Share

HIV risk behaviors among those with and without viral load suppression: findings from population surveys from four African countries

Title
HIV risk behaviors among those with and without viral load suppression: findings from population surveys from four African countries
Presenter
Thomas R Carpino
Authors
T.R Carpino * (1), J.E Mantell (2), N.M Philip (1), W.M El-Sadr (1), A. Low (1)
Institutions
(1) ICAP at Columbia University, New York, United States, (2) Columbia University Medical Center, HIV Center for Clinical and Behavioral Studies, Division of Gender, Sexuality and Health, Department of Psychiatry, New York, United States

BACKGROUND: Individual viral load suppression (VLS) is critical to achieving individual health and population-level prevention of HIV transmission. Yet, the population-level association between successfully controlled viremia and sexual risk behaviors is poorly understood. We assessed behavioral factors associated with unsuppressed viral load (UVL) in Population Based HIV Impact Assessments (PHIA) surveys conducted in four sub-Saharan African countries.
METHODS: We analyzed cross-sectional data from nationally representative, general population surveys conducted in Eswatini, Malawi, Tanzania, and Zambia (2015-2017). Among people living with HIV (PLWH) aged 15-59 years, we focused on four higher risk behaviors during the prior year: multiple sexual partners, unprotected non-marital sex, hazardous drinking (using AUDIT-C), and commercial or transactional sex. We evaluated the correlation between UVL (>1,000 copies/ml) and each risk behavior using logistic regression on weighted data, adjusting for age, household wealth quintile, urban residence, and country, stratified by sex. A risk score was developed by summing the total number of reported risk factors. The score was dichotomized as 0 to 1 versus 2 or more risk behaviors.
RESULTS: Among 73,726 participants, 9,062 were PLWH with available viral load data. UVL varied across countries: Tanzania (48.0%, 814/1,708); Zambia (40.8%, 948/2,413), Malawi (32.0%, 682/2150) and Eswatini (27.7%, 735/2,788). UVL was more common in men (47.8%) than women (37.0%, p<0.0001). Among women, UVL was independently associated with having multiple sexual partners (aOR=1.41, 95% CI 1.05-1.91) and unprotected non-marital sex (aOR 1.64, 95% CI 1.18-2.30). Among men, UVL was independently associated only with hazardous drinking (aOR 2.01, 95% CI 1.48-2.72). Women with UVL were more likely to engage in two or more risk behaviors (aOR 1.47, 95% CI 1.11-1.96), whereas men were not (aOR 1.27, 95% CI 0.86-1.86).
CONCLUSIONS: Study findings suggest that UVL among PLWH remains a substantial challenge. PLWH with UVL in contrast to those with viral load suppression (VLS) are more likely to engage in higher risk behaviors associated with HIV transmission, particularly women. Efforts are needed to disseminate information regarding the importance of VLS for individual benefit and regarding U=U.