute benefit required for progression criterion 5.

Table 20 presents the results comparing the proportion of pupils meeting current daily MVPA guidelines at follow-up for the overall sample and for boys and girls separately across weekdays and weekend days. The proportion of pupils meeting the 60 minutes of MVPA per day did not appear to differ between the intervention arm and the control arm overall and among the boys and girls separately. This is reflected in the adjusted odds ratios and the 95% CIs. In both arms, the proportion of boys who met the guidelines was generally higher than the proportion of girls who met the guidelines.

Complier-average causal effect analysis

A CACE analysis was conducted to explore the treatment effect of those participants who joined at the second enrolment. This analysis included data from children joining at the first or second enrolments in an unbiased evaluation of the intervention effect (primary analysis of children indicating a willingness to join the intervention at the start). By way of context, for the ITT analysis (primary analysis) there were 10 schools: six in the control arm (number of pupils: 27, 21, 23, 25, 19 and 24) and four in the intervention arm (number of pupils: 17, 32, 29 and 35). These pupils are those who said that they were willing to join the intervention at the start of the study and were followed up for the ITT analysis. For the CACE analysis, three schools had complete accelerometer data for the entire year group at baseline and at follow-up, with a further school withdrawing from the study. Of the three schools, two were in the control arm (n = 53 and n = 24 pupils) and one was in the intervention arm (n = 27 pupils). These pupils did not say that they were willing to join the intervention at the start of the study and therefore could have enrolled had they had the chance at the point of second enrolment. In the one intervention school, one additional pupil joined at the re-enrolment point and provided useful data. We therefore examined, using a CACE methodology, whether or not there was a difference between the one pupil who re-enrolled from the intervention arm and those in the control arm who would have enrolled at the second enrolment had they had the chance, excluding pupils who enrolled in the first phase.

The unbiased CACE estimate of the intervention effect in children joining at the second opportunity is presented in Table 21; this is very imprecise as a result of having data from only one child who joined the intervention at this stage. The meta-analysis of the ITT analysis of children joining Action 3:30 at the first opportunity, and the CACE analysis of children joining Action 3:30 at the second opportunity, is also presented in Table 21 as a demonstration of the analysis we had planned were we to proceed to a full trial evaluation of the intervention.

Secondary outcomes

Accelerometer-derived secondary outcomes

Table 22 presents the accelerometer-derived secondary outcomes at follow-up. The overall mean minutes of MVPA were similar between the intervention arm and the control arm, and, when adjusted for local authority, recruitment strategy and baseline outcome, the lack of difference between the arms remained. A similar result can be seen for mean weekend minutes of MVPA, mean weekday c.p.m., mean weekend c.p.m., mean weekday sedentary minutes, mean weekend sedentary minutes and overall sedentary minutes.

Questionnaire- and body mass-derived secondary outcomes

Table 23 presents the secondary outcomes derived from the questionnaire, as well as the measured BMI. No psychosocial outcomes show any notable difference between the control arm and the intervention arm once adjusted for local authority, recruitment strategy and baseline MVPA. Age- and sex-adjusted z-score was slightly lower in the intervention arm than in the control arm at follow-up. The number of active travel days from school and the number of after-school clubs attended (excluding Action 3:30) were slightly smaller in the intervention arm than in the control arm (1.94 vs. 2.35 active travel days and 1.44 vs. 1.70 number of after-school clubs attended respectively).

Adverse events

No adverse events were reported during the intervention delivery.

Sample size

Data from the feasibility study were used, along with those from other studies that detailed an ICC of the same outcome in school-based studies,57 to inform the sample size of a definitive trial to detect a 10-minute difference in weekday MVPA.

The ICC used in the top half of Table 24 was taken from the observed data in this feasibility study. An average cluster size of 27 and an ICC of 0.01 were used in the sample size calculation. Differing levels of alpha and power were investigated. A 20% loss to follow-up rate was conservatively estimated, and the number required per arm was inflated by 20% prior to further inflation by cluster size. The total number of schools needed varies from 8 (4 schools in each arm) to 16 (eight schools in each arm), depending on the power and significance level. The ICC used in the bottom half of Table 24 was taken from the primary outcome of the Bristol Girls Dance Project,57 which was a large cluster randomised trial of an after-school physical activity (dance) programme that focused on weekday MVPA among 11- to 12-year-old girls. The ICC from this study was chosen as an additional measure as this study is the largest UK after-school physical activity study for which an ICC has been reported and it therefore provides the best alternative data available. An average cluster size of 27 and an ICC of 0.001 were used in the sample size calculation. Differing levels of alpha and power were investigated. A 20% loss to follow-up rate was conservatively estimated, and the number required per arm was inflated by 20%. The total number of schools needed varies from 20 (10 schools in each arm) to 38 (19 schools in each arm), depending on the power and significance level.

Data-driven exploratory analysis

After the study findings were discussed with the independent TSC, additional exploratory analyses were conducted to assess whether or not there was any evidence of a potential effect of the intervention on participants who were less active at baseline.

Least active pupils

Analyses to compare weekday MVPA between the control arm and the intervention arm were repeated including only those pupils categorised as least active. Least active pupils are defined as those in the lowest third of the sample in this study in terms of average weekday MVPA. All children who have < 53.44 minutes of weekday MVPA fall into the lowest third. Similar to the primary analyses, Table 25 shows that there was no difference in weekday MVPA between the arms overall or for boys and girls separately. This was also observed for all other secondary accelerometer-derived variables (data not shown).

School context

The context in which an intervention takes place could influence that intervention’s delivery, implementation or functioning58 and, as such, is important to try to understand. Various data were collected during the study to establish school context and its effect on the implementation of the intervention. Details of school size and pupil premium were provided by the school contact or school reception. A school context form enabled members of the study team to collect information on school-level physical activity provisions and school policies on promoting physical activity to staff and students. The components of the form that researchers used were taken from Jones et al.’s48 audit tool, which was created to measure the physical school environment’s suitability for physical activity, and Lounsbery et al.’s59 tool, which was devised to assess school policies regarding physical activity and physical activity throughout the curriculum. One form was completed per school. Each school was given a school physical activity suitability score, which is a sum of the scores from the cycling, walking, and sports and play provision components (Table 26). Scores were adapted from Jones et al.’s48 original audit tool; facility quality was assessed on a more descriptive scale of 1–5 to improve measurement accuracy. Where a facility (e.g. a marked pedestrian crossing) was marked as present, scores were weighted as yes = 1 and no = 0. Where the quantity of facilities was provided, scores were weighted relative to the mean number of facilities across all schools: 0 = none are recorded, 1 = the number is between one and the mean plus one SD, 2 = the number is greater than the mean plus one SD.48 The modal quality of physical activity provision is also reported, by school, in Table 26. Individual school responses measuring school physical activity policies and physical activity throughout the curriculum are reported descriptively in Appendices 6 and 7, but are described in full below. Each school received a ‘policy’ score (maximum score of 13) and ‘curriculum’ score (maximum score of 8). Please note that data are presented only for the six control schools and the four intervention schools that delivered the intervention.

School physical activity provision

As displayed in Table 26, the between-school variation in physical activity provision scores was high across all schools measured. Variation was highest in the cycling provision scale (range 5–48) and in the sports and play provision scale (range 27–81). Intervention and control school clusters did not differ in average scores for cycling provision, walking provision, facility provision, aesthetics or design of school grounds. Schools in the intervention arm had a slightly higher average score for sports and play provision than those in the control arm and, therefore, a higher overall physical activity suitability score. The average quality of facility provision was high in both intervention and control schools, with little variation among individual item quality scores. We may therefore suggest that intervention schools had, on average, an outdoor environment slightly more supportive of active games according to the validation of the original audit tool.48

School policy on physical education provision and physical activity

Appendix 6 presents the scores for the school policy component of the school context audit tool. All schools reported having written or unwritten policies that supported physical activity for their student cohort in some way. Most schools (100% of control schools and 50% of intervention schools) reported having a policy that required PE to follow specific standards. In two control schools, this policy was not formally written/recorded. Two intervention schools had no such policy. Similarly, most schools (control arm, 83%; intervention arm, 75%) had a policy requiring a specific number of minutes be devoted to PE in the curriculum time each week. Half of control and intervention schools also had a written policy encouraging students to walk or cycle to school, and 100% of schools provided cycle training. By contrast, few schools reported consistently supporting staff physical activity through school policies. Only one school (in the intervention arm) reported having a policy supporting active travel for their staff, such as a cycle scheme. Similarly, only two control schools (33.3%) and one intervention school (25%) reported having a policy requiring that all school personnel receive professional development related to physical activity promotion. However, two-thirds of control schools and 100% of intervention schools provided facilities such as showers and cycle parking, which enable staff to travel by active means should they wish. Schools varied greatly in their allocation of budget to PE and PE equipment. All but one control school set aside a budget specifically for PE; however, the level of accountability for that money in different schools appears to differ randomly rather than be in favour of either the control or the intervention arm. PE teachers had some involvement in the spending of the PE budget in all schools; this did not differ by arm. On average, the control and intervention school arms did not differ in school policy scores. It was evident that policies focused mainly on supporting child physical activity, and fewer schools had policies beyond cycle parking and showers that supported or encouraged staff physical activity.

Physical activity in the curriculum

Schools reported a wide variety of other Key Stage 3 subjects that used PE in lessons, including primarily the core subjects. Outdoor learning has become something of an aspiration for all schools across their curricula in recent years, which may explain the high proportion of positive answers to this item on the audit tool. Many schools reported providing regular physical activity breaks during the day beyond PE and usual morning/afternoon breaks, although two of the six control schools did not. Similarly, except in two control schools, teachers were encouraged to promote physical activity to pupils. Pupils were not held back from taking part in curriculum PE in favour of core subjects; however, two control schools reported that sometimes core subjects were prioritised over PE when competing for access to space or facilities. Overall, there was no difference in the mean score between control and intervention schools in terms of how school policy served to promote physical activity throughout the wider curriculum. A full report of the responses to the physical activity in the curriculum section of the questionnaire for each school is presented in Appendix 7.

Table 27 presents a summary of the results presented against the progression criteria and whether or not each criterion was met.