Using sport and physical activity interventions to develop life skills and reduce delinquency in youth: a systematic review

ABSTRACT

Purpose

Disadvantaged youth are confronted with a disproportionate number of risk factors and stressors that can predict youth delinquency and negative behaviours as an adult. Life skills developed through physical activity and sport programmes (PAS) have been identified as a way to provide protective factors for those at risk of, or involved in, youth delinquency.

Method

A systematic search of nine electronic databases was conducted to collate and evidence the impact of life skills developed through sport and physical activity-based programmes targeting youth at risk of or involved in youth delinquency.

Results

The search identified 15 studies across three different intervention levels (primary, secondary, and tertiary), reporting the effects of life skills development on youth at risk of, or involved in youth delinquency. The current evidence base for understanding the development of life skills through PAS programmes is positive but sparse.

Conclusion

Recommendations have been made for future research and include understanding the impact on personal and social identity, exploring the use of gender specific strategies and investigating the effectiveness of delivery staff.

KEYWORDS:

• Adverse childhood experiences
• positive youth development
• sport-based interventions
• sport for development
• life skills

Introduction

Between April 2019 and March 2020 over 19,000 children aged 10–17 were cautioned or sentenced in England and Wales, with 11,100 first time entrants into the youth justice system (Ministry of Justice, 2021). In the same year, 38.5% of those in the youth justice system reoffended with most recent data estimating that reoffending in 2017 cost the county £1.5 billion (Ministry of Justice Analytical Series, 2019). Despite the global pandemic influencing the reduction in youth delinquency (Ministry of Justice, 2022) since the ‘stay at home’ order ended in March 2021 youth offending has increased including recorded rises of 29% for robberies (Youth Endowment Fund, 2022). Youth delinquency is a complex, multifaceted issue affected by the individual, household environment and community (Spruit et al., 2018a). In 2018, the Government’s endorsement in establishing Violence Reduction Units (VRUs) was a positive step in the continuous hardship in tackling youth crime by developing and commissioning local level initiatives. Recommendations have been suggested by The Youth Violence Commission (2020), particularly regarding life skills and sports and leisure. This systematic review provides a critical review into the effects of sports and physical activity based positive youth development programmes in youth at risk of, or involved in, youth delinquency. Youth delinquency refers to a young person who has committed a criminal offence, varying between the ages of 6 and 18 (Young et al., 2017). This paper will summarise and evaluate the impact of life skills developed through physical activity and sport (PAS) and whether these programmes have the ability to facilitate protective factors and mitigate against risk factors linked to youth delinquency.

Adverse Childhood Experiences (ACEs) is an umbrella term referring to the abuse, neglect, and trauma experienced by young people under the age of 18. Identified as a prominent risk factor, ACEs are associated with a wide range of negative life outcomes including higher risk of suicide, cigarette smoking, obesity in adulthood and delinquency (Wolff et al., 2018). The founding ACEs study identified several risk factors including physical, emotional and sexual abuse, neglect, and specific household dysfunction including adulthood mental illness, domestic violence, divorce, having an incarcerated relative and substance misuse (Felitti et al., 1998). Research has highlighted the impact of experiencing these risk factors on young people and their ability to compromise the life expectancy and mental health outcomes throughout adulthood (Merrick et al., 2017). Research indicates that predictions for negative behaviours in adulthood can be identified for those experiencing 4 or more ACEs as a child (Table 1).

Table 1. Predictors of behaviour for those that have experienced 4 or more Adverse Childhood Experiences.

	Compared with people with no ACEs, those with 4+ ACEs are:
4x	More likely to be a high-risk drinker
6x	More likely to have had or caused unintended teenage pregnancy
6x	More likely to smoke e-cigarettes or tobacco
6x	More likely to have had sex under the age of 16
11x	More likely to have smoked cannabis
14x	More likely to have been a victim of violence over the last 12 months
15x	More likely to have committed violence against another person in the last 12 months
16x	More likely to have used crack cocaine or heroine
20x	More likely to have been incarcerated at any point in their lifetime

Disadvantaged youth are confronted with a disproportionate number of risk factors and stressors (Newman, 2020). The likelihood of encountering traumatic experiences (ACES) is higher for those living in disadvantaged communities with income poverty, low quality housing and peers being involved in problem behaviour (Baglivio et al., 2015; Feinberg et al., 2010).

Recently, in the field of criminal justice, in depth understanding has been sought regarding the link between ACEs and youth delinquency. Young people with a higher number of ACEs are more likely to have early-onset offending trajectories (Baglivio et al., 2015), violent offending and be classified as a serious and chronic offender by the age of 18 (Baglivio et al., 2015; Fox et al., 2015; Perez et al., 2018). Research suggests that young people with 4 or more ACEs are 20 times more likely to be incarcerated in their lifetime (Felitti et al., 1998). Young people with an increased likelihood of offending are identified as being ‘at-risk’ or ‘disaffected’ (Lubans et al., 2012). The terms ‘at risk’ has many definitions; however, for the context of this article this term refers to youth who have compromised developmental outcomes who are then vulnerable to poverty, homelessness, and risk-taking behaviours. These culminate in a range of negative outcomes related to unemployment, educational dropout, and negative health behaviours (Kirk & Sampson, 2013; Merrick et al., 2017; Armour & Sandford, 2013) that place young people at a greater vulnerability of problem behaviour.

As a result of the associated costs of youth delinquency, academics and organisations are paying particular interest to the specific activities and interventions that can support ACEs intervention initiatives (Kagi & Regala, 2012). Moving away from the traditional deficit reduction viewpoint of developmental psychology (Johnston et al., 2013), whereby we aim to react to a problem (e.g. obesity, reoffending), there is now an increased focus on preventative measures and early years support for young people at-risk, such as those at risk or living with ACEs (Fraser-Thomas et al., 2005). Although young people with 4 or more ACEs have an increased likelihood of displaying the risky behaviours outlined in Table 1, specific factors can moderate the impact of these behaviours. This provides a focus on interventions that optimize intraindividual (within-the-person) and interindividual (between-people) strengths in a young person to reduce the impact of risk factors, known as protective factors (Craig et al., 2017; Lerner et al., 2019). Protective factors are categorised across 4 dimensions (Craig et al., 2017); Family (e.g. good parental supervision, high family income, parental attitudes against anti-social behaviour), School (e.g. strong commitment to school), Peers (e.g. having peers that do not hold anti-social attitudes) and Community (e.g. strong social support, positive attitude to authority). Protective factors can include the development and management of emotional regulation and processing, resilience, and problem-solving techniques (Cameron & Jago, 2008; Fredrickson et al., 2003; Blanchard-Fields, 2007). Research from several disciplines has explored different ways to promote positive human development among diverse adolescents and in turn foster social justice through opportunities for individuals to enhance their healthy development (Lerner & Overton, 2008). Through public health influence, intervention programmes are defined through three types of prevention methods. Primary interventions focus on intervening with a general population before the effects of criminality occur, usually in areas of high deprivation or prevalence of criminality (Welsh et al., 2014). Whereas secondary interventions target young people at enhanced risk of criminality (e.g. educational dropout, aggressive behaviour) and tertiary interventions target young people where it’s clear offending has occurred, such as serious youth violence (Welsh et al., 2014).

An emerging area of study, Positive Youth Development ([PYD] Lerner et al., 2021), provides a framework for a strength-based approach to foster the assets of young people to avoid complex behaviour and manage life challenges (Dzewaltowski & Rosenkranz, 2014). Positive Youth Development (PYD) can contribute to the promotion of social justice by working towards the equitable treatment and fair allocation of resources for healthy development, regardless of age, gender, or socioeconomic status (Barbarin et al., 2020). One intervention tool used to support youth at risk of involvement in delinquency is sports-based youth development programmes (Spruit et al., 2018b). The emerging field of sport-based youth development (SBYD) utilises frameworks to develop (Dzewaltowski & Rosenkranz, 2014) social identity (Bruner et al., 2017), decision-making (Crean, 2012), emotional development (Cronin & Allen, 2018) and personal and social responsibility (Walsh et al., 2010) by enhancing the development of life skills (Bean & Forneris, 2017; Fraser-Thomas et al., 2005). Life skills have been defined as internal personal assets, characteristics, and skills that can be facilitated or developed in sport and transferred for use in non-sporting settings (Gould et al., 2013, p. 259). It has been widely accepted in research and practice that enhancing life skills through youth PAS can be used to enhance youth development (Wright & Burton, 2008; Weiss et al., 2012). Similarly, at-risk youth participating in sporting programmes has been linked to lower rates of school dropout, criminal activity (Mahoney et al., 2000), career exploration (Walsh et al., 2010) and youth leadership (Martinek et al., 2006). This indicates that PAS can support young people to mitigate harm caused by ACEs and prevent further harm associated with offending behaviour through the facilitation of life skill development.

With the growing body of literature around sports based life skill developmental outcomes for youth, there is still a need to understand how PAS programmes can facilitate, or not, the development of youth at risk in ways that prevent offending behaviours (Whitley et al., 2018). Simply participating in sport is thought unlikely to secure a positive outcome and therefore more understanding of the enablers and attributing outcomes must be sought (Coakley, 2011; Nichols & Crow, 2004). Academics have focused on understanding the design, structure, and management of sport for development with emphasis on the intentional process of delivery and not just the programme outcomes (Bruening et al., 2015; Whitley et al., 2019). When developing an intervention, one of the key variables is intentionality, defined as the deliberate act to create and facilitate developmental outcomes (Cronin & Allen, 2018). Intentionality can be best explained when understanding the implicit and explicit transfer of life skills. Implicit life skills are seen in programmes focusing solely on sport specific skills such as competition, teamwork and leadership that can influence the transfer of life skills into other life domains (Camire & Kendellon, 2016). The explicit approach systematically delivers intentional life skills whereby deliverers connect skills learnt through PAS (e.g. reflection, imagery, and leadership) to everyday life (Jacobs et al., 2018). Sport for Development theory (Lyras & Welty Peachey, 2011) also suggests that interventions should include 5 components (1) impact – assessing the effectiveness, (2) organisational – collaboration between all stakeholders, (3) sport and physical activity – inclusive and high-quality play, (4) educational – including educational programming such as life skills to support transfer of skills into the real world, (5) cultural enrichment – to facilitate long term social capital development (Lyras & Welty Peachey, 2011). However, sports based positive youth development literature has been inconsistent in examining the effectiveness of sport to promote positive change. The absence of scientific evidence overlooking the complexities of social challenges, particularly in young people at risk of delinquency, highlights significant gaps between theory and practice (Lyras & Welty Peachey, 2011). Moreover, few studies have focused on programme characteristics that positively influence outcomes (Lee et al., 2017). Previous systematic reviews have all been a necessary progression in evaluating the current studies available. Three notable reviews have helped shape general understanding of the effects of sport and physical activity. Lerner and Overton (2012) reviews the impact of physical activity programmes on social and emotional wellbeing in at-risk youth. Spruit et al. (2016) provides a meta-analysis on sports participation and its relationship with youth delinquency. Lastly, Hermens et al. (2017) provides an overview of life skill development in various sports programmes that serve socially vulnerable youth. However, the last similar systematic review was run in 2015 (Hermens et al., 2017) yet significant organisations have published reports suggesting the positive impact that sport and activity can play in preventing and reducing youth offending (DCMS, 2019; The Youth Violence Commission, 2020; Youth Endowment Fund, 2022). No recent review has centred their focus on understanding sport and activity that includes intentional life skill facilitation for young people aged 10–17 recruited due to their risk of, or involvement in youth delinquency.

Study aim

There has been an increase in attention on PAS based positive youth development programmes providing protective factors, in the form of life skills, for young people at-risk of or involved in youth delinquency. However, no recent review exists in investigating sport and physical activity based life skill development programmes serving those at risk of, or involved in, youth delinquency. For the purpose of this current research, we adopt the notion that life skills should be intentionally taught and should require as much focus as PAS development (Jacobs et al., 2018). The aim of this present systematic review was to collate and evaluate the impact of intentional life skill developed through sport and physical activity based programmes targeting youth at risk of or involved in youth delinquency. An additional outcome of this review was to describe what is known about conducive conditions from studies on sports and physical activity based PYD. This will provide greater understanding of the similarities in mechanisms used through intentional life skill development and their influences on the prevention and reduction of youth delinquency to support replication in future interventions. Therefore, this review will address the following questions:

What is the evidence regarding the life skills developed through sports based positive youth development programmes and their ability to facilitate protective factors and mitigate against risk factors in youth at risk of, or involved in youth delinquency?

What is known about the favourable conditions of PAS based life skills programmes in the included quantitative and qualitative studies?

By addressing these two research questions, the review consolidates the evidence base, helping policymakers, practitioners, and sporting bodies to develop and implement PAS programmes, targeting youth at risk. It will also present the opportunity to understand the favourable conditions that can strengthen protective factors in young people at risk of or involved in youth delinquency.

Methodology

This review was reported using the updated Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021).

Eligibility criteria

Studies were excluded if they did not present primary data, were not available in English and were published before 1997. Although the systematic review includes studies outside of the United Kingdom, the review focuses on research from 1997 due to key changes in legislation, based on the development of The Crime and Disorder Act 1998. Transformation of youth justice was influenced by a stronger focus on prevention of offending and reoffending through support with education, employment, and health and there was an expectation that a number of studies would originate within the United Kingdom. Study characteristics were defined by PICOS (Population, Intervention, Comparator, Outcome, Study design) to identify relevant sources (Page et al., 2021).

P – Population

The study population included young people aged 10–17 of any sex and ethnicity, who were at risk of youth delinquency. Criminal responsibility begins at 10 and those under 10 cannot be arrested or charged with a crime. Similarly, those 18 and over become part of the adult criminal justice system Therefore young people under 10 and over 17 years old have been excluded. Populations not specifically named as at-risk, socially vulnerable due to their high crime communities or displaying risk-taking behaviour were excluded as the focus remained on those specifically targeted not just for their vulnerability but their risk of problem or offending behaviours.

I – Intervention

Studies were included if a sport or physical activity programme was used to intentionally teach or develop life skills. Sport or physical activity could range between light to vigorous intensity and be of any duration or method (e.g. walking, football, multi-sports). As PYD encompasses a broad range of developmental assets this review focuses on only life skills and the interpersonal and intrapersonal competencies developed through intentional delivery or facilitation through PAS. Any study delivered in school curricular time was excluded as this review focused on programme components related to community sports provision – this is usually voluntary participation, delivered in non-curricular time and open to the community. This style of delivery is one that is funded widely within the UK (Youth Endowment Fund, 2022).

C – Comparison

This review did not restrict inclusion of studies by using a comparison between populations.

O – Outcomes

Primary outcomes included the prevention or reduction in youth delinquency or youth offending and the development of life skills defined as internal personal assets, characteristics, and skills. Examples of this can include critical thinking and decision-making, social identity, emotional development, and personal and social responsibility. Self-esteem outcomes have been included, despite acknowledgement of its links to wider personal development assets, as an indicator of a young person’s ability to cope with emotions and stress (Moulier et al., 2019).

S – Study design

A primary purpose of this review included identifying the different theory base, programme content and instruments used to deliver life skills or positive youth development through sport and physical activity. Therefore, all study designs were included in this review.

Information sources

This systematic review was performed during May 2022 to identify relevant studies, a search strategy was developed through discussions between the authors. A professional librarian developed individualised search strategies for nine different databases (SAGE Journals, WILEY, ProQuest, SportDiscus, PsycINFO, PsychARTICLES, Scorpus, Web of Science and Cochrane Library). Once all appropriate articles were identified a forward-backwards search was conducted for further studies.

Search strategy

A combination of terms relating to (a) youth OR adolescent (b) sport OR physical activity (c) life skills OR positive youth development (d) juvenile delinquency OR youth offending. The full search strategy is available in the Appendix.

Study selection

Records were imported into EndNote X9 software and all duplicates were removed. The lead author screened the titles and abstracts of all unique studies for relevance and assessed for possible inclusion (see Figure 1). Any studies needing further deliberation were brought forward to the research team and screened for eligibility, discussed, and resolved with the lead author.

Figure 1. PRISMA flowchart of the review process.

Data collection process

The lead author independently extracted and inserted data into Excel using a predefined form. The form included study title, authors, publication date, country of study, participant characteristics (age, sex, ethnicity, and at-risk status) and study design characteristics. It was noted whether a control group was used. Outcome data recorded main outcomes and measurements used and relevant information on the intervention was recorded such as procedures, delivery style and type and length of intervention. The type of sport or physical activity and theoretical frameworks were recorded.

Risk of bias in individual studies

To address the rigour of studies, full texts were critically appraised by using the Mixed Methods Appraisal Tool (MMAT) for the remaining studies included (Hong et al., 2018). The MMAT was chosen to develop the appraisal because it provides quality criteria for quantitative, qualitative, and mixed methods studies (Hong et al., 2018). Criteria included appraisals on: appropriate approach to answer the research question; appropriate data collection for the research question; non-response bias; complete outcome data; inconsistencies between quantitative and qualitative data; confounders accounted for and appropriate statistical analysis. To assess the rigour of the studies, the first author ranked the quality of each study from 0% to 100% by answering ‘yes’, ‘no’ or ‘can’t tell’ to a series of questions to identify whether the study satisfied the appropriate criteria or not. The critical appraisal of each study was used to inform on the quality of studies however, the decision was made not to exclude studies with low methodological quality due to the limited number of studies available from the systematic search. Studies with low rigour can provide valuable research, although should be interpreted with caution (Hong et al., 2018).

Synthesis of results

The impact of life skills delivered through sport and physical activity programmes, with any available measurement, for youth at risk of or involved in youth delinquency was reported. A meta-analysis was not possible due to the disparate studies and study designs and therefore a narrative synthesis was performed.

Results

A total of 3,776 titles and abstracts were initially identified from the search criteria, with an additional 33 included from hand searching. After removing 92 duplicates, the first author screened the titles and abstracts of 3,684 unique studies for relevance and assessed for possible inclusion (see Figure 1). Seventy studies remained for full-text assessment after this screening process where a further 55 studies in full were excluded based on the inclusion criteria. A total of 15 studies met the inclusion criteria and were included in the analysis.

Overview of studies

The systematic review identified 15 studies that evaluated the effects of life skills taught through sports-based youth development programmes, based on the three key levels of intervention to prevent or reduce youth delinquency (Table 2). This includes primary, secondary or tertiary interventions. 10 studies delivered primary prevention programmes, working with the general population in communities with a number of environmental risk factors increasing the likelihood of youth delinquency (Armour & Sandford, 2013; Fraser-Thomas et al., 2005; Fuller et al., 2013; Gano-Overway et al., 2009; Goodman et al., 2021; Khoury-Kassabri & Schneider, 2018; Nols et al., 2017; Riley & Anderson-Butcher, 2012; Ullrich-French et al., 2012; Walsh et al., 2010). 2 studies delivered secondary prevention programmes, working with targeted young people displaying risk-taking behaviour and therefore at higher risk of youth delinquency (Bonnette et al., 2001; Spruit et al., 2018a) and 3 studies delivered tertiary prevention programmes, working with young people involved in youth delinquency (Bloemhoff, 2006; D’Andrea et al., 2013; Wu & Hsieh, 2006). Programmes were conducted in 7 different countries, mostly in the United States (n = 9) with settings including sports clubs (n = 4), summer camps (n = 3), after school and community (n = 5), residential settings (n = 2), and outdoor adventure programmes (n = 1). A total of 2,789 participants were included across the studies (male 66.9%) with 4 male focused interventions and 1 female focused intervention.

Table 2. Sports and physical activity programmes to prevent or reduce youth delinquency.

	Primary Interventions
Study	Sample	Theory Base	Programme Content
Armour and Sandford (2013) (UK)	310 boys and 285 girls aged 11–16. Five secondary schools in London characterised by high levels of social deprivation. All took part in an Outward Bound	Positive Youth Development	5-year programme with 4 cohorts to reengage individuals. 150 young people per cohort from multiple schools. Organised programme of activities across the whole year with one week residential for all individuals and one weekend residential for 50 individuals. Volunteers trained by youth workers.
Frazier, Dinizulu, Rusch, Boustani, Mehta, Reitz (2015) (USA)	46 high-risk male and females living in areas with high rate of violence ave. age 13yrs. 7 park staff. After school programme.	Risk and Resilience	3 parks located in high-risk areas (e.g. risky behaviour, gang recruitment) in USA. 10 weeks, 2x week for 90mins. Sessions included didactic instruction, skills demonstration and discussion, role play and sport to provide practice with feedback. Supported by mental health practitioner.
Fuller et al. (2013) (USA)	8 Males aged 10–13, minority ethnicity living in high deprivation area. After school programme	Positive Youth Development (5C’s)	Two 2-hour sessions x 24 weeks. Sport and/or Physical activity sessions, life skills lesson and nutrition lesson. Low ropes course on last session. EDC members of staff.
Gano-Overway et al. (2009) (USA)	395 male and female aged 9–16 M = 11.8, 90% high risk young people from three national youth sports programmes.	Perceived Caring Climate	Free 5-week summer programme. Minimum of 50hrs instructional sports/physical activity and health education. 4hr of sport/PA and 1 hour health education each day with lunch. Staff encouraged to develop caring relationships and character development.
Goodman et al. (2021) (USA)	198 male and female aged 9–15 40.5% female 91.4% Black/African American/ Hispanic. 76.3% low-income. 8 Miami neighbourhoods with high rates of violence. After School programme in 9 parks.	Risk and Resilience (Masten, 2001).	28 weeks. 5 days a week, 3 hours per session. After school programme using sport and recreation. Aimed to facilitate development in communication, emotional literacy and problem solving.
Khoury-Kassabri and Schneider (2018) (Israel)	126 Jewish males aged 13–18 at risk, sports programme.	Positive Youth Development	Youth from 5 sports programme incl. football, horseback riding, basketball, tennis, martial arts. No other information.
Nols et al. (2017) (Belgium)	288 T1 and 160 T2 (44% dropout), Age 12-25, Mage16.5 at risk neighbourhoods. 80 = % male. 6 Sport for Development initiatives.	Self-Efficacy and Self-Esteem	Youth participating in 6 programmes that worked with at risk, working on youth development. Open access with a targeted outreach programme. No other information.
Riley and Anderson-Butcher (2012) (USA)	10 parents from low-income families with child aged 10–12 participating in sports-based PYD summer camp for the first time. 7 female children.	Teaching Personal and Social Responsibility (TPSR)	19 d Learning in Fitness and Education (LiFE) programme. 8am-2pm, free with lunch. Transport to sessions. Designed to promote social competence, social skills, sense of belonging and connectedness. 3 sports activities and 1 educational activity developing social skills.
Ullrich-French et al. (2012) (USA)	197 M = 95 F = 102, low-income families, high risk neighbourhoods PYD summer camp.	PYD: Interpersonal Relationships	4 weeks 8am-2pm. Four physical activity sessions and 1 educational session. Free with food provided. Designed to develop life skills for academic, social and physical domains.
Walsh et al. (2010) (USA)	13 aged 9–11 11M &2F, low income and at-risk community (violent crime, drugs) all African American, After School Club.	Teaching Personal and Social Responsibility (TPSR)	The Coaching Club 2 yrs with 45 sessions and used team sports as vehicle for teaching TPSR. First 8 sessions were formative and week 9 onwards TPSR is delivered. Format incl. Relational time, awareness talk, the lesson, group meeting and reflection time.
	Secondary Interventions
Bonnette et al. (2001) (USA)	Males at-risk. aged 10–13. Treatment group n = 36, control group n = 44, from outdoor summer camp.	Critical Thinking (Cleland, 1994)	3 weeks sports summer camp of 40 min sessions. Treatment group incl. sport skills instructions + critical thinking opportunities (student-centred). Control group incl. sports skills instructions only (teacher-centred).
Spruit et al. (2018a) (Netherlands)	248 intervention group, 120 control group, 88.4% Male. At-risk for developing juvenile delinquency. Sports-based interventions at local sports clubs.	Social Bond Theory Social Vulnerability Theory Experiential Learning Theory	Local sports clubs for youth at risk, 2x a week. Incl. Specific feedback, safe, educational environment, positive relationships between coach and participants.
	Tertiary Interventions
Bloemhoff (2006) (South Africa)	Males at-risk, ave. age 16. Treatment group (N = 46), Control Group (N = 60) from educational youth care centres.	Resilience & protective factors (Jessor, 1993)	One-off programme, 4hr duration and cohorts of 15–20 per group. Participation in three high elements of a ropes course. Participants perform a task while receiving emotional support from remainder of the group.
D’Andrea et al. (2013) (USA)	Females at risk aged 12–21. Treatment group n = 62, control group n = 26 from residential treatment setting.	Positive Youth Development	5 months of weekly basketball competitions against another residential centre. Facilitating life skills through sports-themed goals. Every 6 weeks a skills clinic was hosted.
Wu & Hsieh, 2006 (Taiwan)	12 at risk referred through social services, 11–17yrs 2F &7M. Outdoor adventure programme	Self-Concept & Life Satisfaction	28-day outdoor adventure programme, including low/high ropes, Horse Riding, Sailing, River-trekking, mountain-climbing to promote personal growth and improvement in personal effectiveness.

Study designs and methods varied considerably between 7 non-experimental studies, 4 experimental studies and 4 qualitative studies (Table 3). All quantitative studies used different validated instruments to assess both life skills and behaviours linked to youth delinquency. In some cases, the same life skill was assessed using different instruments. For example, The Shortened Protective Factors Scale (Witt & Crompton, 1997) and Strengths and Difficulties Questionnaire (Goodman, 2001) were both used to measure risk and resilience (Bloemhoff, 2006; Frazier et al., 2015; Goodman et al., 2021). Studies also included interviews and questionnaires with programme staff, leaders and parents (Armour & Sandford, 2013; Fuller et al., 2013; Riley & Anderson-Butcher, 2012; Walsh et al., 2010).

Table 3. Study designs and methodologies for sports and physical activity programmes.

		Primary Interventions
Study	Design & Method	Instrument(s)	Analysis	Results (+/-) significant findings incl. youth delinquency and life skills.
Armour and Sandford (2013)	Case Study/ Qualitative	Student profile, researcher field notes, focus groups, interviews, semi-structured reflective journals, open-ended survey, mid-evaluation survey for mentors, school staff and lead contacts.	Descriptive statistics, grounded theory process	Positive change from student baseline profiles (59%), bigger baseline improvements from weekend residential (63%) and 3-week classic course (75%). Decrease in positive changes once programme ceased. Increase in students taking on other external developmental activities. Developed skills in leadership ability and connections with others.
Frazier, Dinizulu, Rusch, Boustani, Mehta, Reitz (2015)	Pre/post-test, Quantitative	Strengths and Difficulties Questionnaire (SDQ, Age 11-17) Self-Report Behaviour Index Social Skills Improvement System (SSIS)	Paired t-test	Pre/ post follow-up measures. No change in parent reported social skills. Staff reported social skill improved from pre to post. Staff reported no change in problem behaviours, but parent reports showed reduction in problem behaviour. SDQ scored rates low, medium, and high risk. high risk attended more sessions.
Fuller et al. (2013)	Case Study, Qualitative	Interviews participants and parents at 12- & 24-week mark.	Data triangulation, Coding	Attended for attraction to sports-based activities. Valued safe space provided experiences for positive personal development. Programme facilitated development of each % of youth development. Character coded most.
Gano-Overway et al. (2009)	Non-experimental, Quantitative	Caring Climate Scale ASRE Scale (Affective self-regulatory efficacy) ESE Scale (Empathetic self-efficacy) Child Social Behaviour Questionnaires	Descriptive statistics Mediational Model	Perceived caring climate + predicts prosocial behaviour/ – predicts anti-social behaviour. Perceived caring climate + predicts ESE and ASRE and ESE + predicted prosocial behaviour / – predicts antisocial behaviour. ASRE + predicts ESE but ASRE did not predict ESE. Most with moderate-to-large effect.
Goodman et al., 2021)	Pre/ post-test, Quantitative	Emotional Regulation Questionnaire for Children and Adults Self-Efficacy (Dahlberg et al., 2005) Screen for Child Anxiety Related Disorders Short Moods and Feelings Questionnaire SDQ 11–17/ Adverse Childhood Experience Questionnaire	Paired t-test	Pre and post follow up measures. No sig, difference in emotional regulation, self-efficacy, risk for anxiety and risk for depression. Trends did not change at different levels of moderators. Adversity did not moderate relationships when zero to one ACEs were reported or when two or more ACEs were reported.
Khoury-Kassabri and Schneider (2018)	Non-experimental, Quantitative	Delinquency Scale Sport and Organised Physical Activity Programmes (measurement of features of the programme)	Descriptive Statistics ANOVA	Reduction in involvement of delinquent acts explored (crimes against a person, property or public disorder crimes). No interaction effects found between programme features and participation (sport type, programme intensity, training and supervision, interaction with community services).
Nols et al. (2017)	Non-experimental, Quantitative	General Self-Efficacy Scale Rosenberg Self-Esteem Scale Socio-demographic indicators	Descriptive Statistics ANOVA Paired-sample t-tests	Reaching the right population Participants don’t have low perceived self-efficacy or low self-esteem. Sig. Result males scored higher than females. Lower self-esteem in those worried about their future. No sig. Difference in SEst or SEff between T1 & T2. Different results for different initiatives.
Riley and Anderson-Butcher (2012)	Case Study, Qualitative	Semi-structured interviews with parents	Grounded theory	Positive impact on; Biopsycho development, Provided Opportunities to Broaden Horizon, Enhance Affect, behaviours and cognitions. Impact on family interactions and prosocial activities in the community.
Ullrich-French et al. (2012)	Non-experimental Quantitative	Leader Support subscale of Psychological Sense of School Membership Scale Social Competence & Physical Self-Worth subscale of Self-Perception Profile for Children Self-Perception Profile for Children Children’s Attraction to Physical Activity The Children’s Hope Scale	Repeated measures MANOVA Multiple Regression	Sig. Increase in perceived social competence, physical competence, physical self-worth and global self-worth from pre to post programme. Greater change in perceived physical competence predicted greater change in all outcome variable. Increases in social competence and leader support predicted increase in global self-worth and attraction to physical activity. Increases in social competence predicted increases in hope.
Walsh et al. (2010)	Case Study Qualitative	Semi-structured interviews with youth and adult coaches, observations and documents	Deductive/ inductive coding, member checks data triangulation	Evidence to suggest transference of all four primary TPSR goals to the school environment. Evidence supporting transfer of self-direction goal setting couple by explained by focus on possible futures.
		Secondary Interventions
Bonnette et al. (2001)	Quasi-experimental, Quantitative	Teacher Observation Form-Critical Thinking (TOF-CT), New Jersey Test of Reasoning Skills (NJTRS), Self-perception Profile for Children (SPPC)	ANOVA, Z-test and descriptive statistics	No change in global self-worth. Increase in critical thinking. High sig. difference in critical thinking in Sport + group than sport group.
Spruit et al. (2018a)	Quasi-experimental Quantitative	Delinquency rates from official police data Strengths and Difficulties Questionnaire, Overt Aggression Scale from Buss-Durkee Hostility Inventory Dutch Tasks and Skills of Adolescents questionnaire Friends’ participation and peer pressure and support for delinquency subscales Resistance to peer pressure subscale of TVA Academic Engagement	Multilevel Model Intention-to-treat analysis	Intervention group sig. Lower chance of being registered as a suspect 11.8% reduction in suspect registration. Sig. Reduction in peer pressure over T0 and T1 Intervention group higher levels of acceptance to authority. Intervention group reported fewer delinquent friends and more prosocial behaviour.
		Tertiary Interventions
Bloemhoff (2006)	Pre/post-test, quantitative	The Shortened Protective Factors Scales	Paired t-test, ANCOVA	Post-test scores compared against post-test scores of control group. Post-test scores show No sig. difference in 2 variables. Sig. difference in value for achievement (p < 0.05) High Sig. difference in remaining 7 dependant variables (p < 0.01).
D’Andrea et al. (2013)	Quasi-experimental, Quantitative	Achenbach’s Child Behaviour Checklist (CBCL), Observations	ANOVA, post-hoc test, T-tests	Females participating in programme exhibited reduction in restraints and time-outs and reduction in internalising and externalising symptoms. Decrease in CBCL pre to post score.
Wu & Hsieh, 2006	Non-experimental Quantitative	Tennessee Self-Concept Scale Life Effectiveness Questionnaire	Pre/ post-test Paired sample t-test	Sig. Pre/post difference of physical self 8/10 dimensions of life skill effectiveness was sig (p < .05)

Primary prevention programmes

Studies identified as primary prevention programmes focused on intervening before the effects of criminality occur. This is evident in the 10 studies included as all interventions have used a location-based model, focusing on at risk neighbourhoods and communities. In this case, neighbourhoods were identified through levels of deprivation (Armour & Sandford, 2013; Frazier et al., 2015; Fuller et al., 2013; Gano-Overway et al., 2009; Khoury-Kassabri & Schneider, 2018; Nols et al., 2017; Riley & Anderson-Butcher, 2012; Ullrich-French et al., 2012 and Walsh et al., 2010) or high prevalence of criminality (Goodman et al., 2021; Khoury-Kassabri & Schneider, 2018). Four of the studies mentioned using specific and specialist staff to build relationships with their respective communities and included volunteers trained by youth workers (Armour & Sandford, 2013) mental health practitioners (Frazier et al., 2015), ethnically diverse staff (Fuller et al., 2013) and trained recreation leaders (Goodman et al., 2021). Evidence suggested that voluntary engagement from young people identified as at risk was high in these programmes as they were attracted to the sport element (Fuller et al., 2013) and participated in more sessions across the duration of programmes (Frazier et al., 2015). Highlighted in 4 of 9 studies were the endeavours to address social connection and social competence through intervention focus on staff developing caring relationships with participants (e.g. Gano-Overway et al., 2009). Results suggested that primary interventions can support significant positive changes in personal skills (Riley & Anderson-Butcher, 2012), prosocial relationships (Gano-Overway et al., 2009; Riley & Anderson-Butcher, 2012) and social competence (Ullrich-French et al., 2012). Despite all studies referring to outcomes linking to a reduction in youth delinquency, only one referred to these in their results and stated that a significant reduction in involvement of delinquent acts (Khoury-Kassabri & Schneider, 2018). Although there was a reduction in delinquent acts, Khoury-Kassabri and Schneider (2018) found no interaction effects between programme features or participation, meaning it was not clear on which part of the programme was responsible for these changes. Nols et al. (2017) found no significant difference between T1 and T2 for self-efficacy or self-esteem and in fact found that participants did not feel they were in deficit of these in the first place. Similarly, Goodman et al. (2021) found no significant difference in emotional regulation, self-efficacy, risk of anxiety or risk of depression. However, Goodman et al. (2021) argues that this is a positive outcome stating the programme prevented the downward trajectory of academic, social, and emotional problems. Lastly, Armour and Sandford (2013) acknowledged that there was a reduction in positive changes after the programme finished. It is noteworthy that this was the only study that conducted longitudinal research.

Secondary prevention programmes

Secondary prevention programmes offer interventions to young people highlighted as being at high risk of youth delinquency. Only two identified studies focused on this population with one targeting young people due to their low IQ and academic abilities (Spruit et al., 2018a) and the other targeting youth exhibiting at-risk characteristics (Bonnette et al., 2001). A key similarity between the studies was the inclusion of either experiential learning through an educational environment (Spruit et al., 2018a) or experiential learnings by-products (Hamilton & Klebba, 2011) and critical thinking, through additional lessons (Bonnette et al., 2001). Evidence suggests that increasing critical thinking skills does not increase global self-worth in a young person (Bonnette et al., 2001). In contrast, Spruit et al. (2018a) included opportunities for experiential learning with an emphasis on positive staff and peer relationships and found a 11.8% reduction in a youth being registered as a criminal suspect, reduction in peer pressure and an increase in pro-social behaviour and acceptance of authority. Comparably, experiential learning identified greater improvements in young people, but caution must be taken in the reliability of these findings, due to the low number of studies.

Tertiary prevention programmes

Tertiary prevention programmes provide youth delinquents an intervention to reduce the likelihood of re-offending. Only three studies were included, identified through their social services referral system (Wu & Hsieh, 2006), residential facility (D’Andrea et al., 2013) or educational youth care centre (Bloemhoff, 2006). Programmes were delivered over varying durations, lasting between 4 hours and 28 days with two studies including outdoor adventure programmes. All studies focused on the management of deviant behaviour including reductions in restraints and time-outs and internalising emotions (D’Andrea et al., 2013), facilitating personal growth (Wu & Hsieh, 2006), and increasing protective factors (Bloemhoff, 2006). Despite significant findings in all studies, the rigor is questionable with poor data collection, simplicity in data analysis and no longitudinal follow up to substantiate whether reduction in youth delinquency was affected in the short or long term.

Trends across prevention programmes

Across all intervention programmes it is evident that each study provided youth with additional resources for support, advice and personal development. Prominent trends throughout all intervention types highlight the focus on character development and social interaction with others in their community. Primary interventions evidenced their ability to provide social connectedness which directly affected youth perceptions of safety in their community (Fuller et al., 2013), displaying pro-social behaviour in their community (Riley & Anderson-Butcher, 2012) and their feelings of hope (Ullrich-French et al., 2012). Similar findings were visible in secondary interventions whereby PAS based youth development programmes evidenced the reduction of peer pressure, increase of pro-social behaviour and notably, an acceptance of authority (Spruit et al., 2018a). Tertiary programmes provided significant evidence towards improving character and social skills through a sense of acceptance and ability for conflict resolution (Bloemhoff, 2006) and improvements in personal growth and attitudes towards physical self (Wu & Hsieh, 2006).

Risk of bias in individual studies

From the eligible studies, using the Mixed Methods Appraisal Tool (Hong et al., 2018) to assess study quality (see Table 4), 7 articles were rated with a score of 100% (Frazier et al., 2015; Fuller et al., 2013; Gano-Overway et al., 2009; Khoury-Kassabri & Schneider, 2018; Riley et al., 2012; Ullrich-French et al., 2012 and Walsh et al., 2010), 6 articles rated at 80% (Armour & Sandford, 2013; Bonnette et al., 2001; D’Andrea et al., 2013; Goodman et al., 2021; Nols et al., 2017; Spruit et al., 2018a) and 2 rated at 60% (Bloemhoff, 2006; Wu & Hsieh, 2006).

Table 4. Mixed methods appraisal tool (MMAT) scores.

Author (year)	MMAT Scores						Comments
	0%	20%	40%	60%	80%	100%
Armour and Sandford (2013)					X		Divergences and inconsistencies between quantative and qualitative results not adequately addressed.
Bloemhoff (2006)				X			Randomization not appropriately performed; outcome assessors not blinded to the intervention provided
Bonnette et al. (2001)					X		Can’t tell if outcome data is complete
D’Andrea et al. (2013)					X		Outcome data not complete
Frazier et al. (2015)						X
Fuller et al. (2013)						X
Gano-Overway et al. (2009)						X
Goodman et al. (2021)					X		Outcome data not complete (dropout rate 52% in T2)
Khoury-Kassabri and Schneider (2018)						X
Nols et al. (2017)					X		Outcome data not complete (dropout rate 44%)
Riley and Anderson-Butcher (2012)						X
Spruit et al. (2018a)					X		Outcome data not complete (dropout 33%)
Ullrich-French et al. (2012)						X
Walsh et al. (2010)						X
Wu & Hsieh, 2006				X			No information on sample strategy; measures inappropriate – limited information.

Discussion

The aim of this present systematic review was to collate and evaluate the impact of life skills developed through sport and physical activity based PYD programmes, targeting youth at risk of, or involved in, youth delinquency.

Results indicate that sports-based youth development programmes can positively affect youth at risk of, or involved in, delinquent acts. Significant contributions have been made by studies around personal growth including character development, reduction in internalising emotions and well-being. Programmes also demonstrated the ability to impact on pro-social behaviour and perceptions of community to influence the likelihood or involvement in delinquent acts. It is important to acknowledge that previous systematic reviews have omitted key non-significant findings from their results. This restricts researchers and readers understanding the field of PAS based youth development’s significant and non-significant findings. It is key to acknowledge that many studies are not published in relevant sport and physical activity journals but in fields such as adolescence, family violence, mental health, and social work. This could threaten the exposure and impact of research as those working in the field of sport and exercise could miss updates in the field. However, engaging young people at risk of youth delinquency through sporting environments because of research provided in the sports and physical activity field, may produce different and more impactful changes.

This is the first systematic review to describe the evidence from studies on PAS based youth development programmes and their effects on youth at risk of, or involved in, delinquency. This review revealed a surprisingly limited number of studies published in this field, containing only 4 intervention studies, and therefore results should be treated with some caution. Through grouping the programmes constructed on their level of intervention (primary, secondary or tertiary), various positive findings can be found for PAS youth development programmes’ impact on protective factors through life skills developed by youth and their ability to support a reduction in youth delinquency. This provides an alternative and more detailed view in comparison to previous systematic reviews, exploring the impact of PAS participation on life skills development (Hermens et al., 2017) and social and emotional wellbeing (Lubans et al., 2012) on socially vulnerable youth or youth delinquency (Spruit et al., 2016).

Many positive findings were discovered through the analysis of this current review. Conclusions provide evidence that developing youths’ personal identity and their ability to connect with peers and adults could prevent or reduce youth delinquency. Ten out of the 15 studies included an assessment on individual and/or social development including areas such as social connection, sense of acceptance, pro-social behaviour, interaction with family and social competence (e.g. Armour & Sandford, 2013; Gano-Overway et al., 2009; Riley & Anderson-Butcher, 2012). Participation in these programmes may act as a protective tool to connect positive adult role models with youth at risk, giving youth knowledge, skills and to constructively challenge youth (Crean, 2012). Results suggest that primary prevention programmes can be seen to provide safe recreation resources in the community with secondary prevention programmes targeting youths’ social development skills. Additionally, these community-based PAS development programmes could produce higher levels of social trust, encouraging young people to help and not harm others (Flanagan et al., 2005). Notably, Goodman et al. (2021) argues that PAS programmes could disrupt the downward trajectory in social and emotional problems, often seen in communities with systemic inequalities. Finally, tertiary programmes can be seen to focus on reintegrating delinquent youths into the community (Berdychevsky et al., 2022) through more focused interventions.

This systematic review highlighted the infancy of this field, with consistency of development and publication in PAS based youth development tackling youth delinquency emerging from 2010 onwards. However, the review also identified the need to establish clear and confirmatory research into these concepts. The divergences in the theory base used across these 15 studies included general positive youth development but also risk and resilience, critical thinking, self-efficacy and Teaching Personal and Social Responsibility (TPSR). Three studies evaluated similar National summer camps in the USA, providing an opportunity to compare similar programmes with similar youth populations. However, different theory bases and instruments were used to evaluate the programmes. This does not provide any replication of underpinning theories to provide convergence in how to deliver programmes to facilitate life skills and reduce youth delinquency. Similarly, challenges in conceptualizing and understanding life skills by researchers and practitioners are still widely prevalent (Gould et al., 2013; Newman, 2020). It is imperative that steps are taken towards developing a consensus on what life skills are and then how we utilise them to facilitate growth in at risk youth. Few of the studies yielded non-significant findings including no change in global self-esteem, problem behaviours, social skills and no interactions between programme features and participation. This does cast doubt on the overall effectiveness of sports-based youth programmes for addressing youth delinquency but similarly highlights a lack of understanding about the mechanisms that support young people not to offend.

Another challenge identified is the poor data or limited analysis of data in some of the studies. Only 4 out of 15 studies included a control group, reducing the capability of demonstrating comparability between intervention and no intervention. Most studies used small population samples and nearly all used one location, reducing the reliability of research, leading to higher variability and likelihood of increased bias in the results. Due to difficulties in engaging an at-risk population, drop out was high in some studies (Goodman et al., 2021) meaning selection bias could have occurred. For example, Nols et al. (2017) had a 44% unexplained dropout which could have omitted negative experiences of participants, significantly impacting results. Youth exposure to ACEs is associated with reduced odds of sport and physical activity participation (Noel-London et al., 2021) and as this systematic review only included studies with voluntary participation, young people with higher involvement in delinquency may have chosen not to participate. There was no mention of any programme preparation in efforts to engage those more at risk of delinquency and therefore further research should determine how to involve higher-risk youth into voluntary PAS programmes.

There are similar methodological constraints to using such a variety of different study designs and instruments for result synthesis. Firstly, through quantitively assessing the effectiveness of life skill development in youth, studies are only able to assess a few specific constructs, restricting perspectives of life skills with experiences of youth being overlooked (Newman, 2020). Secondly, it is evident in the systematic review that there is no consistency in instruments being used to analyse life skills development in youth. This limits the comparability within the small amount of research currently available in this field and makes it clear that there is no ability to replicate studies to increase rigor. An example of clarity needed is evident across the measurement of offending or problem behaviour within PAS programmes that facilitate life skills. Measurements included examples such as a self-report behaviour index (Frazier et al., 2015), a Child Social Behaviour Questionnaire (Gano-Overway et al., 2009), Delinquency Scale (Khoury-Kassabri & Schneider, 2018).

Results from the systematic review indicate a disproportionate emphasis on the male population, with only one female focused, tertiary programme (D’Andrea et al., 2013). Pierce and Jones (2022) found that accumulation, duration, and timing of ACEs exposure is significantly related to youth delinquency in females but not males. Whereas, Liu and Miller (2020) stated that males are more likely to be aggressively delinquent and females are disproportionately likely to be involved in non-aggressive delinquency. Further research is needed to consider gender-specific responses to ACEs and gender-specific interventions strategies, including different types of conducive environments and a focus on programmes where ACEs have had lesser impact on youth (Pierce & Jones, 2022).

An identified gap in this systematic review is the limited information around programme delivery, reducing future research from understanding conducive environments. There is inadequate understanding in how a programme is delivered and the intentional development in these programmes that facilitate life skills through sport and physical activity. One key area is the lack of information around the coaching staff delivering these programmes with only 5 studies recording staff demographics, training, and experience (Armour & Sandford, 2013; Frazier et al., 2015; Fuller et al., 2013; Walsh et al., 2010). Further research needs to explore the relationship between life skills developed through sports-based youth development and the quality and experience of staff. In social work literature focusing on ACEs and youth sport, there are strong recommendations for data-driven interventions, researching the effects of trauma-informed training for sports and physical activity deliverers. This could enable sport and activity staff to identify ACEs in the community and youth to ensure intentional life skills and protective factors are developed. Noel-London et al. (2021) concur that utilising trauma-informed training in youth sport could bring additional value to reducing youth delinquency.

Additionally, none of the studies made mention on social return on investments (SROI), using financial proxies to monetise outcomes from interventions. Without SROI frameworks, policymakers and stakeholders cannot understand, measure or report on the value created by these interventions (Gosselin et al., 2020). Further research needs to establish actual impact of investments made by programme creators, especially regarding the impact on reducing youth delinquency.

Lastly, despite previous research indicating how youth develop life skills, assessed from an adult-centred approach, little is known about how youth themselves conceptualise life skills. This is evident in the current systematic review as few studies explicitly included interviews with youth to understand their accumulation and transference of implicit and explicitly learnt life skills. Supporting this, Newman (2020) addressed how youth identify, define, and apply life skills development. They found that youth living in deprivation developed and transferred a variety of intrapersonal and interpersonal life skills and were also able to identify, define both explicit life skills taught in the programme and those that were not. Future research needs to include further explorations of who is delivering sports programmes to youth at risk and further understanding on the impact of training, specifically around youth trauma and youth delinquency.

Many sport and physical activity-based organisations delivering programmes are now recognising the impact and effectiveness of PAS programmes on youth that have become disengaged, at-risk, or criminally active (Department of Digital, Culture, Media and Sport, 2019). A significant amount of field research has been developed by organisations such as Alliance of Recreation (2022), DCMS (2019), Get Berkshire Active (2022), and Street Games (2022) to understand how sport can make a positive change to youth at risk. These organisations have shared national, regional and local case studies where PAS programmes have impacted youth delinquency, including the reduction of reoffending (DCMS, 2019). This includes resources providing delivery partners information on the impact of sports-based mentoring in prevention and desistence of youth delinquency, mentoring through PAS and theory of change modules for delivery (Levelling the Playing Field, 2022; Loughborough University, 2020; Street Games, 2020). Despite the increased open dialogue within the field, further academic exploration and collaboration of the current climate in the UK is needed to understand how we can add value to current SBYD theories and practical interventions to support the development of at-risk youth through life skill facilitation.

Limitations

This systematic review does carry some limitations. Firstly, although the search terms used provided depth in reaching relevant studies, the lack of conformity in identifying youth at risk and life skills development could mean that some studies were missed. Similarly, this review focused on voluntary participation and not compulsory participation of youth at risk. This means that some relevant studies identifying evidence of positive development of life skills in at risk youth, may have been omitted. Due to the heterogeneity of study duration, type and variety in the assessment of outcomes, a meta-analysis was not possible. Finally, despite using the MMAT (Hong et al., 2018) to appraise the studies, no exclusions were made for those with low rigour. This was due to the limited number of studies available and the decision was made to enable a comprehensive assessment of the existing literature in this field.

Conclusion

Evidence in this review indicates that PAS based youth development programmes can support the development of life skills in youth at risk of or involved in youth delinquency. The strongest evidence suggests that programmes are most likely to support the growth of youths’ personal identity and social connection, particularly in primary (within at-risk communities) and secondary (targeting specific at-risk youth) interventions. It was established that this field of work is in its infancy, and this can be seen in the small number of studies but also the limited diversity in terms of settings, research methods and reporting of life skills. Future recommendations include the encouragement for both academics and deliverers working in the field to collaborate and contribute to the evidence base. This will provide better synergies in work already being carried out by researchers, government bodies, police and PAS deliverers and further support those at risk of, or involved in, youth delinquency. Further research needs to create consistency in how we identify youth at risk, and how we measure the impact of life skills developed through PAS programmes. It is also recommended that future research continues to understand impact on personal and social identity of youth at risk with particular focus on gender-specific strategies. Lastly, we need to further understand the effectiveness of delivery staff, particularly related to the quality and skills of staff. The current evidence base for understanding life skills developed through PAS programmes is positive but sparse. This review provides evidence that developing youths’ personal identity and their ability to connect with peers and adults could prevent or reduce youth delinquency. Therefore, we believe recommending the use of PAS based youth development programmes to those working with youth at risk of, or involved in, criminal behaviour is justified, despite the further research needed.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, C.L.P, upon reasonable request.