Research and analysis

Improving physical activity and reducing physical activity inequalities: literature review

Published 12 June 2026

Report authors:

• Jovita Leung
• Larissa Stolar
• Benji Horwell
• Filippo Bianchi

Acknowledgements:

We would like to acknowledge the contributions of the DCMS Sport Policy and Analysis Teams, as well as those who took part in the scoping workshops at the early stage of the project from DHSC and Sport England.

Disclaimer:

The conclusions and recommendations presented in this report are those of the authors and do not necessarily reflect the views or policies of the Department for Culture, Media and Sport (DCMS).

Executive summary

Introduction

This rapid evidence review synthesises the latest meta-analyses and systematic reviews on the effectiveness of three priority interventions to promote physical activity:

• School-based: interventions delivered in school settings to increase students’ physical activity.
• Community-based and peer-led: interventions delivered in community settings targeting communities and/or delivered by peers to promote physical activity.
• Active travel: interventions that promote walking and cycling for transport among the general population.

These intervention categories were selected by DCMS as priorities for this evidence review based on the impact on healthy life expectancy, implementation feasibility, and the value-add for DCMS.

While the review looks primarily at interventions to increase physical activity, it also reviews evidence for tackling inactivity by specifically considering the equity impact of interventions i.e. what impact do interventions have on individuals who are less active or less likely to be active by virtue of their demographics.

Aims and research questions

BIT conducted a ‘what works’ evidence review to summarise the evidence on the following research questions:

1. What evidence exists regarding each intervention’s (or policy’s) effectiveness in improving physical activity outcomes?

2. What evidence exists regarding their likely impact on reducing physical activity inequalities?

3. What is the estimated cost of implementation (low, moderate, or high)?

This review was not designed to be a comprehensive or exhaustive search; instead, it aims to provide a rapid synthesis of publicly available evidence to support decision-making within the current policy context.

Findings

School-based interventions

School-based interventions generally show small or no significant effects on increasing physical activity, though physically active lessons (PALs) and classroom-based physically active recess showed more promise (see full definition in ‘School-based interventions’ in the main report).

Studies also highlight potential factors that limit intervention effectiveness, including reductions in non-school-based physical activity to compensate for increased school-time activity, limited teacher engagement and lack of context-specific intervention adaptation.

School-based interventions do not consistently close gaps by socioeconomic status or gender, but there is some evidence they may benefit the least active and overweight children, and targeting lower-SES schools could help reduce population-level disparities over time.

Community-based and peer-led interventions

Aside from peer-led interventions, there is a lack of high quality evidence about the impact of community-based interventions on physical activity. We identified two high quality evaluations of community-based interventions in London and Devon, both of which showed no significant impact on physical activity at the neighbourhood-level. However, both interventions were hampered by low participation rates which may mask individual-level impacts.

Peer-led interventions showed more promising results, with high-quality reviews consistently demonstrating small to moderate improvements in physical activity, particularly when peer leaders were trained and interventions begin with an intensive phase.

Evidence on reducing inequalities is limited; neither of the two community-based evaluations reported impacts on specific subgroups, and peer-led interventions lack robust evidence on impacts across priority subgroups. Some evidence suggests group-based delivery formats and higher intervention intensity may improve outcomes for socioeconomically disadvantaged women and young children.

Active travel interventions

Changes to the built environment may lead to increases in active travel and physical activity, especially when changes enhance the accessibility and perceived safety of walking and cycling. There was only low quality evidence on active school transport initiatives, and this research did not find an effect for these initiatives.

Equity impacts were unclear though some evidence suggests that interventions may disproportionately benefit more advantaged groups, underlining the importance of targeting lower-income and under-resourced communities.

Across the studies we reviewed covering school, community and active travel interventions, costs were only reported in one study evaluating changes to the built environment, highlighting a clear gap in the literature on physical activity interventions.

Recommendations

We make the following overarching recommendations to support DCMS’ work on the Health Mission:

1. Strengthen the evidence base through targeted, robust evaluations- given the range of active programmes already operating across schools, communities, and urban environments, there is a valuable opportunity to embed evaluation into existing delivery structures and ensure future investment is evidence-led.

2. Prioritise delivery models with the potential for population-level impact- Interventions such as peer-led programmes, teacher-delivered PALs and improvements to cycling and walking infrastructure demonstrate potential for improving physical activity.

In respect of specific intervention settings, we recommend:

School-based interventions

1. Developing evidence-based guidance for schools and trainee teachers regarding how to improve pupils’ physical activity

2. Prioritising pilot schools in areas of lower activity

3. Ensuring future evaluations include long-term follow-up and assess compensatory behaviours

Community-based interventions

1. Scaling up peer-led physical activity programmes

2. Prioritising group-based delivery models and peer-led and community programmes

3. Investing in evaluation of promising UK community-based initiatives such as Parkrun, Go! London and Sport England’s Place Partnerships

Active travel interventions

1. Targeting infrastructure changes in areas of high deprivation

2. Ensuring interventions address safety and convenience to encourage uptake

Introduction

One of the three ‘big shifts’ for the UK government’s health mission is to move the emphasis of the UK’s health system from ‘sickness to prevention’ and address the underlying causes of ill-health.

To aid this work, DCMS commissioned BIT to help identify and review existing evidence for promising physical activity interventions and policies that could help to prevent ill-health.

BIT created an initial list of 28 interventions and policies based on interventions identified in existing reports (e.g. ISPAH’s ‘Eight Investments’), examples of physical activity initiatives from around the world, and idea generation grounded in the behaviour change wheel.

Considering factors such as the impact on healthy life expectancy, implementation feasibility, and the value-add for DCMS, BIT then worked with DCMS to identify four priority interventions/policies for a series of rapid evidence reviews.

This included:

1. School-based interventions/policies

2. Community-based and peer-led interventions/policies

3. Active travel interventions/policies; and

4. Interventions involving the co-location of healthcare services and leisure facilities (co-location interventions)

Focus for this scoping review

The focus of the current review is to understand the evidence regarding the impact of school-based, community-based/peer-led, and active travel interventions on physical activity.

This review, along with another rapid evidence review ‘A scoping review on co-location of health and leisure services and their impact on physical activity: a rapid scoping review for DCMS’ explores the impact of a range of interventions on physical activity. Findings from these two reviews will be used to highlight priority areas in which the government could intervene to improve health outcomes through physical activity. As evidence for co-location interventions is more limited compared to school, community and active travel interventions, the co-location review was conducted as a separate scoping review with different research questions.

Achieving population-level change in physical activity requires systemic approaches, with a range of complementary interventions operating across multiple levels. This review is intended to inform where attention and resources might be prioritised to improve health outcomes through physical activity, both in the short and long term.

Aims and research questions

The review is guided by three research questions:

1. Primary research question: What evidence exists regarding each intervention’s (or policy’s) effectiveness in improving physical activity outcomes?

2. Secondary research question: What evidence exists regarding their likely impact on reducing physical activity inequalities?^{[footnote 1]}

3. Exploratory research question: What is the estimated cost of implementation (low, moderate, or high)?

These questions are addressed through a review of up to five meta-analyses or systematic reviews for each intervention type, with a focus on evidence quality, real-world applicability, and relevance to DCMS policy priorities.

This review was not designed to be a comprehensive or exhaustive search; instead, it aims to provide a rapid synthesis of meta-analytic and systematic review evidence to support decision-making within the current policy context.

Methods

We conducted a rapid evidence review to summarise insights from relevant systematic reviews and meta-analyses.

We searched relevant databases using pre-agreed search strategies. Studies were identified using academic databases, web searches, evidence provided by the DCMS, and AI tools.

We reviewed the evidence based on a title and abstract screening. A full-text review was then conducted to compile a final list of relevant evidence to be included in the review. The quality of studies was also assessed using an adapted version of the UK Royal College of General Practitioners (RCGP) framework (see Figure 1). This framework was chosen due to its relevance to meta-analytic and systematic review evidence.

We summarised the findings narratively. Data from the included studies were recorded in a spreadsheet and were summarised narratively.

Full details of our methodology can be found in Appendix 1.

Figure 1. An adapted version of the RCGP’s evidence framework (Joyce et al, 2016)

Evidence	Definition
Strong evidence	Consistent findings provided by a body of multiple high quality scientific studies, or provided by high or moderate quality systematic reviews/ meta-analyses demonstrating consistent results from multiple well-designed studies/ RCTs.
Moderate evidence	Generally consistent findings in fewer, smaller scale or lower quality scientific studies; or provided by high or moderate quality systematic review/meta-analyses demonstrating generally consistent evidence from well-designed studies.
Limited or contradictory evidence	Limited evidence provided by one scientific study or inconsistent findings in multiple scientific studies; high quality evidence lacking.
Research evidence unknown	Inconclusive research evidence at present, but some theoretical support.

What works to promote physical activity

This section summarises the key findings from the rapid literature reviews (see Table 1 below). Findings are presented based on intervention types (school, community-based and peer-led, active travel).

Table 1: Summary of findings

Intervention type/setting	Specific intervention/policy idea	Target population	Quality of evidence	Impact on physical activity outcomes[footnote 2]	Impact on physical activity outcomes	Impact on inequalities	Estimated intervention cost[footnote 3]
School-based	Physically Active Lessons (PALs)	Children and young people	Strong evidence	Small effect (daily minutes of moderate to vigorous physical activity, MVPA)	Large effect (class-time MVPA)	Mixed (some evidence of better results for inactive and overweight/obese children, but also some evidence of less impact for girls regarding after-school programmes)	Low (low-cost training and resources)
School-based	Class-based active recess	Children and young people	Moderate evidence	Small effect	Small effect	Mixed (some evidence of better results for inactive and overweight/obese children, but also some evidence of less impact for girls regarding after-school programmes)	Low (low-cost training and resources)
School-based	Enhanced Physical Education (PE)	Children and young people	Moderate evidence	No significant effect	No significant effect	Mixed (some evidence of better results for inactive and overweight/obese children, but also some evidence of less impact for girls regarding after-school programmes)	Low-to-medium (more PE staff may be required if extending PE time in the curriculum)
School-based	After-school programmes	Children and young people	Limited or contradictory evidence	Negligible effect (some small positive impacts found, not consistent)	Negligible effect (some small positive impacts found, not consistent)	Mixed (some evidence of better results for inactive and overweight/obese children, but also some evidence of less impact for girls regarding after-school programmes)	Medium (likely to be higher staff costs compared to in-school activities and direct and/or opportunity costs for facility hire)
School-based	Enhanced recess/breaks	Children and young people	Limited or contradictory evidence	No significant effect	No significant effect	Mixed (some evidence of better results for inactive and overweight/obese children, but also some evidence of less impact for girls regarding after-school programmes)	Low (low–cost training and resources)
Community-based	Community-based interventions	Community-wide	Limited or contradictory evidence	No significant effect	No significant effect	Mixed (some evidence of positive impacts for socioeconomically disadvantaged children under 5 and women, inconclusive evidence regarding other subgroups)	Medium (highly variable dependent on nature of intervention, but training and co-ordination of volunteers, licensing, insurance likely to impact costs)
Community-based	Peer-led interventions	Community-wide	Strong evidence	Small	Small	Unclear (limited evidence available)	Medium (highly variable dependent on nature of intervention, but training and co-ordination of peer-leaders, licensing, insurance likely to impact costs)
Active travel	Changes to the built environment (e.g. new or improved cycle and pedestrian routes, improved green spaces)	Population-wide	Moderate evidence	Small (positive results consistently reported across different outcome measures of varying quality)	Small (positive results consistently reported across different outcome measures of varying quality)	Mixed (some evidence that benefits are felt by more affluent populations, and theoretical evidence that improvements in safety perceptions could benefit women in particular)	High (high initial costs for infrastructure changes and some costs for ongoing maintenance)
Active travel	‘Carrots’ and ‘sticks’ - positive and negative strategies to improve active travel (e.g. subsidies or infrastructure improvements vs. congestion charges or reduced road space for cars)	Population- wide	Limited or contradictory evidence	No significant effect (non-significant finding that combined ‘carrot-and-stick’ strategies may be more effective)	No significant effect (non-significant finding that combined ‘carrot-and-stick’ strategies may be more effective)	Unclear (limited evidence available)	Estimate not possible (highly variable depending on nature of intervention)
Active travel	Active School Transport (AST)	Children and young people	Limited or contradictory evidence	Negligible effect (some small or non-significant positive impacts found)	Negligible effect (some small or non-significant positive impacts found)	Unclear	Medium-to-high (highly variable depending on nature of intervention, but may require small-scale infrastructure changes and additional staff costs)

School-based interventions

Section summary

• School-based interventions generally show small or no significant effects on increasing physical activity levels, with some specific interventions showing more promise, for example:

  • Physically active lessons
  • Classroom-based physical activity recess

• Studies also highlight potential factors that limit intervention effectiveness:

  • Compensatory behaviours
  • Limited teacher engagement
  • Lack of context-specific intervention adaptation
• Evidence on reducing inequalities is mixed; while school-based interventions do not consistently close gaps by socioeconomic status or gender, they may benefit the least active and overweight children, and targeting lower-SES schools could help reduce population-level disparities over time.
• None of the included studies reported on intervention costs, creating a gap in understanding resource implications.

School-based interventions are programmes implemented within school settings, with the goal of improving physical activity among children and adolescents. The interventions can take a variety of forms, including curriculum changes and teacher training to integrate activity into routines. Examples include integrating movement into academic lessons and enhancing physical education time.

We identified 24 meta-analyses and systematic reviews regarding school-based interventions, of which 5 were selected for this review (see Table 2 below) based on their publication date, relevance to the research questions and location of the reviewed studies. One scoping review, Porter et al. (2024), was also included as a recommendation from DCMS. The majority of interventions were implemented in the US, Australia, the UK and other European countries. All of the reviewed papers included only experimental or quasi-experimental studies within their meta-analyses/systematic reviews, with the exception of Jones et al. (2019), which included qualitative, quantitative and mixed-methods studies.

Table 2: Studies included in the school-based interventions review (organised by publication year, newest to oldest)

Authors (year)	Type	Interventions evaluated	Number of papers, design, primary outcome(s)	Location of reviewed studies	Findings summary
Neil-Sztramko et al. (2021)	Systematic review	Most commonly, interventions were multi-component, whole-school interventions that included a combination of educational materials and changes to the school environment and/or curriculum. Other interventions were focused primarily on providing opportunities for MVPA within, before and after school time, such as active academic lessons and after-school sports programmes.	89 papers, RCT designs only, time spent in MVPA (accelerometer-measured)	Global (mainly from USA, Australia, and UK)	In this review school-based physical activity interventions led to an average increase of 0.73 minutes/day of MVPA. This study did not report any findings on health inequalities or intervention costs.
Jones et al. (2020)	Systematic review and meta-analysis	Interventions that aimed to expand, extend, or enhance opportunities for physical activity in schools. Expanded interventions replaced low-activity time with physical activity (e.g. physically active learning). Extended interventions increased the duration of existing physical activity sessions (e.g. longer PE classes). Enhanced interventions modified current practices to increase activity levels within the same time frame (e.g. making recess more active).	57 papers, mixed-methods designs, time spent in MVPA (accelerometer-measured)	Global (mainly studies from USA, Europe, and UK)	This review found no significant improvement in whole-day MVPA. Interventions that expanded physical activity opportunities showed moderate effects during intervention periods, but were offset by compensatory behaviours outside of school hours, resulting in no overall change in daily physical activity. This study did not report any findings on health inequalities or intervention costs..
Norris et al. (2020)	Systematic review and meta-analysis	Physically active lessons (PALs) integrating physical activity into academic lesson time.	42, experimental designs only (27 of which are RCTs), time spent in MVPA (accelerometer-measured and self-report)	Global (mainly USA, Australia, UK, Europe)	This review found that PALs led to large increases in MVPA during lesson time (SMD=2.33) but smaller increases in total daily MVPA (SMD=0.32). This study did not report any findings on health inequalities. Research samples are predominantly white, with few studies focusing on ethnic minority or disadvantaged populations. No information reported on intervention costs.
Love et al. (2019)	Meta-analysis	Single or multi-component interventions (e.g. interventions including educational component, social environmental changes, and/or physical environment changes) in school-settings of at least 4 week duration aimed at increasing physical activity.	25, RCT designs only, time spent in MVPA (accelerometer-measured)	Global (mainly Australia and Europe)	This review found no effect on daily MVPA on average. There were no differences in impact by gender or socio-economic status, suggesting interventions did not reduce (or increase) physical activity inequalities. No information reported on intervention costs.
Mears & Jago (2016)	Systematic review and meta-analysis	After-school physical activity programs for children aged 5 to 18.	15, experimental designs only (6 of which included for the mata-analysis), MVPA (accelerometer-measured, heart rate monitors and self-report)	USA and UK	The review found mixed evidence on the effectiveness of after-school physical activity interventions, with some indications of short-term improvements in MVPA, based on mid-intervention data from a small number of studies. The meta-analysis found a non-statistically significant increase of 4.8 minutes of MVPA/day (95% CI −0.94 to 10.61). There was some evidence suggesting that overweight/obese children and boys may benefit more from after-school interventions. No information reported on intervention costs.

Authors (year)	Type	Description	Location of reviewed studies	Findings summary
Porter et al. (2024)[footnote 4]	Scoping review	The review identified components of school-based interventions to develop a common framework for designing interventions. The authors included 79 papers assessing 45 different school-based physical activity interventions children aged 7 to 11 years.	Europe	The authors identified 177 unique intervention components which were then grouped into 60 broader component types (e.g. training, incentives, equipment) mapped to 11 different ‘opportunities’ to increase physical activity within school settings (e.g. physical education, active breaks, before/after school clubs). The reviewed interventions most frequently targeted physical education (21%), active and outdoor learning (16%), active breaks (15%), and school-level environment (12%). Many studies failed to report contextual factors, limiting the ability to assess how context influences intervention success.

Effectiveness in improving physical activity outcomes

The reviewed evidence suggests that, on average, school-based interventions result in little to no increase in total daily moderate-to-vigorous physical activity (MVPA). The Cochrane review by Neil-Sztramko et al. (2021), found that school-based interventions resulted in an average increase of 0.73 minutes of MVPA per day, while Jones et al. (2020) and Love et al. (2019) found no evidence of significant effects on MVPA. Although school-based interventions may not be effective on average, there was more promising evidence for some individual interventions including physically active lessons and class-based active recess.

Physically Active Lessons (PALs)

PALs combine physical activity with academic content to increase movement during school hours (e.g. doing star jumps while reciting times tables or using movements to show whether an answer is true or false). 2 out of the 5 included papers under school-based interventions reported effect sizes for this approach, with effects ranging from a small increase in total daily MVPA (SMD = 0.32) to a large increase in lesson-time MVPA (SMD = 2.33).

Norris et al. (2020) found that PALs produced large increases in MVPA during lesson time (standardised mean difference^{[footnote 5]}, SMD=2.33), but more modest increases in MVPA when measured over the entire day (SMD=0.32).^{[footnote 6]} Multiple authors report that this may be a result of a compensatory mechanism, where children reduce their activity levels later in the day to offset the increased exertion during lessons (Norris et al. 2020, Jones et al, 2020). However, given the potential scale of PALs, even a modest improvement in daily MVPA may have substantial population-level impact, especially if sustained over time.

Finally, Norris et al. (2020) report that the effects of PALs tend to diminish when schools do not embed them into routine practice and highlight the need for strong teacher engagement and training to make the intervention a success.

Class-based Active Recess

The evidence reviewed suggests that providing structured active class breaks can lead to increases in MVPA. These types of interventions typically involve providing resources or training teachers to deliver short class recess (usually around 10 min) that can be implemented in their normal classroom setting. Four studies reviewed by Jones et al. (2020) found that recess interventions resulted in statistically significant increases in either MVPA or step counts, although effect sizes were not consistently reported across individual studies, making it difficult to summarise the magnitude of impacts. The high risk of bias (due to low quality designs and/or small sample sizes) in these studies limits the strength of the conclusions and highlights the need for more rigorous research in this area.

Other interventions

Efforts to enhance standard school physical education programmes by increasing the frequency, duration, or intensity of PE classes (Neil-Sztramko et al., 2021; Jones et al., 2020), enhancing non-class based recess by adding or improving resources such as play equipment (Jones et al., 2020), and after-school programmes (Mears & Jago, 2016; Neil-Sztramko et al., 2021; Jones et al., 2020), were all shown to have little to no impact on MVPA.

Why school-based interventions may fail to produce sustained effects

Several factors may contribute to the lack of effectiveness for some school-based interventions:

1. Compensatory reductions in activity: Jones et al. (2020) found that when physical activity was measured across a whole day rather than just during intervention periods, the effects weakened or disappeared​. This suggests that children may reduce their activity levels outside of intervention periods, offsetting the gains made during the intervention.

2. Teacher engagement and implementation challenges: Moderator analysis conducted by Norris et al. (2020) identified that PAL interventions delivered by existing classroom teachers were less successful compared to newly recruited delivery staff. The authors argue that lack of teacher involvement in the design and tailoring of interventions may result in lower motivation and ultimately reduced implementation fidelity and engagement. Porter et al. (2024) state that while school staff are a convenient and cost-effective option for delivering interventions, they often face time and resource constraints that pose challenges to implementing high-quality physical activity programmes.

3. Lack of context-specific adaptation: Porter et al. (2024) argue that tailoring interventions to the unique characteristics of individual schools (such as available resources, priorities, and student demographics) may improve their effectiveness and sustainability. Their scoping review found that many studies lacked high-quality reporting on the context in which interventions were implemented, including factors like geographical location, socio-economic status, ethnicity, and school size. The authors suggest that designing interventions responsive to the realities of current school settings will likely maximise their potential to increase physical activity.

Impact on reducing physical activity inequalities

Taken together, the evidence suggests that school-based interventions appear to equitably engage both inactive and active students, with some studies indicating larger effects among the least active and overweight children. However, there is also some evidence that boys may benefit more than girls from some school-based interventions, highlighting the need for targeted interventions designed to increase physical activity among girls in school settings given that, in the UK, girls are generally less active than boys.

A meta-analysis by Love et al. (2019) examined the differential effects of school-based interventions and found no consistent evidence that they were more beneficial for girls or children from lower socioeconomic backgrounds. The review highlighted that while schools are an equitable setting for intervention delivery, they do not necessarily reduce pre-existing gaps in activity levels between gender and socioeconomic groups​.

Only one of the reviews included in this analysis, Jones et al. (2020), examined results by baseline activity level. Among the 57 studies included in their review, only two reported differential effects for the least active participants, and both found larger effects for this group. This provides an encouraging sign that, at the very least, school-based interventions may engage inactive students just as well as active students.

Some studies suggest that interventions are more effective for boys than girls. The Mears & Jago (2016) review of after-school programmes found that, on average, boys experienced greater increases in MVPA compared to girls, suggesting that school-based interventions may need to be specifically designed to engage girls more effectively, particularly given that girls tend to be less physically active than boys in the UK​.

One encouraging finding from Mears & Jago (2016) is that after-school programmes may particularly benefit overweight/obese children. Among the seven studies included in their review that analysed this subgroup, three reported small but positive effects, ranging from an additional 3.4 to 5.9 minutes of daily MVPA among overweight/obese participants. In contrast, four studies found no differential effect by weight status. This suggests that while results are mixed, some programmes may offer additional benefits for less active children.

While the evidence does not show consistent differential effects within schools, one potential avenue for addressing inequalities at the population level is to prioritise the implementation of effective school-based interventions in schools serving lower socioeconomic status areas. This approach would not guarantee reductions in inequalities within schools, but if interventions are shown to be effective at the school-level overall, targeting schools with the greatest need could help reduce disparities in physical activity between higher- and lower-socioeconomic status populations over time.

Interventions costs

None of the included reviews provided a formal analysis of intervention costs. As a result, there is limited evidence to inform cost-effectiveness comparisons across intervention types, highlighting a gap in the literature and a key consideration for future research and policy decision-making.

Community-based and peer-led interventions

Section summary

• Aside from peer-led interventions, there is a lack of high quality evidence about the impact of community-based interventions on physical activity.

• We identified two high quality evaluations of community-based interventions in London and Devon, both of which showed no significant impact on physical activity at the neighbourhood-level.

• However, both interventions were hampered by low participation rates which may mask individual-level impacts.

• Peer-led interventions showed more promising results, with high-quality reviews consistently demonstrating small to moderate improvements in physical activity, particularly when peer leaders were trained and interventions begin with an intensive phase.

• Evidence on reducing inequalities is limited; neither of the two community-based evaluations evaluated impacts on specific subgroups, and peer-led interventions lack robust evidence on impacts across priority subgroups. Some evidence suggests group-based delivery formats and higher intervention intensity may improve outcomes for disadvantaged populations, particularly women and young children.

• None of the included studies provided formal cost analyses, leaving a gap in understanding the economic case for community-based and peer-led interventions.

For this review, we define community-based interventions as locally-led initiatives that promote and facilitate physical activity within a specific community, often organised by charities, volunteer groups, or local organisations. These interventions can be accessible to an entire community (e.g. Parkrun) or targeted at a specific group (e.g. a walking group for older adults), and typically focus on providing regular free or low-cost opportunities for physical activity.

Peer-led interventions can be considered a subset of community-based interventions and involve the delivery of behaviour change strategies by individuals who share key characteristics or experiences with the target population, such as age, health status, or social background.

We identified 44 meta-analyses and systematic reviews regarding community-based and peer-led interventions, of which 4 were selected for this review (see Table 3 below) based on their publication date, relevance to the research questions, and location of the reviewed studies. The majority of interventions were implemented in the US, Canada, and Europe. All of the reviewed papers included only experimental or quasi-experimental studies within their meta-analyses/systematic reviews, with the exception of one review of systematic reviews and meta-analyses (Craike et al., 2018).

Of the four papers selected for our review; two focused solely on peer-led interventions (Hulteen et al., 2019; Best et al., 2016); one reviewed interventions focusing on socioeconomically disadvantaged populations which included discussion of interventions delivered in ‘community settings’ (Craike et al, 2018); and one was a Cochrane review focused on ‘community-wide’ interventions (Baker et al., 2015). This latter review helped us to identify two high-quality studies (Solomon et al., 2014; Phillips et al., 2014) that met our definition of community-based interventions given the rest of the review was largely out of scope.

The Solomon and Phillips studies found no significant impact on physical activity levels at a neighbourhood level, though the authors attributed the poor results to low awareness and participation rates in the programmes they evaluated, and the chosen outcome measure may have masked individual-level effects.

On the other hand, peer-led interventions emerged as a promising approach to promote physical activity and support sustained improvements over time. To help reduce inequalities, two specific strategies are recommended: (1) prioritising group-based delivery formats over individual approaches, and (2) designing interventions to begin with a more intensive phase, particularly in terms of frequency and level of participant contact.

Table 3: Studies included in the community-based peer-led interventions review (organised by publication year, newest to oldest)

Authors (year)	Type	Interventions evaluated	Number of papers, design, primary outcome(s)	Location of reviewed studies	Findings
Hulteen et al. (2019)	Rapid ‘state-of-the-art’ review	The study reviewed peer-delivered physical activity interventions across different age groups and settings from the 5 years prior to publication.	15 papers, experimental designs only (RCT, non- randomised trial, pre-post experimental design without a control group), PA levels (self-reported and device assessed, and activity log records)	Global (primarily from the US, UK, and Canada)	This review found that while 80% of the included studies reported a significant positive effect of peer-led interventions on physical activity, 53% of studies also reported no significant effects on at least one measure of participant physical activity behaviour. Interventions that included training for peer mentors were more likely to be effective, with 10 out of 11 such studies reporting positive effects, compared to 2 out of 4 studies without peer-leader training. This study did not report any findings on health inequalities or intervention costs.
Craike et al. (2018)	Umbrella review	The study reviewed physical activity interventions targeting socioeconomically disadvantaged populations, including community-based interventions.	17 papers, systematic reviews and meta-analyses, a variety of physical activity behaviour change measures (e.g. accelerometry outcomes, sport participation)	Not reported	This review found that interventions were effective in increasing physical activity among children from socioeconomically disadvantaged backgrounds, particularly in school and early childhood settings. There was limited impact reported for adolescents and adults from low-income groups. No information reported on intervention costs.
Best et al. (2016)	Systematic review and meta-analysis	The study reviewed peer-led self-management programmes grounded in social cognitive theory to promote physical activity among adults.	21 (17 of which included in the meta-analysis), RCTs only, PA duration and PA frequency (most studies used self-reported measures)	Not reported, but evidence mostly came from high-income countries	This review found that peer-led self-management programmes had a moderate pooled effect on physical activity immediately post-intervention (SMD = 0.4, p < 0.001), and a large effect at follow-up in the few studies that included longer-term data (SMD = 1.5, p = 0.03). The meta-analysis of studies using minutes of physical activity showed a small but significant effect (SMD = 0.2, p < 0.001). This study did not report any findings on health inequalities or intervention costs.
Baker et al. (2015)	Systematic review	The study reviewed community-wide interventions targeting the general population. Interventions combined at least two of the following strategies: mass media campaigns, informational materials, individual counselling by health professionals, partnership with organisations, delivery in community settings, environmental improvements.	33, experimental designs (RCTs, non- randomised controlled trials, and quasi- experimental designs), PA levels (self-reported and objective measures like accelerometer data), frequency of PA, percentage of active people	25 in high income countries (mainly from North America and Europe), 8 in low or middle income countries	This review found that community-wide interventions were generally ineffective at increasing physical activity at the population level, with limited reach, low intervention intensity, and variable implementation fidelity identified as key barriers to success. The review did not specifically focus on health inequalities but noted that many interventions failed to engage certain subgroups within the community, potentially leading to unequal benefits. No information reported on intervention costs.

Effectiveness in improving physical activity outcomes

Overall, we found limited robust evidence of non-peer-led community-based interventions for physical activity. The two high quality studies we identified through the Cochrane Review by Baker et al. (2015) found no significant impact on physical activity at the neighbourhood level (Solomon et al., 2014; Phillips et al., 2014). In contrast, there is high quality evidence to suggest that peer-led interventions can effectively promote physical activity (Best et al., 2016; Hulteen et al., 2019).

The Devon Active Villages Evaluation, DAVE (Solomon et al., 2014)

DAVE was a stepped wedge cluster randomised controlled trial funded by Sport England in which 128 rural villages (clusters) across 7 regions of Devon were randomised to receive the intervention in one of four time periods between April 2011 and December 2012.

The intervention comprised an initial needs assessment followed by 12 weeks delivery of at least three types of physical activity opportunities (e.g. fitness classes for adults) and 12 months of post-delivery support.

The authors found no change in the proportion of people active at recommended levels (adjusted odds ratio = 1.02, p-value = 0.8) compared to control, though there was a marginally non-significant finding of a 43-minute increase in moderate-to-vigorous physical activity^{[footnote 7]} per week (p-value = 0.07), as measured through postal surveys.^{[footnote 8]}

The authors attributed the ineffectiveness of the programme to its low reach, driven mostly by lack of awareness. In the intervention groups, only 16% reported awareness of Devon Active Villages, and only 4% reported participation in intervention events. As outcomes were measured at the neighbourhood level, low participation rates may therefore have masked individual-level effects among those that did take part, though this reflects a realistic implementation challenge for any community-based intervention.

Well London Phase 1 (Phillips et al., 2014)

The Well London programme was a community engagement initiative promoting healthy eating, physical activity, and mental well-being in deprived neighbourhoods. Twenty matched pairs of neighbourhoods (clusters) in London were randomised to intervention and control conditions in a cluster randomised trial.

Phase 1 of Well London consisted of 14 interconnected projects co-developed and delivered in each area. A core group of volunteers in each neighbourhood supported residents to participate in the programme, access services, and improve health behaviours.

No evidence of impact was found on the primary physical activity outcome (adjusted odds ratio = 1.01, p-value = 0.9), which was a self-report measure of achieving 30 minutes of moderate-intensity physical activity five times per week, administered via household surveys. There was also no indication of any differential effects within subgroups defined by age, gender, ethnicity, educational attainment, or employment status.

Similar to the DAVE programme, the authors suggest that low participation rates may have compromised any impact of the intervention, though did not elaborate on reasons for low participation. Indeed, follow-up surveys indicated only 3.1% of people participated in any individual Well London project. Again, low participation at the neighbourhood level may have masked individual-level effects for those that did take part.

Peer-led interventions

Peer-led interventions have shown positive results. A meta-analysis by Best et al. (2016) found a modest but statistically significant pooled effect on physical activity immediately post-intervention of SMD = 0.4.^{[footnote 9]} Out of the 21 studies included in their review, 14 reported positive significant effects in physical activity, with effect sizes ranging from 0.1 to 1.1.

Similarly, a review by Hulteen et al. (2019) concluded that peer-delivered interventions were effective across age groups and settings, often performing comparably to professionally-led programmes. Of the 15 studies reviewed, 12 reported significant improvements on physical activity outcomes. Programmes that included training for peer leaders were more likely to be effective, with 91% (10/11) of these studies showing significant improvements in participant physical activity, compared to just 50% (2/4) in studies without such training. However, given the low number of studies that did not include training we would suggest caution interpreting this specific finding.

Interestingly, Best et al. note that changes in physical activity may take time to emerge after peer-led interventions end. They highlight a 16-week intervention which showed no statistically significant differences between the intervention and control groups at the end of the programme (Buman et al., 2011). However, at an 18-month follow-up, the intervention group reported significantly higher levels of physical activity compared to baseline, which the authors speculated may be due to increased self-efficacy to engage in physical activity after the intervention ended. If true, it suggests that peer-led programmes could be particularly effective at supporting sustained behaviour change.

Impact on reducing physical activity inequalities

Evidence on equity impacts is more limited. The Baker et al. (2015) review found that most community-wide interventions did not report outcomes by subgroups, such as gender, socioeconomic status, or cultural groups.

Overall, while peer-led interventions appear effective in improving physical activity on average, evidence for specific priority subgroups was limited across the two papers reviewed (Hulteen et al., 2019; Best et al., 2016). More studies are needed that explicitly measure differential impacts across subgroups.

More targeted evidence comes from the umbrella review by Craike et al. (2018), which focused on understanding the impact of interventions aimed at improving physical activity among socioeconomically disadvantaged groups. The review synthesised findings from systematic reviews of physical activity interventions across a range of settings, but we have only included findings relevant to community-based interventions below.

Community-based interventions targeting children (or parents of children) under 5 years old consistently reported positive and significant impacts on physical activity, particularly through group-based formats (Laws et al. 2014, as cited in Craike et al., 2018).

In contrast, interventions targeting adolescents from disadvantaged backgrounds showed no significant improvements in physical activity across the three relevant reviews included in the umbrella review (Cleland et al., 2012, as cited in Craike et al., 2018).

Among disadvantaged adults, findings were mixed and inconclusive, with some studies indicating positive effects, especially for women, as highlighted in a high-quality review by Cleland et al. (2012) which reported a moderate effect for group-delivered interventions (SMD = 0.36).

For older adults, the Craike et al. review found very limited evidence, making it difficult to draw firm conclusions about the effectiveness of community-based interventions for this population.

Craike et al. also identified intervention intensity as an important factor influencing effectiveness. Across all settings, interventions that began with a higher-intensity phase, involving frequent contact and engagement, were associated with better outcomes (Cleland et al., 2012, as cited in Craike et al., 2018).

In summary, the review concluded that interventions may effectively increase physical activity in socioeconomically disadvantaged children and women. However, evidence for adolescents, adults, and older adults in disadvantaged contexts remains limited and inconsistent, and significant gaps remain in understanding how to support long-term behaviour change in these populations.

Interventions costs

Consistent with the lack of cost analysis in reviews of school-based interventions, our analysis of community-based interventions also revealed no studies providing a formal evaluation of intervention costs. As a result, we lack evidence to compare the cost-effectiveness of different intervention types, including the specific economic advantages or disadvantages of peer-led models within community settings. This gap underscores the need for future research to specifically address the costs associated with such interventions.

Active travel interventions

Section summary

• Changes to the built environment may lead to increases in physical activity, especially when changes enhance the accessibility and perceived safety of walking and cycling.

• Active school transport initiatives typically find little to no impact on physical activity.

• Equity impacts were unclear; some evidence suggests that interventions may disproportionately benefit more advantaged groups, underlining the importance of targeting lower-income and under-resourced communities.

• Cost data were limited and rarely linked to effectiveness, making it difficult to assess value for money or inform decisions on scaling.

Active travel interventions aim to embed physical activity into daily routines by encouraging individuals to take up more active forms of transportation (e.g. walking and cycling). These interventions often involve modifying the built environment to provide greater opportunity for active travel such as constructing cycle trails and enhancing pedestrian pathways. In contrast to school- or community-based interventions, active travel interventions typically operate at the population level and target the general public.

We identified 21 meta-analyses and systematic reviews regarding active travel interventions, of which 5 were selected for this review (see Table 4 below) based on their publication date, relevance to the research questions, and location of the reviewed studies. The majority of interventions were implemented in the US, UK, Europe and Australia. Reflective of the intervention types, most of the evaluations reviewed by the papers were quasi- or natural experiments, but other designs such as controlled before-and-after studies and RCTs were also included.

The reviews suggest that changes to the built environment may lead to increases in physical activity by enhancing the accessibility and safety of walking and cycling, while active school transport initiatives typically find little to no impact on physical activity. The overall quality of evidence is low, largely due to challenges in evaluating interventions using experimental methods. However, higher quality studies tended to report significant positive outcomes. Equity impacts remain unclear, though some evidence suggests benefits may be concentrated among more advantaged populations. Finally, cost data are scarce and rarely linked to effectiveness, highlighting the need for better reporting and economic evaluation to support policy and investment decisions.

Table 4: Studies included in the active travel interventions review (organised by publication year, newest to oldest)

Authors (year)	Type	Interventions evaluated	Number of papers, design, primary outcome(s)	Location of reviewed studies	Findings
Xiao et al. (2022)	Systematic review and meta-analysis	The study reviewed 121 population-level transport interventions targeting the general adult population. The authors included carrots, which are positive interventions (e.g. new bike-share schemes), sticks, which are negative interventions (e.g. congestion charges, increased parking prices), and combined carrot-stick approaches.	102 (64 of which included in the meta-analysis); controlled before-and-after studies of population-level interventions and travel behaviours (57% repeated cross-sectional, 31% longitudinal pre-post, 15% RCT); Changes in travel behaviour including objective indicators (e.g. pedestrian and bicycle counts) and subjective assessments (e.g. self-reported time spent walking or cycling)	Global (50% from North America, 31% from Europe, 8% from Oceania, 9% from Asia, 1% from South America)	There was no evidence to suggest that these policies increased PA. However, in relative terms, this review found that stick and carrot-and-stick interventions showed greater, though not-statistically-significant, effects on increasing active travel than carrot-only strategies (SMD = 0.33 for combined approaches; carrot-only SMD = 0.08). Interventions targeting access, safety, and space functions increase active travel. This study did not directly address health inequalities, but it noted that stick strategies are underused and less studied, potentially due to implementation challenges in more disadvantaged areas. No information reported on intervention costs.
Panter et al. (2019)	Systematic review	The study reviewed environmental interventions targeting the general population, typically in urban settings. Interventions included changes to the physical environment to promote walking and cycling, such as walking and cycling paths, facilities to support cycling, and routes for walking and cycling.	13; studies with a control or comparison group or graded measures of exposure to the intervention; Changes in walking and cycling (e.g. weekly time spent walking/ cycling, share of active commutes). Measures were self-reported, objective or observational (e.g. pedestrian counts).	Not specified, but focuses on relevance to UK policy	This review found that most interventions showed small or inconsistent effects on walking and cycling. Out of the studies reviewed, 6 reported statistically significant positive effects, while 8 reported no significant or inconsistent findings across different outcomes. Stronger outcomes were observed when interventions addressed accessibility and safety, particularly in both supportive and less supportive local contexts. This study did not systematically analyse impacts on health inequalities, but identified that mechanisms and contexts — such as perceived safety and accessibility — may lead to differences in impact across groups. No information reported on intervention costs.
Stappers et al. (2018)	Systematic review	Built environment infrastructure changes to promote physical activity and/or active transportation.	19 (14 unique interventions); Quasi-experimental or natural experiments with pre-post designs; MPA, VPA, MVPA, time spent walking/ cycling and share of active commutes. Measures were objective (e.g. accelerometer and direct observation) or subjective (e.g. phone surveys and travel diaries).	Global (mainly high-income countries including US, UK, Australia, Netherlands, Brazil)	This review found that interventions showed positive effects on cycling, while impacts on walking and overall physical activity were mixed or found no significant effect. Proximity to interventions was associated with greater increases in cycling, suggesting localised benefits for nearby residents. No studies directly assessed sedentary behaviour. This study did not explicitly address health inequalities. However, equity implications were inferred from the findings, as those living closest to the new infrastructure tended to benefit more, raising potential concerns about uneven distribution of impacts. No information reported on intervention costs.
Larouche et al. (2018)	Systematic review	The study reviewed 30 interventions designed to promote active school transport among school-aged children. Interventions included Safe Routes to School, School Travel Plans, Walking School Buses, bike training, and school-based promotional activities.	27; Quantitative empirical study (observational pre–post studies, quasi-experiments, natural experiments, RCTs); Change in physical activity or active transportation (walking and cycling).	Global (mainly from the US, UK, and Canada)	This review found that 13 of 30 interventions reported increases in active school transport (AST), with effect sizes ranging from -0.61 to 0.75 (i.e. small effects). Larger increases were observed in studies with longer follow-up periods. Multi-components interventions (e.g. Safe Routes to School,) which combined infrastructure improvements, education, and community engagement, appeared more effective than single-component approaches. The study noted that implementation was often limited by lack of resources in low-income communities. However, equity was not explicitly addressed in most evaluations. No information reported on intervention costs.
Smith et al. (2017)	Systematic review	Interventions that improve walkability, upgrade parks and playgrounds, and invest in active transport infrastructure (e.g. sidewalks, bike lanes)	28; Quantitative empirical study (natural experiment, prospective, retrospective, experimental, or longitudinal research); Changes in physical activity and active transport.	Global (majority in US and Australia)	This review found that interventions such as walkability improvements, new parks and playgrounds, and active travel infrastructure were associated with increases in physical activity and active travel. Positive effects were observed for walking and cycling, suggesting that these infrastructure changes can contribute to higher activity levels. Some evidence indicated that socioeconomically advantaged groups benefited more from infrastructure changes, with mixed findings regarding equity of access and use. No information reported on intervention costs.

Effectiveness in improving physical activity outcomes

Active travel interventions vary widely in approach, but based on the studies reviewed, they can be grouped into three main categories:

1. Changes to the physical environment

2. Positive and negative strategies to influence travel behaviour, and

3. School-based efforts to promote active travel to school among children

In general, active travel interventions often showed positive results, especially when they included improvements in access, safety, and space for cyclists and pedestrians. However, the quality of the available evidence is generally low, which introduces uncertainties around effectiveness. Evaluation is inherently a complex task for active travel interventions, particularly regarding changes to the physical environment for which it is often difficult to identify a suitable comparison group. Promising interventions may therefore be underrepresented or appear less effective simply because they are harder to evaluate with the same level of rigour.

Additionally, more distal, individual-level outcome measures (e.g. changes in MVPA) can be difficult to obtain, and so many researchers rely on more proximal, population-level outcomes (e.g. changes in bicycle traffic) which can make it more difficult to assess the true impact of interventions on physical activity levels. More flexible and context-sensitive approaches to evidence generation are needed to fully understand active travel interventions’ potential. However, the more robust studies included within the papers we reviewed tended to show more positive and significant results which lends credibility to their effectiveness for improving physical activity.

Changing the built environment to promote physical activity

Three of the reviewed papers looked at built environment interventions to promote active travel and physical activity (Smith et al., 2017; Stappers et al., 2018; Panter et al., 2019). Interventions included improvements to walkability (e.g. new sidewalks and crossings), development of on- and off-road cycling and walking paths, and upgrades to parks, green spaces, and other public infrastructure. Overall, the evidence suggests that these may be a promising avenue for improving physical activity, especially in relation to the creation of on- and off-road cycling and walking paths and interventions which improve the accessibility and perceived safety of active travel. However, the quality of evidence was typically low and there was substantial variation in outcome measures across the included studies, making it difficult to compare interventions and understand their impact on physical activity at an individual level, particularly where population-level measures were used (e.g. observations of walking or cycling traffic).

Smith et al. (2017) reported that neighbourhood improvements were generally associated with increases in physical activity. Of the 28 included studies, two showed no significant impact, one found negative impact, and all others reported significant positive impact on outcomes spanning walking and cycling behaviour, physical activity, and proximal measures such as use of infrastructure or facilities. Notably, significant positive impacts on active travel and physical activity were found in high-quality studies assessing improvements to walking routes, installation of park and playground equipment, and park renovations.

Stappers et al. (2018) found that the implementation of new on- and off-road trails for walking and/or cycling showed positive effects on cycling in six out of eight of the studies that included this outcome, negative effects in one, and no significant effect in the other. Findings for walking and self-reported physical activity were more mixed (two out of five reported increases both in walking and in overall physical activity). Broader active travel interventions affecting the wider infrastructure system (e.g. traffic-free bridges, parallel road and cycling trails) resulted mainly in no significant effect or negative effects on overall PA, walking and cycling.^{[footnote 10]} Notably, most of the included studies found that proximity to the interventions was positively associated with increases in physical activity when this was included in their analysis.

Panter et al. (2019) found that of the 11 built environment interventions included in their review, six reported positive effects on physical activity, with no clear pattern identified regarding the type of intervention and its effect. Encouragingly though, higher-quality evaluations were more likely to show positive effects.

The authors also reviewed how built environment interventions may influence physical activity, and described three central mechanisms which can interact with existing environments to influence physical activity both positively and negatively:

1. Improving accessibility and connectivity- built environment changes can encourage physical activity by making it more desirable to walk or cycle to a location rather than using more passive transport options. This incorporates ideas of time, distance, cost and the physical effort involved in a trip. In environments unsupportive of active travel this can lead to increases in physical activity as more people walk or cycle. Conversely, in already supportive environments, interventions can lead to decreases in physical activity as they may establish more direct or shorter routes to important locations.

2. Improving safety - changes that help segregate pedestrians and cyclists from motor vehicles can help improve perceptions of safety and make walking and cycling a more attractive option.

3. Improving experience - built environment changes can improve the experience of walking and cycling, such as by making it a more aesthetically pleasing transport route or improving the smoothness and comfort of a travel path.

Painter et al. note that mechanisms were rarely described by authors of the individual studies they reviewed, suggesting that a common framework for reporting intervention mechanisms could help improve the quality of the literature. They also attempted to map these mechanisms to the reviewed literature to understand associations with effectiveness, and suggest that the most effective interventions appeared to target accessibility and safety, but this appears to be largely speculative rather than causal evidence.

Promoting active travel through positive (‘carrots’) and negative (‘sticks’) strategies

One meta-analysis categorised active travel interventions based on whether they acted through positive or negative strategies, referred to as ‘carrots’ and ‘sticks’ respectively (Xiao et al., 2022). ‘Carrots’ included interventions such as subsidies for bicycles and increased green space, while ‘sticks’ included interventions like congestion charges or reduced road space. Overall, combined strategies (‘carrot-and-stick’ approaches) appeared to be more effective for active travel measures (SMD = 0.33) than either strategy alone (‘carrot’, SMD = 0.08; ‘stick’, SMD = 0.01). However, the results were not significant and very few studies considered ‘stick’ interventions alone, so this finding should be interpreted cautiously.

Similar to the findings of Panter et al. (2019), interventions that targeted access and safety(as well as increased space for walking or cycling) tended to have significant positive effects on active travel, with SMDs ranging from 0.20 to 0.24 on these specific functions. However, the authors do not report any corrections for multiple comparisons in their methods, and therefore there is a risk that these findings may be due to chance.

Active school transport (AST) interventions

One systematic review looked specifically at interventions to increase active travel for journeys to and from school, referred to as ‘active school transport’ or AST (Larouche et al., 2018). The interventions included programmes like Safe Routes to School (a national US initiative), walking school buses, school-led approaches and promotional campaigns targeting children and adolescents. Overall, AST interventions likely have little to no impact on physical activity, though evidence quality was typically low which makes conclusions uncertain.

Larouche at al’s review found that only 13 of the 30 interventions reported increases in AST, and most effect sizes were small or negligible, with SMDs ranging from -0.61 to +0.75. Two studies evaluating the Safe Routes to School programme found that multi-component interventions combining educational activities with infrastructural changes were more effective than single-strategy interventions, with increases ranging from 5 to 20 percentage points for multi-component interventions compared to 2 to 5 percentage points for single-component interventions.

Qualitative evaluations focusing on the implementation of AST interventions indicated that many interventions were constrained by short follow-up periods, partial delivery of planned components, and limited resources—particularly in low-income communities. The overall quality of evidence was low, with 27 of the 30 interventions rated as weak due to lack of control groups, poor confounder adjustment, and unknown quality of measurement tools.

The review also highlighted important gaps in the existing research. Few studies explored the impact on total daily physical activity beyond the school commute and therefore may overlook the possibility of compensatory effects (as highlighted in ‘School-based interventions’). There was also little consistency in the selection of outcome measures, making it difficult to conduct a meta-analysis and compare interventions across different settings.

Impact on reducing physical activity inequalities

Evidence on the equity impact of active travel interventions is limited and mixed. Smith et al. (2017) explicitly examined differential effects of infrastructure changes by socioeconomic status (SES) and ethnicity. Overall, the authors concluded that there was insufficient evidence to determine whether infrastructure interventions improved outcomes for disadvantaged groups, but they found some indication that these interventions may predominantly benefit more affluent populations. Although the authors did not provide enough detail to understand the underlying reasons, this pattern could reflect greater ability among these groups to use new infrastructure or a tendency for interventions to be implemented in better-resourced communities. If the latter, this impact could be avoided if future interventions were implemented in more deprived communities, noting that interventions would likely need to be tailored to account for specific barriers that may be more prevalent among these communities (e.g. they may have less storage space for bicycles or health conditions that make it difficult for them to participate in active travel).

While Stappers et al. (2018) did not report outcomes by socioeconomic status, ethnicity, or health status, the authors noted that intervention effects were generally strongest among residents living closest to new active travel infrastructure, suggesting that infrastructure changes could benefit groups who are typically less physically active if they are targeted at the right locations.

Panter et al. (2019) did not report outcomes by socioeconomic or demographic subgroup, but highlighted that uptake of active travel interventions is strongly influenced by perceived safety and convenience. These factors may vary across population groups (for example, there is some evidence that women are more likely to report safety concerns as a barrier to cycling^{[footnote 11]}), meaning that some interventions may particularly benefit population groups for whom safety concerns and/or accessibility is a greater concern.

Xiao et al. (2022) did not assess subgroup differences in intervention effectiveness due to limited reporting across studies.

Interventions costs

Only one of the included reviews (Smith et al., 2017) provided cost information for active travel or built environment interventions. Reported costs varied widely, from around USD 45,000 for smaller-scale enhancements (e.g. fitness zones in parks) to over USD 5.5 million for major park renovations. However, even in that study, no formal cost-effectiveness analyses were conducted, and the authors noted that the generally low quality of evidence across studies limits confidence in any conclusions about value for money. This underscores the importance of clear and consistent cost reporting in future studies. Without this, it remains difficult to compare the economic implications of different intervention types or to make informed decisions about scaling or replicating interventions with promising outcomes.

Discussion and recommendations

Caveats and limitation for this review

There are several important caveats and limitations relevant to this review:

• Review scope: This review was intended to be a rapid evidence review of up to five meta-analyses or systematic reviews for each intervention category, where available. This means the review is not comprehensive and may not reflect all relevant evidence, particularly for intervention types with a more diffuse or emerging evidence base. For example, in the community-based interventions, no meta-analysis fully matched our inclusion criteria. In this specific case, we used the closest relevant meta-analysis (the Baker et al. study) to identify individual high-quality studies that more closely matched our definition of community-based interventions. We acknowledge that there may be more recent robust studies of individual community-based interventions, but these studies were not captured by the scope of our search strategy. This limits comparability across intervention types and we acknowledge a more comprehensive review would enhance robustness of the findings. Given the limited scope, our findings should be interpreted as an indicator of the promise of the interventions reviewed, rather than a definitive view on whether or not they are effective.

• Reliance on RCTs and experimental designs: Many of the included systematic reviews and meta-analyses focused solely on randomised controlled trials (RCTs) or other experimental designs. This is understandable given the focus on causal inference, but presents challenges for evaluating certain interventions (e.g. community-based or active travel schemes) where experimental designs are less feasible due to logistical or ethical constraints. As a result, some potentially impactful interventions may be under-represented in this review and a more nuanced interpretation of different types of evidence was not possible for this review.

• Scope of outcomes assessed: This review focused solely on physical activity outcomes and so both wider benefits (such as health or service use outcomes) and more proximal benefits (e.g. changes in attitudes towards physical activity) were not captured in this review. In addition, the evidence included in this review primarily reported on short-term or intermediate outcomes. Few studies assessed whether behaviour change was sustained over time. As such, we are unable to draw conclusions about the long-term impact of most interventions.

Recommendations

To support future decision-making and policy design, we suggest the following overarching recommendations:

Recommendation 1

Strengthen the evidence base through targeted, robust evaluations. We recommend co-funding and co-designing high-quality evaluations of priority interventions, working closely with delivery partners (e.g. Sport England, Active Travel England, Active Partnerships) and local authorities. These evaluations should aim to answer critical questions about what works, for whom, how and at what cost. Where possible, evaluations should:

• use robust study designs, such as cluster randomised trials, stepped-wedge designs, or quasi-experimental natural experiments where appropriate to help establish causal claims about impact;
• incorporate formal theory-based evaluation methods such as contribution analysis to help understand how interventions achieve impact and the relative contribution of each element;
• include objective measurement of physical activity, using validated tools such as accelerometers;
• collect longitudinal data to assess sustained behaviour change and longer-term health outcomes;
• incorporate equity-focused analysis, examining how impacts vary by socioeconomic status, ethnicity, gender, age, disability, and baseline physical activity levels;
• include cost-effectiveness assessments to inform decisions on scaling, replication, and prioritisation between intervention types.

Given the range of active programmes already operating across schools, communities, and urban environments, there is a valuable opportunity to embed evaluation into existing delivery structures and ensure future investment is evidence-led.

Recommendation 2

Prioritise piloting delivery models with the potential for population-level impact. Interventions such as peer-led programmes, teacher-delivered PALs and improvements to cycling and walking infrastructure demonstrate potential to improve physical activity. We recommend prioritising pilots of models that:

• leverage existing community assets (e.g. peer networks, schools, local clubs) to enable broad reach;
• include structured training and support for implementers to ensure quality delivery;
• balance intensity and feasibility by considering initial intensive phases followed by lighter-touch maintenance strategies;
• can be adapted to different local contexts to maximise acceptability, uptake and impact.

Piloting models in a variety of settings, with strong evaluation, will help identify those that can deliver impact at scale.

We have also included the following intervention-specific recommendations:

School-based interventions

1. Consider developing evidence-based guidance for schools and trainee teachers regarding ideas to improve pupils’ physical activity: this could include guidance on implementing or piloting the most promising interventions from this review (e.g. PALs, class-based active recess) as well as more system-wide factors that may influence activity levels (such as school leadership and governance that supports physical activity).

2. Prioritise pilots of promising interventions within schools in areas of lower activity: While schools provide a universal platform, there is limited evidence of impact on reducing inequalities. Prioritising intervention rollout in less active and more disadvantaged areas, as well as tailoring activities to groups who are less active (e.g., girls), could help address participation gaps.

3. Ensure future evaluations include long-term follow-up and assess compensatory behaviours: Many studies show compensatory reductions in activity outside of intervention time. Future evaluations should assess impacts across the whole day and over longer timeframes to capture sustained changes and avoid overestimating short-term benefits.

Community based peer-led interventions

1. Consider UK-based pilots of promising peer-led physical activity programmes, with structured peer leader training: Peer-led interventions show consistently promising results, especially when peer leaders are properly trained and interventions start with an intensive engagement phase. Piloting and potentially scaling these interventions, particularly in underserved communities, could help achieve meaningful behaviour change, noting that interventions will need to be tailored to the local context, rather than taking a one-size-fits-all approach.

2. Prioritise group-based delivery models for peer-led and community programmes: Group formats have demonstrated better outcomes, particularly for disadvantaged women and lower-SES populations, and are likely to be more cost-effective than one-to-one models. Supporting group delivery models can strengthen social support, motivation, and retention.

3. Invest in evaluating promising UK initiatives: Existing large-scale community-based initiatives such as Parkrun, Go! London and Sport England’s Place Partnerships provide promising opportunities to generate evidence about which community-based interventions work as well as how they work and for whom. Our recommendation that future evaluations include a theory-based component is particularly important for these types of initiatives as they are often tailored to their local context. Considering how they work is therefore important for understanding whether and how they are likely to work elsewhere.

Active travel interventions

1. Target infrastructure investments in high-deprivation areas: Infrastructure changes often benefit those living nearby, who are typically more advantaged. Where infrastructure changes are being considered, prioritising lower-income neighbourhoods may improve equitable access and reduce disparities in active travel uptake, so long as these investments are tailored to address barriers to active travel that are relevant to the local community.

2. Ensure interventions address safety and convenience to encourage uptake. Perceptions of safety and convenience are critical drivers of active travel, especially for groups like women and older adults. Designing routes and environments that feel safe and are well-integrated into transport systems is essential.

References

Baker, P. R., Francis, D. P., Soares, J., Weightman, A. L., & Foster, C. (2015). Community-wide interventions for increasing physical activity. Cochrane Database of Systematic Reviews, 2015(1), CD008366.

Best, K. L., Miller, W. C., Eng, J. J., & Routhier, F. (2016). Systematic review and meta-analysis of peer-led self-management programs for increasing physical activity. International Journal of Behavioral Medicine, 23, 527–538.

Buman, M. P., Giacobbi, P. R., Dzierzewski, J. M., Aiken-Morgan, A. T., McCrae, C. S., Roberts, B. L., Marsiske, M., & King, A. C. (2011). Peer volunteers improve long-term maintenance of physical activity with older adults: A randomized controlled trial. Journal of Physical Activity and Health, 8(Suppl. 2), S257–S266.

Cleland, C. L., Tully, M. A., Kee, F., & Cupples, M. E. (2012). The effectiveness of physical activity interventions in socio-economically disadvantaged communities: a systematic review. Preventive medicine, 54(6), 371-380.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Routledge.

Craike, M., Wiesner, G., Hilland, T. A., & Bengoechea, E. G. (2018). Interventions to improve physical activity among socioeconomically disadvantaged groups: An umbrella review. International Journal of Behavioral Nutrition and Physical Activity, 15, 43.

Hulteen, R. M., Waldhauser, K. J., & Beauchamp, M. R. (2019). Promoting health-enhancing physical activity: A state-of-the-art review of peer-delivered interventions. Current Obesity Reports, 8(4), 341–353.

Jones, M., Defever, E., Letsinger, A., Steele, J., & Mackintosh, K. A. (2020). A mixed-studies systematic review and meta-analysis of school-based interventions to promote physical activity and/or reduce sedentary time in children. Journal of Sport and Health Science, 9(1), 3–17.

Joyce, S., Modini, M., Christensen, H., Mykletun, A., Bryant, R., Mitchell, P.B. and Harvey, S.B., 2016. Workplace interventions for common mental disorders: a systematic meta-review. Psychological medicine, 46(4), pp.683-697.

Larouche, R., Saunders, T. J., Faulkner, G., Colley, R., & Tremblay, M. S. (2018). Effectiveness of active school transport interventions: A systematic review and update. BMC Public Health, 18, 206.

Love, R., Adams, J., & van Sluijs, E. M. (2019). Are school-based physical activity interventions effective and equitable? A meta-analysis of cluster randomized controlled trials with accelerometer-assessed activity. Obesity Reviews, 20(6), 859–870.

Mears, R., & Jago, R. (2016). Effectiveness of after-school interventions at increasing moderate-to-vigorous physical activity levels in 5- to 18-year-olds: A systematic review and meta-analysis. British Journal of Sports Medicine, 50(21), 1315–1324.

Neil-Sztramko, S. E., Caldwell, H., & Dobbins, M. (2021). School-based physical activity programs for promoting physical activity and fitness in children and adolescents aged 6 to 18. Cochrane Database of Systematic Reviews, 9, CD010651.

Norris, E., van Steen, T., Direito, A., & Stamatakis, E. (2020). Physically active lessons in schools and their impact on physical activity, educational, health and cognition outcomes: A systematic review and meta-analysis. British Journal of Sports Medicine, 54(14), 826–838.

Panter, J., Heinen, E., Mackett, R., Ogilvie, D., & the iConnect Consortium. (2019). Can changing the physical environment promote walking and cycling? A systematic review of what works and how. Health & Place, 58, 102161.

Phillips, G., Bottomley, C., Schmidt, E., et al. (2014). Well London Phase-1: results among adults of a cluster-randomised trial of a community engagement approach to improving health behaviours and mental well-being in deprived inner-city neighbourhoods. Journal of Epidemiology and Community Health, 68(7), 606–614.

Porter, A., Walker, R., House, D., Salway, R., Dawson, S., Ijaz, S., de Vocht, F., & Jago, R. (2024). Physical activity interventions in European primary schools: A scoping review to create a framework for the design of tailored interventions in European countries. Frontiers in Public Health, 12, 1321167.

Prince, S. A., Adamo, K. B., Hamel, M. E., Hardt, J., Gorber, S. C., & Tremblay, M. (2008). A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. International journal of behavioral nutrition and physical activity, 5, 1-24.

Salway, R., Foster, C., de Vocht, F., Tibbitts, B., Emm-Collison, L., House, D., … & Jago, R. (2022). Accelerometer-measured physical activity and sedentary time among children and their parents in the UK before and after COVID-19 lockdowns: a natural experiment. International Journal of Behavioral Nutrition and Physical Activity, 19(1), 51.

Smith, M., Hosking, J., Woodward, A., Witten, K., MacMillan, A., Field, A., Baas, P., & Mackie, H. (2017). Systematic literature review of built environment effects on physical activity and active transport: An update and new findings on health equity. International Journal of Behavioral Nutrition and Physical Activity, 14, 158.

Solomon, E., Rees, T., Ukoumunne, O. C., Metcalf, B., & Hillsdon, M. (2014). The Devon Active Villages Evaluation (DAVE) trial of a community-level physical activity intervention in rural south-west England: a stepped wedge cluster randomised controlled trial. International Journal of Behavioral Nutrition and Physical Activity, 11, 1–12.

Stappers, N. E. H., van Kann, D. H. H., Ettema, D., de Vries, N. K., & Kremers, S. P. J. (2018). The effect of infrastructural changes in the built environment on physical activity, active transportation and sedentary behavior: A systematic review. Health & Place, 53, 135–149.

Xiao, C., van Sluijs, E., Ogilvie, D., Patterson, R., & Panter, J. (2022). Shifting towards healthier transport: Carrots or sticks? Systematic review and meta-analysis of population-level interventions. The Lancet Planetary Health, 6(11), e858–e869.

Appendix 1: Review methodology

Literature search

We conducted a rapid evidence review to summarise insights from high-quality systematic reviews and meta-analyses. Our goal was to identify up to five studies per intervention type—school-based, community-based/peer-led, and active travel interventions—resulting in a total of up to 15 studies for inclusion.

We applied a structured and transparent search process, including:

• systematic searches of academic databases (primarily PubMed, Cochrane Database of Systematic Reviews, and Google Scholar) using pre-agreed search terms;
• evidence and resources shared by DCMS and other UK government departments;
• AI tools (e.g. ChatGPT, Claude, Consensus, Elicit) to support evidence identification and review.

Through this process, we identified systematic reviews and meta-analyses that synthesised findings from over 100 primary studies—including randomised controlled trials (RCTs), quasi-experimental studies, and other research designs. The selected reviews contributed with complementary perspectives, even within the same broad intervention type. For example, we had reviews covering subtypes of interventions, such as PALs in school settings and active school transport in active traveling interventions. Sometimes, two comprehensive reviews provided different perspectives and frameworks to analyse the results (e.g. the carrot-stick framework used by Xiao et al., 2022 in the active travel section).

Screening

The review followed a structured screening and assessment process, similar to that of a systematic review. Titles and abstracts were first screened, followed by full-text reviews to confirm eligibility. We prioritised studies aligned with the interventions listed in the Intervention Longlist, such as increasing physical activity time during the school day or providing free after-school recreation programmes.

Each study was assessed against the evidence standards set out by Puttick and Ludlow (2013), taking into account the quality of the review, the consistency of findings, and the rigour of the underlying studies.

The screening process followed the pre-agreed inclusion and exclusion criteria:

Category	Criteria
Population	Include: The review will prioritise evidence among the population who are least active, either defined through their baseline physical activity levels (e.g. <30mins moderate physical activity, MPA, per week for adults) or through their association with less active demographics in the population, including: those with disabilities those with long-term health conditions lower socio-economic groups, women and girls, and ethnic minorities. Exclude: policies and interventions solely targeting very physically active populations, either defined through baseline physical activity levels (e.g. >=150mins MPA/week for adults), or through their association with more active demographics in the population, including: Athletes and sportspeople
Policies and interventions	Include: policies and interventions that target the promotion of physical activity
Control groups	N/A
Outcomes	Include: The main outcome of interest is physical activity. The review will include studies which use a range of physical activity-related measurements (e.g. steps, moderate to vigorous physical activity), but we will prioritise aerobic activity over muscle-strengthening and balance-based activities unless the latter two activities are of clear importance to the target population (e.g. balance exercises for older adults). Exclude: Other health-related outcomes (e.g. obesity, cardiovascular diseases) and/or wider outcomes (e.g. education, employment, crime) will be out of scope for this review.
Settings	Include: Any settings
Study design	Include: Systematic reviews and meta-analyses of intervention studies If no systematic reviews / meta-analyses are available, we will look for the next best available evidence (e.g. narrative reviews, experimental or quasi-experimental evidence, robust theory-based evaluations).
Others	Include: The review will prioritise evidence focused on the UK, or studies in the whole or parts of England, Scotland, Wales and/or Northern Ireland. Evidence from comparable developed countries will also be included, especially in cases where UK evidence is not available. Literature that are available in full text Literature must be published in English Literature published in the last 10 years will be prioritised. We may extend this to include studies up to last 20 years if no evidence is available within this timeframe. Policies and interventions implemented between March 2020 and late 2021 will be reviewed with caution due to the impact of the COVID-19 pandemic and the limited generalisability of findings from this period.

Data extraction, assessment, and synthesis

Data were recorded in a spreadsheet and included the following list of variables (where available):

• Title and authors
• Year of publication
• Type of publication (e.g. systematic review, RCT, grey literature)
• Evidence standard assessment (see Figure 1 below for the Nesta Standard of Evidence framework)
• Country of research
• Brief description of the intervention (e.g. format, target audience, duration)
• Outcome measure
• Key findings on intervention effectiveness, impact on health inequalities, and other relevant findings

1. The priority groups in respect of inequalities in physical activity are contained in Sport England’s inequality metric. ↩

2. Where possible, these categories reflect standard interpretations of effect sizes represented by the standardised mean difference, SMD (i.e. 0.2 = small effect, 0.5 = medium, and 0.8 = large). However, as outcomes varied significantly between different studies and effect sizes were not consistently reported, we were not able to report standardised mean differences for all interventions across a consistent outcome measure. These estimates are therefore highly uncertain and should only be used as a rough indication of impact based on the evidence reviewed in this report, rather than a robust estimate of the effect size across a consistent measure of physical activity. Note also that even a small effect can be considered meaningful in this context as studies of physical activity interventions do not typically find large effect sizes. ↩

3. Note that as none of the reviewed studies reported evidence of cost-effectiveness, these are highly uncertain estimates relying on significant assumptions about both implementation, delivery and maintenance costs. They are not a robust cost assessment based on underlying evidence and therefore should not be treated as such. ↩

4. Note that this scoping paper was included at DCMS’s suggestion, but as it is not a systematic review or meta-analysis we have summarised the details separately in the table above. ↩

5. The standardised mean difference (SMD) is a measure of effect size that expresses the difference between two group means in standard deviation units, allowing for comparison across studies using different outcome measures. As a general rule of thumb, an SMD of 0.2 is considered a small effect, 0.5 a medium effect, and 0.8 or above a large effect (Cohen, 1988). ↩

6. Based on data from one study of UK children (Salway et al., 2022), we estimate that for the average child, a SMD of 0.32 would equate to roughly 18 minutes of additional daily MVPA. ↩

7. The authors report this result as 171 metabolic equivalent minutes per week (MET minutes/week), which equates to 43 minutes of moderate-intensity physical activity. ↩

8. Systematic review evidence from Prince et al. (2008) suggests that there is no clear overall correlation between self-reported and objectively measured physical activity. However, they also find that self-report measures generally reported higher physical activity levels than those directly measured by accelerometers, so self-report measures should be interpreted with caution. ↩

9. Based on data from the Health Survey for England (2021) which suggests that the average English adult engages in 396.1 minutes of non-occupational moderate-intensity physical activity per week with a standard error of 9.41, a SMD of 0.4 would equate to approximately 288 minutes of additional moderate-intensity physical activity per week, or 41 minutes per day. ↩

10. It is not clear to us why these interventions might cause decreases in physical activity and the authors do not provide an explanation. Major infrastructure changes might disrupt people’s existing travel routines, leading to temporary decreases in PA before they adjust, but this is highly uncertain. ↩

11. For example, see survey data cited by Sustrans ↩