CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

between HIV status and AAT in a linear regression model adjusting for age, sex, ethnicity and cumulative smoking. Controlling for the same variables, we assessed whether AAT was associated with current CD4 or CD4 nadir. Finally, we assessed whether HIV was independently associated with AAT controlling for the same variables and hsCRP. Results: AAT was measured in 1011 PLWH and 11962 uninfected controls with mean (SD) age of 52 (11.5) and predominantly males (81.7%). PLWH had a higher median (interquartile range) AAT: 1.4 [95%CI: 1.3-1.6] vs. 1.3 [1.2-1-4], (p<0.0001). A total of 4.2% [3.1-5.6] PLWH and 6.6% [6.2-7.1] uninfected controls had low levels of AAT (p<0.01). Low AAT were not more common in PLWH with airflow obstruction assessed by LLN (p=0.57) or the fixed criterion (p=0.80). The effect of AAT on FEV1-predicted was similar in PLWH and uninfected controls (p-interaction=0.66). Also, AAT levels (per 1 g/L) did not affect current CD4 (adjusted β=12.2 [-86.2-61.5], p=0.74) or CD4 nadir (adjusted β= 6.7 [-38.9-52.2], p=0.77). HIV status was independently associated with higher AAT levels adjusted β (0.1 [0.01-1.36, p<0.001]). Conclusion: AAT did not explain the excess pulmonary morbidity previously observed in PLWH. Also, AAT was not associated with immunological status. However, HIV was independently associated with higher levels of AAT and may be an interesting biomarker for other non-AIDS events.

linear regression was used to model the log 10 of HIV viral load and CESD-items (individually) among participants on ARVs for 6 months or more. Covariates included demographics, purchasing power, cognitive impairment and ARV adherence. Results: The point prevalence of depression is 18% to 25% among East African HIV+ AFRICOS study participants-all enrolled in HIV care. Depression is associated with decreased ARV adherence (OR 0.39-0.88, p = 0.01) independently of cognitive impairment. Depression is associated with a nearly double the viral load of non-depressed counterparts, independently of ARV adherence and cognitive impairment (proportional increase 1.42-2.34, p= 0.00) for participants on ARVs > 6 months. Seven of the nine depression symptom clusters are significantly associated with viral load independently of ARV adherence and cognition for participants on ARVs > 6 months. Conclusion: l HIV+ East African AFRICOS participants enrolled in HIV care have high prevalence of depression. l Depression is associated with viral load independently of pathways involved with ARV adherence or misclassification of cognitive impairment. l Depression and HIV viral load is associated across a wide spectrum of depressive symptoms, implying that comprehensive depression treatment is necessary to fully address the relationship between HIV viral load and depression. Future research should include scalable, evidence-based depression treatment for PLWH with assessment of impact on HIV viral load. 753 DEPRESSION IS ASSOCIATED WITH MISSED VISITS, EMERGENCY ROOM UTILIZATION, AND DRUG USE Mamta K. Jain 1 , Yordanos Tiruneh 2 , Piper Duarte 3 , Amneris Luque 1 , Ank E. Nijhawan 1 1 University of Texas Southwestern, Dallas, TX, USA, 2 University of Texas at San Antonio, San Antonio, TX, USA, 3 Parkland Health and Hospital Systems, Dallas, TX, USA Background: The US Preventive Service Task Force recommends screening all patients for depression. Depression has been associated with unsuppressed viral load. We collected data obtained from nurse-administered clinical tools in an urban HIV clinic to characterize the relationship of depression, HIV, and health care utilization. Methods: We retrospectively examined patients who had a Patient Health Questionnaire-9 (PHQ-9) administered during an HIV clinic visit to screen for depression symptoms. We also collected drug screen data from the Substance Abuse and Mental Illness Symptoms Screener (SAMISS). Data from 2014 and 2015 included demographics, HIV viral load, CD4 cell count, number of hospitalizations, number of missed visits, number of emergency room visits, and proportion of overall visits attended. Predictors of depression (defined as PHQ-9 ≥10) were examined by univariate and multivariate logistic regression. Results: Of 5127 HIV-infected patients in our cohort, 21% had mild depression, 11%moderate depression, 7%moderately severe, and 6% severe. Overall the cohort was majority male (69%), black (56%) and 46% had viral suppression. Compared to those with CD4>500 cells/μL, those with CD4 between 351-500 cells/μL (Odds Ratio (OR) 0.89, 95% CI: 0.74-1.06) and 200 to 350 cells/μL ( OR 0.95, 95% CI: 0.79-1.15) had no association with PHQ-9 score ≥10. Those with CD4 between 51-200 cells/μL (OR 1.44, 95% CI: 1.20-1.74) and <50 cells/μL ( OR 1.78,95% CI: 1.42-2.26) were significantly more likely to have depression. Those with depression were significantly more likely to be female, have injection drug use or MSM as an HIV risk factor compared to heterosexual, be white compared to black, have a positive score for substance use by SAMISS, have a CD4 <200 cells/μL, have had one or more emergency room visits, and have missed one or more routine clinic visits (see Table). Depression was not associated with viral suppression. Conclusion: Moderate to severe depression is prevalent among HIV clinic patients. Certain subgroups- women, whites and blacks (compared to Hispanics) and those with active substance use--were more likely to report depression. Depression was associated with missed routine clinic follow-up visits, more

Poster Abstracts

752 DEPRESSION IS INDEPENDENTLY ASSOCIATED WITH NEAR DOUBLING OF HIV VIRAL LOAD Susan M. Meffert 1 , Thomas C. Neylan 1 , Charles E. McCulloch 1 , Jonah Jonah Maswai 2 , John Owuoth 3 , Francis Kiweewa 4 , Lucas Maganga 5 , Victor Valcour 1 , Julie Ake 6 1 University of California San Francisco, San Francisco, CA, USA, 2 KEMRI/Walter Reed Proj, Kericho, Kenya, 3 KEMRI/Walter Reed Proj, Kisumu, Kenya, 4 Makerere Univ Walter Reed Proj, Kampala, Uganda, 5 Mbeya Medical Research Programme, Mbeya, Tanzania, United Republic of, 6 US Military HIV Research Program, Bethesda, MD, USA Background: Depression is a common co-morbidity for People Living With HIV (PLWH) and is associated with non-adherence to antiretrovirals (ARVs). Psychosocial interventions often focus on ARV adherence. However, depression may be associated with HIV outcomes (e.g., viral load) through other pathways, as well (Figure). Methods: The African Cohort Study (AFRICOS) is prospective longitudinal cohort study at eleven HIV care sites in Kenya, Tanzania, Uganda, and Nigeria. For cultural consistency, we examined East African sites (Kenya, Uganda and Tanzania) (n=2,335). Using baseline data from HIV+ AFRICOS participants, all of whom are engaged in HIV care, we assessed cross-sectional relationships between depression, ARV adherence, HIV viral load and cognition. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale Revised (diagnostic cut-off > 16). Logistic regression was used to model self-report of complete/incomplete ARV adherence over the past month, and

CROI 2018 281