This page indexes abstracts, full text articles, and other material about the relationship of circumcision status and the risk of contracting urinary tract infection (UTI).
Urinary tract infections (UTI) are a complication of circumcision. 1 38 40 63 They are bacterial infections of the urinary tract (kidneys, ureters, bladder and urethra). UTIs are usually associated with congenital abnormalities of the urinary tract.5 27 31 These infections can become serious if undetected, and may lead to permanent kidney damage. However, they are generally treated effectively with antibiotics.5,15
Studies have demonstrated that Escherichia coli (E. coli) bacteria, to which the infant has no passive immunity, can be colonized from the glans of circumcised infants and those infants whose foreskins have been forcibly retracted, tearing the protective synechia. Although E. coli is one of the most common bacteria on the surface of human skin, strains found in hospitals tend to be particularly virulent. In infant girls UTIs generally originate in the colon, whereas in infant boys they originate from the external environment, strongly suggesting that for boys such infections are iatrogenic.4 Premature retraction and circumcision expose boys to hospital strains of E. coli that can cause UTI.19
In order for the E. coli bacteria to bind to the glans penis of the infant, it needs an entry such as retraction of the adherent foreskin or circumcision.9 The intact boy has two physical lines of defense that his circumcised counterpart lacks: the preputial sphincter, which closes when a boy is not urinating; and a protected meatus (urinary opening), which is often inflamed and open in circumcised boys.9 In addition, the sub-preputial moisture contains lysosyme, which has an anti-bacterial action.47 Oligosaccharides excreted in the urine of breastfed babies prevent adhesion of pathogens to uroepithelial tissue.21
In the absence of a valid medical indication for circumcision, it would be unethical for physicians to perform it.29
The UTI Scare
Drs. Charles Ginsburg and George McCracken carried out a study of UTI in boys at Parkland Hospital in Dallas. They reported that 95 percent of the boys in their study were not circumcised.5 Parkland Hospital is a public hospital for indigent patients. Although it was not noticed at the time, the hospital did not perform non-therapeutic neonatal circumcisions,6 so most of its young male client population necessarily would have remained intact. The observation that 95 percent of the boys were not circumcised, therefore, signified nothing except that Parkland Hospital did not do unnecessary non-therapeutic circumcisions.
Thomas E. Wiswell, who was a U.S. Army pediatrician, read this study; he was impressed by Ginsburg and McCracken's erroneous observations about circumcision and UTI. In a determined search to show an association between lack-of-circumcision and UTI, Wiswell et al retrospectively examined charts of a number of boys born at U.S. military hospitals.7,11-13,17,25,39 Intact boys were reported to have a slightly higher rate of bacteriuria (bacteria in the urine) than circumcised boys during the first year of life.
Wiswell's sensational statistic, that circumcision resulted in a "ten to hundred times decrease in urinary tract infections in circumcised boys," has often been quoted; however, it is misleading. In fact, UTIs are so rare in any case that, using Wiswell's data, 50 to 100 healthy boys would have to be circumcised in order to prevent a UTI from developing in only one patient. (Using more recent data from a better-controlled study, the number of unnecessary operations needed to prevent one hospital admission for UTI would jump to 195.49)
Wiswell's findings generated a great deal of controversy at the time. The prepuce (foreskin) is a protective organ, and one would not normally expect the removal of a healthy organ to reduce the risk of infections.9 Nevertheless, the apparent correlation of intact foreskin to bacteriuria (and hence UTI) prompted the American Academy of Pediatrics (AAP) to review the evidence available in 1989.
Some points of fact:
Littlewood (1972) found an association of UTI with maternal infection, perinatal anoxia, and high or low birthweight.2 Wiswell failed to account for these confounding factors, although they should have been known to him.
As acknowledged in the AAP's 1989 report, "these studies in army hospitals are retrospective in design and may have methodologic flaws. For example, they do not include all boys born in any single cohort or those treated as outpatients, so the study population may have been influenced by selection bias." Furthermore, the babies in the studies were all sick and hospitalized, so do not represent infants in the general population.
In Wiswell's retrospective reviews of old hospital records, the parents of some of the intact infants may have been instructed to retract the foreskin and scrub beneath. The act of retraction involves forcible separation that destroys a structural defense mechanism. This would allow E. coli to invade where it could not before. The insertion of parents' fingers into the foreskin, or even the handling of the baby's penis, could introduce the bacteria into the preputial space of those uncircumcised infants.
The same criticism applies to the later Wiswell studies and other American studies of infant UTI [Herzog, Roscelli] to date: None have taken care to ensure that there was a control group of infants whose foreskins were simply left alone.
The Wiswell studies considered bacteriuria as diagnostic of UTI. However, a positive urine culture alone is not necessarily indicative of symptomatic UTI requiring treatment.31 There is a significant false-positive rate in diagnosing UTI when urine cultures alone are used.32,33 This criticism was addressed to some extent in Wiswell's second review.12
No information on rooming-in or breastfeeding history for the infants before they were hospitalized was recorded. Breastfeeding is preventative of UTI, and rooming-in with the mother may be (see below)19.
The hospital chart data used in the retrospective studies are unreliable. Hospitals frequently omit to record a circumcision on a baby's chart. In Atlanta, O'Brien found that circumcision was recorded only 84.3% of the time for circumcised boys.34 If the records used in the retrospective bacteriuria studies are similarly inaccurate, then a statistically significant number of the infants with bacteriuria that were claimed to be intact were, in fact, circumcised. This would naturally overstate the rate of infection in intact boys.35
It is very possible that the use of surgical antiseptic (to kill pathogenic organisms during the circumcision procedure itself) was in part responsible for the slight reduction in bacteriuria observed in these studies. This possibility was not accounted for in the studies.
Breastfeeding was determined in the 1990s to substantially reduce incidence of UTI.20-22,36 Therefore breastfeeding is a major confounding factor in any study of the role of circumcision in UTI. The Wiswell studies and all other studies in the literature fail to control for the effects of breastfeeding. The AAP observes that "breastfeeding status has not been evaluated sytematically in studies assessing UTI and circumcision status."52
A number of studies about post-circumcision UTIs, and the role of the surgery in possibly facilitating UTIs, were not able to recommend neonatal circumcision.23,24,26,37 Significantly, a number of recent Israeli studies have reported an increase in urinary tract infection rates in the period following ritual circumcision.10,38,40 62
In a prospective study, Kayaba et al. found a zero incidence of UTI in 603 intact boys, over a range of ages.41 Although this study did not focus on UTI, the Japanese researchers concluded: "Awareness of these findings will eliminate unnecessary circumcision in boys."
Moreover, studies show fairly conclusively that UTIs can be prevented far beyond the extent they are today. While hospital-borne strains of E. coli can attach to the glans and enter the urethra of circumcised boys and those whose foreskins have been retracted, infection can be resisted by certain measures to enhance the immunity of the infant to such pathogens:
Rooming-in permits colonization of the infant's skin and mucosa with the mother's own bacteria. The prepuce and other skin and mucosa of the infant should be specifically brought into contact with the mother's own skin to pass along her flora and initiate the child's natural immunity.3 19
Recurrent UTI. Some doctors recommend circumcision in cases of recurrent UTI. However, there is no medical evidence to support this recommendation. Recurrent UTIs are associated with congenital abnormalities of the upper urinary tract.5,15,28,32,35 McCracken recommends investigation with radiographic and/or sonography.5 In addition, a recent study of mice indicates that p-fimbriated Escherichia Coli, the organism responsible for about 85 percent of UTI, is capable of burrowing into the deeper tissue of the bladder48 or forming pods,56 thus hiding from antibiotics.48 Recurrent infections may actually be recurrence of the original infection, rather than a new infection ascending from the external genitals. If E. Coli behaves in a similar manner in humans, then circumcision would be of no benefit in preventing recurrence of these infections.
Congenital abnormalities of the urinary tract. The term vesicoureteral [ureterovesical] reflux refers to backflow of urine from the bladder to the ureters or kidneys. Ureteropelvic obstruction is a blockage or narrowing of part of the urinary tract. These kinds of congenital abnormalities are known to be the root cause of most UTI, as they may allow pathogens to flow upstream within the urinary tract. Obviously circumcision cannot prevent any such conditions. See Disorders of the Genitourinary Tract at Columbia University for more information.
Comparison of UTI incidence in boys and girls. A recent Scandinavian study found that girls and boys have about the same incidence of UTI in the first year of life. Girls have a four times higher incidence of UTI in the first six years of life than non-circumcised boys. No special concern is manifested regarding this much higher rate of infection in girls, yet surgery has not been proposed to reduce the incidence of UTI in girls.50
The National Kidney and Urologic Diseases Clearinghouse (NKUDIC), a service of the National Institutes of Health, maintains a comprehensive set of pages, listing the known causes of urinary tract infections in adults and children. NKUDIC does not suggest circumcision as a prevention or cure of urinary tract infection.
The American Academy of Pediatrics (AAP) has issued two statements which, when read together, constitute a substantial change in AAP policy toward the prevention of UTI in infants. First, in 1997, the AAP Workgroup on Breastfeeding recommended breastfeeding as highly beneficial in preventing a wide range of infections including UTI.45 Second, in 1999, The 1999 AAP Task Force on Circumcision abandoned the previous stance of the 1989 Task Force on Circumcision that circumcision may provide protection against UTI.53 The 1999 Task Force found that the bulk of the UTI studies were so methodologically flawed—by failing to control for confounding factors such as breastfeeding—that no meaningful conclusions could be drawn from them.53 The 1999 AAP Task Force on Circumcision could not, therefore, recommend circumcision to reduce incidence of UTI (or any other disease).
The 1999 AAP Task Force on Circumcision did, however, declare that breastfeeding produces a three fold reduction in UTI in infants. Two separate panels of the AAP, the Work Group on Breastfeeding and the 1999 Task Force on Circumcision, now recommend breastfeeding to reduce incidence of UTI.45,53 As Outerbridge points out, breastfeeding is very effective in reducing incidence of UTI in both boys and girls.46
The notion that circumcision is a useful prophylactic against disease has been laid to rest by the 1999 AAP Task Force on Circumcision.53 Instead, healthy, natural alternatives such as breastfeeding and rooming-in must be given favour. Breastfeeding offers a wide range of benefits for both mother and baby. Circumcision is surgery, and as such it has attendant risks. Furthermore, circumcision causes a great deal of pain, entails permanent loss of sexual function and sensation, which raises serious ethical questions concerning informed consent.32 Circumcision is not an operation to be performed lightly.
Holdings are listed in the chronological order of publication
back to Hygiene, Disease, and
back to Circumcision Reference Library
(Last revised: 2 February 2009)