Affordable Access

Longitudinal determinants of anal intercourse among women with, and without HIV in the United States

  • Owen, BN
  • Baggaley, RF
  • Maheu-Giroux, M
  • Elmes, J
  • Adimora, AA
  • Ramirez, C
  • Edmonds, A
  • Sosanya, K
  • Taylor, TN
  • Plankey, M
  • Cederbaum, JA
  • Seidman, D
  • Weber, KM
  • Golub, ET
  • Wells, J
  • Bolivar, H
  • Konkle-Parker, D
  • Pregartner, G
  • Boily, M-C
Publication Date
Jun 23, 2022
UPCommons. Portal del coneixement obert de la UPC
External links


Background Anal intercourse (AI) is not uncommon among U.S. women and, when condomless, confers a far greater likelihood of HIV transmission than condomless vaginal intercourse. We aim to identify determinants preceding AI, among women with, and women without HIV. Methods 3708 women living with (73%), and without HIV (27%) participating in the Women’s Interagency HIV Study provided sexual behavior and other data at 6-monthly visits over a median of 9 years (1994–2014). We used generalized estimating equation models to examine sociodemographic, structural and behavioral determinants reported in the visit preceding (1) AI, and (2) condomless AI. Results AI was reported at least once over follow-up by 31% of women without, and 21% with HIV. AI was commonly condomless; reported at 76% and 51% of visits among women living without HIV, and with HIV, respectively. Women reporting AI were more likely to be younger (continuous variable, adjusted odds ratio (aOR) = 0.97, 95% confidence interval (CI):0.96–0.98), Hispanic (aOR = 1.88, CI:1.47–2.41) or White (aOR = 1.62, CI:1.15–2.30) compared to Black, and have at least high school education (aOR = 1.33, CI:1.08–1.65). AI was more likely following the reporting of either (aOR = 1.35, CI:1.10–1.62), or both (aOR = 1.77, CI:1.13–2.82) physical and sexual violence, excessive drinking (aOR = 1.27, CI:1.05–1.66) or any drug use (aOR = 1.34, CI:1.09–1.66), multiple male partners (aOR = 2.64, CI:2.23–3.11), exchange sex (aOR = 3.45, CI:2.53–4.71), one or more female sex partners (aOR = 1.32, CI:1.01–1.75), condomless vaginal intercourse (aOR = 1.80, CI:1.53–2.09), and high depressive symptoms (aOR = 1.23, CI:1.08–1.39). Conclusion AI disproportionally follows periods of violence victimization, substance use, multiple sex partners and depression. Better prevention messaging and biomedical interventions that reduce acquisition or transmission risk are needed, but when AI occurs in the context of violence against women, as our findings indicate, focusing on gender-based violence reduction and immediate treatment to reduce HIV transmission risk is important.

Report this publication


Seen <100 times