Global meta-analysis links violence with HIV outcomes in female sex workers

A global systematic review and meta-analysis found that female sex workers reporting recent or lifetime violence had higher odds of prevalent HIV infection and poorer outcomes in some aspects of HIV care. The analysis combined 226 estimates from 94 studies involving 186,613 female sex workers in 39 countries. Overall, the results associate violence with greater HIV burden, lower recent antiretroviral therapy use and lower viral suppression among those reporting lifetime violence, but do not establish a causal effect.

Researchers searched MEDLINE, Embase and PsycINFO for studies published from January 1, 2010, through June 9, 2026, without restricting language or study design. Eligible studies assessed physical, sexual, emotional or psychological, and financial violence, either during the previous year or over a lifetime. The outcomes were prevalent and incident HIV infection, HIV testing, antiretroviral therapy use, treatment adherence and viral suppression. Multi-level random-effects meta-analyses pooled the available estimates, with adjusted estimates preferred when studies reported both adjusted and unadjusted results.

The clearest association was between violence and HIV prevalence. Female sex workers reporting violence in the previous year had 31% higher pooled odds of prevalent HIV infection, with a pooled odds ratio of 1.31 and a 95% confidence interval of 1.16 to 1.48. This comparison drew on 75 estimates, and statistical heterogeneity was 63%. Lifetime violence was associated with 35% higher pooled odds of HIV prevalence, with an odds ratio of 1.35, a 95% confidence interval of 1.23 to 1.48 and 68 contributing estimates; heterogeneity for this result was 39%.

Stratification by type of violence showed the strongest pooled associations for sexual violence. Recent sexual violence was associated with 55% higher odds of HIV prevalence, producing a pooled odds ratio of 1.55 with a 95% confidence interval of 1.24 to 1.94. The estimate included 28 contributing estimates and had heterogeneity of 69%. Lifetime sexual violence was associated with 40% higher odds, with a pooled odds ratio of 1.40, a 95% confidence interval of 1.21 to 1.63 and 28 estimates. Its heterogeneity was 47%, indicating variation among the pooled results.

Violence in the previous year was also associated with reduced odds of antiretroviral therapy use. Seventeen estimates contributed to a pooled odds ratio of 0.78, equivalent to 22% lower odds, with a 95% confidence interval of 0.64 to 0.94. Heterogeneity was 8% for this association. Among women reporting lifetime violence, the pooled odds of viral suppression were lower, with an odds ratio of 0.88 and a 95% confidence interval of 0.79 to 0.98. Six estimates contributed to that result, which had 20% heterogeneity.

The review found no evidence of pooled associations between violence and HIV incidence, HIV testing or antiretroviral therapy adherence. That finding does not demonstrate that no such relationships exist. The studies differed in design, populations, violence definitions and outcome measurement, while experiences of violence can be difficult to measure consistently. The absence of a detected pooled association therefore should be interpreted separately from the statistically significant findings for HIV prevalence, therapy use and viral suppression.

The evidence base was geographically broad but unevenly distributed. Of the 94 included studies, 47 were based in low- and middle-income countries, representing 50% of the total, while 18 were conducted in low-income countries, representing 19%. Most of the included research was cross-sectional, measuring violence and HIV-related outcomes at approximately the same time. This limits the review’s ability to determine whether violence preceded an infection or treatment outcome. Heterogeneity among the pooled estimates and difficulties in measuring violence were additional limitations, and some studies contributed multiple estimates because they reported several exposures or outcomes.

Taken together, the synthesis identifies experiencing violence as a structural determinant associated with HIV outcomes among female sex workers. The higher pooled prevalence observed after both recent and lifetime violence, together with the strongest type-specific associations for sexual violence, shows that the pattern extended across different exposure windows and forms of violence. The findings support treating violence as part of the conditions shaping HIV prevention and care rather than as an issue separate from them, although the observational evidence does not by itself prove that violence caused the measured differences.

Subject of Research: Associations between violence and HIV outcomes among female sex workers worldwide

Article Title: Impact of violence on HIV outcomes among female sex workers: A global systematic review and meta-analysis

Article References: Dawe, J., Mazhar, K. A., Khan, S. A., Njiro, B. J., Bendaud, V., Sabin, K., Ambia, J., Trickey, A., Barrass, L., Asgharzadeh, A., Stone, J., Artenie, A., & Vickerman, P. (2026). Impact of violence on HIV outcomes among female sex workers: A global systematic review and meta-analysis. PLOS Medicine, 23(10), e1004977. https://doi.org/10.1371/journal.pmed.1004977

Image Credits: AI Generated

DOI: 10.1371/journal.pmed.1004977

Keywords: HIV, female sex workers, violence, systematic review, meta-analysis, HIV prevalence, antiretroviral therapy, viral suppression, sexual violence, global health, Impact, outcomes