Impact statement
Adolescence, a critical stage marked by significant physical, mental, and social changes, often leaves youth vulnerable to mental health challenges. This vulnerability is exacerbated in low- and middle-income countries (LMICs), where there is an increased risk of mental distress due to poverty, environmental instability, and limited access to healthcare services. In LMICs, where 90% of the world’s youth reside, the mental health treatment gap is stark. Task-shifting, which involves redistributing mental health care tasks to less specialised workers, has emerged as a viable solution to address this gap. This scoping review examines the effectiveness of peer-led mental health interventions for youth aged 10–24 in LMICs, highlighting their potential to improve mental health outcomes through relatable, culturally sensitive support. Despite diverse approaches – ranging from psychoeducation and psychotherapy to peer support – the consistent positive impact underscores the potential of peer-led strategies. The success of such interventions hinges on rigorous training, ongoing supervision, cultural sensitivity, and institutional backing to ensure their effectiveness and sustainability. This review highlights the versatility, impacts and challenges of peer-led interventions, advocating for their broader implementation to enhance youth mental health in LMICs.
Introduction
Adolescence is a stage of life marked by pronounced changes to the body, mind, and social environment (Lowenthal et al. Reference Lowenthal, Bakeera-Kitaka, Marukutira, Chapman, Goldrath and Ferrand2014). The intricacies of navigating through this turbulent period often leaves youth vulnerable to mental distress. Approximately 75% of mental health disorders present by the age of 24 years (Das et al. Reference Das, Salam, Lassi, Khan, Mahmood, Patel and Bhutta2016). Youth living in low- or middle-income countries (LMICs) are at an increased vulnerability to mental illnesses due to circumstances such as increased prevalence of poverty, environmental instability, and lack of access to medical and mental health services (Pedersen et al. Reference Pedersen, Smallegange, Coetzee, Hartog, Turner, Jordans and Brown2019). In LMICs, where 90% of the world’s youth live, the mental health treatment gap is the most severe (United Nations 2017). The median number of psychiatrists per 100,000 population in LMICs is 0.05, compared to 8.59 in high-income countries (HICs), resulting in three-quarters of individuals in need of mental health services unable to access them (Luitel et al. Reference Luitel, Jordans, Adhikari, Upadhaya, Hanlon, Lund and Komproe2015; Le et al. Reference Le, Eschliman, Grivel, Tang, Cho, Yang, Tay, Li, Bass and Yang2022).
Task-shifting or sharing has been increasingly cited as a potential approach to respond to unmet needs of countries with a scarcity of specialised mental health care providers (Triece et al. Reference Triece, Massazza and Fuhr2022). Task-shifting or sharing is the redistribution of care usually provided by those with formalised training (i.e. psychiatrists) to individuals with little or no formal training (i.e. lay health workers). The literature has shown that with training and continued supervision, lay health workers can effectively deliver psychological treatments (van Ginneken et al. 2013).
Specifically, there has been a growing body of studies using task sharing with peer groups for intervention delivery (Stokar et al. Reference Stokar, Baum, Plischke and Ziv2014). A peer is defined as a person who shares common sociodemographic characteristics and/or lived experiences of a condition, event, or disorder with the target population (Farkas and Boevink Reference Farkas and Boevink2018; Pedersen et al. Reference Pedersen, Smallegange, Coetzee, Hartog, Turner, Jordans and Brown2019). In the literature, trained peers are referred to with various terms, including “peer-leaders” (Stokar et al. Reference Stokar, Baum, Plischke and Ziv2014), “peer educators,” “peer-volunteers” (Maselko et al. Reference Maselko, Sikander, Turner, Bates, Ahmad, Atif, Baranov, Bhalotra, Bibi and Bibi2020), “peer-facilitators” (Alcock et al. Reference Alcock, More, Patil, Porel, Vaidya and Osrin2009), “peer-counsellors” (Nankunda et al. Reference Nankunda, Tumwine, Nankabirwa and Tylleskär2010), “peer-mentors” (Shroufi et al. Reference Shroufi, Mafara, Saint-Sauveur, Taziwa and Viñoles2013) and “peer-supporters” (Nkonki et al. Reference Nkonki and Daniels2010). For the purpose of this review, we will use the term “peer leaders” to refer to trained peers who deliver interventions. This will also distinguish these interventions from those involving peer support within group therapy sessions led by professional mental health specialists. Peer-led mental health services are grounded on the concept of peer support, which is the provision of emotional appraisal and informational assistance through access to a social network (Dennis Reference Dennis2003; Embuldeniya et al. Reference Embuldeniya, Veinot, Bell, Bell, Nyhof-Young, Sale and Britten2013). The knowledge of a disease stems from lived experiences and peer leaders are typically trained to deliver specific interventions but lack professional status through academic or professional qualifications (Pedersen et al. Reference Pedersen, Smallegange, Coetzee, Hartog, Turner, Jordans and Brown2019). This camaraderie provides qualities unique to peer support by offering validation of lived experiences, building rapport, and establishing bonds (Mancini Reference Mancini2018; Linnemayr et al. Reference Linnemayr, Zutshi, Shadel, Pedersen, DeYoreo and Tucker2021). Individuals who are from the local community are found to relate better to the target group and could share approaches to dealing with distress that were in line with existing sociocultural contexts (Petersen et al. Reference Petersen, Baillie, Bhana and Health2012). This was particularly useful for cultural nuances, that may be delicate for an outsider to address, such as prevailing customs or patriarchal realities. Peer delivery of mental health interventions has also been proposed as a cost-effective yet feasible and acceptable solution to address human resource challenges (Sikander et al. Reference Sikander, Ahmad, Atif, Zaidi, Vanobberghen, Weiss, Nisar, Tabana, Ain and Bibi2019; Maselko et al. Reference Maselko, Sikander, Turner, Bates, Ahmad, Atif, Baranov, Bhalotra, Bibi and Bibi2020) thereby filling a gap in mental health treatment.
There have been several reviews of peer-led mental health interventions, but none specifically explored their impact on youth mental health in LMICs – a unique demographic with a specific set of challenges – and led by peers who are themselves youth (Sikkema et al. Reference Sikkema, Dennis, Watt, Choi, Yemeke and Joska2015; Hoeft et al. Reference Hoeft, Fortney, Patel and Unutzer2018; Bhana et al. Reference Bhana, Kreniske, Pather, Abas and Mellins2021; de Beer et al. Reference de Beer, Nooteboom, van Domburgh, de Vreugd, Schoones and Vermeiren2022). This scoping review profiles the evidence for peer-led interventions to improve the mental health of youth aged 10–24 and living in LMICs with the aim of informing the development of impactful interventions to address this crucial gap. Key research questions we explored include:
-
1. What are the characteristics and outcomes of existing peer-led mental health interventions in LMICs?
-
2. What are the components required to enable peer-leaders to act as key agents in intervention delivery?
-
3. How effective are peer-led interventions in improving youth mental health outcomes in LMICs? What mechanisms make them more effective than if led by adults or non-peers?
-
4. What are the challenges to implementing such interventions?
Methods
Search strategy
The search was developed and conducted by a professional medical librarian in consultation with the author team and included a mix of keywords and subject headings representing adolescents, mental health, peer support, and LMICs. The searches were independently peer-reviewed by another librarian using a modified PRESS Checklist (McGowan et al. Reference McGowan, Sampson, Salzwedel, Cogo, Foerster and Lefebvre2016).
Search hedges or databases filters were used to remove publication types such as editorials, letters, books, book chapters, essays, conference proceedings, and comments as was appropriate for each database. Searches were conducted in MEDLINE via Ovid, Embase via Elsevier, Web of Science via Clarivate, Global Health via EBSCO, and Global Index Medicus via WHO. The searches were executed on 31 August 2023 and found 3,404 unique citations. A search update conducted on 6 September 2024 to identify newly published studies that found an additional 403 studies. Key search terms included “adolescent,” “youth,” “mental health,” “peer” and “LMIC.” Complete reproducible search strategies, including search filters, for all databases are detailed in the Supplementary Materials.
Study selection
All citations were imported into Covidence, a systematic review screening software, which was also used to remove duplicates. Initial inclusion and exclusion based on the title and abstracts were made independently by the first author (DC) and second author (DJM). Potentially relevant studies were selected and the full text reviewed by the first and second authors. Any discrepancies were resolved by the senior author (DD).
Eligibility criteria
Eligibility criteria were specified based on the PICOS model of Population, Intervention, Comparison, Outcome and Study type. Included studies were required to have an intervention that included young people aged 10–24 making up at least 50% of the study population, a peer-led component, has been conducted in any setting (i.e. community, hospital, schools, refugee camps, online etc.), and been delivered in a LMIC. LMIC was defined according to the World Bank List. While we recognise the significant diversity among LMICs, spanning various continents and cultural contexts, we have chosen to include all LMICs as they share common structural and socioeconomic barriers that impact the delivery of health interventions. By including a broad spectrum of LMICs, this study can identify patterns and insights that can inform scalable, context-specific solutions. Both quantitative and qualitative data were acceptable but had to include a measured form of mental health, either distress (i.e. depression, anxiety, suicidal ideation attempt etc.) or wellness (i.e. resilience, self-efficacy etc.). To be included, qualitative outcomes had to specifically document a change in mental health/distress in relation to a peer-led intervention.
Data extraction
Data extraction was conducted using Covidence to support the extraction of relevant and detailed information including author, year of publication, study design, country, and intervention details. The extraction was similarly performed by authors DC and DJM and discrepancies resolved by DD.
Data analysis
The characteristics of the studies were summarised and presented descriptively to illustrate the scope of the included literature. Due to the heterogeneity of studies and scales used, it was not possible to conduct a meta-analysis. Qualitative studies were reviewed for any specific outcomes pertaining to mental health, with relevant quotations from participants extracted from the articles.
Results
Characteristics of included studies
The search identified 5,358 citations, and after excluding duplicates, 3,807 articles were screened using title and abstract. We excluded 3,752 studies for key reasons such as irrelevance to the topic, incorrect target population, intervention and study type. We manually added three studies (Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024). Fifty-eight articles underwent full text screening based on criteria described, and 19 were included (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018, Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a, Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021, Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023; Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024) (Figure 1, Table 1). Included studies were conducted between 2011 and 2024. Six studies were conducted in Asia-Pacific: four in India (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021), one in Iran (Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021), one in Turkey (Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019). The rest were based in Africa: one in Tanzania (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020), two in South Africa (Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023), three in Zimbabwe (Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024), one in Zambia (Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) and six in Kenya (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a, Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023). Seven were randomised controlled trials (RCTs) (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024), three cluster-randomised (Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b) and the rest were quasi-experimental trials (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024). There were 14 quantitative (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024), four qualitative (Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) and one mixed-methods (Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023) study. Five studies were pilot trials (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023). Mathias and (Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018, Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019) and Kermode et al. (Reference Kermode, Grills, Singh and Mathias2021) were studies on a related intervention based on the Nae Disha Curriculum, while Osborn et al. (Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021, Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a) and Venturo-Conerly et al. (Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024) were studies on the Shamiri Intervention. The majority of studies focused on a general population of youth experiencing mental health difficulties, but four had specific focus on youth living with human immunodeficiency virus (HIV) (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023), four on girls and young women (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021), including one on adolescent mothers (Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024), and one on youth refugees (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018).
PRISMA flowchart.
Data extraction of included studies
Types of interventions
The studies were heterogenous in terms of the types of interventions, and fell into three main categories: (1) peer-led psychotherapy and counselling: a therapeutic process facilitated by trained peers, focusing on improving participants’ mental health through structured therapy or counselling sessions, (2) peer education and psychoeducation: an approach where peer leaders teach participants about specific health or psychological topics, with the goal to equip individuals with knowledge and techniques to enhance their mental health and well-being, and (3) peer support: a system of informal, mutual support, where peers provide social connection, encouragement, and a sense of community to foster mental health and well-being. Peer-led psychotherapy and counselling, which require the highest level of specialisation, often include elements of peer education. All three types of interventions integrate components of peer support. Eight studies (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023) incorporated psychotherapy in their intervention, delivered in settings including the community, online, schools, and health clinics. Psychological therapies included trauma-informed cognitive behavioural therapy (TI-CBT), interpersonal psychotherapy (IPT), motivational interviewing (MI), meditation, inquiry-based stress reduction (IBSR), an art-literacy program, and problem solving therapy (PST). All were conducted in a group setting, but Dow et al (Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020) and Ferris France et al (Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023) had also weaved in individual sessions or activity components. Eight studies (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) used peer education and psychoeducation in their interventions. Settings included schools, universities, communities, and an HIV clinic. Im et al. (Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018) explored the Trauma-Informed Psychoeducation (TIPE) intervention to promote refugee resilience, conflict resolution, impact of trauma on the body and emotional coping. Kermode et al. (Reference Kermode, Grills, Singh and Mathias2021) and Mathias et al. (Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018, Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019) ran interventions adapted from the same Nae Disha curriculum involving the formation of Youth Wellness Groups and interactive modules on self-identity, self-esteem, relationship and communication skills, drawing boundaries, and managing emotions. Most were delivered as classroom-based group sessions, with Merrill et al. (Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) additionally incorporating individual meetings and Balaji et al. (Reference Balaji, Andrews, Andrew and Patel2011) organising street plays. Peer support was utilised by three studies (Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024), and involved the creation of spaces to form social networks and facilitate discussions at hospitals, universities, and community clubs.
Description of peer-leaders: Recruitment, training and support
A variety of terms were used to refer to the peer leaders, including “peer-educators,” “community adolescent treatment supporters,” “community youth leaders,” “group leaders,” “youth peer mentors,” “student lay supporters,” “peer mentors,” “lay-providers.” All studies involved peer leaders who were locally based lay-persons, similar in age group and shared geographical, linguistic, cultural and in most cases, health conditions with the target population, along with no formal qualification, qualifying them to be called peers. The age group of peer leaders ranged from 18 to 30 years old, although several studies did not specify it (Table 2).
Description of peer leaders
* Information extracted from protocol paper:
1 Dow DE, Mmbaga BT, Turner EL, Gallis JA, Tabb ZJ, Cunningham CK, et al. Building resilience: a mental health intervention for Tanzanian youth living with HIV. AIDS Care. 2018;30(sup4):12–20.
2 T.L. Osborn, K.E. Venturo-Conerly, G.S. Arango, E. Roe, M. Rodriguez, R.G. Alemu, J. Gan, A.R. Wasil, B.H. Otieno, T. Rusch, D.M. Ndetei, C. Wasanga, J.L. Schleider, and J.R. Weisz, Effect of Shamiri Layperson-provided intervention vs. study skills control intervention for depression and anxiety symptoms in adolescents in Kenya: A randomised clinical trial. JAMA Psychiatry 78 (2021) 829–837. (Supplementary Materials).
3 T.L. Osborn, D.M. Ndetei, P.L. Sacco, V. Mutiso, and D. Sommer, An arts-literacy intervention for adolescent depression and anxiety symptoms: outcomes of a randomised controlled trial of Pre-Texts with Kenyan adolescents. EClinicalMedicine 66 (2023) 102288.
4 K. Venturo-Conerly, E. Roe, A. Wasil, T. Osborn, D. Ndetei, C. Musyimi, V. Mutiso, C. Wasanga, and J.R. Weisz, Training and supervising lay providers in Kenya: Strategies and mixed-methods outcomes. Cognitive and Behavioural Practice 29 (2022a) 666–681.
5 S. Chinoda, A. Mutsinze, V. Simms, R. Beji-Chauke, R. Verhey, J. Robinson, T. Barker, O. Mugurungi, T. Apollo, and E. Munetsi, Effectiveness of a peer-led adolescent mental health intervention on HIV virological suppression and mental health in Zimbabwe: protocol of a cluster-randomised trial. Global Mental Health 7 (2020) e23.
6 K.E. Venturo-Conerly, T.L. Osborn, R. Alemu, E. Roe, M. Rodriguez, J. Gan, S. Arango, A. Wasil, C. Wasanga, and J.R. Weisz, Single-session interventions for adolescent anxiety and depression symptoms in Kenya: A cluster-randomised controlled trial. Behaviour Research and Therapy 151 (2022b) 104040.
7 K.E. Venturo-Conerly, A.R. Wasil, T.L. Osborn, E.S. Puffer, J.R. Weisz, and C.M. Wasanga, Designing culturally and contextually sensitive protocols for suicide risk in global mental health: Lessons from research with adolescents in Kenya. J Am Acad Child Adolesc Psychiatry 61 (2022c) 1074–1,077.
The criteria to be a peer leaders varied according to the target community but shared similarities in terms of language and skills requirements. Four studies (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) with a focus on young people living with HIV (YPLWH) recruited peer leaders who were also living with HIV. Dow et al. (Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020) expanded their recruitment to also include young adults with prior experience with mental health research. At least 10 studies specified requirements for standard of education (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Venturo-Conerly et al. 2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024) of which several also specified criteria for language (Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021), leadership and communication skills, past experiences and interests (Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021; Venturo-Conerly et al. 2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023). Selection and recruitment processes for peer leaders were sometimes detailed. Osborn et al. (Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023, Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021, Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a) recruited from local universities and high school graduate forums using a written application and interview process. Venturo-Conerly et al. (Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024) selected youths based on online applications and in-person semi-structured interviews. Mathias et al. (Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019) invited potential peer leaders for community meetings and interviews. Tinago et al. (Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024) recruited peer educators via snowball sampling and in-person meetings. A few studies selected peer leaders instead, based on scores on a pubertal health questionnaire (Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021), assessment of readiness and capacity (Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022), academic grade and interviews with faculty (Yuksel and Bahadir-Yilmaz Reference Yuksel and Bahadir-Yilmaz2019).
Within the psychotherapy intervention type, the training period spanned from 4 days to 3 weeks. Training topics included specific therapy techniques, interpersonal and leadership skills, and often utilised role playing. Notably, Osborn et al. (Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023, Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a), Simms et al. (Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022), Venturo-Conerly et al. (Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024) trained their peer leaders in clinical risk assessment and making referrals to relevant support (i.e. mental health services, school resources). Studies in the peer education and psychoeducation intervention type had their trainings spanning 1 day to 2 weeks on program content and curriculum. Notably, Merrill et al. (Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) held practice meetings for 1 month after the initial 2 week training period, and Mathias et al. (Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018, Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019) had instated refresher training during the intervention period. Within the three studies utilising peer support as their intervention, only Tinago et al. (Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024) had specified their training type, with a 3 day session on the subject matter. Most trainings were led by the research team. Four studies included a psychologist or mental health specialist (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024), and Ferris France et al. (Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023) used local adult coaches living with HIV and trained in IBSR to conduct the training of trainers.
Nine studies described having regular supervision meetings weekly (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023) and monthly (Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024) to review previous sessions as well as prepare for upcoming sessions. Nine studies explicitly stated the availability of adult coaches or research staff at each intervention session to provide support where needed (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023). This includes youth leaders facilitating alongside community health counsellors (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018), team leaders stepping in if a youth leader was absent (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018) or unable to deal with a situation outside their scope of knowledge (Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023), and readily available care in case of potential risk identified during the intervention (Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024). Two studies regularly utilised fidelity checklists throughout the study to ensure intervention quality (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023). Three studies (Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b) used WhatsApp, an online instant messaging application, as an adjunct to support communications, while Mohamadi et al. (Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021) and Yuksel and Bahadir-Yilmaz (Reference Yuksel and Bahadir-Yilmaz2019) mentioned an ongoing relationship between peer leaders and the research team.
Only six studies (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023; Osborn et al. Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023) detailed a form of remuneration for the peer leaders. Osborn et al. (Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023) and Venturo-Conerly et al. (Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b, Reference Venturo-Conerly, Osborn, Rusch, Ochuku, Johnson, van der Markt, Wasanga and Weisz2024) paid a stipend of $150 and covered transportation costs throughout the intervention period (Venturo-Conerly et al. Reference Venturo-Conerly, Osborn, Alemu, Roe, Rodriguez, Gan, Arango, Wasil, Wasanga and Weisz2022b), Balaji et al. (Reference Balaji, Andrews, Andrew and Patel2011) provided “moderate monetary” and other incentives (i.e. certificates), while Mathias et al. (Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019) and Merrill et al. (Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023) simply stated that peer leaders were paid (without including the amount).
Key findings in mental health outcomes
The studies used various instruments to measure mental health outcomes. General mental health symptoms were measured using the General Health Questionnaire-12 (GHQ-12), Strengths and Difficulties Questionnaire (SDQ), Symptom Checklist-25 (SCL-25), Social Support Questionnaire (SSQ), WHO-5 Well-being Index, Psychological Outcomes Profile, Beck Youth Inventories Second Edition (BYI-II), Child Attitudes Toward Illness Scale (CATIS), Short Warwick-Edinburgh Mental Wellbeing Scale (SWEMWBS), Shona Symptom Questionnaire (SSQ), Persian Standard Symptom Checklist-25 (SCL-25). Anxiety was measured with the Generalised Anxiety Disorder-7 (GAD-7), depression with the Patient Health Questionnaire -9 and -8 (PHQ-9, PHQ-8), and post-traumatic stress using the UCLA PTSD Reaction Index survey, PTSD Check List – Civilian Version (PCL-C). Other mental health scales include the Connor-Davidson Resilience Scale (CD-RISC), Schwarzer general self-efficacy scale, Multidimensional Scale of Perceived Social Support (MSPSS), Perceived Control Scale for Children (PCS), Ways of Coping Inventory (WCI), and Peer and Significant Adult Support (PSAS).
All quantitative studies reported that peer-led interventions led to improvements in at least one mental health symptom scale (Table 1). Studies exploring general mental health symptoms reported an improvement in scores for mental health symptoms, probable depression and suicidal behaviour, as well as an intervention effect on the prevalence and severity of common mental health outcomes. Studies measuring depression and anxiety reported a larger decrease in scores in the intervention than the standard of care or active control groups. Two studies examined PTSD symptoms and reported an improvement in scores in the intervention group (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020). None had reported worsening of mental health as a result of the intervention.
Qualitative studies similarly reported improvements in mental health (Table 3). There were reports of improved self-esteem, confidence and self-worth (Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023), combating feelings of isolation with community building (Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023), and improved mental health (Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019).
Mental health outcomes and quotes from qualitative papers
Cultural considerations during implementation
Several studies specified consideration of the respective country’s sociocultural context during development of their intervention. (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021). This entailed, for example, conducting group sessions with YPLHIV in same-sex groups with sex-concordant peer leaders (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020) or adopting common colloquial terms related to mental health (cultural idioms) identified by local community partners to avoid pathologising trauma responses with western terminology (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018). However, several studies also acknowledged the use of mental health measures that have not been validated or adapted to the cultural context was inevitable due to the lack of alternatives (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021).
The mechanisms that make peer-led interventions effective and their implementation challenges
Several studies reported specific aspects of having a peer-led component, detailing mechanisms that made the intervention more effective than if it had been delivered by adults or non-peers (Table 4). Such mechanisms include the provision of a unique form of emotional support, where youth could relate on a peer-to-peer level (Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022). Participants similarly appreciated the ability to connect with peer leaders given their similar age and lived experiences, for instance, the shared experiences of living with HIV (Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023). There was also a preference for peer delivery when it came to sensitive topics such as Sexual and Reproductive Health (SRH) (Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Merrill et al. Reference Merrill, Frimpong, Burke, Abrams, Miti, Mwansa and Denison2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024). The enhanced sense of community was well received, with participants describing the alleviation of social isolation and increased social support (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018; Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023; Tinago et al. Reference Tinago, Frongillo, Warren, Chitiyo, Jackson, Cifarelli, Fyalkowski and Pauline2024). In some studies, participants praised positive qualities of peer leaders, such as patience, kindness and respect (Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019). Studies also reported the added benefit of peer-led interventions to the peer leaders themselves, that they increased their self-confidence and leadership ability (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023), and allowed them to learn from the participants and gain new experiences (Harrison et al. Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023).
The mechanisms that make peer-led interventions effective and their implementation challenges
However, several challenges to peer-led interventions were also reported. There were issues with adherence of peer leaders to the intervention, with many citing other commitments to school or housework, inconvenient locations and timings (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011). Furthermore, integrating a peer-led system into existing structures is expensive and requires much support from institutions to overcome logistical barriers (Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021; Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023). This was similarly echoed by Kermode et al. (Reference Kermode, Grills, Singh and Mathias2021), who reported that a lot of support was required by the project team to ensure the success of the peer-led intervention. Harrison et al. (Reference Harrison, Mtukushe, Kuo, Wilson-Barthes, Davidson, Sher, Galarraga and Hoare2023) described an emotional toll that peer leaders face when sharing their own stories and striving to understand and support participants, further underscoring the importance of providing adequate support to peer leaders.
Discussion
There is a critical need to address the youth mental health gap in LMICs, and task-sharing with peer leaders present a possible solution. The purpose of this review was to identify intervention trials that utilised peer-led approaches to address youth mental health in LMICs, and to synthesise the types of interventions, delivery components, effectiveness, benefits and challenges encountered during implementation. The inclusion of pilot studies and quasi-experimental trials contributed to an overview of peer-led mental health interventions that are currently in the developmental phase. Our review suggests that a diverse range of peer-led interventions can generate positive improvements in mental health among adolescents in LMICs.
Peer-led interventions can improve mental health outcomes of youth living in LMICs
Peer-led interventions included in this review were highly heterogenous in terms of design, mode of delivery and content. Yet, all reported improvements in different mental health outcomes, highlighting their versatility to be adopted in many contexts. Within peer psychotherapy and counselling interventions, studies used either group and individual therapy or didactic lectures and non-didactic methods (i.e. music reflection, poem, group discussions, podcast making) to increase youth engagement (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Duby et al. Reference Duby, Verwoerd, McClinton Appollis, Jonas, Maruping, Dietrich, LoVette, Kuo, Vanleeuw and Mathews2021; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021). Previous research has demonstrated that peer-delivered psychotherapy is effective for youth living in LMICs (Singla et al. Reference Singla, Meltzer-Brody, Silver, Vigod, Kim, La Porte, Ravitz, Schiller, Schoueri-Mychasiw, Hollon, Kiss, Clark, Dalfen, Dimidjian, Gaynes, Katz, Lawson, Leszcz, Maunder, Mulsant, Murphy, Naslund, Reyes-Rodriguez, Stuebe, Dennis and Patel2021; Tomfohr-Madsen et al. Reference Tomfohr-Madsen, Roos, Madsen, Leason, Singla, Charlebois, Tomasi and Chaput2022) and outperforms waitlist and active control conditions (Venturo-Conerly et al. Reference Venturo-Conerly, Eisenman, Wasil, Singla and Weisz2023). For vulnerable individuals facing stigmatising mental illnesses, peer delivery may be the most acceptable and most culturally appropriate method to receive evidence-based treatment (Tomfohr-Madsen et al. Reference Tomfohr-Madsen, Roos, Madsen, Leason, Singla, Charlebois, Tomasi and Chaput2022). Peer education sessions adopted group classroom sessions, street plays, participation in community activities and one-to-one meetings. Peer education has been reported to be enjoyable and preferred, and can even be more effective than information transmission by professionals, especially when it comes to sensitive topics (Topping Reference Topping2022). This review also suggests that the provision of a peer-support network was enough to alleviate some mental health symptoms. Social support serves as a way to build an in-group community to mitigate social isolation, which is associated with poor mental health outcomes (Grønlie and Dageid Reference Grønlie and Dageid2017; Leigh-Hunt et al. Reference Leigh-Hunt, Bagguley, Bash, Turner, Turnbull, Valtorta and Caan2017). The perception of social support is sometimes more important than the actual support received, and acts as a coping mechanism for daily stressors and a protective mechanism against the development of more severe psychological distress (Casale et al. Reference Casale, Boyes, Pantelic, Toska and Cluver2019). The key benefit of peer support is the mutual benefit to all parties involved in giving and receiving support (Arndt and Naudé Reference Arndt and Naudé2020).
Key components to cultivate effective peer-leaders
The TRUST framework (Training, Referral pathways, Understanding the remit of their role, Supervision, and recognition that Talking helps) outlines the needs of young peer supporters (Wogrin et al. Reference Wogrin, Willis, Mutsinze, Chinoda, Verhey, Chibanda and Bernays2021; Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022). Training of peer leaders was mentioned in all studies included in this review, and was often conducted directly by the research team or by adult coaches (Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023). While training is critical for adult and youth leaders alike, it was especially important for youth leaders as they were often younger and had less education and work experience compared to adults (Maticka-Tyndale and Barnett Reference Maticka-Tyndale and Barnett2010). To mitigate this, it was helpful to provide guidance on the topics of health literacy, cultural practices, mental health, as well as the soft-skills of communication, leadership and conflict resolution. Refresher trainings, included in two studies (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Mathias et al. Reference Mathias, Singh, Butcher, Grills, Srinivasan and Kermode2019), have been shown to be useful for longer interventions to ensure reinforcement of accurate information, adjustments and feedback (Maticka-Tyndale and Barnett Reference Maticka-Tyndale and Barnett2010). Similarly, many studies involved regular supervision sessions to ensure proper conduct of intervention sessions with an opportunity to adapt intervention delivery based on feedback. Effective supervision was critical, and studies without supervision showed diminished quality and fidelity in intervention delivery (Sharma Reference Sharma2002). Youth engagement can be challenging and emotionally draining (Simms et al. Reference Simms, Weiss, Chinoda, Mutsinze, Bernays, Verhey, Wogrin, Apollo, Mugurungi, Sithole, Chibanda and Willis2022), and regular supervision with the research team can help prevent the development of secondary trauma and ensure the welfare of the youth leaders. Osborn et al. (Reference Osborn, Ndetei, Sacco, Mutiso and Sommer2023) provided a relevant protocol for the training and supervision of lay youth providers, which can guide future research and implementation of peer-led interventions.
Cultural appropriateness and generalisability in reporting mental health
Cultural appropriateness allows for the acceptability of an intervention, affecting the effectiveness and the generalisability of study outcomes. There is a lack of culturally validated mental health scales in LMICs, and many studies currently use scales validated in HICs (Kaiser et al. Reference Kaiser, Ticao, Anoje, Boglosa, Gafaar, Minto and Kohrt2022). This makes it difficult to effectively assess and generalise quantitative clinical improvement in mental health in LMICs. Qualitative methods of assessing mental health interventions may be beneficial in such circumstances, and Ferris France et al. (Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023) had justified their intentional use of a qualitative rather than quantitative assessment of their intervention by likening the comparison of health outcomes scales across different cultures to compare chopsticks with forks. Moreover, mental health is still a highly stigmatised topic, which can affect participant reporting of symptoms and therefore study outcomes. Several studies reasoned a low baseline report of symptoms due to reluctance in reporting mental health difficulties (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018) and challenges in articulating feelings and emotions (Ferris France et al. Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023). Venturo-Conerly et al. (Reference Venturo-Conerly, Wasil, Osborn, Puffer, Weisz and and Wasanga2022c) has adapted principles from the Belmont Report (Department of Health et al. 2014) and Declaration of Helsinki (World Medical Association 2013) to form guidelines for addressing risk in high-stigma environments. Methods that can ensure cultural appropriateness would be including local mental health providers familiar with local resources, regulations and norms, who can identify nuances of distress that may indicate risk. Training lay providers in risk assessment and the appropriate channels for escalation and management is also crucial.
Sustainability of peer-led interventions
Ideally, once an intervention is implemented and effective, it would continue without reliance on the research team. For this to happen, it needs to be integrated within existing structures through partnerships with governments and established institutions. This is the goal of many interventions (Im et al. Reference Im, Jettner, Warsame, Isse, Khoury and Ross2018). Some studies reported that benefits of the intervention might revert back to baseline once the group stops meeting regularly (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Kermode et al. Reference Kermode, Grills, Singh and Mathias2021). However, facilitating such institutional or governmental partnerships is often a challenge (Kelly et al. Reference Kelly, Somlai, Benotsch, Amirkhanian, Fernandez, Stevenson, Sitzler, McAuliffe, Brown and Opgenorth2006; Maticka-Tyndale and Barnett Reference Maticka-Tyndale and Barnett2010). This is echoed by several studies that listed integrating peer education into existing educational institutions as a barrier, citing the lack of support for peer educators and the research team, and poor cooperation from teachers and school administrators (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011; Mohamadi et al. Reference Mohamadi, Paryab, Mousavi, Keramat, Motaghi and Garkaz2021). Osborn et al. (Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021) reported an unexpected government ban on research activities in educational institutions that limited the assessment of an extended time point and led to significant attrition in participant attendance. Several studies (Dow et al. Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a; Osborn et al. Reference Osborn, Venturo-Conerly, Arango, Roe, Rodriguez, Alemu, Gan, Wasil, Otieno, Rusch, Ndetei, Wasanga, Schleider and Weisz2021) have integrated peer-led interventions into stepped-care models, where at-risk youth are referred to trained adult clinical providers within established care systems to address elevated symptoms and emergencies. Peer-led models do not function in isolation; their integration into existing care systems ensures a more seamless process for connecting new interventions with established frameworks. However, the reliance of sustainability on external organisations leaves them vulnerable to sudden policy changes and motivations of the existing structures. It is therefore vital for research teams to empower the youth and their communities and establish delivery mechanisms that reduce barriers to these partnerships. To overcome such challenges, Dow et al. (Reference Dow, Mmbaga, Gallis, Turner, Gandhi, Cunningham and O’Donnell2020) developed a protocolised manual allowing for scalability and reproducibility, while Ferris France et al. (Reference Ferris France, Byrne, Nyamwanza, Munatsi, Willis, Conroy, Vumbunu, Chinembiri, Maedziso, Katsande, Dongo, Crehan and Mavhu2023) provided an online Toolkit to offer continuous support and mentoring to peer coaches.
Benefits and challenges to task shifting to peer-leaders
The benefits and challenges of a peer-led intervention reported by studies in this review has similarly been described in the literature. Peer-led interventions allow for the contribution of lived experience, which makes interventions more equitable and contextually acceptable. Peers have been used globally in the delivery of maternal mental health care, and a systematic review by Atif (Reference Atif2015) reported that peers who were mothers themselves were perceived as more trustworthy, friendly and experienced and likely to share personal insights to benefit other mothers (Nankunda et al. Reference Nankunda, Tumwine, Nankabirwa and Tylleskär2010). Moreover, the study reported that peer openness about their health status contributed to overcoming illness-related stigma (i.e. HIV status). Peer-led interventions are beneficial to adolescents as they provide a unique form of rapport that allows discussions of sensitive topics such as sexual health (Visser Reference Visser2007). Adolescents frequently relate more easily with a peer and communicate in a language that is understandable and accessible (Visser Reference Visser2007). Positive peer interactions also allow for role modelling, and often peer leaders are viewed as educators in their communities who can impart contextually relevant information (Atif Reference Atif2015). A key benefit of a peer-led component is the reciprocity and mutual benefit to the giver and receiver of the intervention (Naudé Reference Naal2017). Positive peer interactions are associated with higher self-esteem, wellbeing and health literacy for all parties involved in the delivery of a peer-led intervention (Naudé Reference Naal2017). The act of delivering an intervention can also improve confidence, fulfilment and life skills, and can even improve an individual’s social status and mobility within communities (Alcock et al. Reference Alcock, More, Patil, Porel, Vaidya and Osrin2009; Atif Reference Atif2015).
Barriers to implementing a peer-led interventions include ensuring a stable workforce. Peer leaders may leave for university, seek other employment opportunities or eventually age out of the role, requiring ongoing training of an evolving workforce (Okoroafor and Dela Christmals Reference Okoroafor and Dela Christmals2023). In addition, adolescents often lack full control over their time and environment, with non-adherence commonly resulting from school, home commitments and travel. Balaji et al. (Reference Balaji, Andrews, Andrew and Patel2011) emphasises the importance of incentives such as remuneration, certificates, and prizes to help support the project. Providing compensation also allows lay providers to devote time and energy to intervention training and delivery, freeing them from the need for alternative employment (Venturo-Conerly et al. Reference Venturo-Conerly, Roe, Wasil, Osborn, Ndetei, Musyimi, Mutiso, Wasanga and Weisz2022a). Peer facilitators require significant support from the team, including time for training and supervision, as well as mentorship from trained staff (Kermode et al. Reference Kermode, Grills, Singh and Mathias2021). Although lay providers have been shown to effectively deliver mental health interventions (Singla et al. Reference Singla, Kohrt, Murray, Anand, Chorpita and Patel2017; Sikander et al. Reference Sikander, Ahmad, Atif, Zaidi, Vanobberghen, Weiss, Nisar, Tabana, Ain and Bibi2019), there is reluctance in trusting the ability of task-shifted workers to carry out tasks with the same quality as a trained professionals (Kermode et al. Reference Kermode, Grills, Singh and Mathias2021). A recent meta-analysis indicates that they may not achieve the same level of effectiveness as professionals (Venturo-Conerly et al. Reference Venturo-Conerly, Eisenman, Wasil, Singla and Weisz2023). However, it is crucial to understand that peer-led interventions are not intended to replace those provided by trained professionals, but rather to serve as a scalable, accessible, and affordable (Singla et al. Reference Singla, Meltzer-Brody, Silver, Vigod, Kim, La Porte, Ravitz, Schiller, Schoueri-Mychasiw, Hollon, Kiss, Clark, Dalfen, Dimidjian, Gaynes, Katz, Lawson, Leszcz, Maunder, Mulsant, Murphy, Naslund, Reyes-Rodriguez, Stuebe, Dennis and Patel2021) alternative to address the mental health treatment gap in low-resource settings with insufficient professional human resources. Therefore, despite these existing challenges, peer-led interventions remain a promising public health strategy for improving youth mental health in LMICs.
Limitations
This review is not without limitations. It is possible that eligible articles were left out by our search method, either due to database selection, applied inclusion and exclusion criteria or missed search terms. Moreover, many of the studies included in this reviews are pilot studies or cross-sectional studies without a comparative group. At least three studies cited a small sample size (Mathias et al. Reference Mathias, Pandey, Armstrong, Diksha and Kermode2018; Osborn et al. Reference Osborn, Wasil, Venturo-Conerly, Schleider and Weisz2020a) and inadequate power (Balaji et al. Reference Balaji, Andrews, Andrew and Patel2011), which limits generalisability. There is also a potential publication bias where published literature may over-represent positive outcomes. This, along with the absence of a formal quality assessment of the included studies, can prevent a more accurate appraisal of the value of reported results to the field. Furthermore, this scoping review aimed to include all LMICs, which are inherently diverse in terms of cultural contexts, health systems, and social landscapes, potentially making generalisations among them inaccurate. A fourth limitation arises from incomplete information on the characteristics of certain interventions, where several studies offer a limited description of their intervention components. Combined with the heterogeneity of the studies included, this constituted an obstacle in comparing these studies. While the authors have attempted to minimise this by extracting information from available study protocols, the adoption of standardised presentation and evaluation for such peer-led youth mental health interventions will be helpful to support future meta-analyses and improve the comparability of study results.
Conclusions
This scoping review highlights the breadth of peer-led interventions targeting youth mental health in LMICs, shedding light on their unique mechanisms of promoting mental health with lived experience, camaraderie and reciprocity. Future research should expand to include the perspectives of key stakeholders – peer-leaders, research teams, and regulatory bodies – focusing on factors including fidelity, feasibility and acceptability to enhance implementation insights. While peer-led interventions still rely on adult professionals support; they represent a valuable, scalable and practical strategy to bridge the human resource gap in youth mental health across LMICs.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2024.149.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/gmh.2024.149.
Data availability statement
Additional review data will be shared upon request by inquiry to the corresponding author.
Acknowledgements
We would like to acknowledge and thank the authors of all the studies reviewed. Thank you to Beth Blackwood for helping to develop and execute the search strategy.
Author contribution
DWSC, DJM, DD: refining the scope of the review, data screening, extraction, analysis, literature search, manuscript writing and review; BB: development and execution of the search strategy; PP: refining the scope of the review, manuscript review; DWSC, DJM: formatting tables and figures; all authors read and approved the final manuscript.
Financial support
This study was supported by Duke-NUS Medical School and its Open Access publishing agreement with Cambridge.
Competing interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Open access
Comments
Dear Editor,
We wish to submit a scoping review entitled “A Scoping Review on Peer-led Interventions to Improve Youth Mental Health in Low- and Middle-Income Countries” for consideration by Cambridge Prisms: Global Mental Health.
In this paper, we explore the evidence for peer-led interventions aimed to improve the mental health of youth aged 10-24 years old and living in low- and middle-income countries. To our knowledge, this is the first review describing the delivery, common practices, effectiveness and challenges of such interventions in this setting. Our findings suggest that peer-led models are a valuable strategy for policymakers to leverage in current and future efforts to address the youth mental health treatment gap. The concept of task sharing has emerged as an increasingly viable strategy to addressing global mental health. Extending this approach further to youth peer leaders represents an even greater opportunity, making this paper highly suitable for publication in your journal.
We confirm that this work is original and has not been published elsewhere, nor is it currently under consideration for publication elsewhere. All authors have approved this manuscript and agree with its submission to Global Mental Health.
We have no conflicts of interest to disclose. Please address all correspondence concerning this manuscript to me at danachow@u.duke.nus.edu.
Thank you for your consideration of this manuscript.
Sincerely,
Dana Chow
BSc Biochemistry
Duke-NUS Medical School, Singapore, 4th year Medical Student