Background: Our objective was to assess the efficacy of male circumcision for prevention of HSV-2 in men and vaginal infections in women.

Methods:  Men were randomized to immediate male circumcision (intervention arm) or delayed circumcision for 2 years (control arm). HSV-2^– men (intervention 1400, controls 1387) were followed for 24 months to determine HSV-2 acquisition using Kalon HSV-2 enzyme-linked immunoassay (ELISA) index value ≥1.5. Married women were linked to enrolled men:  825 wives of intervention arm and 783 wives of control arm men were followed at 1 year to assess genitourinary disease (GUD), bacterial vaginosis (BV) by Nugent score and trichomonas by culture.

Results:  Males: HSV-2 acquisition was 7.6% in the intervention arm and 10.1% in the control arm (RR 0.75 (CI 0.59 to 0.96, p = 0.02). The relative risk of HSV-2 incidence was lower in the circumcision arm in all sociodemographic and behavioral subgroups. Multiplex real-time polymerase chain reaction (PCR) for HSV-1, HSV-2, Treponema pallidum and Haemophilus ducreyi identified an etiological agent in 95 (36.3%) of ulcers; of which HSV-2 was detected in 86 (90.5%). Among 62 men who acquired HIV during the trial, 36 (58.1%) had either prior HSV-2 infection (n = 29) or simultaneous HIV and HSV-2 seroconversion (n = 9). Females:  At follow up, among married women who were linked to enrolled men, the rate of symptomatic GUD was 12.5% in intervention and 16.8% in control arm wives (RR 0.78, CI 0.63 to 0.97), but there were no differences in reported symptoms of vaginal discharge or dysuria by study arm. The prevalence of trichomonas was 5.9% in the intervention arm wives and 11.2% in control arm wives (RR 0.52, CI 0.05 to 0.98). The prevalence of BV was 40.3% in intervention arm wives and 50.6% in control arm wives (RR 0.60, CI 0.38 to 0.94). Severe BV (Nugent scores 9-10) was 2.0% in intervention arm wives and 6.5% in control arm wives (RR = 0.39, CI 0.24 to 0.64). Among women with normal vaginal flora at enrollment, progression to BV was reduced in wives of intervention arm men (PRR = 0.80, 95%CI 0.65 to 0.97), and among women with BV at enrollment, persistence was reduced (PRR 0.82, 95%CI 0.72 to 0.96).

Conclusions:  Male circumcision prevents HSV-2 acquisition in men and reduces rates of GUD, trichomonas, and BV in their female partners. These effects of circumcision may influence the protective effect of circumcision on HIV acquisition.