Many Zambian men who undergo circumcision have sex during the healing period, putting themselves and their partners at risk for HIV, according to a recent observational study.1 Despite having been advised to abstain from sex for six weeks following their circumcision, one in four men in the study resumed having sex during the healing period, and substantial proportions of the nonabstainers had unprotected sex (82%) or multiple partners (37%). If the levels of behavior exhibited by study participants are typical, then premature resumption of sexual activity following circumcision may be undercutting the effectiveness of circumcision programs, though the researchers estimate that the HIV transmissions prevented by circumcision still greatly outnumber those resulting from intercourse during the healing period.

The results of three randomized trials conducted in Sub-Saharan Africa suggest that after healing, and in the absence of compensatory behavior, circumcised men are only about a third as likely as uncircumcised men to become infected with HIV. Newly circumcised men are counseled to abstain from sex for at least six weeks, as the unhealed wound provides a route for transmission of HIV; however, data from the three trials indicate that some men ignore this recommendation.

To assess the prevalence and predictors of early resumption of sex and explore its impact on transmission rates, researchers analyzed data from a Zambian study conducted in 2010–2011 to examine risk behaviors following circumcision. Of the appoximately 3,000 men aged 15–29 who participated in the household-based survey, 248 who were about to undergo circumcision were recruited to take part in a substudy focusing on sexual behavior during the immediate postcircumcision period. The men were interviewed at the time of their procedure and again six weeks later. Men answered questions about their social and demographic characteristics, sexual behavior, risk perceptions and STI history; sensitive questions were administered via audio computer-assisted self-interviews. To reduce the likelihood that participants would give socially desirable answers at follow-up, participants were not asked directly whether they had had sex since their circumcision; instead, they were asked whether they were sexually active, and those who responded affirmatively were probed with follow-up questions to determine the timing of resumption of sex.

The researchers conducted multivariate regression analyses to identify characteristics and behaviors associated with premature resumption of sexual activity. They used modeling to estimate the impact of early resumption of sexual activity on HIV transmission at the population level, taking into account such factors as men’s HIV status, their number of sexual partners and their partners’ HIV status.

On average, the 225 men who returned for follow-up interviews six weeks after their circumcision were 21 years old and had had 10 years of schooling; most were unmarried (92%), and fewer than half had a regular sex partner (44%). The vast majority (96%) lived in urban or periurban areas, which were where Zambia’s circumcision programs were being administered at the time. Compared with men in the broader household-based sample, men in the circumcision subgroup were younger, more educated and more likely to have a regular partner.

Prior to their circumcision, men were told to allow six weeks for healing and were counseled to abstain from sex during that time. Nonetheless, at the six-week follow-up interview, one in four men (24%) reported having already resumed having sex; more than one in five (22%) of these sexually active men had had sex in the first week after the procedure, and almost half (46%) during the first three weeks. The vast majority (82%) of men who had had sex during the first six weeks had had unprotected sex at least once during that time, and more than two in five (43%) had had unprotected sex during the first three weeks. Slightly more than a third (37%) had had sex (protected or unprotected) with two or more partners. Thirty-two percent had engaged in both risky behaviors—having unprotected sex and having sex with multiple partners.

In multivariate analyses, few demographic or behavioral characteristics were associated with sexual behavior during the healing period. The odds of having had sex during the first six weeks after circumcision were elevated among older men (odds ratio, 1.2 per additional year) and among those who had had unprotected sex in the four weeks prior to circumcision (2.4); men were more likely to have had unprotected sex within six weeks if they were older (1.2 per year), and less likely to have had unprotected sex within three weeks if they were better educated (0.8 per year). Having had both unprotected sex and multiple partners during the healing period was associated with two measures: lifetime number of partners (1.3 per partner) and having had a casual partner in the four weeks prior to circumcision (5.6).

Using a transmission model, the researchers estimate that in 2010, when Zambia’s circumcision services were scaled up and 61,000 men underwent the procedure, early resumption of sexual activity (assuming it occurred at the levels seen in the current study) resulted in 69 HIV infections that would not otherwise have occurred. They estimate that in 37 of these extra cases, an already infected man transmitted the virus to his female partner during the healing period, while in 32 the partner infected the man. Nonetheless, the model suggests that the infections prevented by the scale-up of circumcision services outnumbered those resulting from premature resumption of sex, yielding a net total of 230 infections averted during the first year alone. The number of infections caused by early resumption of sex, and the proportion of these infections occurring in women, would have been smaller if levels of sex during the healing period had been lower, condom use had been higher, or the prevalence of HIV had been lower among circumcised men than was observed in the program (6%).

The researchers note that the men who took part in the study were “early adopters” of circumcision, and as such may not have been representative of all men in the study regions. Other limitations of the study include that men may have underreported their sexual activity during the postcircumcision period, and that the analysis of infections prevented did not take into account the longer-term benefits of circumcision. These limitations aside, the investigators conclude that “the prevalence of risky sexual behavior during the wound-healing period is relatively high,” and that “programs need to continue to emphasize to clients the risks associated with early resumption of sex.” They add that future research should examine whether interventions to reduce levels of unprotected sexual activity during the healing period would be useful; one approach might be to alert women that they are putting themselves at risk for HIV if they have sex with their partner in the immediate postcircumcision period.—P. Doskoch


1. Hewett PC et al., Sex with stitches: assessing the resumption of sexual activity during the postcircumcision wound-healing period, AIDS, 2012, 26(6): 749–756.