(SLC-007) Longitudinal Changes in Depressive Symptom Severity with Different Antidepressant Classes among People with HIV

Background: Depression is one of the most common psychiatric conditions in people with HIV (PWH) with negative impacts on HIV viral suppression, engagement, and other clinical outcomes. It is important to continue to document real-world effectiveness for these medications in this complex population.

Objectives: To determine absolute response levels (Patient Health Questionnaire-9 [PHQ-9] scores) and compare effectiveness of different antidepressant classes on depressive symptoms in a cohort of PWH in care.

Methods: We conducted a new user analysis of adult PWH in clinical care at 7 US sites in the CFAR Network of Integrated Clinical Systems cohort who were prescribed an antidepressant drug in 2010-2023. PWH were included if they had ≥1 measure of self-reported depressive symptoms (PHQ-9, range 0-27) both ≤2 years before and ≤2 years after prescription date. We considered four antidepressant classes: selective serotonin reuptake inhibitors (SSRI, n=338), serotonin norepinephrine reuptake inhibitors (SNRI, n=47), tricyclic antidepressants (TCA, n=40), and atypical/other antidepressants (e.g., bupropion, mirtazapine) (n=338) combined due to small sample size of individual agents. We used linear mixed models to evaluate antidepressant effectiveness with random intercepts by patient with the primary exposure being receiving antidepressant therapy with interactions between being on treatment and antidepressant class. Models were adjusted for age, birth sex, race and ethnicity, and substance use.

Results: For the 763 PWH in the study, at initial prescription, mean age was 42 years (SD 11), 15% were female, 57% reported non-white race and ethnicity, and mean PHQ-9 was 11 (SD 6). Substance use was common (methamphetamine 14%, cocaine 10%, opioid 4%, cannabis 40%). Over a median follow up time of 8 months (IQR 4-15), adjusted models estimated significant decreases in PHQ-9 scores for those prescribed SSRIs (-1.1; 95% CI: -1.7, -0.4), TCAs (-2.3; 95% CI: -4.1, -0.5), and atypical/other antidepressants (-1.6; 95% CI: -2.2, -1.1); no significant change in PHQ-9 scores was observed for SNRIs (0.4; 95% CI: -1.7. 2.4). We did not find a statistically significant difference in PHQ-9 scores for PWH prescribed SNRIs, TCAs, and atypical/other antidepressants when compared to SSRIs (Interaction P=0.2 for all comparisons).

Conclusions: Findings suggest common antidepressant classes have modest absolute impact on depressive symptoms in real-world settings. We do not observe statistically significant differences in depressive symptoms between antidepressant classes, although confounding by indication may limit inference.