CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

(p=0.061). Those with an unlinked transmission reported more lifetime sexual partners (p=0.008) and had partners with a higher CD4 count (p=0.027) than those with a linked transmission. Conclusions: Nearly a third of the HIV-1 incident infections in this treatment-naïve serodiscordant couple cohort were unlinked transmissions with a high frequency occurring in males consistent with other studies in sub-Saharan Africa. Interventions to prevent HIV transmission in discordant couples should address the risk of infection from an outside sexual partner. This may be particularly important when the uninfected partner is male. 1036 Rising School Enrollment & Declining HIV Risk, 15-19y, Rakai, Uganda, 1994-2013 John Santelli 1 ; Sanyukta Mathur 1 ; XiaoYu Song 2 ;Tzu-Jung Huang 2 ;YingWei 2 ;Tom Lutalo 3 ; Fred Nalugoda 3 ; Ronald H. Gray 4 ; David Serwadda 3 1 New York–Presbyterian University Hospital of Columbia and Cornell, New York City, NY, US; 2 Columbia University–Mailman School of Public Health, New York, NY, US; 3 Rakai Health Sciences Program, Entebbe, Uganda; 4 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, US Background: Economic development, family stability, and social policies influence the ability of adolescents to attend school. Likewise, rising school enrollment may shape social developmental transitions and risk for HIV among youth. Methods: We examined longitudinal quantitative data from the Rakai Community Cohort Study (RCCS) for adolescents (n=11,829 person-rounds for women and 9906 person- rounds for men) from 1994 to 2013 in Rakai, Uganda, and ethnographic data from youth (15-24 years) collected in 2010-11. Longitudinal data were analyzed using logistic and linear regression with robust estimation to identify antecedents and consequences of school attendance. Ethnographic data explored social developmental transitions among HIV-infected and HIV-uninfected youth. Results: School enrollment and household socioeconomic status (SES) rose steadily from 1994 to 2013; orphanhood declined among adolescents after 2004 when ART became available. For young women, school enrollment rose from 25% to 57% and for young men from 34% to 67%. Significant antecedents of school enrollment for adolescent women and men included younger age and (after adjusting for age) higher SES, not being an orphan, being unmarried, and living in a family with fewer children. In qualitative interviews, youth reported lack of money, death of parents, and pregnancy as primary reasons for school drop out. Compared to adolescent women who were not enrolled, school enrollment (adjusting for age) was associated with lower HIV prevalence (6.5% vs. 1.8%) and lower rates of certain HIV risk behaviors including not initiating sexual intercourse (37% vs. 88%), alcohol use in the past 30 days (14% vs. 25%), and consistent condom use with all partners (70% vs. 17%) but not 2+ sexual partners or sexual concurrency. School enrollment has a similar protective association for men, but in addition, young men enrolled is school were less likely to have 2+ sexual partners in the past year (28% vs. 42%) and less likely to be involved currently in sexual concurrency (6% vs. 11%). Conclusions: Social trends such as rising SES and declining orphanhood contributed to rising school enrollment in Rakai. Rising school enrollment was associated with declines in certain HIV behaviors, although these varied by gender. 1037 Alcohol Use and HIV Risk Factors: Results From the 2011 Uganda AIDS Indicator Survey George Aluzimbi CDC Center for Global Health, Division of Global AIDS/HIV, Kampala, Uganda Background: Uganda has one of the highest per capita alcohol consumption worldwide. Alcohol use has been shown to increase the likelihood of high risk sexual behavior and is associated with HIV infection. We evaluated the association between alcohol use during sex and HIV prevalence in a nationally representative household survey in Uganda. Methods: The Uganda AIDS Indicator Survey was a population-based household survey designed to produce national and regional estimates of HIV prevalence and risk behavior, conducted February to September 2011. HIV and syphilis serological tests were conducted on all consenting participants. Respondents who were sexually active in the past 12 months were asked if alcohol was consumed by both, either or none of them during sex with most recent sexual partner. We analyzed data from respondents aged 15-59 who had had sex in the past 12 months and had reported consumption of alcohol during sexual intercourse. Results: A total of 21741 persons were enrolled in the Uganda AIS; 15917 (85.1%) reported being sexually active in the 12 months prior to the survey; and alcohol consumption during most recent sexual activity was reported by 24.0% of men (1719/7133) and 24.1% of women (2123/8775). Condom use at last sexual intercourse was higher among non- drinkers compared to alcohol drinkers (men: 2.6% vs. women: 5.8%). Overall, HIV prevalence was higher in those who consumed alcohol during sex compared to those who did not (men: 10.4% vs. 5.9%; women: 9.5% vs. 7.7%). In adjusted analysis, alcohol use during sex was associated with HIV infection (men: OR= 1.7 [95% CI 1.29-2.15], women: OR=1.3 [CI 1.03-1.59]), syphilis infection (men: OR= 1.7 [95% CI 1.14-2.62], women: OR=1.8 [CI 1.24-2.53]), unprotected sex (men OR=1.7 [CI 1.49-1.98], women OR=1.9 [CI 95% 1.54-2.48]), and paying for sex among men (OR=1.8 [CI 1.42-2.18]). Men who were circumcised (OR=0.5 [CI 0.32-0.78]) were less likely to drink alcohol prior to sexual intercourse. Conclusions: About 25% of Ugandans consumed alcohol during sex. Alcohol consumption was associated with lower condom use, and higher HIV and syphilis infection. Alcohol interventions should be closely integrated into HIV prevention policy and programming. 1038 Population Attributable Fraction of HIV Due to Alcohol in Fishing Communities, Uganda Noah Kiwanuka 5 ; Ismail Ssekandi 2 ; Ali Ssetaala 2 ; Annet Nalutaaya 2 ; Juliet Mpendo 2 ; Paul K. Kitandwe 2 ; Jan D. Bont 3 ; Pontiano Kaleebu 4 ; Nelson K. Sewankambo 5 1 Makerere University College of Health Sciences, Kampala, Uganda; 2 UVRI-IAVI HIV Vaccine Program, Entebbe, Uganda; 3 International AIDS Vaccine Initiative, New York, NY, US; 4 MRC/UVRI Uganda Research Unit on AIDS, Entebbe, Uganda; 5 Makerere University College of Health Sciences, Kampala, Uganda Background: Studies have shown that alcohol drinking is associated with HIV risk behaviours, and HIV prevalent and incident infections. We assessed the population attributable fraction (PAF) of incident HIV infections due to alcohol consumption in fishing communities (FC), along Lake Victoria, Uganda, to determine the potential impact of alcohol interventions on HIV risk. Methods: In a community-based cohort study conducted among participants aged 18-49 years, data on risk behaviours including alcohol consumption and its frequency, were collected at baseline and 12 months follow-up visits. Venous blood samples were collected for HIV serological testing. HIV incidence rates and adjusted incident rate ratios (Adj. IRR) were estimated by Poisson regression models. Crude and adjusted PAFs of incident HIV infections associated with alcohol consumption were calculated as [1-Pr(HIV incident infection given no alcohol consumption)/ Pr(HIV incident infection)] x100, using the Greenland and Drescher method for cohort studies and Stata punaf command. Results: Overall, 48 incident HIV infections occurred giving a cumulative incidence of 3.72% (95% CI, 2.74 - 4.94) and an incidence rate of 3.39/100 pyar (95% CI, 2.55 - 4.49). Of these, 10 (20.8%) occurred among none alcohol drinkers, 12 (25.0%) in occasional drinkers and 26 (54.2%) in regular drinkers (trend p <0.0001). Having 2+ total sexual partners, 2+ new sexual partners, and alcohol drinking before sex were highest among regular alcohol drinkers, followed by occasional drinkers, and least among none drinkers (all trend p values <0.0001). The overall crude and adjusted PAFs were 55.8% (95% CI; 23.3 - 74.6) and 63.8% (95% CI; 23.8 - 82.8) respectively. The lowest adjusted PAF of 52.4% (95% CI; 12.4 - 74.1) was observed among Moslems who drink alcohol while the highest of 70.8% (95% CI; 33.0 - 87.3) occurred among participants who reported 2+ sexual partner in the past 12 months.

Poster Abstracts

608

CROI 2015