Austalian hospital data on UTIs and balanitis for males aged 20-29, a cohort that over the past two decades have gone from majority circumcised to more than 70% intact, show a falling incidence of these conditions. The only plausible explanation seems to be that the foreskin provides a protective effect for balanitis and UTI in adult men.
In a recent series of articles a small group of pro-circumcision activists, led by Professor Brian Morris, have pressed claims about the "protective effects" of circumcision, based on an idiosyncratic reading of the medical literature. The Australian experience is well positioned to evaluate these claims because of the rapid and well-documented fall in the circumcision rate, starting in the 1970s.
Specifically, the cohort of males in their 20s has seen a dramatic rise in the percentage of intact males from 38% in 1994 to 72% in 2010. If the Morris claims are correct, then the effects should be observable in this age group as a rise in incidence of these conditions.
Below I test the claims that circumcision provides both a 5-fold risk reduction in UTI and a 3-fold risk reduction in balanitis for adult males by analysing hospital data for males aged 20-29.
Materials and method
Incidence data was taken from hospital separation statistics in the National Hospital Morbitity Database via the online datacubes provided by the Australian Instute of Health and Welfare. The data series for UTI was obtained from the Australian-Refined Diagnostic-Related Group datacube for items L63 A, B and C (Kidney infection and UTI with catastrophic, severe, and without catastrophic/severe complications). The data series for balanoposthitis was taken from the Principal Diagnosis datacube for item N48.1. Population data is from the Australian Bureau of Statistics. Circumcision prevalence data was extracted from the Australian Studies of Health and Relationships research database and applied to the population data; these are estimates with a margin of error of about ± 3% and do not take account of demographic changes (specifically, the effect of net migration). In all cases, year refers to financial year ending June 30.
UTI and kidney infection
This data series runs from 1999 to 2010. Incidence for 1999 and 2000 was averaged, as was incidence for 2009 and 2010. Similarly, population totals and intact population subtotals were also averaged for the same two periods. The rate of UTI is expressed as cases per 100,000 persons.
For males aged 20-29 years, the number of cases from 1999-2000 to 2009-10 rose 9.3% (to 289), the cohort population rose 20.7% (to 1,669,795), so the rate fell from 19 to 17 per 100,000 males, a decline of 9.5%. The number of intact males rose by 80% (to 1,184,954), so the rate expressed as a proportion of the number of intact male fell 39.3%, from 40 to 24 cases per 100,000.
By way of comparison, the rate for females in the same age group, calculated the same way, rose 42% from 183 to 259 cases per 100,000.
This data series runs from 1994 to 2010. Incidence for 1994 to 1998 was averaged, as was incidence for 2006 to 2010. Similarly, population totals and intact population subtotals were also averaged for the same two periods.
For males aged 20-29 years, the number of diagnoses from 1994-1998 to 2006-10 fell 19% (to 51), the cohort population rose 11% (to 1,574,488), so the rate fell from 4.4 to 3.2 per 100,000 males, a decline of 27%. The number of intact males rose by 95% (to 1,136,930), so the rate expressed as a proportion of the number of intact males fell 59%, from 10.7 to 4.5 diagnoses per 100,000.
This Australian data not only decisively falsify the extravagent claims made by Morris and his pro-circ coterie, but indicate that for adult men the foreskin actually provides a protective effect against urinary tract infection and penile inflammation.
There are clearly many moving parts behind the incidence data for any particular condition, but the fact UTI in the equivalent female cohort rose dramatically while that for males fell, despite an 80% increase in the number of intact males in the same period, indicates that changing diagnostic or treatment standards are unlikely to be the issue.
It should be emphasised that this is Australian population data, ie what has actually happened here, not a projection based on assumption-laden and often flawed studies, conducted in Kenya or Kentucky, about what might happen here.
Similar investigation into other "health benefit" claims seems warranted.