What is Male Circumcision?
Male circumcision is the surgical removal of some or all of the foreskin (or prepuce) from the penis.1
Male Circumcision and Risk for HIV Acquisition by Heterosexual Men
Several types of research have documented that male circumcision significantly reduces the risk of men contracting HIV through penile-vaginal sex.
Compared with the dry external skin surface of the glans penis and penile shaft, the inner mucosa of the foreskin has less keratinization (deposition of fibrous protein) and a higher density of target cells for HIV infection.2 Some laboratory studies have shown the foreskin is more susceptible to HIV infection than other penile tissue,3 although others have failed to show any difference in the ability of HIV to penetrate inner compared with outer foreskin surface.4 The foreskin may also have greater susceptibility to traumatic epithelial disruptions (tears) during intercourse, providing a portal of entry for pathogens, including HIV.5 In addition, the microenvironment in the preputial sac between the unretracted foreskin and the glans penis may be conducive to viral survival.3 Finally, the presence of other sexually transmitted diseases (STDs), which independently may be more common in uncircumcised men, increase the risk for HIV acquisition.6-11
International Observational Studies for Prevention of HIV Acquisition by Heterosexual Men
A systematic review and meta-analysis that focused on male circumcision and heterosexual transmission of HIV in Africa was published in 2000.12 It included 19 cross-sectional studies, 5 case-control studies, 3 cohort studies, and 1 partner study. A substantial protective effect of male circumcision on risk for HIV infection was noted, along with a reduced risk for genital ulcer disease. After adjustment for confounding factors in the population-based studies, the relative risk for HIV infection was 44% lower in circumcised men. The strongest association was seen in men at high risk, such as patients at STD clinics, for whom the adjusted relative risk was 71% lower for circumcised men.
Another review that included stringent assessment of 10 potential confounding factors and that was stratified by study type or study population was published in 2003.13 Most of the studies were from Africa. Of the 35 observational studies in the review, the 16 in the general population had inconsistent results. The one large prospective cohort study in this group showed a significant protective effect: The odds of infection were 42% lower for circumcised men.14 The remaining 19 studies were conducted in populations at high risk. These studies found a consistent, substantial protective effect, which increased with adjustment for confounding. Each of the four cohort studies included in the review demonstrated a protective effect, and two were statistically significant.
Ecologic studies also indicate a strong association between lack of male circumcision and HIV infection at the population level. Although links among circumcision, culture, religion, and risk behavior may account for some of the differences in HIV infection prevalence, the countries in Africa and Asia with prevalence of male circumcision of less than 20% have HIV infection prevalences several times higher than those in countries in these regions where more than 80% of men are circumcised.15
International Clinical Trials for Prevention of HIV Acquisition by Heterosexual Men
Three randomized controlled clinical trials (RCTs) were conducted in Africa to determine whether circumcision of adult males reduces their risk for HIV infection. The controlled follow-up period in all three studies was stopped early, and the control group offered circumcision when interim analyses found that medical circumcision significantly reduced male participants' HIV infection risk. The controlled follow-up period in the study in South Africa16 was stopped in 2005, and the controlled follow-up periods for the studies in Kenya17 and Uganda18 were stopped in 2006.
In these studies, men who had been randomly assigned to the circumcision group had a 60% (South Africa), 53% (Kenya), and 51% (Uganda) lower incidence of HIV infection compared with men assigned to the wait-list group to be circumcised at the end of the study. In all three studies, a small number of men who had been assigned to be circumcised did not undergo the procedure; likewise, a small number of men assigned to the control groups did undergo circumcision. When the data were reanalyzed to account for these occurrences, men who had been circumcised had a 76% (South Africa), 60% (Kenya), and 55% (Uganda) reduction in risk for HIV infection compared with those who were not circumcised.
A 2008 meta-analysis, which examined data from the three RCTs, as well as from cohort and case-control studies, found that HIV risk was reduced 58% in circumcised men (overall risk ratio [RR], 0.42; 95% confidence interval [CI], 0.33-0.53). The authors concluded that the studies provided enough evidence to conclude that circumcision causes a reduction in transmission of HIV-1 infection.19
Male Circumcision and Male-to-Female Transmission of HIV
Studies of whether circumcision of males reduces HIV transmission to their female sex partners overall indicate no protective effect.
A RCT of male circumcision in Uganda found no evidence of reduced HIV transmission to female partners. In the study, 922 HIV-infected men with uninfected partners were assigned to either immediate or delayed circumcision. Overall, 18% of women in the intervention group (partners in the immediate circumcision group) acquired HIV during follow-up, compared with 12% of women assigned to the control group (partners in the delayed circumcision group). There was no difference in HIV incidence between the circumcised and control groups when the couples waited to resume sex until the wound had healed, which in 93% of male subjects was within 6 weeks of circumcision.20 However, women appeared to be at somewhat higher risk for HIV acquisition when the couples resumed sex before the circumcision wounds had healed, although this difference was not statistically significant.
A systematic review and meta-analysis of male circumcision and risk of transmission to women identified 19 studies from 11 populations.21 The meta-analysis of data from the one RCT and six longitudinal analyses showed little evidence that male circumcision directly affects the risk of transmitting HIV to women (RR, 0.80; 95% CI, 0.53-1.36).
Male Circumcision and Male-to-Male Transmission of HIV
Observational studies have yielded mixed results in attempts to detect a protective effect of male circumcision among men who have sex with men (MSM). While some cross-sectional22 and prospective23 studies of MSM have shown statistically significant increases in risk of HIV acquisition by uncircumcised MSM, others have found no evidence that being circumcised was protective against HIV infection among MSM.24 In a recent meta-analysis of 15 observational studies of male circumcision and HIV acquisition by MSM, a statistically nonsignificant protective association was found (OR, 0.86; 95% CI, 0.65-1.13).25
HIV Infection and Male Circumcision in the United States
The United States has a much lower population prevalence of HIV infection (0.4%) than sub-Saharan Africa,53 and an epidemic that is concentrated among men who have sex with men, rather than men who have sex with women.54, 55 In 2006, it is estimated that approximately 56,300 new HIV infections occurred, of which 73% were in males.54 Of all new infections, 53% were in MSM, 31% in heterosexuals with reported high risk of exposure, 12% in injection drug users (IDUs), and 4% in MSM-IDUs. Among men, 72% of estimated new infections occurred in the male-to-male sexual contact transmission category, while heterosexual transmission accounted for 13%.53
In one prospective study of heterosexual men attending an urban STD clinic, when other risk factors were controlled, uncircumcised men had a 3.5-fold higher risk for HIV infection than men who were circumcised. However, this association was not statistically significant due to small sample size.11 And in an analysis of clinic records for African American men attending an STD clinic, circumcision was not associated with HIV status overall, but among heterosexual men with known HIV exposure, circumcision was associated with a statistically significant 58% reduction in risk for HIV infection.56
Male Circumcision in the United States
In national probability samples of adults surveyed during 1999–2004, the National Health and Nutrition Examination Surveys (NHANES) found that 79% of men reported being circumcised, including 88% of non-Hispanic white men, 73% of non-Hispanic black men, 42% of Mexican American men, and 50% of men of other races/ ethnicities.57 It is important to note that reported circumcision status may be subject to misclassification. In a study of adolescents¸ only 69% of circumcised and 65% of uncircumcised young men correctly identified their circumcision status as verified by physical exam.58
According to the National Hospital Discharge Survey (NHDS), 65% of newborns were circumcised in 1999, and the overall proportion of newborns circumcised was stable from 1979 through 1999. In 2007, the NHDS found that 55% of male infants were circumcised.59 Notably, the proportion of black newborns circumcised increased during this reporting period (58% to 64%); the proportion of white newborns circumcised remained stable (66%). In addition, the proportion of newborns in the Midwest who were circumcised increased during the 20-year period—from 74% in 1979 to 81% in 1999—while the proportion of infants born in the West who were circumcised decreased from 64% in 1979 to 37% in 1999. In another survey, the National Inpatient Sample (NIS), circumcision rates increased from 48% during 1988–1991 to 61% during 1997–2000 but declined to 56% in 2008. Circumcision was more common among newborns who were born to families of higher socioeconomic status, born in the Northeast or Midwest, and who were black.60
In 1999, the American Academy of Pediatrics (AAP) changed from a neutral stance on circumcision to a position that the data then available were insufficient to recommend routine neonatal male circumcision. The Academy also stated, “It is legitimate for the parents to take into account cultural, religious, and ethnic traditions, in addition to medical factors, when making this choice.”61 This position was reaffirmed by the Academy in 2005. AAP is currently reviewing and amending their policy in light of new data. Medicaid does not reimburse the costs of neonatal male circumcision in all states, which poses a barrier to male circumcision for individuals without private insurance. In a recent study conducted in 37 states in which the NIS is conducted, hospitals in states where Medicaid covers routine male circumcision had 24% higher circumcision rates than hospitals in states without such coverage.62
Cost-Effectiveness and Ethical Issues for Neonatal Circumcision in the United States
A large, retrospective study of circumcision in nearly 15,000 infants found neonatal circumcision to be highly cost-effective, considering the estimated number of averted cases of infant urinary tract infection and lifetime incidence of HIV infection, penile cancer, balanoposthitis (inflammation of the foreskin and glans), and phimosis (a condition where the male foreskin cannot be fully retracted from the head of the penis). The cost of postneonatal circumcision was 10-fold the cost of neonatal circumcision63. There are also studies showing very marginal cost-effectiveness.
A 2010 study estimated that newborn circumcision reduces a U.S. male’s lifetime risk of HIV acquisition through heterosexual contact by 15.7% overall, by 20.9% for black males, 12.3% for Hispanic males, and 7.9% for white males. In this model, the number of circumcisions needed to prevent one case of HIV was 298 for all males and ranged from 65 for black males to 1,231 for white males. Based on these estimates, the study concluded that newborn male circumcision was a cost-saving HIV prevention intervention.64
Little has been published on the cost-effectiveness of adult circumcision among MSM. A study in Australia found that although a relatively small percentage of HIV infections would be prevented, adult circumcision of MSM could be cost effective or cost saving in some scenarios.65
Many parents now make decisions about infant circumcision based on cultural, religious, or parental desires, rather than health concerns.66 Some have raised ethical objections to parents making decisions about elective surgery on behalf of an infant, particularly when it is done primarily to protect against risks of HIV and STDs that do not occur until young adulthood. But other ethicists have found it an appropriate parental proxy decision.67
Considerations for the United States
A number of important differences from sub-Saharan African settings where the three male circumcision trials were conducted must be considered in determining the possible role for male circumcision in HIV prevention in the United States. Notably, the overall risk of HIV infection is considerably lower in the United States, changing risk-benefit and cost-effectiveness considerations. Also, studies to date have demonstrated efficacy only for penile-vaginal sex, the predominant mode of HIV transmission in Africa, whereas the predominant mode of sexual HIV transmission in the United States is by penile-anal sex among MSM. There are as yet no convincing data to help determine whether male circumcision will have any effect on HIV risk for men who engage in anal sex with either a female or male partner, as either the insertive or receptive partner. Receptive anal sex is associated with a substantially greater risk of HIV acquisition than is insertive anal sex. It is more biologically plausible that male circumcision would reduce HIV acquisition risk for the insertive partner rather than for the receptive partner, but relatively few MSM engage solely in insertive anal sex.68
In addition, although the prevalence of circumcision may be somewhat lower in U.S. racial and ethnic groups with higher rates of HIV infection, most American men are already circumcised; and it is not known whether men at higher risk for HIV infection would be willing to be circumcised or whether parents would be willing to have their infants circumcised to reduce possible future HIV infection risk.