Introduction Although STIs are associated with HIV-acquisition, because they share a causal pathway – sex – how much this risk is independent of sexual behaviour remains unknown.

Methods We conducted a case-control study of MSM STD clinic patients in Seattle, WA, 2001–2014 to evaluate the role of concurrent and prior rectal STIs in HIV-acquisition. Cases were new HIV diagnoses who tested HIV-negative ≤12 months prior. Controls tested HIV-negative and were matched to cases on year. All included men tested for rectal STI and tested negative for syphilis. We used routinely collected condom-use data to create four sexual behaviour categories: no receptive anal intercourse (RAI) in ≤12 months, consistent condom-use for all RAI, condomless RAI only with HIV-negative partners (CRAIneg), and CRAI with HIV-positive/unknown-status partners (CRAIpos/unk). We used logistic regression to estimate odds ratios (OR) of the association between rectal GC/CT and HIV diagnosis.

Results Among 176 cases and 704 controls, concurrent rectal GC (OR3.5 95% CI 2.3–5.5) and rectal CT (OR3.2 95% CI 2.1–5.1) were associated with HIV diagnosis in univariate analysis. Controlling for age, race, number of sex partners, methamphetamine use year and other rectal STI, both rectal GC (aOR2.4 95% CI 1.4–4.0) and CT (aOR2.6 95% CI 1.5–4.4) continued to be associated with HIV diagnosis. Adding sexual behaviours to the model did not change the association between rectal infection and HIV diagnosis (GC aOR2.3, 95% CI 1.4–3.9; CT aOR 2.6 95% CI 1.5–4.3). CRAIneg (aOR3.5 95% CI 1.2–10.4) and CRAIpos/ukn (aOR4.2 95% CI 1.4–12.5) were independently associated with new HIV diagnosis. Rectal  infection in ≤12 months was strongly associated with new HIV diagnosis (aOR3.4 95% CI 1.5–7.4).

Conclusions Concurrent and prior rectal GC/CT are associated with HIV-acquisition independent of sexual behaviour, suggesting a causal role for rectal STI in HIV-acquisition, and supporting STI control as an HIV-prevention strategy.

Disclosure of interest statement This work was funded by the US National Institutes of Health. No pharmaceutical grants were received in the development of this study.