The Short-Term Impact of a Combination Intervention on Depressive Symptoms Among School-Going Adolescent Girls in Southwestern Uganda: The Suubi4Her Cluster Randomized Trial

Abstract

Purpose:

This study aims to examine the short-term impact of a combined intervention consisting of evidence-based family economic empowerment (FEE) and multiple family group (MFG) interventions on depressive symptoms among school-going adolescent girls in southwestern Uganda.

Methods:

We analyzed longitudinal data from a cluster randomized trial. The sample consisted of 1,260 adolescent girls (aged 14–17 years at enrollment) recruited from senior one and senior two classes across 47 secondary schools in the southwestern region of Uganda. Participants were randomized at the school level to either the control condition receiving bolstered standard of care or one of the two treatment conditions—the treatment one condition receiving the FEE intervention or the treatment two conditions receiving both the FEE plus MFG interventions. Descriptive statistics and a three-level mixed-effects model were conducted to examine the effect of a combination intervention on depressive symptoms.

Results:

At baseline, there were no significant differences between the control condition and both treatment conditions. While all three groups experienced a substantial reduction in depressive symptoms from baseline to 12 months, the reductions were stronger for the two intervention groups. However, FEE + MFG was not significantly different from FEE at 12 months.

Discussion:

Results imply that the FEE intervention may be a promising tool in addressing depressive symptoms among adolescent girls. Therefore, to reduce the long-term implications of adverse psychosocial health during adolescence, policymakers and program implementers should explore scaling up economic empowerment interventions in similar settings to bridge the mental health treatment gap for adolescent girls.

Globally, child and adolescent mental health (CAMH) contributes significantly to the burden of diseases [1] with estimates indicating that 20% of children and adolescents experience a disabling mental illness [1]. Particularly, psychiatric conditions are responsible for 16% of the global burden of illness and injury among adolescents aged 10–19 years [1]. In Sub-Saharan Africa (SSA), depression is one of the most common mental health disorders [2–4]. For instance, among a sample of 1,260 adolescent girls in Uganda, 16% and 29% reported severe to moderate depressive symptoms, respectively [4]. Worse still, adolescents in SSA experience additional overlapping stressors including poverty, violence, and health-related difficulties, which have been associated with poor mental health [5–7]. Consequently, these stressors yield cumulative effects, resulting in negative mental health outcomes including depression among adolescents [7,8].

Moreover, most of the people experiencing mental health disorders do not have access to care. In low-income and middle-income countries, studies indicate that the treatment gap is about 93% of individuals without access to mental health services with virtually no coverage of evidence-based interventions—defined as practices or programs that have peer-reviewed, documented empirical evidence of effectiveness [9]. The situation is not different for adolescents living in SSA—a region heavily burdened by poverty and communicable and noncommunicable diseases.

Mental health services for adolescents are virtually nonexistent in SSA; hence, most young people experiencing mental illnesses do not have access to mental health services [10] due to several factors including scarcity of a professional workforce [11] and limited or nonexistent mental health infrastructure and investment [11]. Consequently, alternative approaches may be required in light of the barriers to the formal treatment of depression in SSA. One alternative may be to design combination interventions that are low-cost and accessible, stigma-free, and scalable in low-resource environments.

Within children and adolescents in low-resource settings, gender is a critical factor that influences mental health outcomes. In Uganda for instance, female adolescents are more likely to report mental health challenges compared to their male counterparts [4]. In addition, Rescorla et al. [12] examined mental health symptoms by gender across 31 societies (n = 55,508) including countries in both the developed and developing world. The study found that compared to males, female adolescents reported higher levels of overall emotional distress and more depressive symptoms. In SSA, compared to boys, adolescent girls and young women are three times more likely to have depressive disorders and to attempt self-harm [13]. Specifically, other studies have documented that female orphans in SSA exhibit higher levels of psychological distress, social isolation, loss of education, and risky behaviors compared to boys [14,15].

Even with documented detrimental impacts associated with undiagnosed and/or untreated mental health challenges [16], few services and interventions have addressed adolescents’ mental health needs effectively; even fewer studies have addressed the existing gender gaps and disparities in CAMH [4,17]. Yet, gender is a critical determinant of mental health and psychological wellbeing. Despite mounting evidence demonstrating the effectiveness of mental health interventions in low-resource settings [18], there is still a limited evaluation of effective interventions for adolescent girls in low-resource settings. Therefore, it is critical to creating evidence-based, effective, durable, and implementable mental health interventions for vulnerable subpopulations such as adolescent girls living in low-resource settings [19].

Research [20] has shown family cohesion and support as protective factors in preventing and reducing depression symptoms among children and adolescents. In addition, family social support was also found to be both a moderating and preventive factor [21]. However, adolescent girls in SSA experience multiple adversities including violence, illness, and death of parents who are compounded by poverty resulting in a severe lack of food and access to necessities [6,7]. Hence, this leads to feelings of hopelessness, a lack of empowerment, and poor mental health outcomes [22,23]. Therefore, given the difficulties that adolescent girls face in low-resource settings, further research is needed to assess the short-term and long-term impacts of combination interventions on their emotional and psychological wellbeing.

Toward that end, family-based economic empowerment interventions have been proven to address poverty and improve the psychosocial wellbeing of adolescents [24,25]. This might be attributed in part to the asset effect, as proposed by asset theory [26], which posits that parental/family asset ownership has a beneficial psychological and social influence on adolescents. In the same way, the multiple family group (MFG) intervention has been implemented in both the United States and SSA. Premised on the core components of the four Rs (Rules, Responsibility, Relationships, and Respectful Communication) and two Ss (Stress and Social Support), the MFG targets parenting skills and family processes. Multiple groups and family theories inform the four Rs and two Ss including aspects of psychoeducation and social group work, family systems theory, structural family theory, and social learning theory [27]. In the United States, and South Africa, the MFG intervention has been adapted and deployed among vulnerable adolescents [28,29] with findings indicating improved family processes (support, communication, and parent-child participation), adolescent mental health, self-esteem, and reduced risk behaviors. Despite support for effective adolescent mental health interventions in low-resource settings, there is still limited evaluation of effective interventions for adolescent girls. To address the existing gaps, the present study innovatively adds to the literature in three different ways. First, the study combines two effective interventions that have not been explored together in the SSA context. Second, there is still limited evaluation of effective interventions for adolescent girls; consequently, this study focuses on a vulnerable group of adolescent girls in a low-resource country in SSA. Finally, this study employs a cluster randomized trial design using longitudinal data and applying advanced statistical approaches. Therefore, the study sought to examine the short-term impact of an evidence-based MFG intervention and family economic empowerment (FEE) component tested in Suubi4Her study on depressive symptoms among 1,260 school-going adolescent girls in southwestern Uganda. Based on the theoretical tenets guiding both interventions, we hypothesize that compared to the control condition, and FEE alone, participants receiving both the MFG and FEE intervention will report lower depression symptoms at 12 months postintervention initiation. Moreover, investments in combination interventions are critical to providing an interdisciplinary multilevel response needed to reduce mental health challenges in low-resource settings.

Methods

Longitudinal data from the Suubi4Her (Hope for girls) study funded by the National Institute of Mental Health (Grant #R01 MH113486) were used in this study. The Suubi4Her study is a prospective cluster randomized clinical trial (2017–2022) conducted in 47 secondary schools located in five geopolitical districts of Rakai, Kyotera, Masaka, Lwengo, and Kalungu in southern Uganda. The present study examined the impact of an innovative combination intervention on depressive symptoms among 14-year-old to 17-year-old adolescent girls living in low-resource settings in Uganda. All schools included in the study were matched on the following characteristics. Attracting students from similar socioeconomic status, comparable school size in regard to the total number of students enrolled, same academic level of performance based on the past 3 years of national examination score—the Uganda Certificate of Education (UCE) administered by the Uganda Government’s Ministry of Education and Sports and location (urban vs. rural). Detailed information on study procedures related to participant recruitment and selection processes is described in the study protocol [30].

From a list of 111 potential secondary schools, 47 that met the inclusion criteria previously were selected. To minimize cross-condition contamination, randomization was conducted at the school level using the random assignment feature in SPSS software by the data manager to one of three study conditions: Control condition (n = 16 schools, 408 girls) or one of the two treatment arms, FEE (n = 16 schools, 471 girls) and MFG + FEE (n = 15 schools, 381 girls) (supplementary files provides the Consort Flow Diagram). Included schools were geographically far apart from each other—across five political districts—minimizing the possibility of contamination. To eliminate bias, all eligible participants in the selected schools were recruited.

A total of 1,260 adolescent girls (14–17 years) in their first or second year of secondary school were enrolled in the study. Adolescents were eligible to participate if they met the following inclusion criteria: (1) female; (2) aged 14–17 years; (3) enrolled in the first or second year of secondary school; and (4) living within a family (broadly defined and not an institution or orphanage, as those in institutions have different familial needs).

Data collection procedures

The study used interviewer-administered survey instruments to collect data with each survey lasting between 60 and 90 minutes. The survey instruments were translated into Luganda—the local language spoken in the study area. To ensure consistency and accuracy, the survey instruments were back-translated into English and the process was overseen by trained language experts from Makerere University School of Languages, Literature, and Communication in Uganda. Local research assistants who were fluent in both English and Luganda collected the data. In addition to completing the Collaborative Institutional Training Initiative Human Subjects certificate, research assistants received week-long training on good clinical practice, data collection techniques, and the protection of human subjects.

Ethical considerations

Participation in the study was voluntary. Caregivers and parents who were willing and interested in the study gave a written consent for the adolescent girls to participate in the study. In the same way, an adolescent written assent to participate was obtained once a caregiver’s written consent was received. To avoid potential coercion, the consent and assent activities were done separately for the adolescents and caregivers. Similarly, interviews were conducted in a private setting at the participants’ school. All procedures in the study were approved by the Institutional Review Board at Washington University in St. Louis (IRB- #201703102), the Uganda Virus Research Institute (GC/127/17/07/619), and the Uganda National Council of Science and Technology (SS4406). The parent study is registered in the clinical trials database under registration number NCT03307226.

Ethics and consent

Adolescents provided a voluntary written assent and caregivers provided consent for the adolescent to participate in the study. The recruitment and interaction with human subjects and their health information were completed as per protocols reviewed and approved by Washington University in St. Louis (IRB- #201703102), the Uganda Virus Research Institute (GC/127/17/07/619), and the Uganda National Council of Science and Technology (SS4406).

The study interventions

Participants in the control arm received services provided to all adolescent children in the region. In Uganda, all secondary school students must complete an Adolescent Sexual and Reproductive Health program. As a result, these curricula are considered standard care and were provided to all enrolled individuals regardless of whether they are in the control or treatment groups. The manualized curriculum on adolescent sexual and reproductive health is spread across a variety of academic areas and is given to all teachers and students. Students in each class learn about delaying sex, using condoms and contraception, preventing forced sex, and preventing substance addiction. In addition, gender equality and the importance of postponing marriage are also covered in this curriculum. To ensure uniform delivery of the Ministry of Education–authorized sex education curriculum, the research team performed induction seminars for all teachers participating in the study before its implementation [30].

Participants in treatment condition one received the usual care described above and a packaged FEE intervention. This included an incentivized youth development account (YDA) held in both the child’s and their caregiver’s name in a well-established and recognized financial institution. The child’s family and other relatives were allowed and encouraged to contribute to the YDA. The accumulated savings in the YDA were matched with money from the program by a ratio of 1:1. In addition, participants received four financial literacy training sessions, which have also been found to reduce depressive symptoms and enhance mental health functioning among adolescents in SSA [31]. Both caregivers and children were invited to attend the sessions. This is important because adolescents aged less than 18 years need parental/caregiver as a co-signer to open a bank account in Uganda. This is consistent with the Ugandan banking law which prohibits children aged less than 18 years from independently entering into a binding contract and operating a bank account. Participants in the program can invest up to 30% of their total matched savings in a family-based income-generating business. The remaining 70% of the savings will be used to support the education and skill development of adolescent girls. Participant’s access to the matching funds is conditional on completing financial literacy workshops during the intervention period [30].

Financial literacy sessions were designed to equip families with basic financial knowledge. Specifically, session one focused on the overview of financial literacy and budgeting, session two focused on saving, asset building, and asset accumulation, session three focused on bank services in the community, and session four focused on debt management, borrowing money, cost of borrowing, and the dangers of defaulting. The manualized financial literacy sessions were delivered by trained research assistants. During the training, caregivers and participants were given handouts for home assignments. For instance, during session one, caregivers and adolescents were given assignments to develop a financial plan and set financial goals and a budget template for families to develop their budget.

In addition to the FEE intervention described above, participants in treatment arm 2 (FEE + MFG) received a family-based dialogue and training delivered via the MFG aimed to strengthen family relationships and address mental health challenges that commonly occur in adolescence. The MFG intervention was co-adapted with stakeholders and study participants [32]. The MFG is a 16-week manualized intervention facilitated by trained parent peers and CHWs. The intervention is organized on the four Rs (Rules, Responsibility, Relationships, and Respectful Communication) and two Ss (Stress and Social Support) targeting skills and family processes. One session was delivered each week lasting 45–60 minutes and included role-plays, group discussions, and family activities.

The Suubi4Her MFG 16-session curriculum intervention (Appendix S1) was delivered by trained CHWs and peer parents under the supervision of trained project staff. Each group session involved 12–20 families and other extended family members including uncles, aunts, other siblings, and grandparents were invited to attend the sessions. A total of 12 facilitators both parent peers and CHWs and previously trained in delivering mental health interventions were trained. This is vital because both parent peers and CHWs live within the community and are trusted and respected community members, which facilitates building rapport between families and the facilitators. We used existing structures to recruit CHWs already exiting at the community level trained by the Ministry of Health and other not-for-profit organizations. In the same way, facilitators had to be fluent in writing and speaking Luganda-the local language used in the study area. To be recruited, facilitators had to be trusted and acceptable members of the community (recommended by school leadership), lived within one mile from the participating schools, be willing to committee 60 minutes a week for 16 weeks, willing to attend and complete training as a peer facilitator, and preferably aged between 30 and 60 years. The facilitators received a 2-day training provided by trained research assistants based at the field offices in Masaka and were provided with manuals to deliver the intervention. At the end of the training, facilitators completed the Knowledge Skills and Attitude Test and those with a pass of 85% and more were selected. At the end of the training, facilitators were awarded certificates of completion. For each school, the facilitators were paired (one male and one female) and delivered all the sessions together. At the end of each session, families were given roadwork to complete at home before the next session. The trainers (research assistants) met with the facilitators every 2 weeks to assess progress and address any concerns.

Two research assistants attended all sessions to ensure fidelity across facilitators and facilitate logistics including providing transport reimbursement to the families. The research assistants observed the sessions and shared feedback with the facilitators at the end of the sessions. The MFG intervention focuses on reducing stigma by normalizing shared experiences. The MFG intervention acknowledges poverty as a stressor that may undermine parenting while also recognizing the contextual challenges that contribute to poor mental health functioning for adolescent girls, including high rates of poverty, violence, and family loss due to HIV and other health threats [27].

Sample size

The sample size was determined a priori using NCSS PASS software to compute minimum detectable effect sizes. The sample size was estimated in 1,260 participants evenly distributed across three study groups. Assuming a conservative attrition rate of 20%, data from 1,008 participants will be available for an analysis at all time points. The standardized mean difference d, was computed for continuous mental health outcomes using within-subject correlation ρ, which varied between .20 and .70. Therefore, the study was adequately powered to detect small to medium effects across a variety of analysis conditions and details about the sample size calculation have been published by Ssewamala et al. [30].

Study measures

Dependent variable.

The Beck Depression Inventory (BDI) [33] measured depressive symptoms. The scale measures characteristic attitudes and symptoms of depression, including mood, pessimism and sense of failure, self-dissatisfaction, guilt, punishment, self-dislike, self-accusation, suicidal ideas, crying, irritability, social withdrawal, indecisiveness, body image change, work difficulty, insomnia, fatigability, loss of appetite, weight loss, somatic preoccupation, and loss of libido. The scale consists of 21 sets of statements, ranked based on the severity on a 4-point continuum (0 = least, 3 = most). The theoretical range for the BDI is 0–63 with higher scores indicating higher levels of depressive symptoms. The BDI has been used in many studies to evaluate depression levels among different populations including in Uganda [34–36]. The scale demonstrated a high internal consistency at baseline (Cronbach’s alpha = 0.83) and 12-month follow-up (Cronbach’s alpha = 0.82).

Independent variables.

The primary independent variable was participation in the intervention, coded as 1 for participation in the FEE and two for MFGs plus the FEE + MFG conditions, and 0 for the control condition. In addition, time (baseline or 12 months of postintervention initiation) was included as a categorical variable.

Analytical approach

The first step in the study was to examine the sample’s sociodemographic characteristics. The means and standard deviations, and frequencies and measures of central tendency, are presented. Second, the treatment and control conditions’ baseline characteristics were generated to characterize the sample’s main sociodemographic characteristics. Participants’ depressive symptoms were also assessed to see if there were noticeable differences between the treatment and control groups at baseline and 12 months after the intervention.

To examine group differences between the control, FEE, and MFG plus the FEE + MFG on depressive symptoms, we fitted a three-level linear mixed model. The model included the study group status (FEE, FEE + MFG, or control conditions), a time point variable (baseline and 12-month follow-up), and their interactions. In this study, individuals are nested within each school and observations are nested within each individual, so the multilevel model was useful to account for data clustering. In the case of missing data, a mixed effect model allows for estimates of changes in repeated measures, assuming the data are missing at random. Random intercepts for school number (ID) and unstructured residual variances and covariances for person ID were included to account for clustering [37]. To safeguard against mis-specification of standard errors and possible assumption violations (e.g., nonconstant or non-normal residuals in linear mixed models), the model was fit with robust standard errors (Huber–White “sandwich” variance estimator) clustered by school ID for the composite depression scores [38,39]. Beta coefficients (β) and the 95% confidence interval (CI) are presented in Table 1. Simple main effects comparisons of time within each group and groups within each level of time were used to decompose and interpret a significant group-by-time interaction. In addition, the standardized mean difference (SMD), also known as Cohen’s d, is also included to evaluate effect sizes. All analyses were conducted using Stata SE, version 16 [40].

Table 1.

Beta coefficients and 95% confidence intervals for depressive symptoms among adolescent girls in Uganda, 2017–2018 (n = 1,260)

Outcomes	Depressive symptoms, B (95% CI)	P value
Time
Baseline (reference group)	0
12-months	−2.53 (−3.60, −1.45)	<.001
Group
Control (Reference group)	0
Intervention 1: FEE	−1.34 (−2.78, 0.100)	.068
Intervention 2: FEE + MFG	−.78 (−2.33, 0.76)	.320
Group × time
FEE × 12 months	−.94 (−2.21, 0.34)	.151
FEE + MFG × 12 months	−2.14 (−3.68, −0.61)	.006
Constant	19.21 (18.17, 20.26)	<.001
No. of observations	2,479

STATA code: mixed TS_BDI i.time##i.Study_arm || sch_id:, || STUDYID:, nocons residuals(uns, t(time)) base vce(robust).

FEE = family economic empowerment; MFG = multiple family groups.

Results

At baseline, 1,260 adolescent girls (mean age 15 years) were enrolled in the study with 471 (37.4%) in the FEE group, 381 (30.2%) in the FEE + MFG group, and 408 (32.4%) in the control group. Participants lived on average in a household with seven adults and three children aged less than 18 years. Most participants (N = 965; 77%) were cared for by a biological parent and a total of 215 individuals (17.1%) were orphaned. At baseline, about a quarter of the participants had some savings and the average number of household assets was 11.2. In addition, the average depressive symptom scores at baseline were: control group, 19.18 (SD = 10.29); FEE group, 17.85 (SD = 10.17); and FEE + MFG group, 18.48 (SD = 10.07). Overall, the mean depression score was similar across the study conditions (Table 2). Between baseline and 12-month follow-up, mean depression scores had improved in each study group. However, the FEE + MFG intervention (Δs = −4.68) and FEE (Δs = −3.5) yielded significantly greater decreases in depressive symptoms at 12-month follow-up (Table 2).

Table 2.

Sample characteristics of adolescent girls and young women enrolled in the Suubi4Her trial conducted in southern Uganda at baseline (n = 1,260)

Characteristics	Control (n = 408), mean (SD) or %	Treatment 1 (n = 471), mean (SD)	Treatment 2 (n = 381), mean (SD)
Age (14–17 years)	15.18 (0.86)	15.49 (0.82)	15.43 (0.90)
Household size (Mean, SD)
People in the household	6.81 (2.61)	7.03 (2.66)	7.17 (2.85)
Children in the household	3.39 (2.06)	3.49 (2.17)	3.62 (2.05)
Depressive symptoms at baseline	19.18 (10.29)	17.85 (10.17)	18.48 (10.07)
Depressive symptoms are 12 months	16.62 (9.98)	14.35 (8.87)	13.80 (9.11)
Changes in depressive symptoms (baseline to 12 months within groups)	Δs = −2.56	Δs = −3.5	Δs = −4.68
Any savings (N/%)
No	307 (75.2)	352 (74.7)	298 (78.2)
Yes	101 (24.8)	119 (25.3)	83 (21.8)
Family assets (Mean, SD)	11.2 (3.25)	11.8 (3.12)	11.3 (3.4)
Orphanhood Status, (N/%)
Nonorphan	342 (83.8)	391 (83.0)	312 (81.9)
Orphan	66 (16.2)	80 (17.0)	69 (18.1)
Primary caregiver (N/%)
Biological parent	312 (76.5)	370 (78.6)	283 (74.3)
Grandparent	46 (11.3)	54 (11.4)	40 (10.5)
Other	50 (12.2)	47 (10.0)	58 (15.2)

Coefficients and associated standard errors and test statistics from the linear mixed model appear in Table 1. There was a statistically significant overall main effect of condition (χ² [2] = 11.25, p = .004) and a significant overall effect of time (χ² [2] = 154.12, p < .0001). However, these main effects were qualified by a statistically significant interaction between condition and time (χ² [2] = 7.53, p = .023). Simple main effects comparisons of 12 months versus baseline within each condition revealed substantial and statistically significant reductions in depressive symptoms for all three groups (Δs = −2.53 to −4.67, all p < .001). Simple main effects comparisons of pairwise group differences at baseline were not statistically significant. However, simple main effects comparisons of group differences at follow-up revealed significantly lower mean levels of depression at 12 months for FEE versus bolstered standard of care (Δ = −2.27, p = .001; SMD: −0.24, 95% CI: −0.37, −0.11) and for FEE + MFG versus bolstered standard of care (Δ = −2.92, p < .001; SMD: −0.29, 95% CI: −0.43, −0.15). However, FEE + MFG was not significantly different from FEE at 12 months (Δ = −0.65, p = .303). Taken collectively, these results show that while all three groups experienced a substantial reduction in depressive symptoms from baseline to 12 months, the reductions were stronger for the two intervention groups (Figure 1).

Figure 1.

Depressive symptoms total score from baseline to 12 months.

Discussion

This study examined the short-term (12 months) impact of a combination intervention, including both the FEE intervention and MFGs implemented during the Suubi4her study on depressive symptoms among school-going adolescent girls in southwestern Uganda. The results indicate that both FEE and MFG interventions may be promising tools for addressing depressive symptoms among adolescent girls in Uganda. Specifically, at 12-month postintervention delivery, the FEE intervention alone and the group that received both the FEE and MFG interventions experienced a significant decline in depressive symptoms compared to the control group. However, no significant difference was observed between the FEE and FEE + MFG intervention groups. The findings on FEE are consistent with tenets of the asset theory [26,41], positing that parental/family asset ownership has a beneficial psychological impact on adolescents. Similarly, the results from this study add to the existing literature indicating that FEE has a positive impact on the mental health of adolescents in low-resource settings in SSA [24,42].

In addition, a combination of FEE and MFG interventions significantly reduced depressive symptoms among school-going adolescent girls in rural Uganda. However, the results indicated no significant differences between the FEE alone and FEE + MFG interventions. This is not surprising given the limited social and economic resources in the study area. Therefore, by alleviating family poverty, the FEE intervention may alleviate caregiver stressors that negatively impact adolescent mental health. As hypothesized in the asset theory, greater access to basic resources improves one’s ability to cope with stressful life events, such as a life-threatening disease, and potentially improves mental health outcomes [26,41]. Therefore, there is a need for further research to tease out the impact of the MFG intervention alone among vulnerable adolescents in SSA.

Although the FEE + MFG intervention group was not statistically different from the FEE group, four Rs and two Ss have been associated with positive changes in child and caregivers outcomes in other studies [43]. In addition, other components of the MFG including therapeutic relationships, group mechanisms, and caregiver perceptions and expectations (e.g., acceptability and utility of the intervention) are all relevant factors to building family cohesion and support [28]. In the same way, studies indicate that MFG interventions are associated with improved family processes including support, communication, and parent-child participation [28,29] that are essential for the prevention and promotion of adolescent health and wellbeing. Therefore, future studies should examine the family processes through which the MFG intervention impacts depressive symptoms of adolescents and their families in low-resource settings.

In addition, a combination of FEE and MFG interventions is important because it brings together families to share their challenges and learn from each other. Research indicates that advice and an insight from other families are often seen as less threatening than feedback given by a therapist [28]. MFG further focuses on reducing stigma by normalizing shared experiences. Consequently, enhancing family cohesion and support that are protective factors for adolescent mental health [44]. Aspects of family cohesion and support are critical for successful FEE interventions. Therefore, the MFG can boost and compliment FEE interventions to achieve robust results.

The study findings should be interpreted taking into account the following limitations. The sample was recruited in mainly rural-based communities across public secondary schools. Similarly, the sample of schools recruited into the study was mostly homogenous. Similarly, the study was clustered at the school level across schools with comparable social demographic factors. Therefore, this limits our ability to generalize the findings to urban communities that might be experiencing different stressors and schools with different profiles. Last but not the least, one outcome measure was used to capture depressive symptoms using participant self-reports with a potential for social desirability bias. Future studies should use multiple rates to triangulate the data reported by adolescents.

Therefore, results imply that interventions involving evidence-based FEE may be a promising tool in addressing depressive symptoms among adolescent girls in low-resource settings. To enhance adolescents’ mental health, future studies should incorporate material support. Similarly, to reduce the long-term implications of adverse psychosocial health during adolescence, policymakers and program implementers should explore scaling up the interventions in similar settings to bridge the mental health treatment gap for adolescent girls. Therefore, future research is needed to examine family process characteristics that mediate the intervention and mental health outcomes to better inform and improve the treatment package delivered to families to maximize its relevance and impact.

IMPLICATIONS AND CONTRIBUTION.

Adolescents in Sub-Saharan Africa live in low-resource settings are heavily impacted by poverty and experience multiple stressors that negatively impact their mental health. Therefore, to enhance adolescents’ mental health, future studies should incorporate material support.