Promotive and preventive interventions for adolescent mental health in Sub-Saharan Africa: a combined scoping and systematic review

Psychosocial interventions

Psychosocial interventions (table 2) all had improving AMH as their primary aim. Interventions focused on MH literacy; expression and/or strengthening individual psychosocial skills.

Mental health literacy

Three universal MH literacy programmes were identified. Two of these used educational sessions to target adolescents in a school setting32 33; the third was a global WHO community awareness campaign.34 Quasi-experimental evaluations of these programmes found positive changes in AMH knowledge and understanding, but not in attitudes to AMH.

Expressive-creative interventions

Expressive-creative modalities were at the core of five interventions that aimed to support participants in expressing their thoughts and emotions in the context of bereavement. Targeted at Ebola-affected adolescents, the ‘Playing to Live’ intervention35 combined art, play and yoga therapy with life-skills training, while the ‘Comfort for Kids’ intervention36 used drawing and writing. Both interventions showed positive outcomes in pre–post evaluation studies. Targeted at orphans, a pilot RCT with orphans found that Memory therapy37 38—where participants explore their life story through artwork—strongly reduced general MH symptoms, whereas another small RCT found that the use of Emotional writing was not effective.39

Psychosocial skills-based interventions

Psychosocial skills development was at the core of 25 interventions. Thirteen of these shared a similar aim of equipping adolescents with psychosocial knowledge and skills to reduce distress and increase resilience in a general context of adversity. Combining elements of CBT, social emotional learning, positive psychotherapy and stress management approaches, they varied slightly in their focus to strengthen intrapersonal (eg, recognising and managing emotions, self-esteem, resilience), interpersonal (eg, communication, empathy), cognitive skills (eg, decision-making, problem-solving) and/or relaxation or mindfulness techniques. While most of these interventions aimed to bring about individual-level changes, they were mostly delivered in a group setting, using multiple workshops or therapy sessions. There was also one single-session intervention, focused on character strengths, which was delivered digitally.40 A large RCT (n=1288) to test this intervention in a face-to-face, multisession format is currently underway.41 Quantitative evaluations of these interventions (including six RCTs) reported mixed findings across a variety of outcomes. Six interventions were universal, school-based programmes,42–47 of which the ‘ERASE stress-prosocial programme’, ‘Youth First’, and the ‘Living Well Programme’ demonstrated the most consistent evidence of feasibility and effectiveness. Five interventions targeted orphans47–54 and reported predominantly positive outcomes. Three of these interventions combined therapeutic approaches with peer support modalities. For example, over the last 20 years, the Balekane EARTH programme in Botswana (previously named Ark for Children) has sent groups of orphans to a wilderness-based psychosocial strengthening camp, to build feelings of trust and kinship as a protective factor for mental health.53

Furthermore, an intervention aimed at specifically preventing aggression problems in young males showed positive results,55 while an RCT of an anxiety prevention intervention in children with visual impairments56 found no improvements.

Four interventions were implemented primarily in the context of bereavement, targeting orphans,57 58 ALHIV59 or adolescent females.60 These interventions aimed to support adolescents in expressing loss-related emotions and increase their understanding of grief and support services, in addition to equipping them with psychosocial skills. Evidence for these interventions was limited, as it consisted of two qualitative evaluations and two RCTs, of which one—for a bereavement support group called ‘Abangane’—reported small improvements.60

A substantial number of interventions were implemented in the context of HIV prevention and health promotion. Four of these interventions had mental health as their primary focus, with core components based on psychological therapies. A universal, community-based art programme ‘Make a Difference about Art’,61 where children create ‘hero’ books about their lives; a counselling intervention targeted at migrants62; a mobile-based, counselling and peer-support Whatsapp group for ALHIV63; and Sauti Ya Vijana, which combined several therapeutic approaches targeted at ALHIV,64 65 all aimed to reduce psychosocial problems and increase HIV insight as a mechanism to improve HIV-related outcomes, such as adherence to antiretroviral medication. None of these studies found significant changes in mental health outcomes before and after intervention. Two adolescent-centred health promotion interventions could also be classed under this category, although they had a broader focus of developing psychosocial skills and informational resources to enable adolescents to become effective HIV agents: the Young Citizens Programme increased self-efficacy and collective efficacy through public education and community mobilisation,66 while a peer-led education programme (‘Vhutshilo’), found no changes.67

HIV prevention and management

Seven interventions were multicomponent, family-based HIV prevention programmes. Programmes were either universal,68–70 or targeted ALHIV,71 72 adolescent females,73 OVC,74 those with a HIV positive parent75 76 or mild depressive symptoms.77 Educational components covered topics such as caregiver relationships and family strengths; HIV transmission and treatment knowledge; talking about sensitive topics such as puberty and HIV; stress management and coping; and gender/economic empowerment. These interventions were evaluated rigorously, with six RCTs, but evidence of their impact as mental health prevention interventions was inconclusive: four interventions (‘Our Family, Our Future’, ‘FSI-HIV’, ‘IMARA-SA’ and ‘Let’s Talk’) reported improved outcomes while three others did not.

Four interventions focused on improving health service delivery processes to achieve better health outcomes in vulnerable families78 79 and ALHIV.72 80 These interventions had no didactic element, but instead provided holistic, integrated support by offering home visits, linkages to external providers and livelihood support (in two interventions). Again, the evidence was mixed: two of these interventions, the ‘Bantwana initiative’ and ‘Zvandiri’ found a positive change in MH outcomes, but the other two did not. Finally, HADITHI aimed to increase HIV status disclosure and promote resilience,81 but an RCT found no MH improvements.

Parenting interventions

While improved family connections were addressed as a protective factor in most HIV prevention interventions described above, a further three interventions aimed to strengthen parenting skills as a protective factor for (mental) health problems. Two of these programmes aimed to support early development of children,82 83 and prevent child abuse.84 The third intervention, Sinovuyo Teen85, targeted adolescents and their parents with sessions around stress/anger management, family problem-solving and relationship building. None of these interventions showed improvements in psychosocial outcomes.

Gender equity

Three interventions specifically focused on empowering women and challenging gender norms. One of these was a community-based empowerment intervention for teenage mothers, which aimed to improve their mental health by creating a supportive social environment.86 The two other interventions aimed to build positive mental health as a mechanism to either promote gender equity and transform gender norms,87 or to equip adolescent females with the skills to make strategic life choices and stay safe from sexual abuse.88 Robust quantitative evidence of impact has not been found: an RCT of the latter intervention found no improvements and the first two interventions were evaluated qualitatively.

Economic livelihood interventions

The final category consisted of 10 multiyear economic livelihood interventions. These interventions did not have any psychological content, but used asset-based approaches to promote mental health.

Five interventions in this category were led by the same principal investigator and implemented in Uganda. In Bridges to the Future, Suubi, Suubi-Maka (targeting AIDS orphans) and Suubi+adherence (targeting ALHIV), adolescents were provided with a matched savings account, in addition to financial management training and (peer) mentorship.89–97 The Suubi+Adherence intervention also incorporated HIV prevention elements of the VUKA curriculum (discussed above) to target HIV-related outcomes. An evaluation of Suubi4Her98, which targets females and combines this savings-led economic empowerment approach with a family strengthening component, is currently underway. Furthermore, targeting orphans, SHAZ! provided a vocational training package that included a conditional stipend.99 Trickle Up Plus, OSCAR and CT-OVC also used cash transfers, but these were given to caregivers and not directly to adolescents.100–102 A final intervention covered adolescents’ school fees and uniforms.103 Evidence of possible effectiveness of interventions was strongest in this category. All implemented interventions were evaluated through one or more (cluster) RCTs that that used follow-up measurements over multiple years and showed largely positive results.

Surface structures

Most cultural tailoring took place in the ‘superficial’ surface-structure domain. Examples of this were found in 17 interventions and included the use of local illustrations/cartoons as opposed to written material in a context of low literacy36 58 59 67 69 75 104 105; using indigenous music, folktales/stories, metaphors, proverbs, local legends, games, crafts and/or clothing45 48 51 53–55 59 60 75 87 104 105; or organising community events involving traditional leaders and ceremonies.45 53 54 87 88 Nigerian researchers who implemented the ‘Thinking Group’ for adolescent boys with aggression problems, adapted from the ‘BrainPower programme’ (USA), provided a specific example: in their original intervention, a principle ‘STOP, THINK before ACTING’ was explained with an analogy of traffic lights (red for Stop, amber for Think and green for Act). This was replaced with a reference to football, a game popular among many boys in the country. The referee would STOP the game for a foul, hand out a yellow card (THINK), and the player would subsequently be ACTING properly to avoid a red card and eviction from the game.55

Deep structures

Cultural tailoring at a deep structure level requires an understanding of how sociocultural forces influence mental health, and how the target population understands the cause, course and treatment of mental health issues. Authors of 13 studies described how an understanding of these ‘deep structures’ shaped their intervention, as follows:

For seven interventions, authors explained how their approach aligned with context-specific views and values: Formative work for ‘FSI-HIV’ and ‘Our Family, Our Future’, identified family communication skills, connectedness and good parenting as context-specific strategies that could promote AMH. This led authors to implement a family-based intervention focused on strengthening parent–child relationships and family narrative.23 106 Similarly, authors stated that since SSA societies are often highly collectivistic, group approaches that promote family, mutual unity and collective responsibility, would be most appropriate.54 They highlight that interventions should not over-emphasise Western individualistic values such as independence and critical thinking that might be at odds with local values.45 Strength-based approaches that use conceptualisations of positive mental health rather than psychopathology were also said to be more suitable when a focus on psychopathology might invoke stigma.23 40–42

In terms of context-specific content or focus of interventions, it could be argued that sessions on topics like HIV and orphanhood respond to risk factors for mental health issues that are particularly relevant in SSA contexts. However, specific descriptions of content adaptations to reflect the local context were few: CHAMP-SA and VUKA both had an increased focus on loss and bereavement, as it was felt that this set the tone for all further communications in the intervention.69 104 The cartoon used in the intervention was adapted to include a character who kept a Memory Box, to encourage participants to create one themselves. Similarly, the cartoon in Let’s Talk included a character who was dealing with bereavement and chronic illness, while living in foster care.105 A bereavement intervention was adapted so that it reflected a reality of multiple losses, as opposed to one loss. In addition, as many participants lived in a situation of ongoing adversity, coping techniques focused on dealing with challenges that cannot be changed and are out of one’s control.59 In Rwanda, the FSI-HIV intervention included an optional session on genocide-related trauma psychoeducation.23

Explorations into local understanding, terminologies or expressions of mental health were described as part of developmental work for five interventions: In the adaptation phase of their CBT for South African children with visual impairments, authors explored understanding of the concepts ‘feelings’ and ‘thoughts’. Although participants were found to have an adequate understanding of these concepts through previous exposure within the school curriculum, the authors stressed the importance of such explorations preintervention.107 Formative work for the ERASE intervention described that Tanzanian children had not had this exposure, and therefore, this intervention allowed more time to learn about relevant concepts. In addition, body-oriented techniques were added, since distress was often expressed through somatic complaints.45 Explorations of local conceptualisations and terminologies of mental health issues were specified by some: exploratory work for the FSI-HIV23 identified six locally defined syndromes indicative of mental health needs of children, while others used local terminology such as ‘stress’ and ‘everything is too much’ to refer to depression.58 106

Finally, four groups of authors described how their intervention integrated local traditions and interpretations regarding the causes and treatment of AMH issues: Instead of employing a Western cognitive-behavioural framework for dealing with ‘automatic negative thoughts’ that was part of the original ERASE manual, the adaptation incorporated a local perspective that negative thoughts represent admonitions from ancestors for misdeeds. Students were encouraged to reflect on their conduct and learn from it rather than challenge or dispute negative thoughts. Similarly, in addition to Western desensitisation strategies, students were introduced to traditional ceremonial healing. A collective grief ritual was also performed during a session.45 The Abangane bereavement intervention also merged CBT sessions with discussions around local traditions and myths about death and grief.60 The Balekane EARTH intervention is a wilderness-based retreat which follows rites of passage that shared many similarities to historical initiation as part of Setswana culture in Botswana. In addition, ceremonies are held when children depart and return from the camp.54 The formative work for VUKA showed that in the South African context, interventions need to address social interpretations of mental ill health such as bewitchment and demonisation, yet the authors did not describe how VUKA did this.71

Indicators of positive mental health

Studies showed varying impacts across a wide range of positive mental health indicators, which typically fall within the realm of mental health promotion:

Self-evaluation outcomes were measured by three universal psychosocial skills-based interventions, which all reported moderate improvements. The evaluation of Memory Therapy, a creative-expressive intervention, found improved self-esteem (d=1.33) and self-efficacy (d=1.25). Family-based HIV and MH prevention interventions did not typically measure these outcomes; there was one study that measured self-esteem but found no improvements. Zvandiri, an HIV service delivery intervention found improved confidence, self-esteem and self-worth, although effect size calculations were not possible. Five similar economic livelihood interventions did not find consistent improvements in self-evaluation scores: only Suubi-Maka and Bridges to the Future found improved levels of self-concept and self-esteem (effects size calculations not possible).

Interpersonal skills were measured by two universal psychosocial interventions, of which only the ERASE stress prosocial intervention reported substantial improvements in social difficulties and prosocial skills. Hopelessness was assessed in one universal psychosocial intervention, which found no improvements, and across five similar economic livelihood interventions targeted at OVCs, which all reported reduced hopelessness levels, with small or unknown effect sizes. Two psychosocial and one family-based HIV and MH intervention assessed resilience. Only the universal, school-based RAP-A programme reported a small improvement (d=0.27) in scores. Grief was addressed by two psychosocial interventions targeted at OVCs, of which only Abangane reported small improvements (d=0.14 and d=0.21).

Impact on indicators of negative mental health

In terms of indicators of negative mental health, almost all studies included a measure of general mental health symptoms, depression and/or anxiety. Of four universal psychosocial interventions, the ERASE programme reported an improvement in anxiety scores, while a psychoeducation and life skills programme generally improved internalising and externalising problems (unknown effect size). Shamiri digital moderately improved depression but did not change general well-being or anxiety. Of five psychosocial/bereavement interventions targeted at OVCs, three reported significant changes: Memory Therapy (d=0.80), a peer support group (d=0.84) and Abangane (d=0.21). Of the seven family-based HIV and MH prevention interventions that measured MH symptomology, only three reported improvements in depression/anxiety scores: Our Family, Our Future reported a large effect size (d=0.72), while FSI-HIV and IMARA-SA reported improvements with unknown effect sizes. HADITHI, a HIV status disclosure intervention, found increased rates of MH symptoms at 6 months postintervention, but this reduced at later follow-ups. Economic livelihood interventions showed a more consistent impact, with reduced symptoms of MH conditions reported across all studies. A moderate effect size was reported for the Suubi intervention (d=0.65), while others reported small or unknown effect sizes. However, reductions were not all sustained at follow-ups of more than 2 years,94–96 while SHAZ! only reported a moderate effect size improvement after 24 months (d=0.63). The CT-OVC intervention only showed small improvements (d=0.38) in scores of males.101

Trauma was assessed by four studies. Two MH and HIV prevention interventions (Sauti Ya Vijana and IMARA-SA) did not find improvements in trauma scores, but two economic livelihood programmes with OVCs (OSCAR and Trickle Up Plus) did.