Association between intravaginal practices and HIV acquisition in women: individual patient data meta-analysis of cohort studies in sub-Saharan Africa

Abstract

Background: It is still not clear whether practices such as internal (intravaginal) cleansing and practices used around sexual intercourse, increase women's risk of acquiring HIV infection. Individual studies lack statistical power and direct comparison between studies is hampered by variation in exposure definitions and statistical methods. Methods: A systematic review identified 14 cohort studies in sub-Saharan Africa that measured HIV as an outcome and collected data on vaginal practices; 11 provided individual patient data, and 9 have been analysed to date. We recoded data using standardised definitions of any vaginal practice, intravaginal cleansing, insertion of solid products, and use of products to dry or tighten the vagina. We used univariable Cox proportional hazards models, stratified by study, to examine the association between the use of practices as reported at study baseline and HIV acquisition. Results: A total of 14,766 women contributed 24,587 woman years, with 798 newly-acquired HIV infections. The prevalence of any vaginal practice across cohorts ranged from 18% to 95% (median 76%); intravaginal cleansing was reported by a median of 70% of women (range 7-94%). In univariable analysis, women reporting any vaginal practice had a 1.27 fold higher rate of HIV acquisition (95%CI 0.91-1.77) than never-users, with moderate between-study heterogeneity (I246%). Compared with never-users, hazard ratios for incident HIV in women with different practices were: water only, 1.07 (95%CI 0.83-1.38); other products for intravaginal cleansing, 1.44 (95%CI 1.16-1.79); and practices other than intravaginal cleansing, 1.88 (95%CI 1.25-2.82) Conclusions: Intravaginal practices are common in sub-Saharan Africa. These may increase women's risk of acquiring HIV, particularly practices other than cleansing with water. Additional analyses will clarify the role of confounding in these associations. Common practices that increase HIV risk may also reduce the effectiveness of vaginal microbicides. Interventions to alter potentially harmful intravaginal practices might contribute to combination HIV prevention.