Darcy White 1, Kate S Wilson, Linnet N Masese, George Wanje, Walter Jaoko, Kishorchandra Mandaliya, Barbra A Richardson, John Kinuthia, Jane M Simoni, R Scott McClelland
Affiliation
- 1From the *Department of Epidemiology, †Department of Global Health, ‡Department of Medicine, University of Washington, Seattle, WA; §University of Washington-Kenya Non-Governmental Organization, Nairobi, Kenya; ¶Department of Medical Microbiology, University of Nairobi, Nairobi, Kenya; ‖Department of Biostatistics, University of Washington, Seattle, WA; **Division of Public Health Sciences, ††Division of Vaccine and Infectious Disease, Fred Hutchinson Cancer Research Center, Seattle, WA; ‡‡Department of Research and Programs, Kenyatta National Hospital, Nairobi, Kenya; §§Department of Psychology, University of Washington, Seattle, WA; ¶¶Institute of Tropical and Infectious Diseases, University of Nairobi, Nairobi, Kenya.
PMID: 27631360 PMCID: PMC5026390
DOI: 10.1097/OLQ.0000000000000502
Abstract
Background: Studies of alcohol use and sexual behavior in African populations have primarily been cross-sectional, used nonvalidated measures of alcohol use, or relied on self-reported sexual risk endpoints. Few have focused on human immunodeficiency virus (HIV)-positive women.
Methods: Longitudinal data were collected from a cohort of HIV-positive Kenyan female sex workers. At enrollment and annual visits, participants were asked about past-year alcohol use using the Alcohol Use Disorders Identification Test (AUDIT). The primary endpoint was detection of prostate-specific antigen (PSA) in vaginal secretions at quarterly examinations. Associations between hazardous/harmful alcohol use (AUDIT score ≥7), PSA detection, and secondary measures of sexual risk were evaluated using generalized estimating equations with a log binomial regression model.
Results: A total of 405 women contributed 2750 vaginal samples over 606 person-years of follow-up. Hazardous/harmful alcohol use was reported at 16.6% of AUDIT assessments and was associated with higher risk of PSA detection (relative risk 1.50; 95% confidence interval, 1.11-2.01) relative to no alcohol use. This association was attenuated and no longer statistically significant, after adjusting for age, work venue, intimate partner violence, depression, and partnership status (adjusted relative risk, 1.13; 95% confidence interval, 0.82-1.56). In exploratory analyses, alcohol use was associated with self-report of unprotected sex and with sexually transmitted infection acquisition.
Conclusions: Although hazardous/harmful alcohol use was not associated with detection of PSA in adjusted analysis, associations with secondary outcomes suggest that alcohol use is at least a marker of sexual risk behavior.
Link
