Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 21.
Published in final edited form as: Prev Sci. 2025 Jun 11;26(6):862–872. doi: 10.1007/s11121-025-01821-y

The Effect of Depression on the Pathways Between an Economic Strengthening Intervention and ART Adherence in Youths with HIV: Findings from a Moderated Mediation Model of the Suubi + Adherence Cluster-Randomized Study

Samuel Kizito 1, Fred M Ssewamala 1, Josephine Nabayinda 1, Proscovia Nabunya 1, Ozge Sensoy Bahar 1, Torsten B Neilands 2, Mary M McKay 1
PMCID: PMC13003161  NIHMSID: NIHMS2137952  PMID: 40500555

Abstract

Adolescents living with HIV (ALHIV) have low adherence to antiretroviral therapy (ART). Poverty and mental health challenges remain major drivers of this poor ART adherence. We explored the mediators of the impact of an economic empowerment intervention on ART adherence among ALHIV and assessed the moderating effects of depression. We randomized 39 clinics (702 participants) into the control or intervention groups. Participants were aged 10–16, living with HIV and taking ART. The intervention comprising matched savings account, financial literacy training, and microenterprise workshops. We fitted a sequential structural equation model to examine how the three mediators—HIV stigma, barriers to medical care, and healthcare transition readiness—influenced ART adherence at year seven. Depression was included as a moderator. At baseline, the mean age was 12 years, and only 73.0% achieved good adherence (≥ 90%). The intervention directly improved ART adherence, β = 0.060 (95% CI: 0.038, 0.081), p < 0.001. Also, there was a significant indirect effect of the intervention on ART adherence, mediated through barriers to medical care, β = −0.036 (95% CI: −0.041, −0.032), p < 0.001, and HIV stigma, β = −0.011 (−0.016, −0.007), p < 0.001. Depression reduced the effect of the intervention on ART adherence β = −0.114 (−0.123, −0.104), p < 0.001. Our results showed that providing ALHIV with financial resources improved their ART adherence; however, this was affected by depression. Therefore, programs aimed at improving outcomes in ALHIV should consider incorporating interventions that address mental health challenges in addition to poverty.

Keywords: Adolescents, Antiretroviral therapy adherence, Depression, Economic empowerment, HIV/AIDS, Sub-Saharan Africa

Background

Successful management of HIV/AIDS depends critically on sustained ART adherence, which is essential for achieving viral load suppression—the primary indicator of treatment success and key factor in reducing HIV transmission. ART adherence has additional benefits, such as reduced risk for HIV-associated complications and comorbidities, prolonging survival and enhanced quality of life, and prevention of horizontal HIV transmission (Slogrove et al., 2018). However, to realize these benefits of ART, people living with HIV (PLHIV) must adhere to ART (Byrd et al., 2019). Unfortunately, adolescents living with HIV (ALHIV) have poor ART adherence, especially those in low-resourced settings (Kamau et al., 2024). In Uganda, the location of the current study, ART adherence among ALHIV is low, estimated at ~73% (Vreeman et al., 2018).

Evidence shows that economic instability significantly contributes to the poor ART adherence among ALHIV (Becker et al., 2020; Kalichman & Grebler, 2010; Twekambe et al., 2023), necessitating more efforts towards reducing poverty. For instance, a qualitative study among ALHIV in Kenya identified poverty as a major reason for poor engagement in HIV care and adherence to treatment. In the same study, the participants also expressed a desire for financial interventions (Enane et al., 2020), demonstrating the need for interventions that target household financial instability.

However, few interventions address poverty despite its central role in undermining treatment outcomes among ALHIV. A recent systematic review of interventions to improve ART adherence in SSA found that none of the 31 studies addressed poverty (Damulak et al., 2021). In addition, our literature search on PubMed for randomized clinical trial (RCT) testing economic empowerment (EE) interventions to improve adolescent ART adherence in Sub-Saharan Africa (SSA) yielded only one RCT, the Suubi + Adherence study (Ssewamala et al., 2019), from which the current study is derived. A similar paucity was noted when we searched for economic empowerment interventions targeting viral suppression, a more definitive biomarker of adherence effectiveness. This gap highlights the need for interventions that not only enhance adherence but also directly evaluate downstream effects of adherence. Moreover, the specific pathways through which the EE interventions exert their effect on ART adherence have not been fully explored. To illustrate, although HIV stigma hinders all aspects of the HIV care continuum (Katz et al., 2013), its mediating effect in the relationship between the EE intervention and ART adherence has not been assessed, despite being a significant factor in mediating the effect of EE interventions on other outcomes such as mental health (Cavazos-Rehg et al., 2021). Similarly, the mediating role of other factors that hinder ART adherence, such as barriers to medical care and healthcare transition readiness—defined as how prepared the ALHIV are to autonomously manage their HIV treatment needs as they transition to young adults—needs to be examined (Mbalinda et al., 2021).

Furthermore, ALHIV in SSA face a high risk for mental health challenges, especially depression, affecting over 40% of this population (Kinyanda et al., 2020). It is well established that depression is linked to poor HIV treatment outcomes (Ashaba et al., 2019; Dow et al., 2016; Uthman et al., 2014), and recent work from our team showed that depression mediates the effect of EE interventions on ART adherence(Kizito et al., 2023). However, its moderating effect in the pathways linking EE and ART adherence has not been established. A deeper understanding of these associations could provide valuable insights into the successful implementation of interventions that target poverty and uncover potential targets for future interventions.

Therefore, in this study, we incorporated a moderated mediation model examining the mediators of the effect of an EE intervention on ART adherence among ALHIV in Uganda, while simultaneously examining the moderating effect of depression. The three mediators we tested include perceived HIV stigma, barriers to medical care, and healthcare transition readiness, since literature has shown that these factors negatively influence treatment outcomes in patients ALHIV (Adejumo et al., 2015; Ahmed et al., 2017; Mbalinda et al., 2021). We hypothesized that participants in the intervention group would have (H1) better ART adherence compared to participants in the control group; (H2) lower perceived stigma, better healthcare readiness when transitioning from adolescence to young adults, and would report fewer barriers to medical care, leading to better ART adherence; and (H3) these direct and indirect pathways would be moderated by depression.

These exploratory hypotheses are guided by Asset theory (Sherraden, 1990; Sherraden & Gilbert, 2016). Beyond its role in informing the economic empowerment intervention, asset theory provides a framework for selecting the mediators and moderators in this study. Specifically, access to financial resources not only enhances economic stability but also fosters psychosocial benefits, such as improved mental well-being and self-efficacy. For families affected by HIV, the accumulation of assets provides a crucial safety net to mitigate the negative impacts of health crises, loss of income, and other challenges associated with having HIV. By accessing resources through their assets, these families can cope with such financial and emotional stresses and maintain their well-being. Additionally, asset theory further asserts that owning assets results in additional benefits including increased personal efficacy and social power. These indirect benefits play a crucial role in mitigating perceived HIV stigma. In our study, we extend asset theory to examine how economic empowerment reduces perceived stigma and barriers to medical care—both of which have been linked to better ART adherence. Additionally, we explore the moderating effect of depression, recognizing that while financial stability can mitigate stressors, pre-existing mental health conditions may hinder the full realization of these benefits.

Methods

The study utilized data from the Suubi + Adherence Study, a two-group cluster-randomized controlled trial (Ssewamala et al., 2019). Participants were recruited from 39 health clinics in Southern Uganda. This region has an HIV prevalence of 11.7%—twice the national average of 5.4% (UAC, 2021). Adolescents were included in the study if they fulfilled the following criteria: (1) aged 10 to 16 years; (2) living with HIV; (3) aware of their HIV status; (4) living within a family setting; (5) receiving ART from one of the collaborating clinics involved in the study.

Randomization and Masking

Randomization was conducted at the clinic level by a Columbia University student who had no direct involvement in the study. A 1:1 ratio was used to allocate the clinics to either the intervention (N = 20 clinics, n = 358 ALHIV) or control groups (N = 19 clinics, n = 344 ALHIV). To minimize the risk of contamination, all eligible ALHIV recruited from a particular clinic were assigned the same intervention—corresponding to the study group to which the clinic was assigned. Due to the nature of the intervention, it was not feasible to blind the ALHIV to their assigned groups.

Bolstered Standard-of-Care (BSOC)

As part of routine care, all ALHIV receive comprehensive HIV clinical and psychological care in accordance with the Uganda government Ministry of Health’s recommended guidelines (Health, 2020). These guidelines outline the standard of care that ALHIV should receive, including ART adherence and viral load monitoring, psychosocial support, and adherence counseling. However, routine HIV care is often inconsistent, varying in frequency and quality. Hence, all ALHIV enrolled in our study received bolstered adherence counseling, during which age-appropriate information on various topics related to adherence was provided.

Intervention

Participants in the intervention group received a family EE intervention comprising a youth development account (YDA), microfinance, and financial literacy training. Specifically, a savings account was opened for each participant and activated with an initial deposit of approximately ~ 10 USD sponsored by the study. In addition to the initial deposit, the study matched the amount saved by the participant monthly (up to a maximum of around 10 USD) by depositing an equivalent sum into their account for a duration of 2 years. Importantly, even after the intervention period, participants were encouraged to retain their accounts, with caregivers urged to continue making savings deposits. The participant and the family could use the savings to meet their basic needs, such as medical costs associated with HIV care and treatment (Curley et al., 2016). Also, by introducing the families to formal banking, they had more opportunities, such as creating lines of credit that could be accessed in case families needed to boost their income, including investment in small business development to establish sustainable sources of income for the family. In addition, the intervention incorporated financial literacy and micro-enterprise workshops, which involved active participation from both the ALHIV and their family members. These workshops comprised a series of up to eight didactic sessions, focusing on topics such as initiating a micro-enterprise, effective financial management, setting short- and long-term goals, and the fundamentals of starting a business. By helping families establish businesses, this intervention component ensures that the families had a sustainable source of income to meet the family needs, such as providing basic needs, keeping the children in school, and health care costs.

Study Measures

The outcome was ART adherence, measured at year seven (7) of follow-up. A latent variable for ART adherence was created using ART adherence measured using pill counts and five self-reported items. Pill counts adherence was calculated as the number of pills taken divided by the expected doses. This method has been used elsewhere (Haberer et al., 2012). In addition, three of the five self-reported items were adapted from Wilson’s three-item self-reported adherence tool (Wilson et al., 2016), and these included: (1) In the last 30 days, how many days did you miss at least one dose of your HIV medications? The expected response was the number of days (range 0 to 30); (2) In the last 30 days, how often did you take your HIV medicine in the way you were supposed to? (1 = never, 2 = rarely, 3 = sometimes, 4 = usually, and 5 = always.); and (3) How hard is it for you to take your HIV medicines in the way you are supposed to? (1 = extremely hard to 5 = not hard at all). The two additional items were (1) “When was the last time you missed any of your HIV medication?”—with responses from 1 “never missed” to 6 “more than 3 months” and (2) “How hard is it for you to take your HIV medicines in the way you are supposed to?”, which was scored on a 5-point Likert scale ranging from “extremely hard” to “not hard at all.” Adding more observed items was critical in improving the overall fit of the measurement model for ART adherence. Moreover, using a single item to measure self-reported adherence is prone to recall and desirability biases (Simoni et al., 2006); thus, multiple items are recommended for more accurate assessment. Furthermore, combining self-reported adherence with pill count methods has been shown to improve the accuracy of ART adherence measurement and predict virologic failure more effectively (Wu et al., 2014).

Mediators

Barriers to Medical Care

To understand the ALHIV’s access (or lack thereof) to medical care, we used an 11-item barriers to medical care scale (Kalichman et al., 1999). ALHIV were asked to think of the reasons why they did not get the medical care they needed. A sample item on the scale was “I did not have transportation to medical care.” Each response was coded as 0 if the ALHIV responded “disagree” and 1 if “agree” was selected. A total score was generated by adding up the individual scores from each item, with a possible range of 0 to 11, with higher scores indicating more barriers to medical care (alpha = 0.73).

Healthcare Transition Readiness

This was assessed using a 17-item transition readiness assessment questionnaire (TRAQ) (Wood et al., 2014). Items in this scale examined the participants’ readiness to handle their daily lives and medication adherence. A sample item was “Do you arrange for your transportation to medical appointments?” The items were scored on a 5-point Likert scale ranging from 1 = “No, I do not know how” to 5 = “Yes I always do this when I need to.” Responses from each item were added to generate a total TRAQ score. Higher scores indicated more healthcare transition readiness (alpha = 0.71).

Perceived Stigma

This was measured using 21 items from the social impact scale (Poku et al., 2022). Collectively, the items constitute three sub-scales including social isolation, social rejection, and internalized shame. Each item had responses on a 4-point Likert scale ranging from 1 = “strongly disagree” to 4 = “strongly agree.” The social isolation sub-scale comprised seven items (e.g., due to my illness, I sometimes feel useless). The social rejection sub-scale had nine items with items such as “some family members have rejected me because of my illness.” The internalized shame sub-scale had five items with a sample item reading “I feel others think I am to blame for my illness.” Cronbach’s alpha for the overall measure was 0.86.

Moderator

Depression

Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale (CES-D) (Vilagut et al., 2016). The CES-D consists of 20 items that assess the frequency of symptoms experienced in the past week, rated on a 4-point scale ranging from 1 = “rarely or none of the time” to 4 = “most or all of the time.” An example item is “During the past week, I did not feel like eating; my appetite was poor.” The total score ranges from 20 to 80, with higher scores indicating more severe depressive symptoms (alpha = 0.82).

Data Analysis

All descriptive analyses were conducted using Stata 18.0. Continuous variables were summarized using means and standard deviations, while categorical variables were summarized with frequencies and percentages. Baseline characteristics were compared across the two study groups to ensure the effectiveness of the randomization. To compare the groups, while adjusting for clustering, survey data commands including svy: regress and svy: tab for continuous and categorical data were used, respectively.

Mplus software, version 8.10, was used to fit the moderated mediation models. Specifically, the direct, indirect, and total effects of the EE intervention on ART adherence were assessed, together with the moderating effect of depression. Structural equation modeling (SEM) was employed, applying the cluster command—to address clustering of the participants within the clinics—while applying 2000 bootstrap resamples. The models were fitted such that the temporal order of relationships was preserved. That is, the intervention preceded the mediators and moderator (year six), which also preceded the outcome (year seven). Initially, separate measurement models were run, using confirmatory factor analysis (CFA), to evaluate the factor structure for each latent variable used in the structural model (that is, barriers to medical care, healthcare transition readiness, perceived stigma, and ART adherence). For perceived stigma, second order CFA was performed. Latent variables were created for social rejection, social isolation, and internalized shame, which were then used to measure perceived stigma as a second order latent variable, the latent variable used in the final model. Then, a structural model was fitted by combining the measurement models with the intervention. As shown in Supplementary Fig. 1, the effect of the intervention on the mediators is represented by paths a1 to a3, while paths b1 to b3 represent the effect of the mediators on ART adherence (Gunzler et al., 2013; Hair et al., 2021).

The indirect effect of the intervention on ART adherence was obtained as the product of paths, for example, a1 and b1 (a1*b1). The direct effect, represented by c’, was the effect of the intervention on ART adherence conditional on (i.e., controlling for or in the presence of) the mediators and moderator. The total effect of the intervention on ART adherence, denoted c, was the sum of the direct and indirect effects (Gunzler et al., 2013). The standardized coefficients, β, and their bias-corrected bootstrapped confidence intervals were reported. A significance level of p = 0.05 was considered. The moderating effects of depressive symptoms were assessed by including depression (observed variable) and the depression-intervention interaction terms in the model. In the structural model, we controlled for the participants’ age (measured at baseline) and whether the participants had a treatment buddy. However, for simplicity of the models, these were excluded from the results tables and figures. Additionally, after fitting the structural model, we plotted the moderated mediation region of significance.

Model fit evaluation was done for each measurement model and for the structural model using chi-square statistic for exact global model fit and found all p-value < 0.05. Hence, we relied on approximate model fit indices, including the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). Meeting one of the following two criteria indicated satisfactory approximate model fitness: (a) CFI ≥ 0.95 and SRMR ≤ 0.08 or (b) RMSEA ≤ 0.06 and SRMR ≤ 0.08 (Hu & Bentler, 1999).

To handle missing data while maximizing the efficiency of the available data, the full information maximum likelihood (FIML) method was used in the estimation of SEM parameters. FIML is one of the most used and recommended methods to handle missing data in SEM (Wang & Wang, 2019). Instead of discarding incomplete cases, FIML estimates model parameters using all available data.

Results

The participant flow during the Suubi + Adherence study is presented as supplementary Fig. 1. As shown in Table 1, at baseline, the mean age of the ALHIV was 12.4 years (SD = 1.98), and more than half (56.4%) were females. The average duration of time the ALHIV have been taking ART was 3.9 years (SD = 3.36). At baseline, the mean ART adherence was 77.8%. Overall, there were no statistically significant differences in baseline demographics and mediators across the two study groups.

Table 1.

Characteristics of 702 adolescents living with HIV in Uganda

Characteristics Mean (SD)/number (percentage)
Control Intervention Total
Baseline characteristics
Participant age (min/max: 10 to 16) 12.4 (1.97) 12.5 (1.98) 12.4 (1.98)
Sex assigned at birth (female) 193 (56.1) 203 (56.7) 396 (56.4)
Duration (years) with HIV (min/max: 0 to 16) 3.7 (2.97) 3.6 (3.12) 3.64 (3.05)
Duration (years) on ART (min/max: 0 to 16) 3.9 (3.22) 4.0 (3.51) 3.94 (3.36)
Orphanhood status
 Non-orphan 115 (33.9) 131 (37.3) 246 (35.7)
 Single orphan 129 (38.1) 133 (37.9) 262 (38.0)
 Double orphan 95 (28.0) 87 (24.8) 182 (26.4)
People in the household (min/max: 2 to 18) 5.78 (2.46) 5.70 (2.65) 5.74 (2.56)
Moderator and mediators (at year six)
Barriers to medical care (min/max: 0 to 8) 1.49 (1.91) 1.47 (1.75) 1.48 (1.84)
Healthcare transition (min/max: 33–85) 74.0 (7.44) 73.7 (8.67) 73.9 (7.95)
Perceived stigma
 Isolation (min/max: 7 to 28) 14.4 (4.60) 13.9 (4.79) 14.2 (4.68)
 Rejection (min/max: 8 to 32) 13.7 (4.99) 13.5 (4.92) 13.7 (4.96)
 Shame (min/max: 5 to 20) 11.4 (2.97) 11.12 (2.86) 11.3 (2.93)
Depressive symptoms (min/max: 20 to 67) 34.8 (9.23) 35.0 (9.60) 34.9 (9.37)
ART adherence (at year seven)
Mean pill count adherence (min/max: 0 to 100%) 76.4 (25.7) 80.1 (24.4) 77.8 (25.3)
Mean days without missed dose in the last month 29.4 (1.90) 29.6 (0.84) 29.4 (1.60)
Last time patient missed medication
 Never missed/1–2 weeks ago 66 (19.2) 31 (15.4) 97 (17.8)
 2–4 weeks ago 22 (6.40) 8 (3.98) 30 (5.50)
 1 month-more than 3 months ago 237 (68.9) 149 (74.1) 386 (70.8)
In last 30 days, did a good a job taking ART
 Very poor/poor 4 (1.16) 2 (1.00) 6 (1.10)
 Fair/good 68 (19.8) 33 (16.4) 101 (18.5)
 Very good/excellent 272 (79.1) 166 (82.6) 438 (80.4)
In last 30 days, how did you take ART as recommended
 Never/rarely 11 (3.20) 9 (4.48) 20 (3.67)
 Sometimes 29 (8.43) 13 (6.47) 42 (7.71)
 Usually/always 304 (88.4) 179 (89.1) 483 (88.6)
How hard it is to take HIV medicine as recommended
 Extremely hard/very hard 3 (0.87) 2 (1.00) 5 (0.97)
 Somewhat hard 34 (9.88) 18 (8.96) 52 (9.54)
 Not very hard/not hard at all 307 (89.2) 181 (90.1) 488 (89.5)
Open in a new tab

The ART adherence was measured using pill counts

Direct and Indirect Effects

We observed a significant direct effect of the intervention on ART adherence, β = 0.060 (0.038, 0.081), p < 0.001 (Table 2, Fig. 1). Specifically, ALHIV who participated in the intervention had higher ART adherence levels compared to the control group. This finding supports hypothesis H1. In addition, we identified a significant total indirect effect of the intervention—through two of the mediators—on ART adherence, β = −0.049 (−0.056, −0.042), p < 0.001, as was hypothesized in H2.

Table 2.

Moderated-mediation effects of the economic empowerment intervention on ART adherence

β (95% CI) p - value
Effects on the mediators
Barriers to medical care
 Intervention → Barriers to care − 0.248 (− 0.270, − 0.226) < 0.001
 Depression → Barriers to care 0.097 (0.088, 0.106) < 0.001
 Depression*intervention → Barriers to care 0.452 (0.444, 0.460) < 0.001
Healthcare transition readiness
 Intervention → Healthcare transition readiness − 0.064 (− 0.090, − 0.038) < 0.001
 Depression → Healthcare transition readiness − 0.075 (− 0.080, − 0.069) < 0.001
 Depression*intervention → care transition readiness − 0.147 (− 0.154, − 0.139) < 0.001
Perceived stigma
 Intervention → Perceived stigma − 0.114 (− 0.136, − 0.092) < 0.001
 Depression → Perceived stigma 0.196 (0.187, 0.204) < 0.001
 Depression*intervention → Perceived stigma 0.406 (0.399, 0.414) < 0.001
Mediators → ART adherence
 Barriers to medical care 0.146 (0.130, 0.163) < 0.001
 Healthcare transition readiness 0.016 (0.004, 0.027) 0.006
 Perceived stigma 0.099 (0.084, 0.115) < 0.001
Moderators → ART adherence
 Depression − 0.032 (− 0.042, − 0.022) < 0.001
 Depression*Intervention interaction − 0.114 (− 0.123, − 0.104) < 0.001
Specific indirect effects
 Barriers to medical care − 0.036 (− 0.041, − 0.032) < 0.001
 Healthcare transition readiness − 0.001 (− 0.002, 0.001) 0.154
 Perceived stigma − 0.011 (− 0.016, − 0.007) < 0.001
Effects from intervention to ART adherence
 Direct effect (Intervention → ART adherence) 0.060 (0.038, 0.081) < 0.001
 Total indirect effect − 0.049 (− 0.056, − 0.042) < 0.001
 Total effect 0.011 (− 0.009, 0.030) 0.276
Open in a new tab

Fig. 1.

Fig. 1

Open in a new tab

A SEM for the moderated mediation effect of an EE intervention on ART adherence. Standardized coefficients are presented, and significant associations are represented by an asterisk. Note: Asterisk (*) signifies coefficients with statistically significant p-value

A closer examination of the specific indirect effects revealed that the effect of the intervention on ART adherence was mediated through barriers to medical care, β = −0.036 (−0.041, −0.032), p < 0.001, and perceived stigma, β = −0.011 (−0.016, −0.007), p < 0.001. Specifically, the intervention was efficacious in reducing the barriers to medical care, β = −0.248 (−0.270, −0.226), p < 0.001, and perceived stigma, β = −0.114 (−0.136, −0.092), p < 0.001. However, contrary to hypothesis H2, participants reporting more barriers to medical care and experiencing more perceived stigma also reported higher levels of ART adherence, β = 0.146 (0.130, 0.163), p < 0.001 and β = 0.099 (0.084, 0.115), p < 0.001, respectively.

Moderating Effect of Depression

As shown in Fig. 1 and Table 2, elevated depressive symptoms were associated with increased perceived stigma, β = 0.196 (0.187, 0.204), p < 0.001; more barriers to medical care, β = 0.097 (0.008, 0.106), p < 0.001; poorer ART adherence, β = −0.032 (−0.042, −0.022), p < 0.001; and reduced healthcare transition readiness, β = −0.075 (−0.080, −0.069), p < 0.001.

As hypothesized in H3, depressive symptoms moderated the effect of the intervention on ART adherence, β = −0.114 (−0.123, −0.104), p < 0.001. See Fig. 2.

Fig. 2.

Fig. 2

Open in a new tab

Loop diagram for the moderating role of depression on the relationship between the intervention and antiretroviral therapy adherence. The x-axis represents the different levels of depression, while the y-axis shows how the effect of the intervention on ART adherence changes with the different levels of depression. The negative slope of the red line (the main effect) suggests that as depression increases, the positive effect of the economic empowerment intervention on adherence weakens. In otherwards, at low levels of depression (left side of the graph), the intervention has a stronger positive effect on adherence. As depression increases (right side of the graph), the effect of the intervention declines. Depression begins eroding the benefits of the EE intervention on ART adherence at a centered depression score of −0.6 (raw score: 29). The blue lines represent the confidence intervals of the main moderation effect

Additionally, depression also moderated the effect of the intervention on the three mediators, namely, barriers to medical care, β = 0.452 (0.444, 0.460), p < 0.001; healthcare transition readiness, β = −0.147 (−0.154, −0.139), p < 0.001; and perceived stigma, β = 0.406 (0.399, 0.414), p < 0.001. As shown in Supplementary Table 1, the measurement and structural models achieved good fit (Hu & Bentler, 1999).

Discussion

In this study, we explored whether participating in an EE intervention resulted in better ART adherence among ALHIV in Uganda and the pathways through which this effect was achieved. We found a significant effect of the EE intervention in improving ART adherence, which was sustained at 7 years follow-up. This effect was mediated through reduced barriers to medical care and perceived HIV stigma and moderated by depression. To our knowledge, this is the first study to assess barriers to medical care, HIV stigma, and transition readiness as mediators between EE interventions and ART adherence. Furthermore, no study has examined depression as a moderator of the pathways between the EE intervention and ART adherence among ALHIV in SSA.

Our results align with those from previous studies that found a positive effect of the EE intervention on ART adherence (Bermudez et al., 2018; Kizito et al., 2024; Ssewamala et al., 2020) and support Asset theory—which proposes that the accumulation of assets has non-economic benefits such as improved psychological, behavioral, and health outcomes (Sherraden, 1990). To illustrate, by providing the ALHIV and their families with financial resources, they were able to meet the indirect costs associated with medical care such as transportation costs for clinic appointments, thus overcoming the barriers to medical care—resulting in sustained optimal ART adherence.

This study adds to prior research, which examined factors such as mental health as potential mediators for the effect of the EE on ART adherence (Cavazos-Rehg et al., 2021; Kizito et al., 2023) by providing robust empirical evidence of the mediating effect of barriers to medical care and perceived HIV stigma as specific pathways through which the economic interventions enhance ART adherence among ALHIV in low-resourced settings. A closer examination of the indirect pathways revealed that although the intervention was efficacious in reducing both barriers to medical care and perceived HIV stigma, the participants who reported more barriers to care (and perceived HIV stigma) also reported better ART adherence, and vice versa. This finding is contrary to previous studies, which found that barriers to medical care were associated with reduced ART adherence (Adams et al., 2022; Cluver et al., 2021; MacCarthy et al., 2016). A possible explanation for increased ART adherence amidst elevated barriers to care and HIV stigma could be the alternative coping strategies that these adolescents develop to overcome these challenges. Indeed, Adams’s qualitative study among ALHIV in Kenya found that resilience was a significant driver behind maintaining good ART adherence (Adams et al., 2022). Other studies have reported similar findings (Harrison & Li, 2018; Woollett et al., 2016). Also, there could be other factors, not assessed in the current model, that could be responsible for the negative mediation through interacting with the barriers to medical care. For example, ALHIV facing more barriers could also have strong family and social support systems, which help them navigate these barriers while maintaining or even enhancing their adherence. Indeed, in our recent study, we demonstrated that family support had a direct and indirect effect—mediated through adolescent-caregiver communication—on ART adherence (Nabunya et al., 2023).

On examining the moderating effect of depression, our findings corroborate previous research demonstrating a negative impact of depression on ART adherence (Haas et al., 2023; Uthman et al., 2014), especially for ALHIV with depression scores ≥ 29. These adolescents may require targeted mental health support to preserve intervention benefits. Our results also highlight depression as a critical factor associated with reduced access to health care. Specifically, participants reporting more depressive symptoms reported reduced healthcare transition readiness. Additionally, they faced increased barriers to medical care and reported more perceived stigma. These barriers could stem from factors associated with depression, such as decreased energy and low mood (Haas et al., 2023). Furthermore, the perception of stigma surrounding mental health might have discouraged participants from seeking care. Overall, this study advances our understanding of the effect of depression on the HIV care continuum by revealing how depression hinders access to healthcare and exacerbates HIV stigma and barriers to care, thus influencing treatment outcomes in ALHIV. Our study is among the first to examine depression as a moderator of the pathways through which EE interventions affect ART adherence. These results highlight the importance of integrating mental health interventions into existing economic empowerment programs aimed at improving treatment outcomes for ALHIV. Given that the study was conducted in the Greater Maska region in Southern Uganda, a rural area with a higher HIV prevalence than the national average, our findings should be generalized to other settings with caution. Future research should assess the applicability of these findings in diverse geographic and socioeconomic contexts to enhance external validity.

A particular strength of our study is that it employed robust methodological approaches, including a cluster-randomized controlled study design and structural equation modeling (SEM). The SEM enables the simultaneous investigation of multiple paths within a single model. In addition, by conducting a longitudinal mediation analysis where the participants were followed up for 7 years, the study accounted for temporal effects to provide insights into inference and directionality. Moreover, the study recruited a large sample size of 702 ALHIV. However, some limitations should be noted when interpreting our results. First, the results should be generalized to ALHIV in other settings cautiously. The data used for this study was from ALHIV in a rural area—Greater Masaka region—with an HIV prevalence of 11.7%, which is twice as high as the national prevalence of 5.4% in Uganda (UAC, 2021). Also, we relied on self-report in collecting some data, for instance, barriers to care and perceived stigma. Self-reported data may be subject to respondent bias, where participants may provide responses that they believe are expected or socially desirable. However, this bias was minimized by relying on both self-reports and pill counts in generating the latent variable for ART adherence. Moreover, five items were used in measuring self-reported adherence—including three items from Wilson’s adherence scale (Wilson et al., 2016), which have been validated and found to be more reliable in measuring ART adherence.

Conclusions

Our findings indicate that interventions incorporating financial support have the potential to directly and indirectly influence ART adherence; however, depression undermines these effects. Therefore, future interventions should consider incorporating components that address mental health challenges. We also recommend further qualitative research to gain a deeper understanding as to why ALHIV, facing increased barriers to care and stigma, exhibit higher ART adherence.

Supplementary Material

Consort Flow Diagram
Supplementary Figures and Tables

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s11121-025-01821-y.

Acknowledgements

The Suubi+Adherence study was supported by funding from the Eunice Kennedy Shriver National Institutes of Child Health and Human Development (NICHD), grant number 1R01HD074949–01(PI: Fred M. Ssewamala). The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official views of NICHD. The funders had no role in any of the stages of preparing this manuscript. We express our gratitude to the study participants and their families for their involvement in our research. Our appreciation goes to our collaborative partners including the study clinics, Reach the Youth-Uganda (RYT), and the Masaka diocese. Additionally, we are indebted to the ICHAD team in Uganda for their unwavering support and assistance in the field, which was instrumental in the successful execution of the study. Finally, we thank Prof. Ross Brownson and Prof. Kim Johnson for their input.

Funding

This research was funded by the Eunice Kennedy National Institute of Child and Human Development (NICHD) under grant number 1R01HD074949–01 (Ssewamala, F.M, PI). The content is solely the responsibility of the authors. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Abbreviations

AIDS

Acquired immunodeficiency syndrome

ALHIV

Adolescents living with HIV

ART

Antiretroviral therapy

BSOC

Bolstered standard-of-care

CES-D

Center for Epidemiologic Studies Depression Scale

CFA

Confirmatory factor analysis

CFI

Comparative fit index

CI

Confidence interval

EE

Economic empowerment

GCP

Good clinical practice

HIV

Human immunodeficiency virus

RMSEA

Root mean square error of approximation

SSA

Sub-Saharan Africa

SEM

Structural equation modeling

SRMR

Standardized root mean square residual

TRAQ

Transition Readiness Assessment Questionnaire

UNCST

Uganda National Council for Science and Technology

YDA

Youth Development Account

Footnotes

Conflict of Interest The authors declare no competing interests.

Declarations

Ethical Approval Makerere University School of Public Health Research and Ethics Committee (protocol # 210), Uganda National Council for Science and Technology (UNCST, SS, 2969), the Columbia University Review Board (AAAK3852; 2012–2017), and the Institutional Review Board at Washington University in St. Louis (IRB # 201704066; 2017-current) approved the study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Informed Consent Before enrollment, all participants provided age-appropriate informed assent, which was duly documented. In addition, written informed consent was obtained from the caregivers of the participants.

Clinical Trial The Suubi + Adherence study was registered with ClinicalTrials.gov (Registration number NCT01790373) on February 13, 2013.

Data Availability

The deidentified datasets included in the analysis for this study are available from the principal investigator (Prof. Fred Ssewamala) on reasonable request.

References

  1. Adams C, Kiruki M, Karuga R, Otiso L, Graham SM, & Beima-Sofie KM (2022). “Your status cannot hinder you”: The importance of resilience among adolescents engaged in HIV care in Kenya. BMC Public Health, 22(1), 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Adejumo OA, Malee KM, Ryscavage P, Hunter SJ, & Taiwo BO (2015). Contemporary issues on the epidemiology and antiretroviral adherence of HIV-infected adolescents in sub-Saharan Africa: A narrative review. Journal of the International AIDS Society, 18, 20049. 10.7448/IAS.18.1.20049 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ahmed CV, Jolly P, Padilla L, Malinga M, Harris C, & Mthethwa N,…Preko P (2017). A qualitative analysis of the barriers to antiretroviral therapy initiation among children 2 to 18 months of age in Swaziland. African Journal of AIDS Research, 16(4), 321–328. 10.2989/16085906.2017.1380677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ashaba S, Cooper-Vince CE, Vořechovská D, Rukundo GZ, Maling S, Akena D, & Tsai AC (2019). Community beliefs, HIV stigma, and depression among adolescents living with HIV in rural Uganda. African Journal of AIDS Research, 18(3), 169–180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Becker N, Cordeiro LS, Poudel KC, Sibiya TE, Sayer AG, & Sibeko LN (2020). Individual, household, and community level barriers to ART adherence among women in rural Eswatini. PLoS ONE, 15(4), Article e0231952. 10.1371/journal.pone.0231952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bermudez LG, Ssewamala FM, Neilands TB, Lu L, Jennings L, & Nakigozi G,…Mukasa M (2018). Does economic strengthening improve viral suppression among adolescents living with HIV? Results from a cluster randomized trial in Uganda. AIDS and Behavior, 22(11), 3763–3772. 10.1007/s10461-018-2173-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Byrd KK, Hou JG, Hazen R, Kirkham H, Suzuki S, Clay PG,…Delpino A (2019). Antiretroviral adherence level necessary for HIV viral suppression using real-world data. Journal of acquired immune deficiency syndromes (1999), 82(3), 245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cavazos-Rehg P, Byansi W, Doroshenko C, Neilands TB, Anako N, Bahar O, & S.,…Ssewamala FM (2021). Evaluating potential mediators for the impact of a family-based economic intervention (Suubi+ Adherence) on the mental health of adolescents living with HIV in Uganda. Social Science & Medicine, 280, Article 113946. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cluver L, Shenderovich Y, Toska E, Rudgard WE, Zhou S, & Orkin M,…Armstrong A (2021). Clinic and care: Associations with adolescent antiretroviral therapy adherence in a prospective cohort in South Africa. AIDS (London, England), 35(8), 1263. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Curley J, Ssewamala FM, Nabunya P, Ilic V, & Keun HC (2016). Child development accounts (CDAs): An asset-building strategy to empower girls in Uganda. International Social Work, 59(1), 18–31. 10.1177/0020872813508569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Damulak PP, Ismail S, Abdul Manaf R, Mohd Said S, & Agbaji O (2021). Interventions to improve adherence to antiretroviral therapy (ART) in Sub-saharan Africa: An updated systematic review. International Journal of Environmental Research and Public Health, 18(5), 2477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dow DE, Turner EL, Shayo AM, Mmbaga B, Cunningham CK, & O’Donnell K (2016). Evaluating mental health difficulties and associated outcomes among HIV-positive adolescents in Tanzania. AIDS Care, 28(7), 825–833. 10.1080/09540121.2016.1139043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Enane LA, Apondi E, Toromo J, Bosma C, Ngeresa A, Nyandiko W, & Vreeman RC (2020). “A problem shared is half solved”–A qualitative assessment of barriers and facilitators to adolescent retention in HIV care in western Kenya. AIDS Care, 32(1), 104–112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gunzler D, Chen T, Wu P, & Zhang H (2013). Introduction to mediation analysis with structural equation modeling. Shanghai Archives of Psychiatry, 25(6), 390–394. 10.3969/j.issn.1002-0829.2013.06.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Haas AD, Lienhard R, Didden C, Cornell M, Folb N, Boshomane T, & M.,…Wettstein AE (2023). Mental health, ART adherence, and viral suppression among adolescents and adults living with HIV in South Africa: A cohort study. AIDS and Behavior, 27(6), 1849–1861. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Haberer JE, Kiwanuka J, Nansera D, Ragland K, Mellins C, & Bangsberg DR (2012). Multiple measures reveal antiretroviral adherence successes and challenges in HIV-infected Ugandan children. PLoS ONE, 7(5), Article e36737. 10.1371/journal.pone.0036737 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hair JF, Hult GTM, Ringle CM, Sarstedt M, Danks NP, spsampsps Ray S (2021). Mediation analysis. In Partial least squares structural equation modeling (PLS-SEM) using R: A workbook (pp. 139–153). Springer International Publishing. 10.1007/978-3-030-80519-7_7 [DOI] [Google Scholar]
  18. Harrison S, & Li X (2018). Toward an enhanced understanding of psychological resilience for HIV youth populations. AIDS Care, 30(sup4), 1–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Health UMO (2020). Consolidated guidelines for the prevention and treatment of HIV and AIDS in Uganda. https://differentiatedservicedelivery.org/Portals/0/adam/Content/HvpzRP5yUUSdpCe2m0KMdQ/File/Uganda_Consolidated%20HIV%20and%20AIDS%20Guidelines%202020%20June%2030th.pdf. Accessed 19 May 2024.
  20. Hu L, & t., & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. [Google Scholar]
  21. Kalichman SC, Catz S, & Ramachandran B (1999). Barriers to HIV/AIDS treatment and treatment adherence among African-American adults with disadvantaged education. Journal of the National Medical Association, 91(8), 439–446. [PMC free article] [PubMed] [Google Scholar]
  22. Kalichman SC, & Grebler T (2010). Stress and poverty predictors of treatment adherence among people with low-literacy living with HIV/AIDS. Psychosomatic Medicine, 72(8), 810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kamau SG, Akatusasira R, Namatovu A, Kibet E, Ssekitto JM, Mamun MA, & Kaggwa MM (2024). The level of antiretroviral therapy (ART) adherence among orphan children and adolescents living with HIV/AIDS: A systematic review and meta-analysis. PLoS ONE, 19(2), Article e0295227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Katz IT, Ryu AE, Onuegbu AG, Psaros C, Weiser SD, Bangsberg DR, & Tsai AC (2013). Impact of HIV-related stigma on treatment adherence: Systematic review and meta-synthesis. Journal of the International AIDS Society, 16, 18640. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kinyanda E, Salisbury TT, Muyingo SK, Ssembajjwe W, Levin J, & Nakasujja N,…Araya R (2020). Major depressive disorder among HIV infected youth in Uganda: Incidence, persistence and their predictors. AIDS and Behavior, 24, 2588–2596. [DOI] [PubMed] [Google Scholar]
  26. Kizito S, Nabayinda J, Neilands TB, Kiyingi J, Namuwonge F, & Damulira C,…Ssewamala FM (2023). A structural equation model of the impact of a family-based economic intervention on antiretroviral therapy adherence among adolescents living with HIV in Uganda. Journal of Adolescent Health, 72(5), S41–S50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kizito S, Namuwonge F, Nabayinda J, Nalwanga D, Najjuuko C, Nabunya P,…Ssewamala FM (2024). A cluster-randomized controlled trial of an economic strengthening intervention to enhance antiretroviral therapy adherence among adolescents living with HIV. AIDS and Behavior, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. MacCarthy S, Hoffmann M, Nunn A, Silva LAV, & d., & Dourado I (2016). Barriers to HIV testing, linkage to care, and treatment adherence: A cross-sectional study from a large urban center of Brazil. Revista Panamericana De Salud Pública, 40(6), 418–426. [PMC free article] [PubMed] [Google Scholar]
  29. Mbalinda SN, Bakeera-Kitaka S, Lusota DA, Musoke P, Nyashanu M, & Kaye DK (2021). Transition to adult care: Exploring factors associated with transition readiness among adolescents and young people in adolescent ART clinics in Uganda. PLoS ONE, 16(4), Article e0249971. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Nabunya P, Samuel K, & Ssewamala FM (2023). The effect of family support on self-reported adherence to ART among adolescents perinatally infected with HIV in Uganda: A mediation analysis. Journal of Adolescence. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Poku O, Attoh-Okine N-D, Corbeil T, Chen Y, Kluisza L, & Ahmed A,…Robbins RN (2022). Assessing the validity of the social impact scale among a longitudinal cohort of adolescents and young adults living with perinatally-acquired HIV. JAIDS Journal of Acquired Immune Deficiency Syndromes, 10, 1097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sherraden M (1990). Stakeholding: Notes on a theory of welfare based on assets. Social Service Review, 64(4), 580–601. 10.1086/603797. [DOI] [Google Scholar]
  33. Sherraden M, & Gilbert N (2016). Assets and the poor: New American welfare policy. Routledge. [Google Scholar]
  34. Simoni JM, Kurth AE, Pearson CR, Pantalone DW, Merrill JO, & Frick PA (2006). Self-report measures of antiretroviral therapy adherence: A review with recommendations for HIV research and clinical management. AIDS and Behavior, 10, 227–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Slogrove AL, Schomaker M, Davies MA, Williams P, Balkan S, & Ben-Farhat J,…Collaboration, C. I. f. P. H. E. a. R. C. G. C. (2018). The epidemiology of adolescents living with perinatally acquired HIV: A cross-region global cohort analysis. PLoS Medicine, 15(3), Article e1002514. 10.1371/journal.pmed.1002514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ssewamala FM, Byansi W, Bahar OS, Nabunya P, Neilands TB, & Mellins C,…Nakigozi G (2019). Suubi+Adherence study protocol: A family economic empowerment intervention addressing HIV treatment adherence for perinatally infected adolescents. Contemp Clin Trials Commun, 16, Article 100463. 10.1016/j.conctc.2019.100463 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ssewamala FM, Dvalishvili D, Mellins CA, Geng EH, Makumbi F, Neilands T, & B.,…Namuwonge F (2020). The long-term effects of a family based economic empowerment intervention (Suubi+Adherence) on suppression of HIV viral loads among adolescents living with HIV in southern Uganda: Findings from 5-year cluster randomized trial. PLoS ONE, 15(2), Article e0228370. 10.1371/journal.pone.0228370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Twekambe E, Epuitai J, Kawala EK, Namanya V, Ayikobua ET, Obakiro SB,…Kiyimba K (2023). A qualitative exploration of the psychosocial factors affecting antiretroviral therapy adherence among HIV infected young adults in Eastern Uganda. medRxiv, 2023.2002. 2004.23285423. [Google Scholar]
  39. UAC UAC (2021). 2021 factsheet: Facts on HIV and AIDS in Uganda 2021. https://uac.go.ug/media/attachments/2021/09/13/final-2021-hiv-aids-factsheet.pdf. Accessed 19 May 2024.
  40. Uthman OA, Magidson JF, Safren SA, & Nachega JB (2014). Depression and adherence to antiretroviral therapy in low-, middle-and high-income countries: A systematic review and metaanalysis. Current Hiv/aids Reports, 11, 291–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Vilagut G, Forero CG, Barbaglia G, & Alonso J (2016). Screening for depression in the general population with the Center for Epidemiologic Studies Depression (CES-D): A systematic review with meta-analysis. PLoS ONE, 11(5), Article e0155431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Vreeman RC, Ayaya SO, Musick BS, Yiannoutsos CT, Cohen CR, & Nash D,…Wiehe SE (2018). Adherence to antiretroviral therapy in a clinical cohort of HIV-infected children in East Africa. PLoS ONE, 13(2), Article e0191848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wang J, & Wang X (2019). Structural equation modeling: Applications using Mplus. John Wiley & Sons. [Google Scholar]
  44. Wilson IB, Lee Y, Michaud J, Fowler FJ, & Rogers WH (2016). Validation of a new three-item self-report measure for medication adherence. AIDS and Behavior, 20(11), 2700–2708. 10.1007/s10461-016-1406-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wood DL, Sawicki GS, Miller MD, Smotherman C, Lukens-Bull K, Livingood W, & C.,…Kraemer DF (2014). The Transition Readiness Assessment Questionnaire (TRAQ): Its factor structure, reliability, and validity. Academic Pediatrics, 14(4), 415–422. [DOI] [PubMed] [Google Scholar]
  46. Woollett N, Cluver L, Hatcher AM, & Brahmbhatt H (2016). “To be HIV positive is not the end of the world”: Resilience among perinatally infected HIV positive adolescents in Johannesburg. Children and Youth Services Review, 70, 269–275. [Google Scholar]
  47. Wu P, Johnson BA, Nachega JB, Wu B, Ordonez CE, Hare A, & Q.,…Marconi VC (2014). The combination of pill count and self-reported adherence is a strong predictor of first-line ART failure for adults in South Africa. Current HIV Research, 12(5), 366–375. 10.2174/1570162×1205141121102501. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Consort Flow Diagram
NIHMS2137952-supplement-Consort_Flow_Diagram.pdf (157.4KB, pdf)
Supplementary Figures and Tables
NIHMS2137952-supplement-Supplementary_Figures_and_Tables.pdf (124.9KB, pdf)

Data Availability Statement

The deidentified datasets included in the analysis for this study are available from the principal investigator (Prof. Fred Ssewamala) on reasonable request.

RESOURCES