1 Introduction

Depression prevalence is on the increase with global estimates of 280 million people living with depression in 2019, including 23 million children and adolescents [1]. In Kenya, studies have reported high depression prevalence rates [2,3,4,5,6,7].

Depression is associated with suicidal thoughts [1, 8]; poor antiretroviral therapy adherence among people living with HIV [2, 9]; poor physical functioning [4]; heightened stigma and other emotional challenges [8]; and lower quality of life [10] among others.

People suffering mental ill-health require social support in educational programmes, employment, housing, and participation in other meaningful activities [1] to deal with associated psychosocial challenges. This is important to prevent the use of dysfunctional coping strategies including isolation and substance use [8].

Due to limited budgets devoted to mental health care in low- and middle-income countries (LMIC), the responsibility of detection and treatment of mental disorders is shifted to primary health care [11]. However, inadequate training of primary care providers often leads to delays in diagnosis and management of mental disorders [1]. Where such services are available in primary health care facilities, they are underfunded and of poor quality. This calls for innovative and cost-effective approaches in the management of mental health in sub-Saharan Africa countries.

The need for task shifting in mental health care has seen a range of community-based interventions from the use of peers [12]; brief training [11]; guidance and counselling and life skills education [15]; talk therapy [16, 17]; digital interventions [18, 19] and multisectoral approaches including cash transfer programs [13, 14, 20], among others. These interventions are provided by lay professionals without prior formal training in mental health who undergo short duration training [21, 22].

Following the interventions, a number of outcomes have been reported that include increased ability for depression diagnosis [11]; improvement in skill sets among rural primary care providers [17]; reduced distress and HIV-related stigma [17]; clinically significant reduction in depression symptoms [16, 20], and decreases in cortisol [13]; and improvement in quality of life [13, 14, 23]. Some interventions have been reported to be cost effective [19]; and feasible and acceptable [12, 18, 23] while others report no measurable effects on mental health outcomes [21].

Whereas Kenya is ranked fifth in Africa in the number of depression cases, many individuals do not seek help due to stigma and discrimination, inadequate skilled human resources for mental health and lack of integration of mental health within primary care [24, 25]. One in four persons seeking healthcare in Kenya have a mental health condition [26] while only 27% of women and 21% of men, respectively with depression and anxiety are receiving treatment [27]. That more than two-thirds of depressed people are left without treatment is an urgent problem that requires effective responses especially with inadequate funding for mental healthcare. Additionally, mental health conditions increase healthcare costs besides posing a significant drain on meagre family and state resources, hence requiring urgent cost-effective responses.

Therefore, this review seeks to synthesize the existing literature on non-pharmacological interventions and their effectiveness against depression for possible adoption and upscaling in primary health care.

2 Methodology

The scoping review was guided by the five stages of Arksey and O’Malley framework [28] and aimed to synthesize existing research on non-pharmacological interventions for depression in Kenya.

2.1 Identifying the research questions

The scoping review aimed to answer two questions: “What is the existing literature on non-pharmacological interventions for persons with depression in Kenya?” and “What is the effectiveness of non-pharmacological interventions for persons with depression in Kenya?”

2.2 Identifying relevant studies

We developed a search strategy that was used to conduct the review from six databases—AJOL, EBSCOhost, PubMed, ScienceDirect, Cochrane Library and International Bibliography of the Social Sciences (ProQuest). Databases were searched between 7th and 14th June 2023. A combination of search terms used incorporated Medical Subject Headings (MeSH) terms and keywords related to “depression” and “intervention” and their synonyms combined by Boolean operators (S1 File). The search strategy was adapted accordingly for all the databases. We also manually searched for relevant studies from reference lists.

2.3 Study selection

Studies were selected based on the following inclusion criteria: Published in English from 2000 to May 2023; based on a sample of any age from Kenya; and described a non-pharmacological intervention for depression/depressive symptoms. We excluded editorials, commentaries, conference presentations, review articles and studies whose outcomes did not target depression/depressive symptoms or reported pharmacological interventions for depression/depressive symptoms.

After articles were identified by title and abstract at screening, they underwent full text review by all reviewers. Cases of disagreement on inclusion of articles were resolved by HO independently reviewing abstracts to resolve the conflict. The process of study selection is presented in the PRISMA-ScR flow chart in Fig. 1.

Fig. 1
figure 1

PRISMA-ScR flow diagram

Full size image

2.4 Charting the data

HO designed an extraction template that was used to extract data. Data extracted included author, year of study and type of study, population, intervention and intervention outcomes and was guided by the research questions. Data extraction for the first two articles was done simultaneously by all reviewers for uniformity in extraction. Data for the remaining articles was extracted by JN, AM and FK. All papers were incorporated into the review without regard to methodological quality because quality appraisal is not mandatory in scoping reviews. A total of 24 articles were included in the final review (Table 1).

Table 1 Characteristics of included studies
Full size table

2.5 Data analysis

Narrative synthesis was used to group findings in the context of the intended outcomes set out in the research questions [29]. We sought and summarized data on non-pharmacological interventions and their primary and secondary outcomes. Emerging themes were then presented narratively.

2.6 Reporting

Reporting was guided by the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist (S2 File).

3 Results

3.1 Characteristics of the included studies

The database and manual search yielded 4009 studies and after removing duplicates (n = 679) and screening by title and abstract (n = 3209), 121 studies remained for full text review.

A total of 97 full-text articles were excluded and 24 included in the final review (Table 1). The included studies were all peer reviewed articles published between 2013 and 2023.

Twelve studies used experimental designs including pretest–posttest control group [26], single case with repeated measures [31], randomized trials [32,33,34,35,36,37,38,39,40] and cohort [41]. Quasi-experimental design was used in six studies [42,43,44,45,46,47]. Other designs include secondary analysis [48, 49]; longitudinal [50,51,52]; and mixed methods [53].

Twelve studies reported on comparison groups [30, 32,33,34,35, 37,38,39,40, 42, 45, 46]. In some studies, the control group involved treatment-as-usual (TAU) in the form of routine postnatal care [42] or antenatal care [45] or as waitlist [35, 40]; enhanced usual care (EUC) in primary healthcare facilities [32]; usual school activities [33]; standard care consisting of antiretroviral therapy adherence counselling and participation in support group [48] or antiretroviral therapy care in HIV-outpatient clinics [30]; and Early Childhood Development (ECD) only content [43].

Two studies reported on group interventions covering large geographical zones. In the first study, schools in the control arm were provided cash incentives of $240 annually to use for their school development projects [34]. Control locations in the second study were placed on a waiting list to enter the program at scale up [50].

3.2 Description of sample

The sample of 11,208 participants comprises adult women who had experienced gender-based violence [32], whose children were transitioning back to life from the streets [41], pregnant and new mothers [31, 40, 42, 43, 45, 53], HIV-positive [35] and those seeking fertility treatment [37]; adolescents and youth [33, 38, 39]; orphans and vulnerable children [34, 44, 49, 50, 52]; HIV-positive Gay, Bisexual and other Men who have Sex with Men (GBMSM) [48] and family caregivers [47]. Eleven studies targeted only women [31, 32, 35, 37, 40,41,42,43,44,45, 53].

3.3 Depression measures

In majority of studies, the primary outcome of interest was change in severity of depression/depressive symptoms. A variety of depression measures were used including the Patient Health Questionnaire-9 Depression Scale (PHQ-9) [30, 31, 43, 48, 53]; PHQ-8 [38, 39]; Depression Self-Rating Scale for Children (DSRS) [33]; Beck’s Depression Inventory-II (BDI-II) [36, 41, 42, 46, 47, 51]; Centre for Epidemiologic Studies Depression Scale Revised (CESD-R) [34]; Centre for Epidemiologic Studies Depression Scale (CES-D10) [50]; Youth Self-report [44, 52]; Mini International Diagnostic Interview (MINI 5.0) [35]; Edinburgh Postnatal Depression Scale (EPDS) [40, 45]; Zung Self-Rating Depression Scale (T-SDS) [37]; and Child Depression Inventory (CDI) [49]. The General Health Questionnaire (GHQ-12) assessed psychological distress (anxiety and depression) in one study [32] while another used the Hospital Anxiety and Depression Scale to screen for depressive symptoms [47].

3.4 Interventions

Twenty studies reported psychosocial interventions [30,31,32,33, 35,36,37,38,39,40, 42,43,44,45,46,47,48, 51,52,53] while four reported socioeconomic support interventions [34, 41, 49, 50]. These interventions lasted from 2 weeks to 4 years.

Interventions were provided by community health workers/volunteers (CHW/CHV) [32, 43]; community health nurses [42]; a group of lay professionals [33, 35, 39] supervised by qualified therapists [36, 53]; traditional health providers [51]; multidisciplinary teams of professionals [45, 52]; digital platforms [31, 37] peers [48]; and researchers [30, 44, 46].

3.4.1 Psychosocial interventions

Six studies explicitly used the term ‘cognitive behaviour therapy’ (CBT) [30, 31, 43, 47, 51, 53]. These interventions lasted from 6 weeks to 2 years. Group cognitive behavioural therapy consisted of psychoeducation, behaviour analysis, unhealthy thought patterns and problem solving. One study [53] reported a Learning Through Play Plus (LTP +) intervention combining Learning Through Play and Culturally adapted Cognitive Behaviour Therapy (CaCBT). The intervention comprised four components: supportive, educational, problem-solving techniques and a parenting program. Another study [51] reported on the mental health Global Action Programme Intervention guide (mhGAP-IG) which combined Cognitive Behavioural Therapy and training of community health workers. Finally, Green et al. [31] reported on the use of automation to deliver cognitive behavioural therapy through text-based conversations over short messaging service (SMS) and Facebook Messenger in addition to live support.

Fourteen studies [32, 33, 35,36,37,38,39,40, 42, 44,45,46, 48, 52] reported the use of broad psychosocial interventions including motivational interviewing, stress management, problem solving and provision of specific information. Two studies [37, 38] reported on the use of digital platforms to provide psychoeducation.

The use of peers to provide informational, empathetic and motivational support was reported by one study [48] while three studies reported on the use of interpersonal therapy (IPT) [35, 40, 46]. Finally, one study reported the use of group multisystemic therapy and psychoeducation [44].

3.4.2 Socioeconomic support interventions

Four studies reported on socioeconomic support interventions including cash transfer (CT) for poor households with vulnerable children [49, 50]. One study [34] reported on a longitudinal school support intervention comprising payment of school tuition fees, provision of school uniform and nurse visits. The final study reported on a group lending microfinance approach where group members used $0.20 each week to loan to other members of respective groups at 10% interest [41].

3.5 Outcomes

3.5.1 Clinical outcomes

Studies unanimously reported significant reduction in depression scores and symptom severity posttreatment [30, 32, 33, 35, 37,38,39,40, 42,43,44,45,46,47,48, 51,52,53]; buffering the increase in depression severity over time and achieving remission [34]; reduction in the likelihood of having depressive symptoms [50]; reduction in the prevalence of severe and moderate depression [36, 47]; and an increase in depression scores posttreatment in control groups [33, 46].

One study [49] found no significant association between depression and a cash transfer intervention. However, the study also found that orphaned and vulnerable adolescents (OVA) living in cash transfer households were on average 23% less likely to have depressive symptoms.

3.5.2 Non-clinical outcomes

Studies reported that interventions provided were acceptable and feasible in retention and feasibility. This was measured in terms of completion rates across both intervention and control groups [33]; high attendance [53] and adherence [35]; low attrition/loss to follow up [31, 34]; and high ratings of usefulness of intervention [39]. Session attendance and program participation was associated with significantly lower depression [32, 41]. Participants also reported positive experiences regarding shared experiences [43], self-care, ability to replace negative thoughts and improvement in interpersonal relationships [34].

4 Discussion

This scoping review aimed to summarize and present current literature on non-pharmacological interventions for depression in Kenya. The review identified available interventions, intervention providers and outcomes that targeted mostly women, adolescents and youth and family caregivers. Majority of interventions (n = 20) were psychosocial in nature incorporating cognitive behavioural therapy, play, automation and psychoeducation while four reported socioeconomic support in the form of cash transfers, microfinance and school support.

To enable task shifting, interventions were provided by lay health workers including community health workers/volunteers and community health nurses; traditional health providers and peers; and digital platforms. In all cases, lay professionals received limited training, which corroborates previous reviews that show the effectiveness to fidelity, acceptability and affordability of providing limited training for lay professionals in low- and middle-income countries [54]. Considering limited funding for mental healthcare in Kenya, these findings indicate that the use of lay professionals is a cost-effective approach to management of depression. With minimal training, task shifting can be achieved by having lay professionals and peers provide therapeutic interventions ranging from cognitive behavioural therapy, interpersonal therapy, multisystemic therapy and psychoeducation. Task shifting to lay professionals and peers is likely to address stigma that is a major barrier to mental healthcare access. Additionally, to leverage on technological developments in sub-Saharan Africa, the use of digital approaches including mobile telephones, internet accessibility and automation will likely increase access to mental healthcare among young populations. This will also provide an opportunity to compare outcomes of technology-supported interventions vis-à-vis other approaches.

Overall, findings show significant reduction in depression scores and symptom severity posttreatment, reduction in the odds of having depressive symptoms and in prevalence of severe and moderate depression. These findings support previous reviews that reported positive outcomes of using task shifting approaches in depression interventions [54,55,56,57] suggesting that diverse populations can benefit from non-pharmacological interventions. Findings also report feasibility and acceptability of interventions in terms of high attendance and completion rates and ratings of usefulness of intervention. This is especially so since interventions provided spaces for individual and group expression and addressed broad underlying issues. Ultimately, such approaches are aware that people suffering depression lack access to mental health care in their healthcare system, suffer poverty and hence, have little social capital and means of livelihood. Evidence suggests that providing information using health promoters [32, 33, 35, 39, 42, 43] and providing socioeconomic support [34, 41, 49, 50] are cost-effective and sustainable solutions in managing depression. These findings shed light on the socioeconomic dynamics of mental health, that is, the central role of poverty in access to treatment, and suggest that local community approaches are appropriate because of their all-inclusive nature.

This review also highlights cultural strengths that interventions can take advantage of. For example, existing community bonds allow for the successful use of microfinance among women groups to improve their economic and social condition [41]. With less focus on financial profiteering, microfinance group meetings are opportunities for social interaction. Such bonds enable sharing of the depression burden during group therapy which helps to reveal underlying issues e.g., gender-based violence and institutionalized stigma; and improves interpersonal relationships and social functioning. Mental healthcare providers can take advantage of cultural connections as a strength for both group- and person-centred therapy more so in sub-Saharan Africa where such bonds promote group solidarity, compassion and mutual respect.

Economic vulnerability has been cited as a risk factor in mental health outcomes working through social capital [41]. Evidence provided suggests the cost-effectiveness, feasibility and acceptability of socioeconomic support interventions. It can be argued that by utilizing multiple approaches involving entire families, socioeconomic support interventions are more likely to result in better mental, social, economic and other outcomes including buffering adolescents from the onset or worsening of depression symptoms and building resilience [34, 50]. For instance, a combination of school support and microfinance programs that provide both social protection and economic empowerment improves parents’ psychological functioning while keeping children in school. Children who would otherwise have dropped out of school are able to complete their education and avoid isolation and stigma while ensuring psychological wellbeing for the parents.

Although the body of knowledge remains limited, available literature suggests that non-pharmacological interventions have been implemented in low- and middle-income countries with varying levels of success [55,56,57,58]. The current findings highlight the emerging need for situational adaptations and integration into local primary healthcare. With slight variations, these interventions can be scaled up and integrated into primary care to manage depression in Kenya. Finally, findings suggest that socioeconomic approaches hold more promise in sub-Saharan Africa since they are broad-based and address the mental health consequences of poverty.

5 Strengths and limitations

The major strength of this study is that it is to our knowledge, the first scoping review exploring non-pharmacological interventions for depression conducted in Kenya. Findings provide evidence of the efficacy of psychological and socioeconomic approaches in depression management. However, our search was limited to six databases and for papers published in English which may have excluded papers published in other languages. Except for a few studies, the majority had short-term follow up periods which suggests the need for long-term monitoring of intervention-related positive mental health outcomes. Finally, the wide variety of interventions used, depression assessment instruments and study designs make it difficult to generalize findings or conduct a meta-analysis.

6 Conclusions

This scoping review aimed to synthesize research on non-pharmacological interventions for depression/depressive symptoms in Kenya and their effectiveness. This review responded to the urgent need for knowledge on available interventions and the limited literature available. Whereas available evidence suggests positive mental health outcomes in the short-term, cost-effectiveness, acceptability and feasibility, there’s need for further research on the longer-term success of interventions. Overall, these findings show potential for the integration and scalability of lay-professional-facilitated non-pharmacological interventions in existing mental health care in low resource settings.