Social and economic equity and family cohesion as potential protective factors from depression among adolescents living with HIV in Uganda

Abstract

Introduction:

Adolescents living with HIV in Uganda are impacted by poverty and face a number of health and social challenges including access to medication, health complications, and social stigma. These stressors have been linked to depression, which can lead to lower HIV treatment adherence. This study seeks to determine how social and economic equity, family cohesion, and social supports may be related to depression among adolescents living with HIV.

Methods:

We used baseline data from the Suubi+Adherence study, a 5-year longitudinal randomized controlled trial among adolescents living with HIV in southwestern Uganda (N = 675; ages 10–16 years). Hierarchical logistic regression models were conducted separately among in-school adolescents and out-of-school adolescents to assess the hypothesized associations between economic and social equity, social support, and depression.

Results:

About half of the participants meet the criteria for depression. Adolescents with depression were found to have fewer economic and social supports. Our findings indicate that social and economic equity (OR = 0.85, 95% CI: 0.74, 0.99), family cohesion (OR = 0.94, 95% CI: 0.91–0.96), and social support from friends (OR = 0.95, 95% CI: 0.91–0.998) are associated with depression for in-school HIV infected adolescents and could be protective factors.

Discussion:

The results of this study suggest that social and economic equity may play a protective role against depression and other poor mental health outcomes. Potential interventions for adolescents living with HIV should consider these social and familial factors as they may be protective of depression in this population.

Introduction

Studies in Sub-Saharan Africa (SSA), including Uganda, have reported increasing numbers of perinatally HIV-infected children, with slightly over 190,000 children aged 0–17 affected (1) and 9,629 new infections diagnosed in 2013 (2). There are a number of HIV-related health and social challenges that intersect with poverty and occur within low resource settings including limited access to healthcare and medications, persistent and untreated health complications, and social stigma (3–5). These stressors are often compounded by struggles with access to food and other basic resources, and can result in feelings of hopelessness, a lack of empowerment, and even clinical depression - a topic severely understudied among adolescents living with HIV in SSA (6–8). There is some evidence that prevalence of depression among HIV-infected youth in SSA is high. In fact, depression is the most common psychiatric disorder among those living with HIV and the prevalence of depression among HIV-infected youth in SSA is estimated at over 40% (9–13). Both negative health outcomes and social stigma have been related to higher depression, which in turn, leads to lower HIV treatment adherence (14, 15). Furthermore, depression reduces ART adherence, which impacts the health of the individual and increases risk of HIV transmission. It is therefore critical to delineate protective factors that can facilitate positive mental health outcomes and improve one’s capacity to manage an HIV diagnosis and treatment in an environment with severe economic hardships.

The Conservation of Resources Theory of Stress and Coping postulates that a lack of resources, especially those most essential to survival (i.e., housing, services, and support), can impede one’s ability to cope with stressful life situations, including a life-threatening illness (16). Without one’s basic needs being met, an individual is likely to experience detrimental health-related consequences, including severe psychological distress when faced with stressors. Within this context, one factor that can potentially protect against poor mental health outcomes is social equity, defined as the absence of unjust or unfair social or economic disparities (17). Social and economic inequalities have been found to be related to risk factors for many negative mental health outcomes, including depression (18, 19). Several studies have also noted family cohesion and overall family support as protective factors that prevent and moderate symptoms of depression among children and adolescents (20–22). Relatedly, existing literature examining the effects of social support on depression in children and adolescents have found social support to be both a moderating and preventative factor (23, 24). In fact, social support has been found to moderate the impact of food insecurity on depression for individuals living with HIV in Uganda (25).

While there is some research evidence examining the impact of social and economic factors on depression among individuals with HIV, these relationships have not yet been explored among adolescents living with HIV in Uganda. More information may be especially relevant for these young individuals who live in low resource settings and often struggle to have their basic needs for survival met. Therefore, the purpose of the current study is to explore associations between social and economic equity, family cohesion, and social supports and incidences of depression, utilizing data from the Suubi+Adherence study among adolescents living with HIV in Uganda. Guided by the Conservation of Resources Theory of Stress and Coping, we hypothesize that key social and economic determinants, including economic and social equity, family cohesion, and social support impact the depression of our participants.

Methods

Study design and participants.

We utilized baseline data from the Suubi+Adherence study, a five-year (2012–2017) longitudinal randomized controlled trial (R01HD074949) funded by the National Institute of Child Health and Human Development (NICHD). The study recruited adolescents living with HIV between 10–16 years (n = 702) across 39 health clinics in southwestern Uganda. The overall aim of the Suubi+Aherence study was to determine the impact of a family-based economic empowerment intervention on treatment adherence as well as medical and psychosocial outcomes. Data was collected using evidence-based clinical measures and standardized, culturally-adapted assessments. Due to missing data on the depression scale, the current study utilized responses from the 675 participants with the effective information on depression screening questions. The low sample size of the excluded participants (n = 27) precluded further analysis and were likely to be missing at random.

To be included in this study, adolescents had to meet the following inclusion criteria: 1) being between the ages of 10 to 16, 2) being HIV positive and having this status disclosed to them, 3) receiving antiretroviral therapy and care from one of the participating clinics in Southern Uganda, and 4) living with a family (not necessarily biological parents). This study’s recruitment and interaction with human subjects and their health information was completed according to protocols reviewed and approved by the Columbia University (Protocol AAAK3852), the Makerere University School of Public Health (Protocol 210) and the Uganda National Council for Science and Technology (Protocol SS 2969) Ethics and Institutional Review Boards. All participants provided written assent and caregivers gave written informed consent for their adolescents’ participation. Experienced research assistants (RAs) with certifications in Good Clinical Practice and CITI Human Subjects Protection collected the data.

Measures

Depression.

The primary outcome of interest was adolescents’ depression, assessed by the Children’s Depression Inventory (CDI). The CDI has proven successful in several different cultural contexts and is one of the most widely utilized standardized self-report instruments for assessing adolescent depression (26–30). The short version of the CDI, including the 10-item version used in the current study, is adapted from the original 28-item scale. Each item on the CDI has three response options that correspond to varying levels of symptomology for clinical depression (31). Coded 0–2, 0 represents no symptom, 1 represents a mild or probable symptom, and 2 represents a definite symptom (31). Items in the inverse direction were reverse coded to create a summated score. Higher scores indicate higher levels of depression. The theoretical score range for this variable is 0 to 20. The actual score range for the sample in this study was 0 to 20. A score of greater than or equal to 3 indicates clinical depressive symptomology (sensitivity 93.3%, specificity 70.7%) (32).

Economic equity.

Economic equity was measured from two aspects: household financial status using family assets and employment, and food security (33). Composite scores for each aspect were created by quantifying the number of positive response items coded “1”. Specially, family assets were measured using a 20-item index (0–20) assessing the availability of tangible household assets (e.g. house, livestock, garden and transportation). The dichotomous variable was crated as low possession and high possession (6 or fewer reported assets = 0 vs. 7 or more reported assets = 1). Other items for the family assets and employment variable included caregiver employment in the formal labor market (yes = 1 vs. no = 0); available cash savings (yes = 1 vs. no = 0); caregiver participating in a formal banking institution (yes = 1 vs. no = 0); and material housing value (low value including mud or hut = 0 vs. high value including brick = 1).

Food security was assessed using three questions: number of meals per day (1 or fewer = 0 vs. 2 or more = 1); frequency of eating meat or fish in the prior week (1or fewer = 0 vs. 2 or more = 1); and having breakfast on the day of interview (yes = 1 vs. no = 0).

Social equity.

As in our existing study on this topic, social equity was measured using 6 items about access and proximity to community resources and availability of social support for participants (33). Composite scores for each aspect were created by quantifying the number of positive response items coded “1”. The 6 items are school enrollment (yes= 1 vs. no = 0); proximity to school, water source and health clinic (far, very far or no = 0 vs. near or very near = 1); electricity in the home (yes vs. no); and social support for medication adherence (yes = 1 vs. no = 0).

Social support.

Both social support from families (family cohesion and social support related to caregivers) and friends (social support related to friends) was measured. Family cohesion was measured using 8 items adapted from the Family Environment Scale (34) and the Family Assessment Measure (35), both of which have been used in Uganda in the past (28, 36–41). Adolescents were asked to rate how often each item occurred in their family on a 5-point Likert scale, with a range from 1= never to 5= always. Items were reverse coded to create summated scores, with higher scores indicating lower levels of family cohesion. The theoretical score range for this variable 8 to 40. The actual range for the sample in this study was 8 to 36, with a Cronbach’s alpha of 0.80.

Social support from caregivers, classmates, teachers, and friends was measured using 24 items adapted from the Friendship Qualities Scale (28, 40–43), which has been used in Uganda in the past to assess the quality of children’s friendships and relationships with their caregivers, classmates, and teachers using 6 items for each relationship on a 5-point Likert scale, with a range from 1= never to 5= always. For the adolescents who did not enroll in school, all the questions related to classmates and teachers were coded as never. Summary scores were created for each relationship with higher scores indicating higher levels of social support. The theoretical score range for each relationship was for 6 to 30. The actual score range for the sample in this study was the same as the theoretical range, with a Cronbach’s alpha of 0.68.

Covariates.

Other sociodemographic covariates included age, gender (female vs. male), household size (number of people and children in the household), type of the primary caregiver (parents vs. grandparents vs. other relatives), and years since they have known their HIV status which may be associated with depression among African adolescents living with HIV as supported by previous study findings (29).

Statistical analyses

We first conducted bivariate analyses to test for differences in sample characteristics by depression status using survey commands to adjust within-clinic clustering. Under the survey commands, we report Rao-Scott F-statistic (44) for categorical variables and adjusted Wald F-statistics (design-based F) for continuous variables to examine individual-level gender differences on study characteristics. P-values of less than 0.05 were considered statistically significant. As aforementioned, the Conservation of Resources Theory of Stress and Coping is the theoretical framework guiding our study and posits that psychological adjustment to major stress and trauma (i.e., HIV diagnosis and related consequences) is in large part due to one’s preexisting social and economic resources (16, 45, 46). Relatedly, our statistical approach was informed by Lund et al. (2010) who recommend that indicators of deprivation be sequentially added into regression models in order to delineate how each indicator uniquely modifies the strength of the association. Accordingly, a hierarchical logistic regression was conducted to assess the associations between economic and social equity, social support, and depression. Because out-of-school adolescents were not asked the social support questions related to classmate and teachers we built the logistic regression models separately for the in-school-adolescents and out-of-school adolescents. Standard errors and test statistics were adjusted for within-clinic correlation using robust Huber-Whiter sandwich variance estimation (48). Specially, Model 1 controlled for sociodemographic factors, Model 2 controlled for economic and social equity factors, and Model 3 controlled for social support. The regression odds ratio (OR), and 95% confidence interval (CI) for each predictor, are presented in Table 3 and 4. All analyses were conducted using Stata SE. Version 15 (49).

Results

Participant characteristics.

More than half of the participants met the criteria for clinical depression (52.30%). For in-school adolescents, about half met criteria for clinical depression (50.34%) versus out-of-school adolescents where 66.26% met criteria for clinical depression. In-school-adolescents with depression had statistically significantly lower scores for assets & employment (M = 2.08 vs. M = 2.34, p = 0.003), food security (M = 2.13 vs. M = 2.29, p = 0.006), family cohesion (M = 30.36 vs. M = 33.43, p < 0.001), and social support related to friends (M = 15.69 vs. M = 16.82, p = 0.007) (Table 1). Although it is not statistically significant, out-of-school adolescents with depression had substantial lower scores for assets & employment (M = 1.60 vs. M = 1.82, p= 0.459), food security (M = 1.65vs. M = 1.93, p = 2.44), social equity (M = 3.09 vs. M = 3.25, p = 0.551) (Table 2).

Table 1.

Sample characteristics among in-school adolescent subgroup and by depression (n = 592)

	Total	Depression (n = 298)	No depression (n = 294)	Design-based F	p
	n (%) or mean [95% CI]
Age	12.13 [11.81, 12.46]	11.86 [11.57, 12.15]	12.40 [11.99, 12.82]	8.74	0.005
Gender				17.4	< 0.001
Male	257 (43.41)	151 (50.67)	106 (36.05)
Female	335 (56.59)	147 (49.33)	188 (63.95)
Primary caregiver				4.71	0.014
Parent	294 (49.66)	138 (46.31)	156 (53.06)
Grandparent	170 (28.72)	102 (34.23)	68 (23.13)
Other relatives	128 (21.62)	58 (19.46)	70 (23.81)
Household composition
Number of people in household	5.72 [5.50, 5.95]	5.72 [5.44, 6.01]	5.72 [5.38, 6.06]	< 0.001	0.986
Number of children in household	2.35 [2.19, 2.51]	2.39 [2.17, 2.60]	2.31 [2.10, 2.52]	0.25	0.619
Years since they have known their HIV status	3.44 [3.00, 3.869]	3.09 [2.76, 3.43]	3.78 [3.16, 4.404]	6.39	0.016
Economic equity
Assets and employment	2.21 [2.02, 2.40]	2.08 [1.90, 2.26]	2.34 [2.12, 2.56]	9.95	0.003
Food security	2.21 [2.15, 2.27]	2.13 [2.05, 2.22]	2.29 [2.22, 2.36]	8.68	0.006
Social equity	4.55 [4.43, 4.66]	4.46 [4.34, 4.58]	4.64 [4.46, 4.81]	3.05	0.089
Family cohesion	31.89 [31.33, 32.44]	30.36 [29.58, 31.14]	33.43 [32.80, 34.06]	36.26	< 0.001
Social support
Related to caregivers	17.70 [17.32, 18.07]	17.39 [16.92, 17.87]	18.00 [17.50, 18.50]	3.59	0.065
Related to friends	16.25 [15.83, 16.67]	15.69 [15.01, 16.36]	16.82 [16.36, 17.27]	8.01	0.007
Related to classmates	17.32 [17.04, 17.60]	17.01 [16.47, 17.54]	17.64 [17.23, 18.04]	2.93	0.095
Related to teachers	13.20 [12.86, 13.54]	13.24 [12.8, 13.68]	13.16 [12.72, 13	0.1	0.752

Bold text indicates statistical significanct values

Table 2.

Sample characteristics among out-of-school adolescent subgroup and by depression (n = 83)

	Total	Depression (n = 55)	No depression (n = 28)	Design-basedF	p
	n (%) or mean [95% CI]
Age	14.17 [13.58, 14.75]	14.14 [13.57, 14.72]	14.21 [13.23, 15.20]	0.02	0.884
Gender				0.05	0.817
Male	37 (44.58)	25 (45.45)	12 (42.86)
Female	46 (55.42)	30 (54.55)	16 (57.14)
Primary caregiver				1.63	0.203
Parent	24 (28.92)	13 (23.64)	11 (29.29)
Grandparent	28 (33.73)	21 (38.18)	7 (25)
Other relatives	31 (37.35)	21 (28.18)	10 (25.71)
Household composition
Number of people in household	5.90 [5.29, 6.52]	5.60 [4.82, 6.38]	6.50 [5.54, 7.46]	2.45	0.126
Number of children in household	2.39 [1.94, 2.83]	2.27 [1.66, 2.89]	2.61 [1.82, 3.39]	0.4	0.529
Years since they have known their HIV status	4.22 [3.37, 5.06]	4.49 [3.35, 5.63]	3.68 [2.54, 4.81]	0.98	0.328
Economic equity
Assets and employment	1.67 [1.40, 1.94]	1.60 [1.39, 1.81]	1.82 [1.21, 2.43]	0.56	0.459
Food security	1.75 [1.57, 1.92]	1.65 [1.41, 1.90]	1.93 [1.58, 2.28]	1.4	0.244
Social equity	3.14 [2.91, 3.38]	3.09 [2.78, 3.40]	3.25 [2.85, 3.65]	0.36	0.551
Family cohesion	30.52 [29.00, 32.03]	30.55 [28.55, 32.54]	30.46 [27.83, 33.10]	< 0.001	0.963
Social support
Related to caregivers	17.31 [16.55, 18.07]	17.33 [16.37, 18.29]	17.29 [16.37, 18.20]	< 0.001	0.944
Related to friends	15.24 [14.01, 16.47]	15.35 [13.64, 17.05]	15.04 [13.46, 16.62]	0.07	0.792

Logistic regression findings.

Tables 3 and 4 show the results of logistic regression models assessing the associations between economic and social equity, social support, sociodemographic characteristics and depression. For in-school adolescents (Table 3), in Model 1 adjusting for sociodemographic characteristics, gender, age and primary caregiver were found to be associated with adolescent depression. Specifically, the odds of having depression was 1.83 times higher in male adolescents compared with female adolescents (OR = 1.83, 95% CI: 1.37, 2.76). Additionally, each additional year increase in age was associated a 13% decrease in the odds of having depression (OR = 0.87, 95% CI: 0.78, 0.97). Furthermore, the odds of having depression was 1.85 times higher in adolescents whose primary caregiver was grandparent compared with adolescents whose primary caregiver was parent (OR = 1.85, 95% CI: 1.24, 2.76).

Table 3.

Hierarchical logistic regression on sociodemographic characteristics, economic and social equity, and social support predicting depression among the in-school adolescents (n = 592)

	Model 1			Model 2			Model 3
	OR	95% CI		OR	95% CI		OR	95% CI
Sociodemographic characteristics
Age	0.87	0.78	0.97	0.88	0.79	0.97	0.87	0.77	0.98
Gender
Male	1.83	1.37	2.46	1.8	1.31	2.48	1.62	1.15	2.27
Female		Ref			Ref			Ref
Household composition
Number of people in household	0.94	0.85	1.04	0.95	0.85	1.06	0.95	0.85	1.06
Number of children in household	1.1	0.97	1.24	1.09	0.95	1.25	1.07	0.93	1.23
Primary caregiver
Parent		Ref			Ref			Ref
Grandparent	1.85	1.24	2.76	1.77	1.16	2.71	1.83	1.16	2.88
Others	1.03	0.68	1.55	0.92	0.59	1.45	0.86	0.55	1.36
Years since they have known their HIV status	0.96	0.91	1.02	0.96	0.9	1.01	0.96	0.9	1.02
Economic equity
Assets and employment				0.85	0.74	0.98	0.85	0.74	0.99
Food security				0.83	0.71	0.97	0.88	0.73	1.06
Social equity				0.92	0.76	1.12	0.92	0.75	1.13
Family cohesion							0.94	0.91	0.96
Social support
Related to caregivers							1.04	0.99	1.09
Related to friends							0.96	0.91	0.998
Related to classmates							0.99	0.95	1.03
Related to teachers							1.05	0.999	1.1

Bold text indicates statistical significanct values

Table 4.

Hierarchical logistic regression on sociodemographic characteristics, economic and social equity, and social support predicting depression among the out-of-school adolescents (n = 83)

	Model 1			Model 2			Model 3
	OR	95% CI		OR	95% CI		OR	95% CI
Sociodemographic characteristics
Age	0.94	0.68	1.31	0.96	0.68	1.35	0.96	0.68	1.35
Gender
Male	0.87	0.31	2.42	0.88	0.33	2.37	0.86	0.3	2.46
Female		Ref			Ref			Ref
Household composition
Number of people in household	0.79	0.6	1.02	0.8	0.6	1.06	0.79	0.59	1.07
Number of children in household	1.24	0.86	1.8	1.23	0.86	1.77	1.24	0.85	1.81
Primary caregiver
Parent		Ref			Ref			Ref
Grandparent	3.22	1.16	8.96	3.11	1.06	9.13	3.15	1.05	9.43
Others	1.68	0.57	4.91	1.55	0.54	4.44	1.58	0.52	4.79
Years since they have known their HIV status	1.12	0.95	1.32				1.12	0.95	1.32
Economic equity
Assets and employment				1.01	0.57	1.79	1.01	0.58	1.76
Food security				0.81	0.43	1.5	0.8	0.43	1.49
Social equity				1.002	0.61	1.65	1.01	0.61	1.66
Family cohesion							1.01	0.93	1.09
Social support
Related to caregivers							1.02	0.88	1.18
Related to friends							1.002	0.88	1.13
Related to classmates
Related to teachers

Bold text indicates statistical significanct values

Values for social support from classmates and teachers were not included for out-of-school adolescents

In Model 2 adjusting for sociodemographic characteristics plus economic and social equity factors, we found that in addition to age and gender, economic equity was associated with adolescent depression. Specifically, each additional asset and employment unit was associated with a 15% decrease in the odds of having depression (OR = 0.85, 95% CI: 0.74, 0.98) and each additional food security unit was associated with a 17% decrease in the odds of having depression (OR = 0.83, 95% CI: 0.71, 0.97). In Model 3, after further controlling for social support factors, each additional family cohesion unit was associated with a 0.6% decrease in the odds of having depression (OR = 0.94, 95% CI: 0.91, 0.96). Additionally, each additional unit of social support related to friends was associated with a 0.4% increase in the odds of having depression (OR = 0.96, 95% CI: 0.91, 0.998). Food security however, was no longer associated with adolescent depression.

For out-of-school adolescents (Table 4), primary caregiver was the only factor found to be associated with adolescent depression. Specially, the odds of having depression was 3.22 times higher in adolescents whose primary caregiver was grandparent compared with adolescents whose primary caregiver was parent (OR = 3.22, 95% CI: 1.16, 8.96).

Discussion

This study examined the relationship between potential protective factors and prevention of poor mental health outcomes, specifically depression, among adolescents living with HIV in Uganda. Our findings indicate that social and economic equity can be protective factors against depression for adolescents living with HIV. It is notable that for out-of-school adolescents living with HIV, primary caregiver was the only factor found to be associated with adolescent depression and no protective factors against depression were observed among this subgroup. This indicates the potential that these two subgroups of adolescents (in-school adolescents versus out-of-school adolescents) may experience unique etiological factors that impact their depression symptoms. In fact, it may be that out-of-school adolescents are not attending school for numerous reasons including struggle with their mental health, which indicates an important area of future study.

For in-school adolescents, our findings are in line with the Conservation of Resources Theory of Stress and Coping that postulates greater access to basic resources increases one’s ability to cope with stressful life events, including a life-threatening illness, and can improve one’s mental health outcomes. Specifically, our results signal that the more equitable the social and economic environments are, the fewer depression symptoms one experiences. This is consistent with literature that states that reducing disparities and promoting equity increase help-seeking behaviors as well as treatment interest and motivation (50). This, in turn, promotes better treatment adherence and outcomes for adolescents living with HIV and their families (51). Hence, not only is it important to work directly with populations adversely affected by HIV and depression, but it is also vital to address systemic and policy factors that cause disproportionate access to treatment and support (52, 53).

Similar to existing literature (20, 22, 54), family cohesion was also shown to be a factor associated with better mental health outcomes. As indicated in the findings of this study, family cohesion is associated with depression in adolescents living with HIV and should be considered when assessing and delivering treatment for both the physical and mental health needs of this population. Community and family interventions should strive to teach and promote behaviors and interactions that lead to greater family cohesion, such as adaptability, family functioning, and family support (20, 21). Including the familial and social aspects in treatment and intervention strategies have the potential to address not only depression, but also other long term health and mental health needs to promote improved outcomes for both adolescents and their caregivers (55–57).

The study findings did not indicate that social supports either inside or outside of the family were a protective factor for depression for adolescents living with HIV. These findings differ from the previous literature, as both emotional and instrumental social support have been found to be protective factors for depression (4, 25, 58). Other research, however, has shown the effects of social supports to be dampened by perceived and enacted HIV stigma (59). Due to HIV-associated stigma, it may be that adolescents are not likely to disclose their HIV status outside of their immediate family. Thus, support may come solely from caregivers and immediate family members rather than friends, classmates and teachers. More work may be needed in relation to psychoeducation about HIV and mental health within the community in order to reduce stigma and increase accessibility and availability of overall social supports (23, 60).

While equity and family cohesion can be protective factors for experiencing depressive symptoms among adolescents living with HIV, this study also found being male and being a younger adolescent to be risk factors for experiencing symptoms of depression. Our findings show that male adolescents are more likely to be depressed than female adolescents in this population, and that younger male adolescents (age 10–13) are more likely to be depressed than older adolescents (age 15–17). These findings are consistent with other literature examining adolescent depression in sub-Saharan African populations living with HIV (29, 30). In addition, our findings corroborate existing studies that indicate men and boys’ general reluctance to seek out help when struggling with mental health issues due to being socialized that expression of emotions and psychological pain can be a sign of weakness (61–63). Given our findings that females experience significantly more equity, family cohesion, and support related to caregivers, interventions may be needed that specifically address support in these areas for male adolescents as well.

The results of this study should be taken into consideration in the context of certain limitations. First, the use of self-report measures to gather information on depression and other variables of interest is subject to misclassifications, such as social desirability bias. It is also possible that the adolescents may not be fully aware of some of the economic and social conditions that impact their families. Further research should consider gathering information from both adolescents and their caregivers. Additionally, survey items regarding equity, family cohesion, and social supports were designed to gather broad information on these topics. Future research should consider the use of standardized measurement tools to specifically evaluate these variables in the context of depression. Further, limitations to the current statistical analysis include limited statistical power for analyses among out-of-school adolescents due to a low sample size, and a low Cronbach’s alpha for one of the measures used to assess social support (i.e., FQS). Finally, this study used only baseline data, thus causal associations cannot be determined.

This study compared relative levels of equity, family cohesion, and social support for adolescents living with HIV from Uganda and their depression. The more equitable their circumstances and the more cohesive their family unit were, the lower their likelihood for experiencing depression. Understanding the role that social and economic determinants have on depression and potential outcomes is an important step toward developing policy and treatments that can be utilized with this population. Potential interventions for adolescents living with HIV should carefully consider these social and familial factors as results indicate they may be vital to the treatment of depression in this population. In addition, our findings signal the critical need for further exploration to focus on unique subgroups of adolescents with HIV in SSA who have pronounced vulnerability. In particular, the characteristics of being male, a younger adolescent, and out-of-school were observed to especially challenge one’s emotional wellbeing and there is a need for future efforts to consider these risk factors.