Prevalence of probable depression and factors associated with mean Hopkins Symptom Checklist (HSCL) depression score among young women at high risk aged 15–24 years in Kampala, Uganda

Abstract

Background In populations at high risk of HIV infection, rates of depression can be elevated with far-reaching effects on overall well-being. There is limited research on depression among young women engaged in high-risk sexual behaviour in low and middle-income settings. We investigated the prevalence, correlates and factors associated with mean HSCL depression score among young women at high risk (aged 15–24 years old) in Kampala, Uganda. Methods We conducted a baseline analysis of a randomized controlled trial. Probable depression was measured using the 15-item Hopkins Symptoms Checklist for depression (HSCL). This checklist has been validated in Ugandan populations, and our reliability test gave a Cronbach alpha coefficient of 0.89. The test was administered to all the participants. Participants whose HSCL mean score was greater than 1.75 were categorized as having probable depression. Socio-demographics and behaviour data were collected and factors associated with mean HSCL depression score were analysed using multiple linear regression. Results Data was available for 600 participants, mean age 20.4 (SD±2.44) years. The prevalence of probable depression was 56% (95% CI, 52%-60%). Probable depression symptoms were most prevalent among those who reported ever-experiencing violence from a sexual partner (64.7%), those aged between 20–24 years (58.2%) and those who reported more than 10 sexual exposures in the month prior to the interview (56.8%). At the adjusted analysis level, condom use during their last sexual intercourse prior to the survey decreased probable depression symptoms by 0.147 units compared to those who never used condoms (β = -0.147, 95% CI -0.266–0.027). Having experienced physical violence by a sexual partners increased mean HSCL depression score by 0.183 units compared to those who have never experienced violence (β = 0.183, 95% CI 0.068–0.300). Participants who reported ever using drugs of addiction had their mean HSCL depression scoreincrease by 0.20 units compared to those who have never used (β = 0.20,95% CI 0.083–0.317). Conclusions Probable depression is high in this population and increased mean HSCL depression score is related to violence. Periodic screening for depression and interventions targeting depression, partner violence and risky sexual behaviours are recommended.