JB 25/0/2001

Medical Risks and Benefits

Does circumcision protect against STDs?
Is it more hygenic?
Is it true that it prevents cancer?
Are circumcised men less likely to get AIDS?

Unless circumcision is a requirement of your religion, the potential health benefits are likely to be the major factor in deciding whether it is a good thing for your sons. Many conflicting claims and counter-claims are aired by those with axes to grind, which is regrettable since the serious medical literature offers clear answers to these questions

Non-sexual diseases      Cancer      STDs      Risks

Non-sexual diseases

Two benefits of circumcision are clear without the need for any research:
circumcised males cannot suffer from phimosis (inability to retract the foreskin) or paraphimosis (a tight foreskin which becomes stuck in the retracted position). Both these conditions are reasonably common.

Phimosis is, in medical jargon, a chronic problem without primary acute effects. In other words, it persists for a long time but does not neccesarily cause any serious illness or injury in the short term. However, if the sufferer is an adult it will to be disastrous for his sex life, and this in turn is likely to cause psychological problems. It also predisposes him to other infections. Phimosis is quite common, affecting at least 10% of uncircumcised men in the West, and much higher figures in some Asian countries.

Paraphimosis, on the other hand, is a medical emergency and failure to seek prompt attention can have serious consequences. The strangulation caused by the tight skin can lead to swelling, and ultimately possibly to loss of the glans of the penis. Application of iced water to reduce the swelling will often enable the skin to be pulled forwards again, though a cut is sometimes required. In either case circumcision is almost always recommended. Paraphimosis is often said to be a rare problem, but 8% of the males in our surveys who were not circumcised as infants had suffered from paraphimosis.

Non-sexual urogenital infections are much more common among uncircumcised men and boys. In our survey they were five times as common in those who were not circumcised at a young age, and for many men this was a good reason for getting circumcised. These infections fall into two groups, external inflammation and uninary tract infections.

Balanitis - inflammation of the glans - is virtually unknown among circumcised males, and reasonably common among uncircumcised. These infections are not usually life threatening, but are most uncomfortable. In our study 14% of natural males had suffered from such infections, and about threequarters were subsequently circumcised.

Urinary Tract Infections (UTIs) are more common in uncircumcised males. (They are also much more common among gay men since the common gut bacterium Eschericia coli (E. coli) is the commonest cause of UTIs. A series of large scale studies (one involved 427,000 infants!) have been carried out by Thomas E. Wiswell and colleagues on children born in American Army hospitals. They looked at urinary tract infections in the first year of life, and showed that they were 11 to 20 times more common in uncircumcised boys. In one of Wiswell's studies urinary infections affected 4.12% of uncircumcised boys before age 1, as opposed to 0.21% of circumcised. Our more modest survey of Australian men shows that this is also true of the later years of life - natural men were 4 times as likely to have suffered from such infections.

Meatal Ulcer, the formation of ulcers at the 'pee-hole', is the one medical problem which is more common among circumcised males, though only in the first years of life. Normally all that is necessary is application of an ointment such as Vaseline to seal off and soothe the ulcer. If they are ignored ulcers can lead to narrowing of the urinary orifice. Properly treated, though, there will be no lasting ill effects. For this reason it is commonly recommended that circumcision should be carried out either in the first few days of life, when the urine is very dilute, or else not until the child has full bladder control. It has been claimed that narrowing of the orifice is normal in a circumcised penis, but a small survey I conducted found no sign of this - men with smaller meatuses were more often uncircumcised.


Cancer of the penis is virtually totally prevented by circumcision at a young age. Carcinoma of the penis is not invariably fatal - as with other cancers, if it is caught before it has spread the chances of survival are reasonable. However, the treatment involves amputating all or most of the penis, though recently radiation therapy has offered a more conservative approach. It is not a common disease in developed countries, accounting for only 12% of all cancer deaths in the US. (In Bali, on the other hand, it is the second most common cancer among men). An American study of 219 cases, spanning the period 1936 to 1975, found that none had been circumcised in infancy or childhood, though seven were found to have the disease when circumcised later in life. An Indian study of 84 patients seen in the 10 years 1967-1977 found that 82 were uncircumcised Hindus, and the remaining two were also not circumcised. A Canadian study covering all cases in one province over ten years found that, of the 37 cases diagnosed, 35 were in natural men, while the other two were in men who had been cut as adults. The astounding fact about the two North American studies was that, in countries where at least threequarters of the male population are circumcised, they didn't find a single case of penis cancer in men circumcised at birth or as children. Isolated cases have been reported, but they are so rare as to defy statistical analysis.

The Canadian survey found that one man in 125,000 will contract this cancer every year and one man in 290,000 will die from it each year. Since it is in fact only uncircumcised men who are dying, if we take it that only 15% of the men in Canada are left natural, and that the average lifespan is 70 years, we end up with the conclusion that one in every 600 uncircumcised North American males will die from cancer of the penis. This is probably unduly pessimistic, at least for Australia, since in 1990 there were only 7 cases recorded by the NSW Cancer Registry. With a male population of around 2.5 million, this converts to one man in 360,000 suffering each year - less than half the Canadian figure. Doing the same sums, on the basis that 30% of Australian men are not circumcised, we get a lifetime risk of contracting penile cancer of 1 in 1,500 for natural men, or better than 1 in 5,000 for all Australian males. The incidence in the largely uncircumcised Scandinavian countries is also around 1 in 1,500, so we can take that as the likely risk for a natural man in the developed world.

Cervical cancer is less common among Jewish women. Cancers of both the penis and the cervix are much more common among the uncircumcised Hindu inhabitants of Bali than among the ethnically similar Muslim inhabitants of the rest of Indonesia, who circumcise their boys. These and other similar studies have for a long time suggested that circumcision provides some protection against cervical cancer. Recently a major, multination study by Xavier Castellsague and his colleagues have provided strong confirmation of this point. They showed that the cause is not smegma, as had once been thought, but transmission of carcinogenic strains of papilloma (genital wart) virus. It follows from this that the risk is only present if the uncircumcised man has had sex with more than one woman - there is no increased risk for the wife of an uncircumcised man if she is the only sexual partner he's ever had.

Sexually transmitted diseases

Proponents of circumcision often claim that it reduces the incidence of traditional sexually transmitted diseases - syphilis, gonorrhoea, herpes, nongonoccal urethritis (NGU) and warts - but for many of these diseases there is little firm evidence relating to developed, Western nations. Looking at the medical literature, a Canadian team were claiming as recently as 1982 that it was 'accepted' that circumcision decreases the incidence of STDs. However, few studies back up this claim. One study appeared to imply that circumcised men were actually slightly more likely to suffer from NGU but the authors suggested that in fact the presence of a foreskin made the symptoms of a mild case less noticeable, so that such sufferers were less likely to seek treatment.

A recent Australian study, of 300 consecutive heterosexual patients at a Sydney STD clinic, found no signs of a connection. 62% of their sample were circumcised, and 38% natural, and gonorrhoea, NSU, herpes and warts all showed no correlation with foreskins. Syphilis and other diseases such as chancroid were too rare to quantify, and this may be the problem - these diseases are now, fortunately, rare in the Western world. There does seem to be firm evidence from African studies that chancroid - a common STD in developing countries, but rare in the West - is less likely to afflict circumcised men.

It is now quite clear that circumcision provides protection against infection by the human papilloma virus, which causes genital warts. A recent major study of 1913 couples, in 5 countries, showed a very clear connection. This also seems to be the link with both penile and cervical cancer, which commonly seem to be caused by 'rogue' strains of the normally benign wart virus.

The AIDS question

Recently the important discovery that circumcised men stand less risk of catching AIDS has received much publicity. Many studies have looked at this, and a major reanalysis of the data has shown that the link is undeniable. However the substance of this claim has not always been very clear, and this has led to widespread misunderstanding. What has been discovered refers only with heterosexual transmission, and only in one direction - circumcised men run less risk of catching the virus from intercourse with an infected woman. Nevertheless it is hard to underestimate the significance of this finding, since to date it provides the only demonstrably useful prophylactic medical intervention against the spread of HIV infection.

There has always been a hypothetical basis for this effect, since the AIDS virus cannot survive drying; the dry environment of a bare glans will give the virus less time to penetrate than the moist skin of a covered one. But the breakthrough has been the discovery of the site of this penetration. It had been widely thought that in uncircumcised men the skin of the glans has no keratin layer - the 'horny' layer that toughens and strengthens most skin. Keratinisation has been supposed to occur after circumcision. Szabo & Short showed that this is not so - there is an identical keratin layer on both circumcised and uncircumcised penises. The only penile skin which lacks keratin is the inner foreskin layer. Furthermore this skin is, unlike the rest of the penis, rich in Langerhans cells. These cells are normally part of the body's mechanisms for resisting infection, but the HIV virus tricks them into providing a portal for infection, and they seem to be the key factor in HIV penetration.

The major routes of HIV infection in developed, Western, nations are homosexual practices and needle sharing. So the major significance of this discovery will be in the developing nations, where most HIV infections are acquired in heterosexual intercourse.

Risks of the operation

Circumcision is a surgical operation, and like any other it can have complications. Haemorrhage (excessive bleeding) and infections can occur, and can be life threatening. The risk of death or serious trauma from circumcision, however, is extremely low. Harold Speert found ony 1 death in the half a million circumcisions carried out in New York in 10 years, and that was carried out by an unlicensed mohel (Jewish circumciser) not a doctor. Wiswell's more recent US studies, covering several hundred thousand infant boys, found no deaths or serious trauma resulting from circumcision. There seems in fact to have been only 1 documented death resulting from complications of circumcision during the 1980s in the USA.

In Australia there were only two deaths resulting from circumcision between 1960 and 1966. 120,000 boys were born per year in Australia during this period, and if we assume a circumcision rate of 70% this translates to one death per 294,000 circumcisions. In thirty years improvements in medical facilities must mean that circumcision in contemporary Australia is even safer, but it is hard to evaluate the risk when the number of deaths is so small.

Major hazards such as loss of the glans or the entire penis seem to be extremely rare, though some cases, such as the boy who received $22.8 million damages, have had wide publicity. No such cases occurred in either Speert's or Wiswell's large-scale studies, and so far as I can discover only 3 occurred in the entire USA during the 1980s. The risk, then, as with death, is almost too low to calculate. Furthermore, as with fatalities, most of the reported cases of major damage have occurred in religious circumcisions carried out by non-medical operators. The risk of death or major damage to the penis from cancer is at least two hundred thousand times greater than the risk from circumcision.

Less drastic trauma is not quite so rare. In Thomas Wiswell's study surgical trauma was reported in 25 cases out of 100,000. We are talking here about trauma sufficiently severe to require remedial action, and this translates to one boy in 4,000. Speert's figures were rather better, at one case in 10,000.

Then there is the risk of complications from the operation - infections, haemorrhages and so on. These are the most common circumcision problems, affecting up to one child in 1000 in Wiswell's study. In contrast to this, by logging all complications, however trivial, in a small sample of 100 circumcised infant boys, Patel found a much higher complication rate: "In these 100 infants, complications, usually minor, were very common, and included hemorrhage (35), meatal ulcers (31), infection (eight), phimosis (one) and meatal stenosis." However very few of Patel's complications involved any medical intervention. For example of his 35 haemorrhages, almost all were just a slight oozing which would not normally be regarded as a complication. Only 1 haemorrhage required intervention, only 1 of the meatal ulcers was reported as causing discomfort to the infant, only 1 infection required antibiotics. In fact there were only 3 cases that would have been recognised as complications in most studies. One was serious - the infection (in a child with a very incomplete circumcision) led to secondary phimosis, and the boy had to be re-circumcised. Even so, a 3% complication rate is higher than other studies have reported. Patel's sample was much smaller than the others, but it is clear that in some cases both the care taken over the operation and the follow-up care can be inadequate.

This includes care over the cosmetic outcome. In our surveys around 10% of the circumcised males reported objective faults with their circumcisions and a similar proportion of both men and women were unhappy about their sons' circumcisions. Women reported such problems in their lovers just as often (21% of the sample had, at some time in their life, had boyfriends who had been badly circumcised). For many people the risk of a badly done cut versus the risk of a problem foreskin will be more relevant factors than the likelihood of death from a relatively rare disease. Non-retractable foreskins are likely to affect about one in ten European males, but may be much more common among some Asian groups. This a similar figure to the number of reports of unsatisfactory circumcisions in my surveys, though these figures are not really comparable since men who were dissatisfied with their (or their son's) circumcision are more likely to respond to a survey.

Improving post-operative follow-up and accepting that cosmetic defects which will not affect function nevertheless demand remedial action are the key ways in which the medical profession could improve the acceptance of circumcision and make its medical benefits available to more of the population.

  From the medical point of view the benefits of circumcision hugely outweigh the risks. Nevertheless these benefits are not overwhelming - the serious problems, such as cancer of the penis, are rare and the common problems, such as phimosis and balanitis, are rarely serious.


Copyright (c) 2001