Background: Lack of circumcision and condom failure have independently predicted risk of HIV seroconversion in different prospective cohorts of men who have sex with men (MSM). We sought to determine whether condom failure explains the association between lack of circumcision and HIV risk for the insertive anal sex partner when condoms are used. For this to be true, condom failure must be associated with both lack of circumcision and risk for HIV seroconversion of the insertive partner.

Methods:  Condom failure, defined as breakage or slippage during insertive anal sex, was assessed by circumcision status over 18 months of follow-up among 2161 MSM in the National Institute of Allergy and Infectious Diseases (NIAID) -sponsored Vaccine Preparedness Study.

Results:  Of included participants, 12% reported being uncircumcised. Overall, these men reported condom failure in 1.7% of 12,021 insertive anal sex episodes, as compared to 1.5% of 74,160 among circumcised men. Using generalized estimating equation (GEE) binomial models to account for frequency of insertive anal sex, which was slightly higher among uncircumcised men, risk of condom failure was similar in the 2 groups (OR 1.11, 95%CI 0.81 to 1.51). Across 5092 6-month participant follow-up periods where at least 1 episode of protected insertive anal sex was reported, 51 HIV seroconversions were detected. A mean of 11.7 (range 1 to 85, total 59,648) episodes of protected insertive anal sex were reported in these periods, while slippage or breakage was reported in an average of 0.16 insertive anal sex episodes (range 0 to 10, total 857). Using pooled logistic models for seroconversion, we found no evidence that condom slippage or breakage increased risk of HIV infection for the insertive partner (p = 0.53). Furthermore, controlling for slippage or breakage did not attenuate the increased risk among uncircumcised men previously reported from this cohort.

Conclusions: Similar condom failure rates in both circumcised and uncircumcised men, and the lack of an independent association between condom failure and HIV seroconversion for the insertive partner, suggest that increased condom failure neither explains nor mediates the increased risk of HIV seroconversion observed among uncircumcised MSM practicing insertive anal sex.