BackgroundPrior research suggests that a high prevalence of depression, with a detrimental impact on treatment outcomes exists among HIV-infected youth. Data on potential risk factors of depression among HIV-infected youth in sub-Saharan Africa are scarce. This cross-sectional study aimed to identify contributory/protective factors associated with depression in Malawian adolescents 12–18 years old living with HIV.MethodsDepression was measured by a validated Chichewa version of the Beck Depression Inventory version-II (BDI-II) and the Children’s Depression Rating Scale-Revised (CDRS-R). Data on variables thought to potentially be contributory/protective were collected and included: socio-demographics, past traumatic events/stressors, behavioural factors/social support, and bio-clinical parameters. Chi-square test or two-sample t-test was used to explore associations between factors and depression. Additional testing via linear/logistic regression, adjusting for age and sex, identified candidate variables (p < 0.1). Final regression models included variables with significant main effects and interactions.ResultsOf the 562 participants enrolled (mean age, 14.5 years [SD 2.0]; 56.1 % female), the prevalence of depression was 18.9 %. In multivariate linear regression, the variables significantly associated with higher BDI-II score were female gender, fewer years of schooling, death in the family/household, failing a school term/class, having a boyfriend/girlfriend, not disclosed or not having shared one’s HIV status with someone else, more severe immunosuppression, and bullied for taking medications. Bullying victimization was reported by 11.6 % of respondents. We found significant interactions: older participants with lower height-for-age z-scores and dissatisfied with their physical appearance had higher BDI-II scores. In multivariate logistic regression, factors significantly associated with depression were: older age, OR 1.23 (95 % CI 1.07-1.42); fewer years of schooling, OR 3.30 (95 % CI 1.54-7.05); and bullied for taking medications, (OR 4.20 (95 % CI 2.29-7.69).ConclusionHaving fewer years of schooling and being bullied for taking medications were most clearly associated with depression. Programmes to support the mental health needs of HIV-infected adolescents that address issues such as disclosure, educational support, and, most notably, bullying may improve treatment outcomes and are recommended.

IntroductionThere is a remarkable dearth of evidence on mental illness in adolescents living with HIV/AIDS, particularly in the African setting. Furthermore, there are few studies in sub-Saharan Africa validating the psychometric properties of diagnostic and screening tools for depression amongst adolescents. The primary aim of this cross-sectional study was to estimate the prevalence of depression amongst a sample of HIV-positive adolescents in Malawi. The secondary aim was to develop culturally adapted Chichewa versions of the Beck Depression Inventory-II (BDI-II) and Children's Depression Inventory-II-Short (CDI-II-S) and conduct a psychometric evaluation of these measures by evaluating their performance against a structured depression assessment using the Children's Rating Scale, Revised (CDRS-R).Study designCross-sectional study.MethodsWe enrolled 562 adolescents, 12–18 years of age from two large public HIV clinics in central and southern Malawi. Participants completed two self-reports, the BDI-II and CDI-II-S, followed by administration of the CDRS-R by trained clinicians. Sensitivity, specificity and positive and negative predictive values for various BDI-II and CDI-II-S cut-off scores were calculated with receiver operating characteristics analysis. The area under the curve (AUC) was also calculated. Internal consistency was measured by standardized Cronbach's alpha coefficient, and correlation between self-reports and CDRS-R by Spearman's correlation.ResultsPrevalence of depression as measured by the CDRS-R was 18.9%. Suicidal ideation was expressed by 7.1% (40) using the BDI-II. The AUC for the BDI-II was 0.82 (95% CI 0.78–0.89) and for the CDI-II-S was 0.75 (95% CI 0.70–0.80). A score of ≥13 in BDI-II achieved sensitivity of >80%, and a score of ≥17 had a specificity of >80%. The Cronbach's alpha was 0.80 (BDI-II) and 0.66 (CDI-II-S). The correlation between the BDI-II and CDRS-R was 0.42 (p<0.001) and between the CDI-II-S and CDRS-R was 0.37 (p<0.001).ConclusionsThis study demonstrates that the BDI-II has sound psychometric properties in an outpatient setting among HIV-positive adolescents in Malawi. The high prevalence of depression amongst HIV-positive Malawian adolescents noted in this study underscores the need for the development of comprehensive services for HIV-positive adolescents.