Unpacking gender-specific risk and protective factors for mental health status among Congolese refugees in Uganda

ABSTRACT

Background: Research indicates refugees from the Democratic Republic of Congo (DRC), particularly females, have a higher incidence of mental health problems compared to the global norm for conflict-affected populations.

Objective: This study aimed to unpack gender differences in the mental health of Congolese refugees by examining specific risk (trauma exposure, adaptation challenges, and discrimination) and protective factors (marital status, literacy, and social resources) in relation to posttraumatic stress symptoms (PTSS) and depression. Method: Utilizing systematic random sampling, we surveyed 667 adult refugees (370 females, 297 males) in the Kyaka II refugee settlement in Western Uganda. A multi-group path analysis was conducted, initially allowing for variance between genders and subsequently comparing a constrained model, where paths were set equal across groups.

Results: The unconstrained model presented an excellent fit to the data. When paths were set to be equal across groups, the decline in model fit, confirmed by a chi-square difference test, indicated differences in the model for males and females. A series of Z-tests were used to compare individual paths. Experiencing discrimination was a stronger risk factor for depression among men, whereas a history of rape was more strongly associated with depression for females. Being literate and a member of a social group in the settlement were stronger protective factors for depression among men, whereas living with a partner and a felt sense of connection to their community was more important for women. Associations between risk and protective factors and PTSS were more similar across groups, only membership in a group was significantly moderated by gender; with group membership being more impactful for males.

Conclusion: The results highlight similarities and differences in predictors of distress for male and female Congolese refugees and point to potential avenues for tailoring programming to be gender sensitive.

HIGHLIGHTS

• This research identified key differences in how specific traumatic experiences and social factors correlate with psychological distress for male and female Congolese refugees, underscoring the need for tailored support strategies.

• The study reveals that while discrimination is a stronger predictor of depression in men, experiences of rape are more closely linked to depression in women. Conversely, social factors like literacy and group membership offer more protection to men, whereas relationship status and community connection are more protective for women.

• The findings shed light on the importance of gender-sensitive mental health and psychosocial support interventions.

Antecedentes: La investigación indica que los refugiados de la República Democrática del Congo (RDC), particularmente mujeres, tienen mayor incidencia de problemas de salud mental en comparación con la norma mundial para las personas afectadas por conflictos.

Objetivo: Este estudio tuvo como objetivo investigar las diferencias de género en la salud mental de los refugiados congoleses mediante el examen de riesgos específicos (exposición a trauma, desafíos en la adaptación y discriminación) y factores protectores (estado civil, alfabetización y recursos sociales) en relación con síntomas de estrés postraumático (PTSS por sus siglas en inglés) y depresión.

Método: Utilizando un muestreo aleatorio sistemático, encuestamos a 667 refugiados adultos (370 mujeres, 297 varones) en el asentamiento de refugiados Kyaka II en Uganda Occidental. Se condujo un análisis de rutas de múltiples grupos, que inicialmente permitió la variación entre géneros y posteriormente comparando un modelo restringido, donde las rutas se establecieron iguales entre los grupos.

Resultados: El modelo sin restricción presentó un ajuste excelente a los datos. Cuando las rutas se establecieron iguales entre los grupos, hubo una disminución significativa en el ajuste del modelo, confirmada por una prueba de diferencia de chi cuadrado, lo que indica relaciones diferentes en el modelo para hombres y mujeres. Se utilizó una serie de pruebas Z para comparar las rutas individuales en el modelo. Experimentar discriminación fue el mayor factor de riesgo para depresión en los hombres, mientras que una historia de violación estuvo fuertemente asociada con depresión en las mujeres. Ser alfabetizado y miembro de un grupo social en el asentamiento fueron los factores protectores más poderosos para depresión en los varones, mientras que vivir con una pareja y tener una sensación de conexión con su comunidad fueron los más importantes para las mujeres. Las asociaciones entre los factores de riesgo y protectores y PTSS fueron más similares entre los grupos, solo la pertenencia a un grupo fue moderada significativamente por el género; la pertenencia a un grupo tiene más impacto para los varones.

Conclusión: Los resultados enfatizan similitudes y diferencias en los predictores de malestar para refugiados congoleses hombres y mujeres y señalan vías potenciales para adaptar la programación para que sea sensible al género.

KEYWORDS:

• Congolese refugees
• mental health
• Uganda
• social determinants
• gender

PALABRAS CLAVE:

• Refugiados Congoleses
• salud mental
• Uganda
• determinantes sociales
• género

1. Background

The Democratic Republic of Congo (DRC) has been plagued by cycles of conflict and political instability for over 25 years (Kivu Security Tracker, 2022; UNHCR, 2021). Civil wars from 1996 to 1997 and again from 1998 to 2003 have been followed by outbreaks of violence perpetrated by more than 120 militias active in the country (Stearns, 2011). Throughout the DRC, widespread human rights violations have been documented, including abductions, torture, and the systematic use of sexual violence by armed personnel (Johnson et al., 2010; Eriksson Baaz & Stern). Rampant violence, coupled with political and economic instability, has resulted in waves of people fleeing the country over the past several decades. According to the United Nations High Commissioner for Refugees (UNHCR), there are over 962,000 registered Congolese refugees and asylum seekers, primarily hosted in neighbouring countries, with approximately 44.9% currently residing in Uganda (UNHCR, 2021).

Research with Congolese refugees in Uganda (Ainamani et al., 2020; Bapolisi et al., 2020; Ssenyonga et al., 2012) has shown a higher prevalence of mental health disorders, including posttraumatic stress disorder (PTSD) and depression, than the global average for conflict-affected populations (Charlson et al., 2019). There is evidence of a gender disparity in mental health among this population, with female refugees reporting more severe symptoms than males (Ainamani et al., 2020); however, both genders experience elevated levels of mental health problems. The high rates of mental health problems have been attributed to the severity of pre-migration exposure to potentially traumatic events (PTEs) (Ainamani et al., 2020; Chiumento et al., 2020; Kolassa et al., 2010; Neuner et al., 2004; Onyut et al., 2009) coupled with the daily stress of life in displacement contexts (Chiumento et al., 2020). Limited research suggests that factors such as social support, cultural identity, and access to healthcare may also play a role in the mental health outcomes of Congolese refugees in Uganda (Chiumento et al., 2020).

Existing studies have illuminated the critical mental health challenges faced by Congolese refugees in Uganda. However, studies are relatively few and have not comprehensively addressed the nuanced interplay between pre-migration traumas, post-migration stresses, and the protective factors that mitigate mental health outcomes. Understanding the complexities of factors correlated with mental health problems in this population and how they might differ by gender is essential to developing effective, targeted mental health and psychosocial support (MHPSS) interventions for treating distress and promoting well-being.

The purpose of our study was to identify risk and protective factors associated with depression and posttraumatic stress symptoms (PTSS) among Congolese refugees living in the Kyaka II settlement in Uganda and to explore how these factors might differ for males and females. We chose to focus on PTSS and depression because previous studies have identified these as common concerns among the population (Ainamani et al., 2020; Bapolisi et al., 2020). Clarifying the significant risk and protective factors of mental health issues was recently identified as a critical research priority in humanitarian settings (Tol et al., 2023). Recognising the distinct factors linked with mental health for males and females is one crucial step in developing gender-sensitive programming. We based our hypothesised model on previous research with this (Ainamani et al., 2020; Bapolisi et al., 2020; Chiumento et al., 2020; Eriksson Baaz & Stern; Seruwagi et al., 2022; Ssenyonga et al., 2012) and other displaced populations (Farahani et al., 2021) and interviews with key informants in the settlement.

Research has typically shown a dose–response relationship between trauma exposure and psychological distress, with exposure to more types of PTEs associated with increased vulnerability for mental health problems (Kolassa et al., 2010; Neuner et al., 2004). We included the number of conflict-related PTEs to which participants were exposed as a risk factor for elevated PTSS and depression. Sexual violence is particularly likely to lead to poor mental health (Cortina & Kubiak, 2006; Kessler et al., 2017). Although much of the literature on Congolese refugees has focused on the impacts of rape on women (Ainamani et al., 2020; Familiar et al., 2021), population-based research with conflict-affected people in DRC (Johnson et al., 2010) indicated that 23.6% of males had also experienced sexual violence. For this reason, in addition to the number of conflict-related PTEs, we looked specifically at the association between sexual violence and distress for both males and females.

Contextual problems stemming from lack of basic resources, inequalities based on refugee status, and inadequate social networks are associated with mental health problems among Congolese (Bapolisi et al., 2020) and other refugee populations (Miller & Rasmussen, 2010). In this study, we focused on perceived difficulties adjusting to life in the settlement and discrimination based on being a refugee as risk factors for elevated PTSS and depression because qualitative research with the population identified these as critical concerns (Chiumento et al., 2020).

We also evaluated potential protective factors, which have largely been absent in previous studies. Identifying modifiable protective factors is essential for developing programming that supports the well-being of populations living in adversity (Hall et al., 2015), where resources for mental health treatment are typically limited. We tested several psychosocial resources as potential protective factors based on previous research (Hall et al., 2015; Villalonga-Olives et al., 2022) and consultations with key informants working with the population in Uganda. Three of the hypothesised protective factors were conceptualised as indicators of social capital. Social capital has been broadly defined as encompassing an individual's social relationships and participation in social networks (Putnam, 2001). One component of social capital is referred to as structural, comprised of the observable aspects of social relationships, such as networks, connections, and participation in formal organisations or informal social groups. Structural social capital emphasises the roles, rules, and resources available through one's social connections (Grootaert, 2004). Under this component, we evaluated marital status and participation in a social group in the settlement. The second component of social capital is cognitive social capital, reflecting a perceived sense of belonging in the community (Putnam, 2001). We also included literacy as a potential protective factor in the analysis because it has been highlighted as an important factor for adjustment among refugees (UNHCR, 2012).

In summary, the study aimed to advance our understanding of factors associated with PTSS and depression of male and female Congolese refugees in Uganda. We hypothesised that the number of war-related traumatic events, exposure to sexual violence, and contextual problems would be associated with higher symptoms. In contrast, we hypothesised that different social resources (marital status, group membership, cognitive social capital) and literacy would be associated with lower symptoms. We also examined whether the risk and protective factors would differ for males and females, given evidence of gender differences in mental health and the impacts of trauma (Olff, 2017). Because the test for gender differences was exploratory, we did not advance specific hypotheses about differences.

2. Method

2.1. Setting

Kyaka II refugee settlement in southwestern Uganda is one of the primary sites for hosting Congolese refugees (Danish Refugee Council & Danish Demining Group (DDG), 2018). Kyaka II is classified as a settlement rather than a refugee camp per Uganda's national policy of self-reliance and freedom of movement for refugees (Danish Refugee Council & Danish Demining Group (DDG), 2018). The settlement covers an area of approximately 27 square kilometres and is home to over 118,000 refugees and asylum seekers; 95% of the population are from the DRC, with smaller numbers from Burundi, Rwanda, and South Sudan (UNHCR Operational Data Portal (ODP), 2023). Over 80% are women and children, and 51% of the total population is female. The refugees are provided with land plots to build houses and to farm, with food rations decreasing over time. Despite this, refugees living in the settlement face numerous challenges, including limited access to essential services such as health care and education, as well as high levels of poverty and food insecurity. The present study was informed by field studies carried out by organisations working in the Kyaka II settlement that suggested high rates of psychological distress, low uptake of available MHPSS (UNHCR, 2018), and gender differences in vulnerability (Agency for Technical Cooperation and Development (ACTED), 2019).

2.2. Design and procedures

Before beginning data collection, ethical approval was obtained from the Makerere University School of Social Sciences Research and Ethics Committee in Uganda and the Uganda National Council for Science and Technology. Additionally, permission we obtained permission from the Office of the Prime Minister in Uganda to conduct the research in the Kyaka II Settlement.

The study involved a correlational design, with data collected using surveys administered by ten trained enumerators (5 male and 5 female) who were bilingual in English and Kiswahili. Five of the nine zones in the settlement were randomly selected; each zone was sampled proportionately to the zone's population based on the number of households in the zone reported by UNHCR in September 2020. Because lists of residents could not be obtained to generate a random sample, systematic random sampling was used, wherein the enumerator started from a landmark in the zone, counted five dwellings, and approached residents of the 6th to request participation. When multiple people were home, the enumerator asked for the head of the household. If the head of household was not home, the spouse or other eligible adult was asked to participate. After informed consent was obtained, the survey was administered verbally, and responses were recorded digitally using tablets.

2.3. Participants

Participants were 667 adult Congolese refugees (370 female, 297 male). Key informants in the settlement advised that it was better to assess age within ranges, as many people would not know their exact birth date. Approximately 48.6% of the sample was under the age of 35, 36.6% were between the ages of 35 and 49, and a smaller percentage (16.8%) was 50 and older. Participants reported having 0–11 children, an average of approximately four children.

Among females in the sample, 60% reported having no formal education, 31.8% had a primary education, and 8.2% had a secondary education. Only 36.5% of females were literate. Males reported a slightly higher level of formal education; 35.4% had no formal education, 45.8% completed primary school, and 14.5 completed secondary school. More than half of the males (58.2%) reported being able to read and write. The modal number of years living in the settlement was two. Approximately 83% of males and females lived in the settlement for less than five years.

2.4. Measures

The survey questionnaire was comprised of existing scales of the respective constructs described below. A professional translation company back-translated scales that were not available in Congolese Kiswahili. The entire questionnaire was reviewed by two other translators for accuracy prior to beginning data collection. Five MHPSS providers working in the settlement also reviewed questions to evaluate relevance. The final questionnaire was field-tested with members of the target group, who were explicitly asked to inform the interviewers when they did not understand a question. Necessary adjustments were made to ensure items were understood. The means and standard deviations, along with Cronbach’s alpha for continuous variables, and the percentages of participants who responded yes to yes/no items, are shown in Table 1.

Table 1. Measurement scales.

	Females				Males
Variable	α	Min.-Max.	M (SD)	% Yes	α	Min.-Max.	M (SD)	% Yes
1. Married, living with spouse	–	0–1	–	45.7	–	0–1	–	83
2. Literate	–	0–1	–	37.9	–	0–1	–	57.5
3. Group Membership	–	0–1	–	48.6	–	0–1	–	49.8
4. Cognitive SC	.70	0–15	8.86 (3.44)	–	.75	0–16	6.78 (3.80)	–
5. Conflict-related PTEs	–	0–19	14.92 (4.20)	–	–	0–19	14.39 (3.67)	–
6. Rape	–	0–1	–	38.6	–	0–1	–	8.4
7. Adjustment	–	0–3	1.79 (1.00)	–	–	0–3	1.82 (0.95)	–
8. Discrimination	–	0–3	1.29 (1.21)	–	–	0–3	1.22 (1.11)	–
9. Depression	.95	1–4	2.86 (.69)	–	.95	1–4	2.65 (.74)	–
10. PTSS	.96	0–76	47.34 (18.52)	–	.95	0–76	43.66 (18.50)	–

Note: Relationship status coded 0 = not living with a partner (single, spouse elsewhere, widowed), 1 = living with partner. Literacy coded 0 = not literate, 1 = literate. Conflict-related PTEs = number of conflict-related potentially traumatic events. Rape coded 0 = no, 1 = yes. Cognitive SC = cognitive social capital. PTSS = Posttraumatic Stress Symptoms.

Exposure to conflict-related PTEs was assessed with a 20-item questionnaire developed for research with Congolese refugees in Uganda (Ainamani et al., 2020). The items were based on research aimed at identifying the common types of PTEs to which individuals are exposed in conflict-affected areas of the DRC (Ainamani et al., 2020). The items included experiencing and witnessing violence from members of armed groups, experiencing or witnessing robbery, looting, or destruction of property by armed personnel, experiencing, or witnessing torture, and other violent acts. Participants were asked to respond yes or no to each item, and we created a sum score as a proxy for severity of conflict-related trauma exposure. We also assessed rape using the same scale with a single item.

Contextual problems were assessed with five items modified from the Humanitarian Emergency Settings Perceived Needs Scale (HESPER) Scale (Semrau et al., 2012) that assessed the extent to which lack of basic resources (food, clean water, shelter), adjustment to life in the settlement, and discrimination for being a refugee were a problem in everyday life. The measure was found to have good inter-rater and test-retest reliability (Semrau et al., 2012). Participants are read the following prompt: ‘I am going to ask you about the serious problems that you may currently be experiencing. We are interested in finding out what you think − a serious problem is a problem that you consider serious. There are no right or wrong answers.’ Participants responded on a 4-point scale ranging from 0 (no problem) to 3 (very serious problem). The scores for the single items on adjustment and discrimination were used in the analysis.

The Adapted Social Capital Assessment Tool (De Silva et al., 2006) was modified to assess structural and cognitive social capital. For structural social capital, we developed a list of 13 groups present in the settlement with key informant interviews. These were based on themes including farming, religious/spiritual, music/dance, and we included the option to report on a group that was not included in our list. Participants were asked if they were a member of each group (yes/no) and whether they participated regularly (yes/no). For the analysis, we used responses to whether they participated regularly, creating a dichotomous variable about whether or not they participated regularly in groups. Cognitive social capital was assessed with 4-items from the Adapted Social Capital Assessment Tool that measured sense of trust and belonging including ‘Do you feel you are really a part of this community?’ and ‘In general can the majority of people in this community be trusted.’ Participants responded on a 4-point scale ranging from 0 (not at all) to 4 (extremely). We created a mean score for the cognitive social capital scale.

Symptoms of post-traumatic stress (PTSS) were assessed using the 20-item PTSD Checklist for DSM-5 (UNHCR, 2012). Items were summed to create a total score. Symptoms of depression were assessed with the 15-item depression scale from the Hopkins Symptom Checklist (Blevins et al., 2015). The measure is scored by creating a mean score from all items.

We also included a brief demographic questionnaire to obtain information on age, education level, literacy, and marital status. Literacy was assessed with one item that asked participants if they could read and write in any language, to which they responded yes or no. For marital status, we asked if participants were currently married and whether they lived with their spouse in the settlement (yes/no).

2.5. Data analysis strategy

We conducted descriptive and bivariate correlational analyses using SPSS version 28. To address the primary research questions, we conducted a multigroup path analysis using AMOS software (Arbuckle, 2014) to evaluate the significance of hypothesised risk and protective factors and to compare the models between males and females. Multigroup path analysis is a statistical technique used to examine the differences in the relationships among variables across two or more groups (Arbuckle, 2014).

We constructed our theorised model such that risk factors (total conflict-related PTEs, exposure to rape, difficulties adjusting to life in the settlement, and perceived discrimination) and protective factors were predictors of variation in PTSS and depression. Correlations between predictor variables were placed in the model, and covariance was placed between PTSS and depression. After the model setup was complete, it was fitted simultaneously to the two groups to test multigroup moderation based on participant gender. A chi-square difference test was used to compare the overall model between males and females. A series of Z-tests followed this to compare individual paths in the model, as recommended by Gaskin (Gaskin, 2016).

3. Results

Bivariate correlations among study variables are shown in Table 2.

Table 2. Bivariate associations of study variables for females (N = 370) and males (N = 297).

Scale	1	2	3	4	5	6	7	8	9	10
1. Marital Status	–	−.04	.10	−.01	−.15**	−.18**	−.18**	−.03	−.26**	−.20**
2. Literacy	.01	–	.17**	−.09	−.16**	−.02	−.12*	−.03	−.15**	−.13*
3. Group Membership	−.07	.31**	–	.10	−.17**	−.10	−.31**	−.17**	−.19**	−.21**
4. Cognitive SC	.01	.04	.18**	–	.07	−.18**	−.23**	−.37**	−.19**	−.25**
5. Conflict-related PTEs	.07	−.19**	−.18**	.19**	–	.32**	.32**	.16**	.36**	.48**
6. Rape	.01	−.04	−.03	.01	.24**	–	.21**	.20**	.35**	.33**
7. Adjustment	.02	−.06	−.22**	−.19**	.22**	.12*	–	.46**.	.41**	.49**
8. Discrimination	−.03	−.15**	−.16**	−.34**	.13*	.24**	.44**	–	.28**	.34**
9. Depression	.06	−.34**	−.36**	−.28**	.32**	.10	.38**	.43**	–	.84**
10. PTSS	.01	−.24**	−.32**	−.13*	.47**	.17**	.38**	.38**	.74**	–

Note: The correlations in the upper half of the matrix are for females and the correlations on the bottom half of the matrix are for males. Relationship status coded 0 = not living with a partner (single, spouse elsewhere, widowed), 1 = living with partner. Literacy coded 0 = not literate, 1 = literate. Conflict-related PTEs = number of conflict-related potentially traumatic events. Rape coded 0 = no, 1 = yes. Cognitive SC = cognitive social capital. PTSS = Posttraumatic Stress Symptoms. *p < .05, **p < .01, ***p < .001

The unconstrained model, where paths were allowed to vary between males and females, presented an excellent fit to the data χ²(12, 667) = 13.44, p = .338 (GFI = .99, CFI = .99, RMSEA = .01). In the fully constrained model, where all paths were set to be equal across groups, there was a notable decline in model fit, χ²(61, N = 667) = 251.53, p < .001, with fit indices indicating a less than optimal fit (GFI = .94, CFI = .88, RMSEA = .07) compared to the unconstrained model. We conducted a chi-square difference test to compare the fit of the constrained model against the unconstrained model, resulting in Δχ² = 238.09 with a change in degrees of freedom of Δdf = 49, which was statistically significant, p < .001. This significant result from the chi-square difference test indicates that the constraints imposed on the model (i.e. the paths being equal across groups) significantly worsened the model fit. Therefore, it suggests that the relationships in the model differ between males and females, and the hypothesis of equal paths across the two groups is rejected. As such, path coefficients were examined separately for males and females. Unstandardised path coefficients are shown in Table 3. Females in the sample (N = 370) had significantly higher depression scores when they were living without a spouse, experiencing more problems adjusting to life in the settlement, were exposed to a higher number of conflict-related traumatic events, and had a history of experiencing rape. Cognitive social capital and literacy were protective factors associated with lower depression scores. Problems adjusting to the settlement, discrimination for being a refugee, conflict-related traumatic events, and having experienced rape were risk factors for PTSS. Cognitive social capital was also a protective factor for PTSS for females in the sample.

Table 3. Results of multigroup path analysis.

Depression
	Female			Male
	Estimate	SE	p	Estimate	SE	p	z-score^a
Relationship Status	0.204	.060	<.001	0.071	.090	.428	−1.231
Literacy	−0.140	.063	.026	−0.305	.071	<.001	−1.746*
Adjustment	0.145	.036	<.001	0.124	.040	.002	−0.385
Discrimination	0.025	.029	.378	0.151	.036	<.001	2.725***
Conflict-related PTEs	0.035	.008	<.001	0.048	.010	<.001	1.024
Rape	0.240	.066	<.001	−0.121	.126	.337	−2.545**
Group Membership	−0.015	.063	.811	−0.218	.072	.003	−2.121**
Cognitive SC	−0.034	.010	<.001	−0.035	.010	<.001	−0.068
Posttraumatic Stress Symptoms
	Female			Male
	Estimate	SE	p	Estimate	SE	p	z-score
Relationship Status	2.500	1.543	.105	−1.313	2.275	.564	−1.387
Literacy	−1.913	1.603	.233	−3.049	1.800	.090	−0.471
Adjustment	5.153	.913	<.001	3.178	1.014	.002	−1.448
Discrimination	1.494	.736	.042	3.227	.905	<.001	1.485
Conflict-related PTEs	1.431	.201	<.001	1.955	.254	<.001	1.618
Rape	4.272	1.681	.011	0.065	3.183	.984	−1.169
Group Membership	−0.433	1.596	.786	−5.363	1.833	<.001	−2.029**
Cognitive SC	−0.508	.244	.037	−0.374	.246	.129	0.389

Note: Unstandardised estimates are shown. Relationship status coded 0 = not living with a partner (single, spouse elsewhere, widowed), 1 = living with partner. Literacy coded 0 = not literate, 1 = literate. Conflict-related PTEs = number of conflict-related potentially traumatic events. Rape coded 0 = no, 1 = yes. Cognitive SC = cognitive social capital.

^a z-score is a test of the difference in the estimates for females and males on each parameter. *p < .05, **p < .01, ***p < .001.

Males in the sample (N = 297) were more depressed when they had problems adjusting to life in the settlement, experienced discrimination for being a refugee, and had a higher number of different conflict-related traumatic events. Being a member of at least one social group in the settlement and cognitive social capital were protective factors for depression. Problems adjusting to life in the settlement, discrimination for being a refugee, and conflict-related traumatic events were also statistically significant risk factors for PTSS. Being a member of a social group was the only protective for PTSS.

As shown in Table 3, associations between literacy, discrimination, rape, group membership, and depression were significantly moderated by gender. The associations between literacy and discrimination had significantly stronger associations with depression for males than for females, whereas having experienced rape was more strongly associated with depression for females than for males. Further, the association between group membership was significantly stronger for males (and non-significant for females). The association between group membership and PTSS was also moderated by gender, such that the association was significantly stronger for males.

4. Discussion

In this study, we sought to delineate both risk and protective factors associated with PTSS and depression symptoms among a sample of adult Congolese refugees living in Uganda, with an emphasis on gender differences. In line with previous research on this population (Ainamani et al., 2020; Bapolisi et al., 2020; Ssenyonga et al., 2012) and refugees from other countries (Farahani et al., 2021), women reported higher distress than men. The sample reported higher rates of psychological distress than the global average for conflict-affected populations (Charlson et al., 2019). Our findings underscore the profound impacts of complex pre-migration PTEs, ongoing challenges in daily life, and, in some cases, the protective roles of social resources and literacy. The gender-specific insights highlight the importance of addressing social factors associated with distress in MHPSS programming, in addition to trauma-related mental health symptoms of this population.

For the female Congolese refugees in our study, the correlates of depression and PTSS, though overlapping in places, exhibited some distinct patterns. Specifically, exposure to a higher number of conflict-related PTEs, having endured the trauma of rape, and difficulties adjusting to life in the settlement were associated with higher depression scores. Statistically significant protective factors for depression were living with a spouse, having a felt session of connection and belonging in the community (i.e. cognitive social capital), and literacy. Difficulties adjusting to life in the settlement, conflict-related PTEs, and the experience of rape were also risk factors for elevated PTSS symptoms. Additionally, experiencing discrimination due to being a refugee was positively correlated with PTSS for females, but literacy status was non-significant. Cognitive social capital was the only significant protective factor for PTSS.

Among the male refugees in our study, there were also differences and similarities in the pattern of risk and protective factors associated with depression and PTSS. For depression, there were three statistically significant risk factors: the number of conflict-related PTEs, difficulties in adjusting to life in the refugee settlement, and experiencing discrimination based on their refugee status. Being a member of at least one social group within the settlement and cognitive social capital were associated with lower symptoms of depression, underscoring the therapeutic potential of communal ties and a feeling of connection (Chiumento et al., 2020). Risk factors for PTSS among males were the same as those for depression, namely, conflict-related PTEs and difficulties adjusting to life in the settlement were associated with heightened symptoms. However, only membership in a social group within the settlement served as a protective factor, indicating the distinct importance of such group affiliations in mitigating the sequelae of trauma for men.

It is not possible to determine from the data why some of the variables correlated significantly with one symptom category and not another. For example, literacy was identified as a significant protective factor against depression but not for PTSS. This discrepancy might suggest that literacy imparts a sense of agency that specifically buffers against depression symptoms by enhancing coping strategies, communication skills, and access to information. However, the nuanced impact of such factors, given the high correlation between symptom categories, suggests that statistical significance may not always translate into clear distinctions in real-world outcomes. Future research should include qualitative analyses to explore how individuals perceive the role of protective factors like literacy in their mental health. Such insights could illuminate the complex interplay between various factors and their unique contributions to mental well-being.

We found patterns of similarities and differences between males and females concerning risk and protective factors for depression and PTSS. A greater number of conflict-related PTEs and difficulties adjusting to life in the settlement were associated with a heightened risk of depression and PTSS symptoms for both males and females. The experience of rape had a notably stronger association with depression for females compared to males. The observed difference was potentially due to the higher prevalence of rape among the female sample, consistent with other research on the population (Eriksson Baaz & Stern), coupled with the prevalent stigma and discrimination towards survivors (Schmitt et al. 2021). Discrimination was a significantly more potent risk factor for males than for females in relation to depression but was a risk for elevated PTSS for both genders in the sample. One potential explanation for this difference lies in the traditional gender roles prevalent in DRC culture, where men are customarily the primary providers for their families (Lwambo, 2013). Discrimination likely hinders employment opportunities, creating substantial barriers for men in fulfilling this conventional role. This dynamic can lead to feelings of inadequacy or failure, potentially triggering or exacerbating depression. The stress of encountering persistent discrimination may exacerbate symptoms of PTSS for both males and females (Chiumento et al. 2020), highlighting the complex interplay of sociocultural factors and mental health within this population. However, these propositions require empirical evaluation.

Living with a spouse was only a significant protective against depression for females. In the present study, it is possible that having a spouse provided additional emotional support and safety for women in the settlement. Similarly, results of a study with Somali refugees in Ethiopia found that being divorced was a risk factor for depression for the total sample of males and females (Feyera et al., 2015). Additional research is needed to unpack the role of partner relationships in supporting or detracting from the mental health of this and other refugee populations.

Literacy was a protective factor for depression for males and females in the sample, although the association was significantly stronger for males. One explanation is that literacy provides a sense of agency, allowing refugees to navigate life in displacement better. Research with refugees in low and high settings has demonstrated that individuals with more formal education, presumably indicating literacy, reported fewer psychological difficulties (Farahani et al., 2021).

Group membership emerged as a significantly stronger protective factor for males, with stronger correlations with both depression and PTSS. This finding suggests that belonging to and participating in social groups offer a critical buffer for men in this context. In contrast, for females, at least in relation to depression, a sense of connection and belonging was more salient. In their systematic review of risk and protective factors among forced migrants, Farahani et al. (2021) found that social networks, assessed in myriad ways, were a critical protective factor for diverse populations of refugees and asylum seekers. The finding that the type of support (participation in a group vs. felt sense of connection) differed for males and females echoes findings of a qualitative study with Congolese refugees in Kenya (Tippens, 2020), indicating that women utilised social bonds in times of stress. In contrast, men reported turning to formal networks for instrumental support. These findings point to a potentially significant difference in how males and females in this population experience and benefit from social relationships.

Results shed light on the complexity of factors that correlate with PTSS and depression symptoms among Congolese refugees in Uganda and the similarities and differences for males and females. While the dose–response relationship between trauma exposure and psychological distress is well-established for this (Ainamani et al., 2020; Ssenyonga et al., 2012) and other populations (Neuner et al., 2004), our findings highlight the varying risk of ongoing challenges like adjusting to living in a refugee settlement and experiences of discrimination. Findings underscore the importance of literacy, particularly for males, which is important in this context where a substantial part of the population had limited access to formal education in DRC. Results also point to different implications for males and females, described below.

4.1. Limitations

Results must be interpreted together with the limitations in study design, the timing of data collection, the measures, and the data collection procedure. First, the correlational design prohibits causal inferences. There is probably a bidirectional relationship between certain protective factors and symptoms of distress; longitudinal data would be needed to clarify the direction of influence.

A significant methodological concern arises from the timing of our data collection, which occurred exclusively during the daytime. While necessary for moving around the settlement for practical and safety reasons, such a sampling strategy might have inadvertently skewed our sample toward those more likely to be home, potentially those experiencing heightened distress. This selection bias is corroborated by the discrepancy in our findings related to income-generating activities. In contrast, only 10% of males and 5% of females in our study reported such engagement; UNHCR estimated a rate of 28% for the broader Kyaka II population.

Additionally, the timing of our data collection coincided with the pandemic in December 2020, a global crisis associated with elevated distress across numerous populations. It remains plausible that the pandemic could have exacerbated the levels of distress observed, intertwining with the existing traumas and daily stressors faced by the refugee community.

Reliance on self-reported data invariably introduces a degree of subjectivity, where under-reporting may result from stigma or fear while over-reporting might be motivated by perceived benefits. Furthermore, PTSD and depression serve as foundational constructs in Western psychiatric paradigms, their full applicability to the Congolese refugee context remains an open question. Recognising that these constructs may not wholly encapsulate the rich tapestry of emotional and psychological experiences inherent to this population is essential. Future research would benefit from a deeper exploration of cultural concepts of distress. It is also important to note the high correlations between PTSS and depression, which calls into question whether these are distinct categories of symptoms.

In our statistical analysis, we could not evaluate the association between economic stress and mental health outcomes because of the ceiling effect, given that more than 95% of participants reported a lack of access to basic needs. Thus, our results do not reflect the likely profound association between extreme economic deprivation and mental health status. Lastly, our assessment tool also bore certain omissions. For instance, we did not account for disability status, a significant oversight given its implications for marginalisation within the Kyaka II settlement.

4.2. Implications and future research

From a practical standpoint, our results highlight the importance of addressing trauma-related mental health symptoms among this population. Refugees from DRC have typically been exposed to multiple conflict-related to PTEs in addition to years of adversity from poverty and uncertainty in the context (Ainamani et al., 2020; Bapolisi et al., 2020; Kivu Security Tracker, 2022; Ssenyonga et al., 2012). Results highlight potential areas for adapting programming to be gender sensitive. It is imperative that providers are aware of the high probability that females from this population have experienced sexual violence (Eriksson Baaz & Stern). At the same time, 10% of our sample of males reported having experience rape, and this could be an underestimate given the stigma of reporting and previous studies of the population in DRC (Johnson et al., 2010). Thus, the potential that male refugees have experienced sexual violence should not be overlooked.

Given the limited resources in refugee settlements in Uganda and other low-income countries, it is essential to incorporate supportive programming aimed at building resilience among the population, as mental health treatment is often not available. Our findings suggest that it is more important for females to create social programming that fosters a sense of connection. In contrast, males may benefit more from group membership with shared activities and goals. That finding that literacy was protective against depression symptoms, particularly for men, highlights the importance of advocating for educational opportunities for refugees in this context.

Difficulties adjusting to life in the settlement were significant for both males and females. Because this was asked generally in our study, future studies must determine the specific challenges in adjustment to intervene appropriately. Results also suggest that discrimination is related to psychological distress, particularly for men. Programming might address this through peacebuilding intervention between refugee and host populations to counteract some of the tensions that have emerged in recent years in Uganda (Bjørkhaug, 2020).

Findings point to avenues for future research. It will be essential to determine if there are differing outcomes for males and females in MHPSS and if implementing gender-specific supportive strategies leads to better mental health outcomes. Mixed methods or qualitative research could further unpack the nuances of gender differences and how they relate to mental health.

The potential biases introduced by our sampling method and the self-report nature of our data collection underscore the value of employing diverse methodologies in future studies. For instance, utilising mixed methods approaches could meld the richness of qualitative insights with quantitative robustness, capturing the complexities of the refugee experience more holistically. Additionally, the temporal limitations of cross-sectional designs, like our study, highlight the need for longitudinal research. Such endeavours would allow for a more dynamic understanding of how refugee mental health trajectories evolve, especially in the context of ongoing crises like the pandemic.

5. Conclusion

In summary, with this study, we explored risk and protective factors for mental health among Congolese refugees in Uganda. Results highlight the importance of addressing the lasting impacts of past trauma and ongoing life difficulties and the importance of considering gender when developing supportive interventions. As the global community grapples with the refugee crisis, research on risk and protective factors is pivotal in guiding interventions and policies to alleviate displaced individuals’ profound challenges. Through our findings, we aim to contribute to efforts that ensure their mental well-being is prioritised and addressed.

Disclosure statement

No potential conflict of interest was reported by the author(s).