Circumcision - Sexually Transmitted Infections

There has been an alarming increase in STIs in recent years, reaching 19 million cases in 2006 in the USA according to the Centers for Disease Control (CDC) [Altman, 2007]. This included 1,030,911 cases of Chlamydia, a record high, with the actual number estimated as 2.8 million [Altman, 2007]. There were 358,366 reported cases of gonorrhoea, the actual number being twice this, and 9,756 cases of syphilis in the USA in 2006 [Altman, 2007]. Genital herpes (HSV-2), human papillomavirus (HPV) account for the vast majority of STI cases, although doctors are not required to report them nationally [Altman, 2007]. Nevertheless, it has been estimated that 25 million Americans (17% of the adult population) and worldwide, 536 million people (16% of those aged 15–49 years) are infected with HSV-2 [Golden & Wasserheit, 2009]. Although seldom fatal in adults, HSV-2 is associated with substantial morbidity, and direct medical costs in the USA are approx. $1 billion annually. It also increases the risk of HIV infection. HSV-2 can be passed on to infants during birth. There are 2800 neonatal herpes cases per year in the USA and these often result in severe disability or death [Golden & Wasserheit, 2009]. In the Australian state of New South Wales 17,000 people tested positive for Chlamydia, gonorrhoea, syphilis and HIV in 2008, but the number is thought to be much higher because many who have an STI show no symptoms.

Ulcerative STIs (particularly chancroid and syphilis) are associated with lack of circumcision.

The present section deals with all STIs with the exception of human immunodeficiency virus (HIV) and HPV, which are dealt with in separate sections to follow.

STIs have an enormous adverse impact on global public health [Low et al., 2006]. For example, HPV accounts for 3.3 million disability adjusted life years (DALYs) from cervical cancer caused by high-risk HPV types [Low et al., 2006]. Syphilis, which is associated with fatalities, stillbirth in pregnant women, prematurity and congenital infection, is responsible for 4.2 million DALYs [Low et al., 2006].

Chlamydia and gonorrhoea, which account for 7 million DALYs, cause infertility by blocking fallopian tubes in women and the vas deferens in men, as well as being responsible for pelvic pain, pelvic inflammatory disease and ectopic pregnancy, which can lead to death of the mother [Low et al., 2006]. A commitment to control all STIs needs dispassionate public health action to replace ‘moral prophylaxis’ [Low et al., 2006]. Condoms, while helpful, are not completely effective, even when use consistently [Low et al., 2006]

The link between circumcision and protection from STIs has a history going back to over 150 years. So let’s walk through the research findings over time.

In 1855 syphilis was discovered to be associated with lack of circumcision [Hutchinson, 1855]. Then in 1891 Romondino confirmed this finding and also noted the possible protection afforded by circumcision against genital herpes and urethritis [Remondino, 1891].

In 1947, a study involving 1,300 consecutive patients in a Canadian Army unit, showed that being uncircumcised was associated with a 9-fold higher risk of syphilis and 3-times higher gonorrhea [Wilson, 1947]. A report in 1949 found higher syphilis, chancroid and gonorrhoea [Hand, 1949]. Higher chancroid was also reported in 1952 [Asin, 1952].

In 1967 higher HSV-2 was found in men who were not circumcised [Parker & Banatvala, 1967]. Then, in the mid-70s work by the London Hospital showed higher chancroid, syphilis, papillomavirus and herpes in uncircumcised men [Taylor & Rodin, 1975]. Higher chancroid was also seen in a 1980 report [Hammond et al., 1980]. Subsequent to this, a study in 1983 at the University of Western Australia, showed twice as much herpes and gonorrhea, 5-times more candidiasis and 5-fold greater incidence of syphilis [Parker et al., 1983]. In South Australia, a study in 1992 showed that uncircumcised men had more chlamydia (odds ratio 1.3) and gonoccocal infections (odds ratio 2.1) [Hart, 1993]. Others have reported higher rates of non-gonococcal urethritis in uncircumcised men [Smith et al., 1987].

In 1988 a study in Seattle of 2,776 heterosexual men reported higher syphilis and gonorrhoea in uncircumcised men, but no difference in herpes, chlamydia and non-specific urethritis (NSU) [Cook et al., 1994]. Like this report, a study in 1994 in the USA, found higher gonorrhoea and syphilis, but no difference in other common STIs. An earlier (1987) study of 9,514 sexually transmitted infection patients from a US military base found higher non-gonococcal, but not gonococcal, urethritis in those who were uncircumcised [Smith et al., 1987].

In 1994, Donovan and associates reported the results of a study of 300 consecutive heterosexual male patients attending Sydney STI Centre at Sydney Hospital [Donovan et al., 1994]. They found no difference in NSU, genital herpes (24% having a history of this [Basset et al., 1994]) or seropositivity for HSV-2 (65% [Basset et al., 1994]) and genital warts (i.e., the benign, so-called 'low-risk' HPV types 6 and 11, which are visible on physical examination, unlike the 'high-risk' types 16 and 18, which are not). As mentioned earlier, 62% were circumcised and the two groups had a similar age, number of partners and education. Gonorrhoea, syphilis and hepatitis B were too uncommon in this Sydney study for them to conclude anything about these other STIs.

A word of caution is needed in evaluating data from STI clinics in that circumcised men, if protected from any particular STI, are less likely to attend the clinic. This means data from STI clinics is biased towards a lesser ability to detect a protective effect of circumcision. Studies in general populations are less likely to suffer from such bias and so would be more reliable.

Similar findings were obtained in the National Health and Social Life Survey in the USA, which asked about gonorrhoea, syphilis, chlamydia, non-gonococcal urethritis, herpes and HIV (a virus more often acquired intravenously in heterosexual i.v. drug-using men in the USA) [Latif et al., 1989], although some under-reporting by uncircumcised men was likely as they tended to be less educated. Also, circumcision at birth was assumed, so that the number who sought circumcision later in life for problems, such as STIs and/or other infections, and therefore had switched group, was not taken into account.

In a cross-sectional and cohort study from a multicenter controlled trial involving 2021 men in the USA from 1993 to 1996, and using multiple logistic regression to compare STI risk among circumcised and uncircumcised men adjusted for potential confounding factors, uncircumcised men were significantly more likely to have gonorrhoea in the multivariate analysis adjusted for age, race and site (odds ratio 1.3 and 1.6 for each respective study) [Diseker et al., 2000]. This was also the case for syphilis (odds ratios 1.4 and 1.5), but not chlamydia.

The warm moist environment under the prepuce favours bacterial replication. The delicate inner lining’s mucosal nature and risk of tearing it and the frenulum during intercourse are other factors. Chancroid is more likely to present on the inner and outer prepuce, whereas syphilis and herpes simplex type 2 (HSV-2) tend to infect the genitalia more widely.

For genital herpes a 1998 review of 11 studies [Moses et al., 1998] noted 2 studies that showed an association with lack of circumcision [Taylor & Rodin, 1975; Parker et al., 1983] and 4 that found no association [Basset et al., 1994; Cook et al., 1994; Donovan et al., 1994; Laumann et al., 1997].

For gonorrhoea 5 reported significant association [Wilson, 1947; Hooper et al., 1978; Parker et al., 1983; Hart, 1993; Cook et al., 1994] and 2 no association [Smith et al., 1987; Laumann et al., 1997]. For chlamydial, non-gonococcal or other types of urethritis 2 studies reported association with lack of circumcision [Hart, 1993; US, 1998], 3 with circumcision [Hooper et al., 1978; Newell et al., 1993; Laumann et al., 1997] and 3 no association [Smith et al., 1987; Cook et al., 1994; Donovan et al., 1994]. Similarly, no association was found in a 2005 report [Dickson et al., 2005]. An Australia-wide telephone survey found no differences in genital warts, Chlamydia, genital herpes, gonorrhea, non-specific urethritis or public lice between circumcised and uncircumcised men [Richters et al., 2006].

A New Zealand study saw no difference in frequency of serum antibodies to HSV-2 (7%) between men aged 26 who had been circumcised prior to the age of 3 compared with those who were uncircumcised [Dickson et al., 2005]. Similarly, the NHANES study by the CDC found no significant association between HSV-2 and circumcision status, seroprevalence being 13.7% in uncircumcised and 11.6% in circumcised [Xu et al., 2007]. Data discussed below for randomized controlled trials did, however, find HSV-2 to be lower (by one-third) in those in the circumcised arm. A small study in India in 2009 of patients with recurrent herpes genitalis found recurrence was reduced 20-fold during the 11-27 years post-circumcision [Jerath & Mahajan, 2009]. Circumcision also prolonged the disease-free period in between recurrences.

In the case of HSV-1, a study in the USA by groups in Alabama and Tennessee of Black heterosexual men aged 18–25 years who were attending an STI clinic in the USA found HSV-1 seroprevalence to be 2.8 times higher in those who were uncircumcised [Van Wagoner et al., 2009].

Meta-analyses:

Results of the first meta-analyses of ulcerative STIs were reported in 2006 by Weiss and co-workers in London. This was based on 26 research articles (from the USA, UK, Australia, Africa, India and Peru) [Weiss et al., 2006] and are summarized in the Table below. The analyses established that circumcised men were at very much lower risk of chancroid and syphilis. The association with HSV-2 was weaker, genital herpes being only 12% lower in circumcised men.

For the syphilis studies in this Table, adjusted odds ratios or prevalence ratios in each were:  0.54 [Buve et al., 2001], 0.69 [Bwayo et al., 1994], 0.25 [Cook et al., 1994], 0.52 [Diseker et al., 2000], 1.01 [Gray et al., 2004], 0.64 [Lavreys et al., 1999], 0.60 [Newell et al., 1993], 0.19 [Parker et al., 1983], 0.63 [Reynolds et al., 2004], 0.78 [Tabet et al., 2002], 0.70 [Todd et al., 2001], 0.95 [Urassa et al., 1997], and 0.71 [Vaz et al., 1995].

The risk ratios in the chancroid studies in the Table were: 0.13 [Hand, 1949], 1.11 [Rakwar et al., 1999], 0.40 [Lloyd, 1934], 0.04 [Barile et al., 1962], 0.66 [Nasio et al., 1996], 0.62 [Cameron et al., 1989], 0.21 [Hart, 1975].

In the case of the HSV-2 studies, rate, prevalence or odds ratios were: 0.88 [Gottlieb et al., 2004], 1.20 [Auvert et al., 2001], 0.81 [Gray et al., 2004], 0.56 [Kapiga et al., 2003], 1.18 [Lavreys et al., 1999], 0.39 [Obasi et al., 1999], 0.91 [Reynolds et al., 2004],  0.84 [Suligoi et al., 2001], 0.73 [Weiss et al., 2001], and 1.04 [Weiss et al., 2001].

Another meta-analysis, published in 2007, found uncircumcised men were 2.3 times more likely to have genital ulcer disease [Van Howe, 2007].

This meta-analysis, by Van Howe, also reported higher sexually acquired urethritis in circumcised men [Van Howe, 2007]. The findings should, however, be disregarded, since it has been revealed that much of the data Van Howe included in his analysis bore little resemblance to the actual data that can be found in the source publications he used [Waskett & Morris, 2008]. In fact an Erratum published by the journal discredits Van Howe's findings and neutralizes his claim. His entire paper should in fact have been retracted.

In the case of non-specific urethritis (NSU), a correct meta-analysis of the actual data in the 10 studies sourced in Van Howe’s report showed NSU to be 8% lower in circumcised men: summary OR = 0.92 (95% CI 0.64–1.3), which was not statistically significant [Waskett & Morris, 2008].

Design aspects of a number of the case-control studies have been criticized. As a result, until recently, there was no overwhelming agreement, as highlighted in the meta-analyses [Weiss et al., 2006] referred to in the Table above. Nevertheless, on the bulk of evidence, it seemed that at least some STIs were more common in the uncircumcised. The association of several STIs with lack of circumcision was not as strong in Western settings, and the prevalence may have been influenced by factors such as the degree of genital hygiene, availability of running water and socioeconomic group being studied. In some more recent studies in developed nations, in which hygiene is good, little difference was apparent in some of the more common STIs such as gonorrhea and herpes.

Longitudinal studies:

Longitudinal studies are regarded as superior to case-control association studies. One survey, in Christchurch, New Zealand, involving a birth cohort born in the early 1970s, found that to age 25 the uncircumcised reported having had a 3.2-fold higher rate of STI when compared with those who were circumcised, after adjustment for the higher number of sexual partners and of rate of unprotected sex in the 30% who were circumcised [Fergusson et al., 2006].  The frequency of STIs amongst the participants were 52% Chlamydia, 31% genital warts, 12% non-specific urethritis, 10% genital HSV-2 and 5%, gonorrhea. It was concluded that if all had been circumcised the rate of STI would have been reduced by 48%. For Chlamydia the OR was 2.5 (CI 0.73-8.5). A subsequent longitudinal study, in Dunedin, New Zealand, involving a similar birth cohort, found that to age 32 years 23% reported ever having had had an STI irrespective of circumcision status [Dickson et al., 2008]. An earlier report on this cohort found that, to age 26, HSV-2 seroprevalence was similar [Dickson et al., 2005]. (Similar HSV-2 seropositivity was also seen in a US study of Black heterosexual men, but in the case of HSV-1 seroprevalence was 2.8-fold higher in those aged 18-25 who were not circumcised [Van Wagoner et al., 2009]). Similar seroprevalence of HPV, as well, was reported in 2009 [Dickson et al., 2009]. Despite this, penile HPV is cleared 6 times faster by circumcised men, so accounting for them having a much lower HPV prevalence on the penis [Lu et al., 2009]. Each of these longitudinal cohorts was, however, relatively small, which might have contributed to the discrepancies between them.

Randomized controlled trials:

The randomized controlled trial (RCT) is, however, the ‘gold standard’ for study design in epidemiology. A RCT in Uganda found that the protective effect of circumcision against genital ulcer disease was 48% [Gray et al., 2007a]. Subsequently, two large RCTs found lower genital herpes in men allocated to the circumcised arm of the trial. The seroprevalence of HSV-2 was 45% lower in the trial involving 2974 men in South Africa [Sobngwi-Tambekou et al., 2009a] and was 30% lower in the trial involving 6396 men in Uganda [Tobian et al., 2009a].

The data emanating from Rakai, Uganda, has in fact now been from two RCTs in that locality. One of these (RCT-2) involved men with a higher sexual-risk profile. After the initial report referred to above, in a trial focused primarily on HIV prevention, and that found circumcision afforded 48% protection against GUD [Gray et al., 2007a], a subsequent report noted that GUD was 39% lower in circumcised men in RCT-2 (5.6% vs 9.2%), but did not differ in RCT-1 (4.6% vs 4.3%) [Tobian et al., 2009b]. The point prevalence of GUD at the time of a study visit was found to be 1.9% in the uncircumcised men and 0.8% in those who had undergone circumcision, and period prevalence of GUD was reported as being 46% lower in the men who had been circumcised [Gray et al., 2009b]. Among those with positive syphilis serology during follow-up, GUD symptoms (preceding serological syphilis detection) were 4.5% in the circumcised men and 13.3% in the uncircumcised men (PRR = 0.33), i.e, was 67% lower [Gray et al., 2009b]. In the Rakai RCT-1 trial, circumcision reduced HSV-2 by 23% and in RCT-2 by 41% [Tobian et al., 2009b]. Circumcision was associated with a reduction in GUD of 49% in those who were HSV-2 seronegative [Gray et al., 2009b]. No difference in syphilis seroprevalence was found in this trial [Tobian et al., 2009b], but only 2% of men were infected with this STI.

Although it was suggested at the time that the RCT might have been insufficiently powered to reach a valid conclusion [Golden & Wasserheit, 2009], positive syphilis serology was, however, found in 7% of men in each group in a later report, which also tested for the DNA of the syphilis organism, T. pallidum, and found positive results in 7/56 swabs from genital ulcers in uncircumcised men, but none of the ulcers in 25 circumcised men [Gray et al., 2009b]. In this report HSV-2 seroprevalence was 27% and 28% in each respective group, and in men with genital ulcers, HSV-2 DNA was found in 48% and 39% of swabs (P=0.62) [Gray et al., 2009b]. None of the ulcers in either group contained DNA for H. ducreyi (chancroid) or HSV-1. A large proportion of those whose test was STI negative had a non-STI as a cause of their penile ulceration. It was suggested that most of the ulcers were a result of infection, by non-STI pathogens, of tears in the foreskin and its attached frenulum, the authors pointing out that tearing occurs commonly in uncircumcised men during intercourse [Gray et al., 2009b], thus highlighting an added benefit of circumcision.

Interestingly, as mentioned above, although others have reported no difference in HSV-2 seroincidence (i.e., antibodies to HSV-2 in the bloodstream, indicative of current or prior infection) between circumcised and uncircumcised men [Dickson et al., 2005; Van Wagoner et al., 2009], the incidence of genital ulcer disease (GUD), including herpetic lesions, is twice as high in uncircumcised men [Bailey & Mehta, 2009], suggesting that circumcision may reduce the RECURRENCE of genital lesions arising from HSV-2 infection. As mentioned earlier, a study in India, albeit small, found recurrence was 20 times lower in men who underwent circumcision when compared with genital herpes in those who remained uncircumcised, and circumcision prolonged the interval beteen recurrences [Jerath & Mahajan, 2009]. As also mentioned above, in the case of HSV-1, seroprevalence was found to be 2.8 times higher in uncircumcised Black heterosexual men aged 18–25 years who were attending an STI clinic in the USA [Van Wagoner et al., 2009].

An added benefit of circumcision in reducing HSV-2 is that it should also contribute to a lowering of HIV infection [Bailey & Mehta, 2009], even though the latter appeared independent of HSV-2 serostatus [Sobngwi-Tambekou et al., 2009a]. A synergy between HIV and HSV-2 infections has also been reported by this group [Mahiane et al., 2009]. Female-to-male HSV-2 infection per unprotected sex act was found to be 0.013 in uncircumcised men, compared with 0.0074 in circumcised men (RR 0.56; P = 0.005) [Mahiane et al., 2009]. For HIV, infection per unprotected sex act was 0.016 in uncircumcised men and 0.0036 in circumcised men (RR 0.23; P < 0.001) [Mahiane et al., 2009]. In partnerships, these values were, for HSV-2, 0.048 and 0.029 (RR 0.60; P = 0.001), and, for HIV, 0.054 and 0.014 (RR 0.26; P < 0.001), respectively. The modelling used suggested that the presence of HSV-2 in either partner would increase HIV infection from the woman to the man by 3.0-fold [Mahiane et al., 2009]. HSV-2 suppressive therapy has, however, failed to decrease HIV acquisition, as seen in two RCTs [Tobian & Quinn, 2009]. Importantly, the protective effect of circumcision against HIV acquisition appears independent of HSV-2 serostatus [Sobngwi-Tambekou et al., 2009a].

In a RCT in South Africa, the prevalence of gonorrhea was 9% lower in circumcised men, but this difference was not statistically significant (adjusted odds ratio (AOR) = 0.91) [Sobngwi-Tambekou et al., 2009b].

In the South African RCT, Chlamydia was 42% lower and Trichomonas vaginalis was 46% lower in the men who had been circumcised [Sobngwi-Tambekou et al., 2009b]. In an as-treated analysis, T. vaginalis was even lower, 51%, AOR being 0.41 [Sobngwi-Tambekou et al., 2009b]. This explains why women with circumcised male partners are less at risk of T. vaginalis infection.

In the Kenyan RCT, circumcision did not protect against either gonorrhea, Chlamydia or Trichomonas [Mehta et al., 2009b]. The data on gonorrhoea and Chlamydia is consistent with most earlier observational findings, and is to be expected because the preferred host site for these bacterial STIs is the internal urethral cuboidal or columnar epithelium. In this regard, although the Kenyan RCT data differ from the South African RCT data for Chlamydia and Trichomonas, it should be noted that the prevalence of Trichomonas in the Kenyan study was, however, lower than in other African countries.

The various newer findings for STIs were the subject of a review [Tobian et al., 2010]. The authors encouraged an increase in neonatal circumcision to reduce STI prevalence.

Studies in men who have sex with men (MSM):

The incidence of syphilis was 2.9 times lower in a study in Sydney comparing circumcised MSM with uncircumcised MSM, and was 10 times lower in the 33% who only ever engaged in insertive anal intercourse [Templeton et al., 2009b]. Similarly, a study of MSM in Seattle, found diagnosis of syphilis to be 2.0 times higher in uncircumcised men, and was completely absent from the 11% who said they were insertive-only [Jameson et al., 2009]. The adjusted OR, did not, however, reach significance. Seroprevalance of HSV-2 and HSV-1 were not lower, however, in either of these studies, although any association would have been missed because the findings were based on serological testing rather than viral testing of genital swabs. In MSM who were insertive-only, HSV-2 seroprevalence was 34% lower, albeit not reaching statistical significance either [Jameson et al., 2009]. The protection afforded by circumcision was similar to that seen in two randomized controlled trials of heterosexual men in Africa mentioned above. In this regard, herpetic lesions are twice as high in uncircumcised (heterosexual) men [Bailey & Mehta, 2009], and there is evidence that circumcised (heterosexual) men are much less prone to recurrence of genital herpes (by 20-fold) and had longer intervals between bouts compared with uncircumcised men [Jerath & Mahajan, 2009]. In the study of MSM in Sydney, prevalence of self-reported genital warts (which are caused by LOW-risk human papillomavirus (HPV)) was similar, but no information was obtained for high-risk HPV. The fact that low-risk HPV types infect the penis generally, including the shaft, means circumcision would be unlikely to reduce prevalence. In contrast the high-risk types are more prevalent near the tip of the penis and have been shown to be at least half as common in circumcised (heterosexual) men (see later section of penile cancer and HPV infection). In men aged 20-49 who had had sex with men, the dual biomedical factors that can lead to HSV-2 acquisition were only 0.8% in whites, but 8.5% in blacks and 7.5% in Hispanics [Xu et al., 2007].

Mechanism:  

One might ask why the prepuce increases risk of an STI? It may be that it is because the warm moist environment under the foreskin favours bacterial replication. The foreskin traps microorganisms in a pool of smegma, so facilitating transmission. The inner preputial lining, being mucosal, is delicate, so it and the frenulum can tear during intercourse. As well, the prepuce presents a larger area for infection. This would make it more prone to Chancroid, although syphilis and HSV-2 infect the genitalia more widely.

A report in 2010 identified the entire microbiome of the penis of 12 men before and after circumcision [Price et al., 2010]. This adds to data from decades earlier by Wiswell and other groups. Among the 42 unique bacterial families identified in the 2010 study, Pseudomonadaceae and Oxalobactericeae were the most abundant irrespective of circumcision status. Circumcision was associated with a significant change in the overall microbiota (P = 0.007) and with a significant decrease in putative anaerobic bacterial families (P = 0.014). Two families in particular - Clostridiales Family XI (P = 0.006) and Prevotellaceae (P = 0.006) - were uniquely abundant before circumcision. Within these families the authors identified a number of anaerobic genera previously associated with bacterial vaginosis including: Anaerococcus spp., Finegoldia spp., Peptoniphilus spp., and Prevotella spp. The researchers concluded that the anoxic microenvironment under the foreskin may support pro-inflammatory anaerobes that can activate Langerhans cells to present HIV to CD4 cells in draining lymph nodes. They suggested that the reduction in putative anaerobic bacteria after circumcision may play a role in protection from HIV and other sexually transmitted infections.

Promiscuous teenagers and early 20s are contributing to an epidemic of STIs such as Chlamydia, gonorrhoea and syphilis, with the number of new infections in many western countries like Australia and the USA soaring to record highs [Weaver, 2007; Hook, 2008]. In Australia Chlamydia infections more than tripled between 1999 and 2006 [Kang et al., 2006; Anonymous, 2007c]. In the USA in 2006 Chlamydia surpassed 1 million cases for the first time [Hook, 2008]. Chlamydia is 2.5 times higher in females than males [Weaver, 2007]. Its rise coincides with an increase in the numbers of uncircumcised males in these sexually more promiscuous younger people. A 10-fold rise in syphilis has moreover been reported in Australia in recent years. Knowledge about STIs amongst high school students in later years is poor and condom use remains low and inconsistent [Agius et al., 2006]. In the USA 1 in 4 adolescents can expect to have an STI, showing just how widespread STIs are in the population and the need to act [Hook, 2008]. The rising incidence has led to calls by experts for urgent action [Mindel & Kippax, 2005]. Circumcision is one such action that should be implemented.

In a global society risk of contracting an STI cannot be ascribed parochially. Travellers are particularly vulnerable to the different risk in a new country they may visit, particularly when holiday-making is associated with consumption of alcohol and other drugs, as well as an attitude of having a good time, which can lead to sexual relations with the locals, often with no condom [Marrazzo, 2005]. Circumcision offers a permanent protection, albeit not 100%, and therefore represents just one strategy to protect against STIs. This has led to calls for professional organizations to provide a leadership role in educating parents of all newborn sons about the benefits of circumcision in prevention of STIs and other problems [Golden & Wasserheit, 2009].

THRUSH

Penile candidiasis (thrush) is also significantly less common in circumcised men. It was 5 times lower in a study in Perth, Western Australia [Parker et al., 1983] and 60% lower in a large Australia-wide survey (odds ratio 0.40) [Richters et al., 2006]. This yeast (fungal) infection can occur from contact with a female sexual partner who has it. Men with diabetes are at increased risk. Symptoms can be none, a transient rash or a severe burning sensation after intercourse.