Study design

The Health and Activity of Pupils in the Primary Years (HAPPY) study is a quasi-experimental observational study comprising two components: a quantitative school survey and a cross-sectional study. This paper presents the results of the quantitative survey using the STROBE reporting guidelines (S1 File). Details of the survey development and administration are provided below, while full methods and procedures for the broader HAPPY study, including the cross-sectional component, are available in the published study protocol [20]. Our quantitative survey targeted all state-funded primary schools in Greater London and reported on each school’s implementation of PA policies and practices in line with the WHO’s recommendations for a whole-school approach to PA, specifically assessing schools’ practices against each of the WHO’s six specified domains.

Survey development

Our survey was developed in consultation with three primary school teachers, recruited through the NIHR Clinical Research Network, not part of the study sample, and the survey included questions adapted from previous studies [21–24]. The teachers were invited to pilot the survey through a Qualtrics link and were given one week to provide feedback. A response box under each item allowed them to comment on the ease of understanding, language, response options, overall flow, and completion time, as well as to note anything the survey did not capture but should be included. Teachers suggested minor changes, which we incorporated by clarifying one item and adding an ‘other’ option to another, and they indicated that the survey was appropriate for teacher and staff respondents.

Feedback was shared with the study team, and the final 20-item version of the survey (S2 File) was agreed upon by all authors and consulted teachers. Questions were designed with either single- or multiple-response formats, with the exception of the last question, an open-ended text box inviting optional free-text comments. The survey was made available in two formats, an online version (via Qualtrics, accessed by link or QR code) and a paper-based version. Eleven of the 20 survey questions were mapped to the WHO domains: three questions each were included under Domains 1 (QPE) and 4 (PA during scheduled breaks); two questions under Domain 5 (active classrooms); and one question each under Domains 2 (active travel), 3 (PA outside of school hours), and 6 (inclusive PA) (S2 Table).

Example questions included: “Does your school have any of the following in place for facilitating physical activity?” (items relating to Domains 1-5); “When does your school (or any other organisation) provide any extracurricular physical activity or sports programmes for children at your school?” (Domains 3 and 4); and “Is there sufficient provision (staff/equipment) to support participation for children with specific disabilities or impairments in physical activities?” (Domain 6). The remaining survey questions collected basic school information and broader school characteristics related to PA provision. In this paper, we report findings from the WHO domain-specific survey items only.

Procedure

On 6 November 2023, all state-funded primary schools in Greater London (1826 at the time of this study) were sent a paper version of the survey by post, addressed to the headteacher, and included a prepaid return envelope to return the survey. For consistency, no questions were mandatory in either format. Any staff member best placed to answer the questions could complete the survey, and respondents were asked to indicate their role within the school. School contact details (headteacher name, postal address, and email address) were obtained from publicly available UK government websites [25]. To increase response rates, schools were contacted by email at three time points (January, April, and July 2024) to remind them to complete the survey. The fourth and final reminder, by post, was sent in September 2024. The survey closed on 31 December 2024.

Statistical analysis

We compared the characteristics of responding and non-responding schools (school type, size (pupil numbers), gender, ethnic composition (based on the dominant ethnic group per school), Ofsted rating, number of pupils eligible for free school meals, number of pupils with English as an additional language, number of pupils with special educational needs and disabilities (SEND)) obtained from UK government websites [25–27]. We also compared area-level deprivation measured using borough-level quintiles of the Income Deprivation Affecting Children Index (IDACI), derived from average scores for each borough [28]. All data were categorical. Chi-squared or Fisher’s exact statistical tests were used to assess differences in characteristics between responding and non-responding schools. We report the percentage of schools that provided positive responses to each survey item included in each WHO domain. Positive responses indicated that specific practices, initiatives, or facilities were in place at the school. Responses where items were not ticked were treated as valid responses, indicating that the practice was not in place. Missing or unclassified responses were excluded from percentage calculations. All analyses were conducted using R software, version 4.4.2 [29].

Ethics

Ethical approval was granted by Imperial College London’s Research Ethics Committee (reference: 6800895). Each school received the survey with an introductory page outlining the purpose of the study, that participation was voluntary, and assurances of confidentiality. Completion of the survey was not compulsory for the school. Formal written consent was not required to use data; return of a completed survey was considered implied consent for the data to be used for analysis. All data were anonymised for analysis (i.e., school names, locations, and teacher names were removed).

Response rates

We received 201 responses from 1826 primary schools across all 32 Greater London boroughs, yielding an 11.0% response rate. Of these, 18 surveys were excluded from the analysis: two were from unidentifiable schools, eleven were duplicate surveys from the same school (with the most recent and complete response retained), and five lacked sufficient completion (online responses ended prematurely, typically less than 50% complete). The final sample included 185 surveys, corresponding to 10.1% of the total schools contacted. We received 115 (62.2%) survey returns by post and 70 (37.8%) through the online link. The surveys were completed by 85 (45.9%) class teachers, 74 (40.0%) headteachers or deputy headteachers, 7 (3.8%) school administrators, and 19 (10.3%) other school staff (e.g., PE/Sports Leads).

School characteristics

Schools that completed the survey were similar in all characteristics to those that did not with the exception of the number of pupils eligible for free school meals (Table 1). Schools with more than 100 pupils eligible for free school meals were significantly underrepresented among responding schools compared to non-responding schools (X² = 5.34, p = .021).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of Greater London schools that completed the survey compared with schools that did not.

https://doi.org/10.1371/journal.pone.0352283.t001

Implementation of the WHO domains

A total of 76.0% (133/175; 10 missing responses) of schools reported having a PA policy in place, but the degree of reported implementation differed across the WHO domains (Table 2). In Domain 1 (QPE), nearly all schools (98.4%) reported delivering curricular PE, and over two-thirds (76.2%) reported access to an internal or external specialist PE teacher. However, just over half of schools (53.5%) met the curricular guidelines of providing an average of at least two hours of PE per week across year groups. In Domain 2 (active travel), 78.9% reported having an active travel plan, and 77.3% had a separate entrance for pedestrians or cyclists to encourage safe travel by foot. However, less than a quarter of schools (24.3%) promoted initiatives to support walking part of the school journey for children who travel by vehicle (‘Park and Stride’). For Domain 3 (PA opportunities outside of school hours), almost all schools (98.4%) provided PA opportunities after school, while PA opportunities before school and during weekends or holidays were less commonly reported (48.6% and 28.4% respectively). In Domain 4 (PA opportunities during scheduled breaks), 92.4% of schools reported providing PA during breaks and 82.0% during lunchtimes. Schools also reported having access to playgrounds (99.5%), portable play equipment (97.8%), access to permanent play structures (89.6%), and hardcourts (72.0%). In contrast, access to playing fields was considerably lower (48.1%). In Domain 5 (active classrooms), 81.7% of schools reported integrating PA into classroom lessons, and over half (55.1%) reported implementing active mile initiatives [30]. Lastly, in Domain 6 (inclusive PA), sufficient provision to support children with SEND was reported by 81.5% of schools. Overall, 68.5% of schools (122/158; excluding seven with missing data for one or more domain items) reported implementing all six WHO domains (at least one item under each domain).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Number of schools reporting implementation of the WHO domains per corresponding survey item.

https://doi.org/10.1371/journal.pone.0352283.t002

Optional free-text comments

Just under half of responding schools provided additional free-text comments through an open-ended text box at the end of the survey, beyond what was captured by the survey questions. These included challenges in supporting physically active school environments and a lack of dedicated funding and tight budgets, which limited some schools’ ability to purchase specialised or permanent equipment, improve outdoor spaces, and subsidise extracurricular opportunities. Several schools commented that changes to the PE and Sport Premium funding criteria (a UK government fund introduced in 2013 to improve the quality of PE in primary schools) [31] made it difficult to use funds flexibly to meet their specific needs. They also highlighted time-related constraints in delivering PE in a crowded curriculum and a lack of staff capacity to support off-site activities. Some school staff expressed concern about the sustainability of quality PA provision due to over-reliance on a small number of unspecialised staff.

In contrast, many schools identified key enablers that supported an active school environment. These included strong leadership support, peer collaboration among staff, and targeted efforts to increase PA participation among some groups. For example, girls-only sports clubs, free clubs for pupils eligible for the pupil premium, and inclusive competitions adapted for children with SEND. Expanding the range of sporting opportunities (e.g., access to swimming facilities and inclusive inter-school competitions) through external partnerships with local sports networks, charities, and community providers was welcomed by several schools. Some schools also described themed health weeks and event days to promote healthy lifestyles.

Strengths and limitations

A key strength of our study is that it is the first, to our knowledge, to examine schools’ physical activity (PA) policies and practices assessed against global recommendations for promoting a whole-school approach to PA. Another major strength is the use of a piloted survey instrument informed by the WHO’s whole-school PA promotion framework, enabling structured data collection across all six recommended domains. Unlike national surveys, which are the main sources of data assessing children’s engagement with and attitudes to PA [6], our approach captured the presence and variation of institutional-level PA policies, infrastructure, and practices aligned with the WHO framework. The distribution of our survey to schools in an ethnically and socioeconomically diverse urban conurbation is another important strength of our study. Our findings may be representative of other major, diverse conurbations, as we received responses from at least one school in every Greater London borough with varying levels of deprivation.

Our study’s strengths should be considered alongside its limitations. Our response rate may be considered low; however, challenges with obtaining survey responses from schools are widely recognised [48]. Input from both classroom and leadership staff captured diverse perspectives, but role-specific knowledge gaps (e.g., policy-level decisions) may have introduced inconsistencies or misreporting, potentially affecting the reliability of responses across schools. A significant difference between responding and non-responding schools in the proportion of children eligible for free school meals, which serves as a proxy for socioeconomic status, may limit the generalisability of our results. Schools with fewer disadvantaged pupils typically have greater resources and capacity to implement PA-supportive policies, which could inflate our estimates of implementation relative to the wider population. Additionally, the number of questions included under each domain varied, with three domains assessed using only a single question, while others included two or three questions. Domains with fewer questions may have limited the depth and rigour of evaluation.

Future surveys should target respondents with role-specific expertise or adopt consensus-based responses using a more extensive survey. They could also refine and expand domain-specific indicators by incorporating rating scales or quality assessments. A richer dataset could be obtained by integrating qualitative methods and observational approaches alongside survey responses, thereby providing clearer and more actionable criteria for policymakers and school authorities to monitor progress and support efforts to overcome persistent barriers, such as limited curricular PE time and inadequate access to facilities. Our findings further underscore the need to identify these barriers among more deprived schools, which were less likely to respond and may face distinct challenges in implementing whole-school PA approaches. Nonetheless, the study offers valuable insights for public health practitioners, education policymakers, and in-school providers seeking to embed equitable and sustainable approaches to promoting PA in primary schools.