The Experience of Abuse and Depressive Symptoms Among Adolescents in a Post-Conflict Setting: A Cross-Sectional Study

Abstract

This study aimed to explore the relationship between the experience of abuse and depressive symptoms among children and adolescents in a post-conflict setting while controlling for perceived support and individual background factors. This cross-sectional study was conducted in the post-conflict setting of Borama, Somaliland. A total of 645 adolescents that attended upper primary and lower secondary schools were included. The exposure variables were (a) experience of physical abuse, (b) experience of psychological abuse, and (c) experience of either or both forms of abuse. The outcome was depressive symptoms as measured by the Patient Health Questionnaire-9. Physical and psychological abuse were associated with higher levels of depressive symptoms (B = 2.74, p < 0.001 and B = 1.62, p < 0.001, respectively). The girls had higher levels of depressive symptoms than the boys (B = 0.71, p = 0.004). Moreover, a greater age was associated with higher levels of symptoms (B = 0.17, p = 0.033). Higher social support levels were associated with lower levels of depressive symptoms (B = −0.08, p = 0.001). Our findings suggest that interventions providing social support to adolescents who experience abuse may be beneficial in reducing depressive symptoms.

1. Introduction

The abuse suffered by children takes many forms, including physical, psychological, and sexual. Globally, half of all children aged 2–17 years have experienced abuse or neglect. In Africa, it is estimated that 200 million children have been abuse victims [1]. A study conducted in Kenya, Zambia, and the Netherlands reported higher rates of physical abuse and neglect among children living in Kenya and Zambia than among those living in the Netherlands [2]. Findings from Tanzania show that 90% of children had experienced physical discipline at the hands of their parents [3], and in other studies, 39.6% of children and adolescents in Nigeria and 100% of children and adolescents in Uganda reported experiencing some form of abuse at the hands of parents and teachers [4,5,6]. Gender has also been found to be related to childhood abuse. Males in Kuwait and Malawi have reported higher levels of physical abuse than females [7], whereas females in Kuwait and Swaziland have reported more emotional abuse than males [7,8,9].

Physical and psychological abuse against children and adolescents is associated with various societal factors, including economic stress, cultural norms, gender inequality, low literacy levels, conflict, war, and other disruptions to society [10]. Physical punishment is also used as a disciplinary approach at home and school in several cultures and countries [6,7]. Previous studies have demonstrated that abuse, in any form, impacts the mental health of children and adolescents [11,12,13]. For example, a recent meta-analysis showed that children and adolescents who were exposed to physical or emotional violence were more likely to be diagnosed with major depressive disorder [12]. Similarly, it has been shown that children and adolescents that suffered abuse are more likely to report mental health problems in adulthood [7,13,14].

Conflict-induced humanitarian crises, such as prolonged war and armed conflict, affect the social determinants of mental health, the well-being of families and communities, and their access to basic healthcare services and education [10,15]. Armed conflict occurs when there is a use of armed force between two or more parties, resulting in the death and displacement of civilians, as well as violations of human rights [16]. “Post-conflict” is another term used to define the period after an armed conflict has ended, where this period is characterized by the rebuilding of the country and its healthcare system, as well as economic instability [17]. Research has highlighted the impact of experiencing abuse on the mental health of children and adolescents living in post-conflict settings [10,15,18]. A recent systematic review and meta-analysis reported that the mental health problems in conflict and post-conflict settings are enormous, underscoring the urgent need for mental health interventions [19]. Populations in post-conflict settings—such as Somalia and Somaliland, which have experienced violence, conflict, displacement, instability, poverty, and famine over the past three decades—are likely to experience serious mental health issues [20]. A WHO report [20] estimated that one in every three people in Somalia suffers from mental health problems.

Somaliland declared its independence from Somalia in 1991, and it is politically stable and safer than the remainder of the Somali region. Nevertheless, in a study that involved approximately 3000 households in Somaliland interviewed about mental disorders, 12% of the respondents reported caring for someone with a severe mental illness [21].

Mental health in post-conflict settings is threatened by exposure to violence and other risk factors, such as poverty and a lack of access to essential services (e.g., healthcare, education, housing, water, and sanitation) [19]. Research has also shown that parental income and employment are associated with depressive symptoms [22]. Globally, 10–20% of children and adolescents experience mental health problems, and depression places a heavy burden not only on the young people who suffer from it but also on their families and countries [23]. Depressive symptoms in adolescents are linked to both age and gender. Girls are more likely to report several symptoms [22,24], and the 2020 Somaliland Health Demographic Survey [25] reported that approximately 12% of women aged 15–49 years had experienced physical abuse. However, these statistics must be understood within the specific cultural and social dynamics of Somaliland. In many societies, including the Somali context, physical discipline at home and school can be normalized, which may lead to the underreporting of physical abuse [6,7]. Furthermore, significant stigma surrounds mental health problems, which are often perceived as permanent and incurable [20]. This can create reluctance among adolescents to acknowledge or report depressive symptoms. The post-conflict environment itself, characterized by decades of instability and displacement, creates a backdrop of chronic stress that can exacerbate the impact of abuse on mental health [20]. It is also noteworthy that these dynamics are situated within a society undergoing change, as evidenced by the implementation of Somaliland’s National Child Protection Policy in 2015 [26], signaling a growing national awareness of these issues. Nevertheless, the prevalence of mental health problems in Somalia and Somaliland remains poorly understood. Only one study conducted in Somaliland has investigated mental health problems, and that study concerned university students [24]. It found that the prevalence of mental health problems among university students was approximately 20% [24]. This study’s conceptual approach is informed by Bronfenbrenner’s Social–Ecological Model, which emphasizes that individual development is influenced by dynamic interactions within multiple, nested environmental systems [27] and supports our analysis of factors operating at various levels, including the individual (e.g., age, gender), the microsystem (e.g., experiences of abuse, support from family and peers), and the macrosystem (e.g., the broader post-conflict cultural context of Somaliland). To fill this knowledge gap, this study aimed to investigate the relationship between the experience of abuse and depressive symptoms among children and adolescents in a post-conflict setting in Somaliland while controlling for perceived support and individual background factors. Our hypothesis was that physical and psychological abuse would be associated with higher levels of depressive symptoms.

2. Methods

2.1. Study Design and Setting

This study’s cross-sectional design was employed in Borama, which is the capital and largest town of the northwestern Awdal region of Somaliland and has a population of 278,000 [28]. As an urban center in a post-conflict society, Borama serves as a vital educational hub for a diverse student body. Its public schools primarily enroll children and adolescents from lower socioeconomic backgrounds, including a significant number of students who commute daily from surrounding rural areas following the introduction of free primary and intermediate education [29]. This unique demographic mix and its role as a key access point to a vulnerable population make Borama an ideal setting for this study. The Republic of Somaliland is a post-conflict, self-declared independent country with a population of 3.5 million people. The country introduced free primary and intermediate education in January 2011. Borama now has 15 public schools: 3 upper secondary schools (grades 9 to 12) and 12 primary and middle schools (grades 1 to 8). Students are placed in grades according to their academic performance rather than their age. Thus, some older students may still attend primary or middle school. There are also private schools, to which families that can afford it send their children. Therefore, most students who attend public schools have a low socioeconomic status. The Somaliland Education Statistical Yearbook 2014/2015 [30] reports that girls’ involvement and retention in education stands at merely 38.8% compared with 48.2% for boys. The dropout rate is particularly high for both genders in grades 5 to 8, with girls exhibiting a more substantial dropout rate of 11.6%. Consequently, girls are more prone to early dropout from school than boys. According to a report by the Awdal Regional Education Office, in the 2019/2020 academic year, 1402 students attended the three upper secondary schools, and 5957 students attended the 12 primary and middle schools. Nine schools were involved in this study.

2.2. Participants and Sampling

The study population comprised children and adolescents that attended upper primary and middle schools. The participants were selected using a combination of convenience and random sampling. Six middle school classes (grades 6 and 7) and three upper secondary school classes (grades 9 and 10) were chosen for convenience, and random sampling was used to recruit students from each class using attendance lists under the guidance of the participating schools’ management. The samples were proportional to each middle and upper secondary class. Students who were on sick leave during the data collection were excluded.

A total of 645 children and adolescents provided informed consent to participate in the study. Three participants were excluded due to missing data regarding the variables of interest. Thus, 642 participants were included in the analysis, which represented 99.5% of the initial sample.

3. Measures and Variables

3.1. Exposure Variables

The experience of abuse was assessed through two yes-or-no questions: “Have you experienced physical abuse in the past six months?” and “Have you experienced psychological abuse in the past six months?” The responses were combined into a yes/no variable concerning either or both forms of abuse. The respondents were provided with examples of what physical and psychological abuse entail. For instance, physical abuse was illustrated with examples such as being hit with a stick, belt, or other hard objects, or being slapped, punched, or otherwise physically assaulted. Psychological abuse included examples such as being called names, threatened, or insulted.

3.2. Covariates

Social support was measured using the Multidimensional Scale of Perceived Social Support (MSPSS) [31], which was adapted to the African context using Mels et al.’s [32] cultural adaptation procedure. The MSPSS measures the perceived adequacy of support from family, friends, and significant others. It comprises 12 items scored on a 4-point Likert scale ranging from 1 (not at all) to 4 (a lot), and it is divided into three subscales: family, friends, and significant others. In this study, we used the family and friends subscales, which comprised eight items in total, to assess the perceived social support from these sources. The total score for the two subscales ranged from 8 to 32, with a higher score indicating greater perceived support. The MSPSS, including its three subscales, has established psychometric properties when used among adolescents in both Western [33] and African contexts [34]. In this study, the MSPSS had a Cronbach’s alpha value of 0.86.

We controlled for several potential covariates that have previously been shown to be related to depressive symptoms in children and adolescents [22]: gender (boy/girl), age (continuous variable), and parental employment (“both parents employed,” “one parent employed,” “neither parent employed,” or “I do not know”). We also adjusted for school level (middle/upper secondary school), as this factor was complementary to age given that in this setting, older age did not necessarily correspond to a higher school level.

3.3. Outcome Variable

The adolescent version of the Patient Health Questionnaire-9 (PHQ-9) [35] was used to measure depressive symptoms. This scale is a widely used instrument that has been translated into more than 70 languages and dialects. It includes nine items measuring factors such as sleep, concentration and energy problems, low self-esteem, and anhedonia, which are rated according to their occurrence frequency over the previous 2 weeks on a 4-point scale ranging from 0 (not at all) to 3 (nearly every day). Thus, the total score ranges from 0 to 27, with higher scores representing more severe symptoms of depression. The scale can also be divided into categorical depression levels, with scores of 0–4 representing minimal depression, scores of 5–14 indicating moderate depression, and scores of 15 or higher representing severe depression. The adolescent version of the PHQ-9 has established psychometric properties for adolescents in both Western [36] and African contexts [37]. In this study, the PHQ-9 had a Cronbach’s alpha of 0.63, which is slightly below the commonly accepted threshold of 0.70 [38]. However, values above 0.60 are considered acceptable in exploratory studies.

3.4. Data Collection

The original language of the questionnaire was English, but we used a translated version of the PHQ-9 and MSPSS taken from previous study [39]. However, we also discussed the translated questionnaire with our research team, who are native Somali speakers, and addressed cultural appropriateness; consequently, we made some further adjustments by using synonyms for certain terms. Prior to data collection, a pilot survey was conducted with 50 students from two other public schools in the Borama district (grades 7 and 10). The pilot study resulted in no further revisions to the questionnaire. However, the data collectors reported that some students needed support in reading some of the questions. Therefore, during the actual survey, the data collectors provided support as needed. For example, the data collectors offered explanations when adolescents did not understand the questions, such as questions related to psychological abuse.

The data were collected in November and December 2020. During the beginning of the COVID-19 pandemic, from March to June 2020, schools in Somaliland were closed, as in many other countries around the world. However, they reopened again for the fall semester of 2020, which made it possible to conduct data collection. The respondents completed the translated questionnaire using the Open Data Kit (ODK) Collect app. The data collectors, 50 in total, were recruited from among fifth-year medical students who had mental health knowledge and experience in collecting survey data. The data collectors received three days of refresher training and were supervised during the data collection by one of the first authors and the last author.

3.5. Statistical Analysis

Means, standard deviations, ranges, frequencies, and proportions were used for descriptive purposes. Independent-sample t-tests were used to perform comparisons between the participant groups. In addition, when the participant groups were highly divergent in size, we performed Kruskal–Wallis tests, along with pairwise comparisons. The relationship between experiencing abuse and exhibiting depressive symptoms was explored using linear regression models. The exposure variables were (a) experience of physical abuse, (b) experience of psychological abuse, and (c) experience of either or both forms of abuse. The outcome was depressive symptoms as measured via the PHQ-9.

The relationship between abuse and depressive symptoms was explored using unadjusted and adjusted models, to which the aforementioned potential confounders were added. Prior to performing the regression analyses, the data were checked for intercorrelations—that is, relationships between the independent variables—using the variance inflation factor (VIF), with a VIF of 10 or greater indicating multicollinearity [40]. The VIF values of the independent variables ranged from 1.04 to 4.36. Therefore, no multicollinearity was present in the data.

The results of the regression analyses were expressed as unstandardized beta coefficients (B) with 95% confidence intervals (CIs) and R² values as measures of the explained variance. p-values less than 0.05 were considered statistically significant. All analyses were performed using IBM SPSS Statistics Version 27.

3.6. Ethics

Ethical approval for this study was obtained from the School of Postgraduate Studies and Research at Amoud University (Ref: 2020-0123AU) and the Swedish Ethical Review Authority (Ref: 2021-02814). In addition, the Ministry of Education and Science regional office in Borama district granted permission to conduct the research in the participating schools. All the participating adolescents were provided with information about the study’s aim and procedures and informed that their participation was voluntary. All the participants provided oral and written informed consent. The data was collected by trained data collectors who were medical students with prior experience working with vulnerable populations. These data collectors were instructed to offer support or refer adolescents in need of further assistance to Amoud University’s Community Service Center.

4. Results

4.1. Participants’ Characteristics

The participants’ demographic characteristics are shown in

Table 1. Participants’ characteristics (n = 642).

Variable	n (%)
Age, M (SD, min–max)	14.3 (1.9, 10–19)
Gender
Girl	359 (55.9)
Boy	283 (44.1)
School level
Middle school	387 (60.3)
Upper secondary school	255 (39.7)
Parental employment
Both parents employed	114 (17.8)
One parent employed	401 (62.4)
Both parents unemployed	38 (5.9)
I don’t know	89 (13.9)
MSPSS family and friends subscales total score, M (SD, min–max)	23.6 (5.0, 8–32)
Experience of physical abuse 1
Yes	56 (8.7)
No	586 (91.3)
Experience of psychological abuse 1
Yes	51 (7.9)
No	591 (92.1)
Experience of abuse 1
Both physical and psychological abuse	7 (1.1)
Either physical or psychological abuse	93 (14.5)
No experience of physical or psychological abuse	542 (84.4)
PHQ-9 total score, M (SD, min–max)	3.3 (3.2, 0–19)
PHQ-9 diagnostic categories
Minimal depression	480 (74.8)
Moderate depression	159 (24.8)
Severe depression	3 (0.5)

¹ During the last 6 months.

. The mean age of the participants was approximately 14 years; more than half (56%) were girls, and 60% attended middle school. Most participants (62%) had only one employed parent, while almost one-fifth (18%) had two employed parents. Some participants (6%) reported that both their parents were unemployed, and 14% reported that they did not know their parents’ employment status.

Over the previous six months, 56 participants (9%) had experienced physical abuse and 51 (8%) had experienced psychological abuse. Among the boys (n = 283), 27 participants (10%) had experienced physical abuse and 24 (8%) had experienced psychological abuse. The corresponding numbers among the girls (n = 359) were 29 (8%) and 27 (8%). Seven participants (1%) had experienced both physical and psychological abuse. The mean PHQ-9 score was 3.3 points. Approximately 25% of the participants were classified as moderately depressed.

Table 2. PHQ-9 mean total scores in participant groups (n = 642).

Variable	PHQ-9	t/χ2	p	Pairwise Comparisons
Gender	M (SD)	2.13	0.034
Girl	3.5 (3.3)
Boy	3.0 (2.9)
School level		4.49	<0.001
Middle school	3.7 (3.2)
Upper secondary school	2.6 (3.0)
Parental employment		0.44	0.720
Both parents employed	3.0 (3.1)
One parent employed	3.3 (3.2)
Both parents unemployed	3.7 (4.3)
I don’t know	3.0 (3.1)
Experience of physical abuse
Yes	5.9 (3.9)	6.51	<0.001
No	3.0 (3.0)
Experience of psychological abuse
Yes	4.9 (3.8)	3.19	<0.001
No	3.1 (3.1)
Experience of abuse
Both physical and psychological abuse (B) (n = 7)	4.3 (2.3)	42.04	<0.001	E > N
Either physical or psychological abuse (E) (n = 93)	5.6 (4.0)
No experience of physical or psychological abuse (N) (n = 542)	2.9 (2.9)

Bold values show a higher meaning.

shows the mean total PHQ-9 scores of the participant groups. The girls had a significantly higher mean total score than the boys (p < 0.05), and the middle school students had a significantly higher mean total score than the secondary school students (p < 0.001). The participants who had experienced physical or psychological abuse had a significantly higher mean total score than those who had experienced no abuse (p < 0.001). Moreover, the participants who had experienced physical or psychological abuse had a higher level of depressive symptoms than those who had not experienced abuse (p < 0.001).

4.2. Relationship Between the Experience of Abuse and Depressive Symptoms

Table 3. Unadjusted linear regression models for exploring physical and psychological abuse in relation to depression symptoms according to the Patient Health Questionnaire-9 (PHQ-9). Significant associations (p < 0.05) are marked in bold (n = 642).

	Depression Symptoms According to PHQ-9
Unadjusted Model I, Independent Variables	B (95%CI)
Experience of physical abuse
No (ref)
Yes	2.75 (1.90–3.59)
Experience of psychological abuse
No (ref)
Yes	1.62 (0.73–2.50)
R2	0.08
Unadjusted Model II, Independent Variable	B (95%CI)
Experience of both/either physical and/or psychological abuse
No (ref)
Yes	2.61 (1.96–3.26)
R2	0.09

shows the relationships between physical and psychological abuse and depressive symptoms. Physical abuse and psychological abuse were associated with higher levels of depressive symptoms (B = 2.74, p < 0.001 and B = 1.62, p < 0.001, respectively).

Model I in

Table 4. Adjusted multiple linear regression models for exploring physical and psychological abuse in relation to depression symptoms according to the Patient Health Questionnaire-9 (PHQ-9). The results are adjusted for covariates. Significant associations (p < 0.05) are marked in bold (n = 642).

	Outcome: Depression Symptoms According to PHQ-9
Adjusted Model I, Independent Variables	B (95%CI)
Experience of physical abuse
No (ref)
Yes	2.37 (1.52–3.22)
Experience of psychological abuse
No (ref)
Yes	1.29 (0.42–2.17)
Gender
Boy (ref)
Girl	0.71 (0.23–1.18)
Age	0.17 (0.01–0.33)
School level
Middle school (ref)
Upper secondary school	−1.27 (−1.92–−0.62)
Parental employment
Both parents unemployed (ref)
Both parents employed	−0.59 (−1.70–0.52)
One parent employed	−0.34 (−1.34–0.67)
Don’t know	−0.62 (−1.77–0.52)
Social support according to the MSPSS	−0.08 (−0.13–−0.03)
R2	0.13
Adjusted Model II, Independent Variables	B (95%CI)
Experience of both/either physical and/or psychological abuse
No (ref)
Yes	2.26 (1.60–2.92)
Gender
Boy (ref)
Girl	0.75 (0.27–1.22)
Age	0.17 (0.02–0.33)
School level
Middle school (ref)
Secondary school	−1.27 (−1.92–−0.62)
Parental employment
Both parents unemployed (ref)
Both parents employed	−0.53 (−1.67–0.54)
One parent employed	−0.34 (−1.34–0.66)
Don’t know	−0.66 (−1.79–0.48)
Social support according to the MSPSS	−0.08 (−0.13–−0.03)
R2	0.13

shows the relationships between physical and psychological abuse and depressive symptoms after adjusting for covariates. Both physical and psychological abuse remained significantly associated with higher levels of depressive symptoms (B = 2.37, p < 0.001 and B = 1.29, p = 0.004, respectively). Moreover, the girls had a significantly higher level of depressive symptoms than the boys (B = 0.71, p = 0.004). Likewise, older age was associated with a significantly higher level of depression symptoms (B = 0.17, p = 0.033). The upper secondary school students had significantly lower levels of depressive symptoms than the middle school students (B = −1.27, p < 0.001). Higher social support levels were significantly associated with lower levels of depressive symptoms (B = −0.08, p = 0.001). There was no significant relationship between parental employment and depressive symptoms.

Model II in Table 4 shows the relationship between the experience of either or both forms of abuse and depressive symptoms. Physical and/or psychological abuse was significantly associated with higher levels of depressive symptoms (B = 2.26, p < 0.001) after adjusting for covariates. The relationships between the covariates and the outcome were like those identified using Model I.

5. Discussion

This study aimed to explore the relationship between the experience of abuse and depressive symptoms among children and adolescents in a post-conflict setting while controlling for perceived support and individual background factors. Our findings show that approximately 9% of the participants experienced physical abuse, while 8% experienced psychological abuse. Although the prevalence of abuse in our sample was lower than in studies conducted in other African countries [3,4,5,6], this finding must be interpreted with cultural caution. The lower prevalence may not reflect a lower occurrence but rather the influence of social dynamics unique to the post-conflict setting of the Somaliland context. Several studies have emphasized that violence against children and adolescents may be underreported [6,7,22] because, in many societies, physical abuse is used as a disciplinary approach at home and school and is thus normalized [6,7], and this may be the case in the Somali society. This is compounded by the significant stigma surrounding mental health issues in Somaliland, where they are often viewed as permanent and incurable [20]. This dual challenge, the normalization of physical discipline and the stigma of mental illness, likely creates a barrier against both the experience and reporting of abuse and its psychological impact, suggesting our findings may represent the “tip of the iceberg” of a deeper issue. It should also be noted that abuse was self-reported, and some adolescents may not feel comfortable answering such questions. However, in 2015, Somaliland implemented the National Child Protection Policy [26], which may have increased awareness of the impact of violence against children in the country.

In line with previous studies [7,8], our findings also show that more boys (10%) than girls (8%) reported experiencing physical abuse. This rate among the girls is somewhat lower than that reported in the Somaliland Health Demographic Survey, which found that approximately 12% of women aged 15–49 had experienced physical abuse [25]; however, that report concerned an older demographic of girls and women. Interestingly, in our study, equal proportions of girls and boys (8%) reported psychological abuse.

Despite this context of similar or even lower reported exposure to abuse, a crucial finding is that girls still reported significantly higher levels of depressive symptoms (B = 0.71, p = 0.004). This suggests a greater vulnerability among girls to the psychological impact of adversity, which may be due to gendered socialization, where girls are more likely to internalize distress. This finding underscores the need for gender-sensitive mental health interventions that recognize these different patterns of risk and response. The main finding of this study is that experiences of both physical and psychological abuse were associated with higher levels of depressive symptoms. This association remained significant after controlling for covariates. These results are in line with those of previous studies that have reported relationships between various forms of abuse and depressive symptoms [7,9,11,12,13]. Moreover, our findings suggest that the girls and older participants who experienced abuse had higher levels of depressive symptoms than the boys and younger participants. This is in line with a systematic review that found a higher prevalence of depression among girls and older adolescents [22]. The same review also found that adolescents living in conflict or post-conflict settings, as is the case in Somaliland, were prone to mental health problems [22]. In conflict and post-conflict settings, children and adolescents may suffer from both abuse and depression [18]. Our results also show that higher social support levels were associated with lower levels of depressive symptoms and that the upper secondary school students had lower levels of such symptoms than the middle school students. This is consistent with previous research showing that perceived social support from family and friends acts as a buffer for adolescents’ mental health [41].

Our findings have several implications. First, abuse, in any form, impacts children’s and adolescents’ mental health and academic achievement and may lead to intergenerational abuse, constituting a risk factor for intimate partner violence in adulthood [42,43]. Therefore, interventions intended to reduce child and adolescent abuse must be implemented. Parenting support programs have been shown to improve parent–child relationships and reduce child maltreatment [44]. Interventions targeting parents may help them develop skills with which to manage their children’s behaviors without resorting to abusive disciplinary practices [44].

Second, in Somaliland, as in many other countries, mental health issues carry a stigma, as they are believed to be permanent and uncurable [20]. Therefore, it is important to raise awareness of mental health among adolescents and communities. Schools can be considered ideal arenas for mental health interventions for children and adolescents because they can be provided to entire classrooms or schools, which may help reduce the associated stigma [45].

Finally, our finding that social support from friends and family is a significant protective factor against depressive symptoms provides strong empirical support for the stress-buffering hypothesis, suggesting that robust social connections can mitigate the harm of adversity, such as abuse [46,47]. The critical importance of this immediate social environment is further underscored by longitudinal research in other high-adversity contexts. For instance, [48] found that mothers’ mental health in early childhood was a significant predictor of adolescents’ later exposure to adversity in Malawi. Taken together, our findings on social support’s protective factor and the evidence on maternal well-being’s predictive factor converge on a crucial point: interventions aimed at strengthening caregiver well-being and the immediate support systems around children are vital for preventing adversity and its mental health consequences.

At the most immediate level, the microsystem, interventions should focus on strengthening family bonds through evidence-based parenting support programs that reduce harsh discipline [44]. Simultaneously, efforts within schools—a key part of the adolescent mesosystem—can foster structured peer-support networks and a sense of community belonging, creating a protective environment outside the home [45]. These informal supports should be reinforced by strengthening the exosystem, ensuring access to formal institutional resources, such as non-stigmatizing school counselling services and community-based youth programs. Critically, because the pervasive stigma surrounding mental health operates at the macrosystem (cultural) level, these targeted interventions must be coupled with broader awareness campaigns to be fully effective. Ultimately, our findings advocate for a holistic, social–ecological approach that simultaneously strengthens an adolescent’s personal relationships while building a resilient and supportive community and institutional framework around them.

Strengths and Limitations

Several limitations should be considered when interpreting our findings. First, the cross-sectional design does not allow us to draw conclusions about causality. Thus, the relationship between abuse and depressive symptoms can only be considered correlational. Second, our study was subject to potential selection bias. Our sampling frame was limited to adolescents attending public schools, meaning our findings are not generalizable to out-of-school adolescents, who may face different or more severe risks of mental health problems. The focus on public schools also means our sample predominantly comprised students from lower socioeconomic backgrounds in both urban and neighboring rural areas, potentially limiting the generalizability to adolescents from higher-income families attending private schools. Third, a significant limitation was the study’s reliance on self-reported data for abuse and depressive symptoms, which introduced the risk of self-report bias. The response validity may have been affected by several factors. For instance, even though the concept of psychological abuse was introduced, adolescents may have been hesitant to report it due to the normalization of certain disciplinary behaviors within the cultural context. This social normalization, combined with the stigma surrounding mental health problems in Somaliland, likely contributed to a social desirability bias, where participants may have underreported both their experiences of abuse and the severity of their depressive symptoms to avoid shame or judgment. It is also crucial to acknowledge that our data were collected shortly after nationwide school closures in Somaliland from March to June 2020 due to the COVID-19 pandemic. This timing may have influenced the responses, with the pandemic potentially increasing depressive symptoms and exposure to abuse due to isolation and household stress. As such, the reported experiences may reflect a heightened risk period. While this may have affected the baseline prevalence, the strong association between abuse and depression remained significant. However, we cannot separate the pandemic’s effects from ongoing post-conflict stressors. Fourth, a further limitation was that we did not collect data on comorbid mental health problems. Factors such as substance abuse, poor academic functioning, or challenges with school adaptation are known to be associated with depressive symptoms and may also be linked to experiences of abuse. The absence of this data prevents us from exploring the complex interplay between these variables and understanding the full spectrum of challenges faced by these adolescents. To overcome these limitations, future longitudinal research is essential to establish causality by examining the temporal sequence between abuse and depression and to track how the protective role of social support evolves throughout adolescence, an approach shown to be feasible in similar high-adversity contexts [48].

Despite these limitations, this study has several notable strengths. First, the large sample size (n = 642) provided considerable statistical power to detect significant associations between abuse, social support, and depressive symptoms, which is particularly important given the relatively low reported prevalence of abuse in our sample. This large sample, drawn from a challenging and under-researched post-conflict setting, offers one of the first quantitative insights into this issue among in-school adolescents in Somaliland. Second, a significant strength was that, to our knowledge, this was the first study to specifically explore the relationship between the experience of abuse and depressive symptoms among children and adolescents in Somaliland. As such, our study contributes crucial and relevant knowledge to a significant evidence gap for this specific population. Finally, the study benefitted from the use of standardized and psychometrically sound instruments (PHQ-9 and MSPSS) that have been used in previous research in African contexts. The cultural adaptation and rigorous data collection protocol, which involved 50 trained medical students and on-site supervision, further enhanced the quality and reliability of the data collected.

6. Conclusions and Implications for Practice

Although the rates of abuse and the severity of depressive symptoms in this study were relatively low, the experience of physical and psychological abuse was associated with higher levels of depressive symptoms among children and adolescents in Somaliland. Conversely, social support was associated with lower levels of depressive symptoms. Although the findings of our study are from a cross-sectional study, we hypothesize that multi-layered interventions are needed. To reduce abuse at its source, there are some interventions that are recommended, such as Parenting for Lifelong Health for Teens [49], which focuses on positive parenting and has been adapted for low-resource settings; this could be a suitable approach to reduce harsh disciplinary practices [44]. Additionally, to leverage social support’s protective power, a multi-pronged approach within schools is crucial, including fostering structured peer support networks, training teachers in psychological first aid, and implementing mental health literacy campaigns to reduce stigma [45]. Given that girls show greater vulnerability to depression, all interventions must be gender-sensitive, offering safe spaces for girls to build supportive networks. We are currently working with an advisory group comprising adolescents, teachers, community members, and healthcare providers to co-develop a culturally appropriate intervention based on these findings. This collaborative process, combined with advocacy for the full implementation of Somaliland’s National Child Protection Policy [26], will ensure that research translates into a meaningful, protective framework for all adolescents.