Every claim, collapsed to a verdict you can expand
This page merges two things people ask for separately: what the research actually shows about circumcision (HIV, UTIs, cancer, sensation, complications), and how the country-level numbers — STI rates, meatal stenosis — hold up under scrutiny. Both are real, sourced, and easy to misread in isolation, so each claim below starts closed, showing only what it is and why it doesn't settle the argument. Expand any of them for the studies, the numbers, and the counter-studies, presented fairly on both sides.
Before the studies: the core argument doesn’t actually rest on any of this. Every claim below could break our way or against us and the underlying case — a healthy body part shouldn’t be removed from someone who can’t consent, without medical necessity — wouldn’t change. See the Bodily autonomy page for why. One number is still worth sitting with, though: where adult men are actually free to decide for themselves, they overwhelmingly don’t choose this. Denmark’s circumcision rate is about 0.4%.[38] Given the choice, at an age they can actually make it, that’s the answer people give. That alone should be close to the end of the discussion.
National medical bodies
Presented without editing out the ones that disagree with us.
| Body | Position | Recommends routine infant circumcision? |
|---|---|---|
| American Academy of Pediatrics (2012, expired 2017) | Benefits outweigh risks; justifies access for families who choose it[3] | No — explicitly stops short of recommending it |
| Canadian Paediatric Society (2015, reaffirmed 2024) | Does not recommend routine circumcision of every newborn male; where medical necessity is absent, defer to the individual[4] | No |
| WHO (2007–) | Recommends voluntary medical male circumcision as part of HIV prevention in specified high-prevalence settings[5] | Not universally — targeted, and framed as voluntary |
| Royal Dutch Medical Association (KNMG) (2010) | No convincing evidence of usefulness or necessity; regards non-therapeutic circumcision of minors as conflicting with autonomy and physical integrity; urges a policy of deterrence[6] | No — actively discourages |
| Danish Medical Association (2016) | Should only be performed with the individual's informed consent[7] | No |
| Nordic children's ombudsmen (2013) | Joint statement: non-therapeutic childhood circumcision violates fundamental medical-ethical principles[7] | No |
| Royal Australasian College of Physicians | Does not support routine infant circumcision[8] | No |
| British Medical Association | No medical consensus; emphasizes consent and the child's best interests[9] | No |
Note the shape of this table. The clearest institutional support is for voluntary circumcision of adults in specific epidemic conditions. Not one of these bodies recommends routine infant circumcision — and the American statement most often cited as pro-circumcision explicitly declines to recommend it.
Two things worth knowing first
How to read a risk reduction
Nearly every argument on this subject is won or lost on one distinction: a relative risk reduction ("50% lower!") and an absolute one (0.1 percentage points) can describe the exact same study. When a source gives you only the relative figure, that omission is itself information.
See the worked example
Suppose a condition affects 2 people in 1,000, and an intervention halves that to 1 in 1,000.
- The relative risk reduction is 50%. This is the number that appears in headlines.
- The absolute risk reduction is 0.1 percentage points. This is the number that describes your actual life.
- The number needed to treat is 1,000 — a thousand people undergo the intervention so one avoids the condition.
All three describe the same study. Both sides of this debate quote whichever one flatters them. When you see a percentage, ask which kind it is.
Two bodies of evidence that don't agree
Almost every argument on this page — and especially the HIV claim below — comes down to a conflict between two kinds of study: African randomized trials that found large protection, and developed-country national cohorts that later found none. Both are real; the difference is where and in whom they were measured.
Read the full framing
Randomized controlled trials were run in sub-Saharan Africa in the 2000s. They assign men at random, which removes confounding, and they found substantial protection against HIV. Randomization is why these are treated as the strongest form of evidence.
National cohort studies have since been run in developed countries — Denmark and Canada in particular — following hundreds of thousands of men through national health registries. They found no protective effect at all.
Both are real. Both are published in serious journals.
Where we think the resolution is
The African trials are probably internally valid — they measured what they measured, in the population they measured it in. What they are not is generalizable. A surgical intervention against a generalized heterosexual epidemic with high background prevalence does not automatically do anything useful in Copenhagen or Ontario, where the epidemiology, access to condoms, STD testing, the transmission routes and the baseline risk are all different.
The Danish and Canadian cohorts are the empirical test of that transfer. It failed.
Six common claims
Every card below opens the same way: what the claim is, and the short reason it doesn't settle the question of cutting a child who can't consent. Expand a card for the full studies, the numbers behind them, and — where one exists — the counter-study.
Claim 00 · The cost side
Risks and complications Not as rare as advertised
What it is: the reassurance that complications are rare enough not to weigh much in the decision — bleeding, infection, meatal stenosis, skin bridges, over- or under-removal, lost frenulum, and rarely, serious injury or death.
Why "rare" undersells it: studies that actually screen every boy, not just the ones who complain, find meatal stenosis alone in 7–18% of circumcised boys — a complication intact boys essentially never get — and a 2024 database of 1.7 million boys found penile problems at nearly three times the rate in circumcised infants.
How to read that honestly
Studies that wait for a boy to complain of symptoms will always find less than studies that physically examine every boy in the cohort, because a meaningful share of narrowing never gets reported unless someone looks for it. The safe, honest summary: meatal stenosis is a real, circumcision-specific complication, and studies that actually check for it consistently find it in a clear minority to a substantial share of circumcised boys. Boys circumcised in the first week of life were about twice as likely to develop it as those circumcised later[35] — useful for parents weighing timing, and it isn't disputed.
Meatal stenosis is a narrowing of the urethral opening. It causes a misdirected or hard-to-aim urine stream and, in more serious cases, painful or obstructed urination; correcting it usually means a minor surgical procedure (meatotomy). The relevant fact: it occurs almost exclusively in circumcised males. In intact boys it is rare to essentially absent, because the foreskin protects the glans and meatus from the chronic irritation thought to drive it.[1]
Beyond meatal stenosis, every intervention has a cost side, and it is frequently summarised as "rare" without a number attached.
- Bleeding and infection — the common early complications. Most are minor and manageable.
- Meatal stenosis — see above. The AAP's own 1984 parent brochure said it plainly — meatitis and meatal stenosis of the exposed glans "virtually never occur" in intact males — before that line was quietly dropped from later editions.[1]
- Adhesions and skin bridges — tissue reattaching abnormally; sometimes requires a second procedure.
- Too much or too little skin removed — leading to painful tightness on erection, or an unsatisfactory result. Revision rates are not trivial.
- Loss of the frenulum — often not discussed as a complication because it is often not discussed at all.
- Serious injury and death — rare, and genuinely rare, but documented. You will see specific annual death-toll figures circulating in advocacy material; those are modeled estimates rather than counts, and we don't repeat them here.
Overall complication rates for newborn circumcision in well-resourced settings are usually reported in the range of a fraction of a percent to a couple of percent for minor complications. Rates rise for procedures done later in childhood, and rise sharply in settings without surgical infrastructure.
A 2024 analysis of roughly 1.7 million boys in a US commercial claims database found penile problems recorded in the first five years of life at nearly three times the rate in circumcised boys (1.7%) as in intact ones (0.5%) — and the rate rose further when a surgeon rather than a pediatrician performed the procedure.[2] That cuts directly against the popular claim that circumcision leaves boys with fewer penile problems, not more.
The comparison that matters: a healthy foreskin has a complication rate of zero, because it is not a procedure. For the open question this section can't yet answer — what "too much skin removed" or "botched" even means without an agreed metric — see Open questions.
Claim 01
HIV Depends entirely on where
What it is: three African randomized trials found circumcision cut female-to-male HIV transmission by roughly half — the strongest single card in the pro-circumcision hand.
Why it doesn't justify cutting an infant: when Denmark (810,719 men) and Ontario ran the same test in their own developed-country populations, decades later, both found no protective effect at all — and even in the original trials, the absolute reduction behind that "roughly half" was only about 1.3 percentage points.
Orange Farm (Auvert et al., 2005)
Roughly 60% relative reduction in female-to-male HIV acquisition. Stopped early for efficacy.[10]
Kisumu (Bailey et al., 2007)
Roughly 53–60% relative reduction. Published in The Lancet. Also stopped early.[11]
Rakai (Gray et al., 2007)
Roughly 51–55% relative reduction. Published alongside Kisumu.[12]
What the headline numbers leave out
These are the strongest studies in the debate and they are still not clean. The critiques below are drawn largely from Boyle and Hill's peer-reviewed review in the Journal of Law and Medicine.[13]
The absolute reduction was about 1.3 percentage points. Every figure above is relative. Across the three trials the absolute risk reduction over roughly two years was in the region of 1.3%. By the standard set out in the risk-literacy section at the top of this page, quoting only the relative figure is exactly the move we told you to distrust — including when we do it.
All three stopped early. Trials halted early for benefit systematically overestimate effect size; this is a documented statistical phenomenon, not an accusation.[14] After correction for lead-time bias, critics put the pooled relative reduction nearer 49% than 60%.[13]
Follow-up was around two years. Nobody has shown the effect persists across a sexual lifetime, which is the timeframe an infant circumcision decision is actually being made over.
Blinding was impossible and men in the circumcision arms received additional safe-sex counselling — a co-intervention that was not matched in the control arms.
The healing period is an unmatched exposure gap. Circumcised men were instructed to abstain for roughly six weeks. The arms therefore differed in sexual exposure time, not only in foreskin status.
Substantial loss to follow-up, with critics arguing the dropout patterns could move the result.
Selection. Participants volunteered for a circumcision trial in cultures where the practice carries social meaning.
Wawer et al., 2009 — the one that gets left out
A fourth randomized trial in Rakai tested whether circumcising HIV-positive men reduced transmission to their female partners. It was stopped for futility. At 24 months, cumulative HIV infection among female partners was 21.7% in the circumcision group versus 13.4% in the control group (adjusted HR 1.49, 95% CI 0.62–3.57, p=0.368).[15]
Read that carefully in both directions. The result was not statistically significant — the confidence interval crosses 1, the sample was small, and it does not establish that circumcision increases risk to women. But the point estimate ran the wrong way, and the trial was terminated because further data was unlikely to show benefit. Concern focused partly on couples resuming sex before the six-week healing period was complete. This trial is part of the same research program as the three above and is cited far less often.
None of this makes the trials worthless. Three randomized trials converging is still the most substantial evidence anyone has produced here, and a site that dismissed them would deserve to be ignored. But "gold standard" is a description of study design, not a certificate of perfection, and these particular trials carry documented problems that are usually omitted when the 60% figure gets quoted.
Frisch & Simonsen, 2021
810,719 non-Muslim Danish males born 1977–2003, followed across 17.7 million person-years. Non-therapeutic circumcision showed no statistically significant reduction for any STI, including HIV. Overall STI rate was in fact 53% higher in circumcised men (HR 1.53, 95% CI 1.24–1.89).[16]
Weaknesses, stated plainly: only 3,375 men (0.42%) were circumcised, so power is limited. The elevated overall STI figure is driven by anogenital warts; the syphilis signal rests on just four cases. Critics argue the analysis doesn't account for sexual orientation, which matters because circumcision status cannot affect risk for the receptive partner in anal sex.[17] The defensible claim here is no protective effect found — not that circumcision increases STIs.
Nayan et al., 2022
Population-based matched cohort of males born in Ontario and circumcised at any age between 1991 and 2017. Circumcision was not independently associated with HIV acquisition, and no association appeared in any sensitivity analysis. Published in The Journal of Urology.[18]
Worth noting who wrote this: urologists at the University of Toronto and Mount Sinai, publishing in a mainstream urology journal. This is not advocacy research, which makes it considerably harder to dismiss.
Cochrane review (Wiysonge et al.)
21 observational studies, 71,693 participants. No reduction in HIV acquisition risk among men who have sex with men.[19]
The comparison nobody makes
The United States has both a high circumcision rate and a higher HIV prevalence than Western European countries where circumcision is rare. Ecological comparisons prove nothing on their own — too many variables — but if circumcision were doing meaningful population-level work in a developed country, this is not the pattern you would expect.
Our read
The trials happened, the effect was real, and it was measured in consenting adults during a generalized heterosexual epidemic. Two large developed-country cohorts then looked for the same effect in their own populations and did not find it.
That is not a contradiction requiring one side to be wrong. It is what a context-dependent effect looks like. And it means the HIV argument — the single strongest card in the pro-circumcision hand — does not support operating on infants in Denmark, Canada, or the United States.
One further asymmetry deserves stating, because it usually gets buried under the word "randomized". The trials have higher internal validity: randomization removes confounding. But the cohort studies have higher construct validity for the question actually being argued about. The trials measured voluntary adult surgery, over two years, in a generalized heterosexual epidemic. Denmark and Ontario measured infant circumcision, across decades, in developed countries — which is the intervention and the setting under debate.
Claim 02
Urinary tract infections Real, small, treatable
What it is: circumcised infant boys get fewer UTIs — the most solidly established benefit in the literature, and the only one on this page for which waiting genuinely forfeits it.
Why it doesn't justify cutting: it takes roughly 111 circumcisions to prevent one UTI, a condition normally cleared with a course of antibiotics.
Circumcised infant boys get fewer UTIs. This is consistent across studies and we accept it without argument — it is the most solidly established benefit in the literature.
Weaknesses: the evidence base here is largely observational rather than randomized, so it carries the confounding problems that randomization exists to remove. The meta-analysis pooling it graded the underlying studies as generally low quality.[20] The direction of effect is consistent enough that we accept it, but the precise magnitude is softer than the confident numbers suggest.
The scale is the whole argument. Meta-analysis puts the number needed to treat at roughly 111 circumcisions to prevent one UTI in a boy at normal risk.[20]
So: 111 surgeries, some fraction of which will have complications, to prevent one infection that is in most cases treated with a course of antibiotics. Reverse the framing and ask whether you would accept a drug with that profile.
Note also that this is an argument for infant circumcision specifically — the UTI risk window is infancy. Unlike the HIV argument, waiting for consent genuinely does forfeit this benefit. It is the only benefit on this page for which that is true, which is precisely why it deserves an honest hearing rather than dismissal.
One more piece of context, in both directions: in the first few months of life, uncircumcised boys' UTI risk actually runs slightly above girls' — that elevated early window is the whole reason this benefit exists. Past infancy the pattern reverses hard, and UTIs become markedly more common in girls than in boys of any circumcision status for the rest of childhood. Nobody proposes surgery on girls to close that much larger gap.
Claim 03
Penile cancer True but nearly irrelevant
What it is: circumcision does appear to further reduce the risk of an already-rare cancer — on the order of one case per 100,000 men per year in developed countries.
Why it doesn't justify cutting: numbers needed to treat start around 900 and run far higher — hundreds to thousands of infant surgeries to prevent one case, decades later.
Penile cancer is rare — on the order of one case per 100,000 men per year in developed countries. Circumcision does appear to reduce the risk further.
Numbers needed to treat in the literature start around 900 and run very much higher depending on assumptions and baseline rates.[21] To prevent one case of a rare cancer late in life, you operate on hundreds or thousands of infants.
Breast cancer is dramatically more common than penile cancer, and prophylactic mastectomy is dramatically more effective at preventing it than circumcision is at preventing penile cancer. We do not perform it on infant girls. The reasoning that rules that out is the reasoning we are applying here.
Claim 04
HPV, chlamydia & other STIs Contradictory by setting
What it is: trial data found reduced HPV and HSV-2 acquisition alongside the HIV results, and advocates cite country-level circumcision rates against national STI rates as further proof.
Why it doesn't hold up: a national cohort of 810,719 Danish men found no significant reduction in any of the eight STIs tracked, over 17.7 million person-years — and the country-level charts cited instead are misleading anyway: Denmark's circumcision rate is 0.4% with one of Europe's highest reported chlamydia rates, purely because it tests far more. An HPV vaccine also already outperforms circumcision without removing anything.
The one rule: a rate measures testing as much as infection
A country's reported STI rate reflects how many people are infected and how hard the country looks. A place that screens aggressively finds and reports more cases; a place that barely tests reports few, no matter the true burden. So a high reported rate can mean "lots of infection" or "excellent surveillance" — and you usually can't tell which from the number alone. Anyone who hands you a country league table as proof of anything is skipping this step. We won't.
The Denmark paradox — why this chart proves nothing on its own
Denmark has a circumcision rate near 0.4% and one of the highest reported chlamydia rates in Europe — higher than the heavily-circumcised United States. Taken at face value, this chart would "show" that being intact causes chlamydia. That conclusion is false, and seeing why is the whole point.
The ECDC states it plainly: a handful of Nordic countries plus the UK account for the overwhelming majority of Europe's reported chlamydia cases because they run national screening programs and test far more than everyone else.[37] Denmark's number is high because Denmark looks hard, not because Danes are unusually infected. The chart is largely a map of testing intensity wearing the costume of an infection map.
So what can you take from cross-country STI data? Only the modest, defensible thing:
- There is no visible population-level STI advantage for high-circumcision countries. If routine infant circumcision meaningfully protected against STIs in the real world, the heavily-circumcised US would sit far below intact Europe. It doesn't — it sits right in the middle of the European pack, and below several intact countries.
- That absence of a visible benefit is a fair observation, because it cuts against the strong "circumcision prevents STIs" claim without needing the numbers to be perfectly clean.
- What you cannot claim is the reverse — that intactness raises STI risk — because the same testing confound that protects us from the first error would produce exactly this artifact.
Trial data from the African cohorts found reductions in HPV and HSV-2 acquisition alongside the HIV findings.[22] Syphilis results were less consistent, and no protective effect is established for gonorrhea or chlamydia.
These are the same three trials as the HIV claim, so they carry the same strength — and every weakness listed under Claim 01: early stopping, roughly two years of follow-up, unblinded design, the unmatched healing period, and loss to follow-up.[13] The HPV and HSV-2 findings were also secondary outcomes rather than what the trials were designed and powered to measure, which weakens them further.
The Danish national cohort found no significant reduction for any of the eight STIs evaluated, and a higher overall rate driven by anogenital warts.[16] See the caveats under Claim 01 — that finding is weaker than it first appears, but "no protection found" survives them. Country-level STI comparisons, above, point the same way at the population level.
And the argument that actually settles this one: HPV vaccination exists. It is far more effective against HPV than circumcision, it protects against the strains that cause cancer, it works for both sexes, and it removes nothing. When a non-surgical option outperforms a surgical one on the same indication, the surgical one has lost the argument on its own terms.
Claim 05
Sensation and sexual function Genuinely unresolved
What it is: mapping fine-touch pressure thresholds across 19 penile sites, Sorrells et al. (2007) found the foreskin measures as the single most touch-sensitive part of the penis — the tissue removed by circumcision.
Why it doesn't settle anything: whether that translates into lived sexual experience is genuinely unresolved — studies on both sides are small, self-reported, or confounded, which on its own means no benefit large enough to justify cutting an infant has been established either.
This is the messiest literature on the page, and the place where advocacy on both sides is least careful. We are flagging study weaknesses in both tabs, including in the studies that favor our position.
Sorrells et al., 2007
Mapped fine-touch pressure thresholds across 19 penile sites. The most sensitive regions measured were on the foreskin. The authors' conclusion: circumcision removes the most sensitive regions of the penis.[23]
Taylor et al., 1996
The original description of the ridged band and its concentration of Meissner's corpuscles — the anatomical basis for expecting a sensory difference at all.[24]
Frisch, Lindholm & Grønbæk, 2011
Danish national health survey, 5,552 participants. Circumcised men reported frequent orgasm difficulties more often than intact men; female partners of circumcised men reported more sexual difficulties, including pain during intercourse.[25]
Cross-sectional and self-reported, and only about 5% of Danish men are circumcised, so the exposed group is small.
Bronselaer et al., 2013
1,369 men (310 circumcised) completed a self-assessment instrument covering sensitivity, pleasure, orgasm intensity, numbness and discomfort. Circumcised men reported lower sensitivity and more unusual sensations across a number of measures.[26]
This one has a real problem and we are not going to hide it. The study ran 42 comparisons without correcting for multiple testing, which substantially inflates the chance of false positives — with that many tests, some "significant" results are expected by chance alone. Recruitment was at railway stations, and the work originated as a student thesis. Treat it as suggestive at best.
Kim & Pang, 2007
Survey of men circumcised as adults, reporting decreased masturbatory pleasure after the procedure.[27]
Small, self-reported, and retrospective. Useful mainly because adult-circumcision studies avoid the comparison problem below — these men actually experienced both states.
Bossio, Pukall & Steele, 2016
The most methodologically careful study in this area. Tested touch, pain, warmth and heat-pain thresholds in 30 circumcised and 32 intact men, demographically matched. Concluded that neonatal circumcision has minimal long-term implications for penile sensitivity.[28]
The complication: the same paper's data found the foreskin significantly more sensitive to touch than every other genital site tested, and more sensitive to warmth than the glans. Commentators including Brian Earp have pointed out the tension between that measurement and the paper's headline conclusion.[29] Both things are in the same study.
Sexual function in the RCT cohorts
Follow-up studies within the Rakai and Kisumu trials reported no adverse effect on sexual satisfaction or function, with some men reporting improvement.[30] Because these follow randomized cohorts, they avoid the self-selection problem that affects most surveys.
These men chose to enrol in a circumcision trial, were circumcised as adults in a culture where the practice carries positive social meaning, and were surveyed by the teams running the trial. Expectation effects are difficult to rule out.
Our read
Nobody has won this outright. The measurement studies suggest the removed tissue is the most touch-sensitive part of the organ; the studies that translate that into lived sexual experience are, on both sides, small, self-reported, or confounded.
Three problems make it structurally hard to resolve: men circumcised as infants have no comparison to report; adult-circumcision studies recruit men who chose the surgery, usually to fix a problem; and fine-touch thresholds measured in a flaccid state may simply not predict sexual experience.
The conflict of interest nobody declares
When you read a medical opinion on circumcision, it helps to know something about how such opinions are formed — and by whom. This is not a claim that any individual doctor is corrupt. It is a claim about incentives, which shape honest people without their noticing.
Three quiet pressures
Consider the physician most likely to be quoted as an expert on infant circumcision. He has probably performed hundreds or thousands of them. He is very likely circumcised himself. He may have circumcised his own sons. Now ask what it would cost him — not financially, but internally — to conclude that the procedure is ethically indefensible.
- Self-justification. To decide the operation is wrong is to accept that a great deal of one's own professional life was spent removing healthy tissue from infants who could not consent. That is an enormous thing to ask a person to conclude about themselves, and the mind is very good at not concluding it. The more procedures performed, the higher the wall.
- The body in the mirror. Concluding the foreskin has real value is, for a circumcised doctor, also a conclusion about his own body and his own parents' decision. That is precisely the ingratitude frame from the rhetoric page, operating on the expert instead of the layperson — and experts are not immune to it.
- Money, at the level that actually matters. A single newborn circumcision earns a physician a modest fee — roughly $250 to $600 privately, often much less through Medicaid.[31] No individual doctor is getting rich per case. But most physician compensation now runs on productivity models that pay per unit of work performed, so volume is the variable, not the single fee.[32] And the incentive is visible at the institutional level too: the American Academy of Pediatrics has actively lobbied Medicare to increase the work value assigned to the circumcision billing code.[33] That is a documented financial interest held by the professional body itself, not a claim about anyone's private motives.
Why "most experienced" can mean "most invested"
Here is the part that turns an ordinary observation into a real problem. Medical literature and expert panels lean, quite reasonably, on the most experienced practitioners. For nearly every procedure, that is exactly who you want: the surgeon who has done a thousand gallbladders knows things the textbook doesn't.
But that selection rule assumes the expertise in question is technical — how to do it well, safely, with the fewest complications. On the ethical question circumcision actually raises, the most experienced practitioner is also the person with the most invested in the answer being "yes." He is the last person likely to entertain the possibility that the whole category is a mistake, because he has the most to lose by entertaining it. On the technical question his experience is an asset. On the ethical one it is a conflict.
The distinction that carries the argument
For almost any other operation this conflict is defused by two facts: the patient chose the procedure and needed it. The surgeon's enthusiasm is checked by a consenting adult weighing their own risks and benefits. Circumcision of an infant removes both checks at once. There is no consent, and in the ordinary case there is no medical necessity — so the practitioner's investment in the procedure runs unopposed by the one person whose body it is. The safeguards that make "trust the experienced expert" reasonable everywhere else are exactly the safeguards that circumcision is missing.
How to use this — and how not to
This is a reason to read expert claims carefully, not to dismiss them. A circumcised urologist who has done a thousand procedures can still be completely right about the HIV data; conflicts of interest bias, they do not automatically falsify. The correct response is not "so ignore the doctors" — that would be the appeal-to-authority tactic run in reverse.
The correct response is to notice who is unusually well positioned to be wrong in one specific direction, and to weight the sources who don't carry that conflict accordingly — the national committees that reviewed the same evidence and declined to recommend the procedure, the researchers in countries where neither the doctors nor the reviewers were themselves circumcised, and the clinicians who perform the operation and oppose it anyway, since they are the ones arguing against their own incentive.
The key figures in one place
| Figure | Value | Source & caveat |
|---|---|---|
| Meatal stenosis, circumcised — on screening | 7–18% | Van Howe; Acimi; includes asymptomatic[34][35] |
| Meatal stenosis, intact males | rare / near-absent | The comparison that matters[1] |
| Penile problems, first 5 years (circ. vs. intact) | 1.7% vs. 0.5% | 2024 US claims database, ~1.7M boys[2] |
| HIV, Denmark & Ontario cohorts — the Western data | No effect | 810,719 Danish men; Ontario registry; decades, developed countries[16][18] |
| UTI, number needed to treat | ~111 | Normal-risk boys; meta-analysis[20] |
| Penile cancer, number needed to treat | 900+ | Rises further with baseline assumptions[21] |
| Fine-touch threshold, foreskin vs. glans | ~5–10× lower (more sensitive) | Sorrells et al. 2007; most sensitive foreskin site needed roughly a tenth the pressure of the glans to detect touch — the ridged band specifically, about a fifth[23] |
| Global male circumcision prevalence | ~37–39% | WHO; mostly religious/cultural[38] |
Rates move year to year and reporting methods differ between agencies and study designs, so treat cross-source comparisons as approximate. Where a precise, current figure matters, go to the primary source — links on the Sources page.
What all of this does and doesn't establish
- Every claimed benefit is real in some form and small in absolute terms. UTIs and HIV in a specific epidemic are the strongest; penile cancer and general population-level STI protection are weak to nonexistent as reasons to operate on an infant.
- The risk side is genuinely under-measured, not merely small. Meatal stenosis alone runs far above the "rare" framing once anyone actually looks for it.
- Country-level STI data shows no real-world population benefit from circumcision, and can't show much more than that in either direction — the Denmark paradox is the proof.
- Sensation is the one claim where nobody has won, which by itself means it can't carry the argument for cutting.
That's a narrower set of conclusions than an advocacy site might want on either side. It's also the set the data can actually carry, which is the only kind worth putting your name to. This discrepancy shows these arguments are not clear and while you may come to your own conclusions, there is no guarantee your son will come to the same ones.
Sources cited on this page
- American Academy of Pediatrics, Care of the Uncircumcised Penis parent brochure — comparison of the 1984 edition (which described the foreskin's protective function and stated that irritation, meatitis and meatal stenosis of the exposed glans "virtually never occur" in intact males) with later editions from which that passage was removed. Documented, with copies of the correspondence in which AAP staff stated they were unable to determine why the passage was deleted, by the Circumcision Resource Center (circumcision.org, "AAP Pamphlet: Care of the Uncircumcised Penis Included Foreskin Functions"). Primary correspondence dated 1996;
- Fendereski K, Horns JJ, Driggs N, Lau G, Schaeffer AJ. "Comparing Penile Problems in Circumcised vs. Uncircumcised Boys: Insights From a Large Commercial Claims Database With a Focus on Provider Type Performing Circumcision." Journal of Pediatric Surgery, 2024;59(11). PMID 39084960. Roughly 850,000 circumcised boys matched against 850,000 uncircumcised controls in a US commercial claims database. Penile problems within the first five years of life: 1.7% of circumcised boys versus 0.5% of uncircumcised boys — a 2.9-fold higher risk (95% CI 2.8–3.0, p<0.001). Boys circumcised by surgeons had a 2.1-fold higher rate of problems in the following year than those circumcised by pediatricians (95% CI 2.0–2.3, p<0.001). Procedure-related complications within 28 days were less common (0.5%), most often penile edema.
- American Academy of Pediatrics Task Force on Circumcision. "Circumcision Policy Statement." Pediatrics, 2012;130(3):585–586, with an accompanying technical report. Concluded that health benefits outweigh risks and justify access for families who choose it, while explicitly declining to recommend routine circumcision for all newborn males.
- Canadian Paediatric Society. "Newborn male circumcision." Position statement, 2015, reaffirmed 2024. Does not recommend routine circumcision of every newborn male; where medical necessity is not established, interventions should be deferred until the individual can choose.
- WHO / UNAIDS. "New data on male circumcision and HIV prevention: policy and programme implications." Montreux consultation, March 2007, and subsequent guidance on voluntary medical male circumcision (VMMC) in high-prevalence settings.
- Royal Dutch Medical Association (KNMG). "Non-therapeutic circumcision of male minors." Position statement, 27 May 2010, endorsed by Dutch associations of paediatricians, urologists and paediatric surgeons. Finds no convincing evidence of usefulness or necessity; regards the practice as conflicting with the child's autonomy and physical integrity; calls for a policy of deterrence.
- Danish Medical Association statement (2016) that circumcision should only be performed with informed consent; and the 2013 joint statement of the Nordic children's ombudsmen that non-therapeutic childhood circumcision violates fundamental medical-ethical principles.
- Royal Australasian College of Physicians. "Circumcision of Infant Males." Position statement. Does not support routine infant circumcision.
- British Medical Association. Guidance on non-therapeutic male circumcision. Notes the absence of medical consensus and emphasises consent and the child's best interests.
- Auvert B, Taljaard D, Lagarde E, et al. "Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 trial." PLoS Medicine, 2005;2(11):e298. Orange Farm, South Africa.
- Bailey RC, Moses S, Parker CB, et al. "Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial." The Lancet, 2007;369(9562):643–656.
- Gray RH, Kigozi G, Serwadda D, et al. "Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial." The Lancet, 2007;369(9562):657–666.
- Boyle GJ, Hill G. "Sub-Saharan African randomised clinical trials into male circumcision and HIV transmission: methodological, ethical and legal concerns." Journal of Law and Medicine, 2011;19(2):316–334. PMID 22320006. Peer-reviewed critique arguing inadequate equipoise, selection bias, inadequate blinding, problematic randomisation, early stopping with exaggerated treatment effects, high loss to follow-up, and failure to investigate non-sexual transmission. Also the source for the ~1.3% absolute risk reduction figure and the lead-time-corrected pooled estimate of ~49%.
- On trials stopped early for benefit systematically overestimating treatment effects — see Bassler D, Briel M, Montori VM, et al. "Stopping randomized trials early for benefit and estimation of treatment effects: systematic review and meta-regression analysis." JAMA, 2010;303(12):1180–1187. A general methodological finding, not specific to circumcision.
- Wawer MJ, Makumbi F, Kigozi G, et al. "Circumcision in HIV-infected men and its effect on HIV transmission to female partners in Rakai, Uganda: a randomised controlled trial." The Lancet, 2009;374(9685):229–237. DOI 10.1016/S0140-6736(09)60998-3. 922 men randomised; trial stopped early for futility. Cumulative female HIV infection at 24 months 21.7% (intervention) vs 13.4% (control); adjusted HR 1.49, 95% CI 0.62–3.57, p=0.368 — not statistically significant. See also the accompanying Lancet comment by Baeten et al. and subsequent correspondence.
- Frisch M, Simonsen J. "Non-therapeutic male circumcision in infancy or childhood and risk of human immunodeficiency virus and other sexually transmitted infections: national cohort study in Denmark." European Journal of Epidemiology, 2022;37(3):251–259. DOI 10.1007/s10654-021-00809-6 (published online 26 September 2021). 810,719 non-Muslim Danish males born 1977–2003; 17.7 million person-years; 3,375 (0.42%) non-therapeutically circumcised. No statistically significant reduction for any STI; overall STI rate 53% higher in circumcised males (HR 1.53, 95% CI 1.24–1.89), driven by anogenital warts (HR 1.51) and syphilis (HR 3.32, but only four cases).
- Meyrowitsch DW, Andersen LS, Lohse N, von Euler-Chelpin M, Hankins C. Letter re Frisch & Simonsen. European Journal of Epidemiology, 2022;37:651–652. Argues the Danish analysis does not account for sexual orientation. See also the authors' reply, same volume, 653–654.
- Nayan M, Hamilton RJ, Juurlink DN, Austin PC, Jarvi KA. "Circumcision and Risk of HIV among Males from Ontario, Canada." The Journal of Urology, 2022;207(2):424–430. DOI 10.1097/JU.0000000000002234. Population-based matched cohort; circumcision not independently associated with HIV acquisition, with no association in any sensitivity analysis.
- Wiysonge CS, Kongnyuy EJ, Shey M, et al. Cochrane systematic review of male circumcision for prevention of HIV acquisition in men who have sex with men. 21 observational studies, 71,693 participants; no significant reduction in acquisition risk. Confirm the current version and citation details before linking.
- Singh-Grewal D, Macdessi J, Craig J. "Circumcision for the prevention of urinary tract infection in boys: a systematic review of randomised trials and observational studies." Archives of Disease in Childhood, 2005;90(8):853–858. Reports a number needed to treat of approximately 111 in boys at normal risk, falling sharply in high-risk groups.
- Numbers needed to treat for penile cancer prevention vary enormously with assumptions and baseline incidence; published figures start in the high hundreds and run far higher. Cite a specific analysis rather than a round number.
- HPV and HSV-2 findings derive largely from the same African trial cohorts as [10]–[12]; results for syphilis have been less consistent, and no protective effect is established for gonorrhoea or chlamydia.
- Sorrells ML, Snyder JL, Reiss MD, et al. "Fine-touch pressure thresholds in the adult penis." BJU International, 2007;99(4):864–869. Semmes-Weinstein monofilament testing across 19 sites (lower gram-force threshold = more sensitive). The most sensitive site measured was the rim of the preputial orifice (mean 0.093g, the only foreskin-vs-scar comparison reaching significance in the paper's own table, p=0.03); the ridged band measured 0.205g. The glans, combining both measured sites, worked out to roughly 1g in an adjusted model comparing circumcised and intact men (0.161g difference, p=0.04) — the paper's own stated conclusion is that "the transitional region from the external to the internal prepuce is the most sensitive region of the uncircumcised penis and more sensitive than the most sensitive region of the circumcised penis." Caveats worth stating plainly: the examining physician could not be blinded to circumcision status, some individual foreskin-site comparisons in the paper's own table don't reach significance despite being described as significant in the text, and a couple of the lowest-n site averages carry wide error margins — treat the site's own summary sentence, quoted above, as the safest claim to lean on.
- Taylor JR, Lockwood AP, Taylor AJ. "The prepuce: specialized mucosa of the penis and its loss to circumcision." British Journal of Urology, 1996;77(2):291–295. The original description of the ridged band and its Meissner's corpuscles. Note that the functional significance of this structure is contested — see [29].
- Frisch M, Lindholm M, Grønbæk M. "Male circumcision and sexual function in men and women: a survey-based, cross-sectional study in Denmark." International Journal of Epidemiology, 2011;40(5):1367–1381. DOI 10.1093/ije/dyr104. n=5,552.
- Bronselaer GA, Schober JM, Meyer-Bahlburg HFL, et al. "Male circumcision decreases penile sensitivity as measured in a large cohort." BJU International, 2013;111(5):820–827. Note the methodological caveat given at the point of use: 42 comparisons without correction for multiple testing.
- Kim D, Pang MG. "The effect of male circumcision on sexuality." BJU International, 2007;99(3):619–622. South Korean survey of men circumcised as adults.
- Bossio JA, Pukall CF, Steele SS. "Examining Penile Sensitivity in Neonatally Circumcised and Intact Men Using Quantitative Sensory Testing." The Journal of Urology, 2016;195(6):1848–1853. Concluded minimal long-term effect on sensitivity; the same paper reports the foreskin as significantly more touch-sensitive than other genital sites tested.
- Earp BD. "Infant circumcision and adult penile sensitivity: implications for sexual experience." Trends in Urology & Men's Health, 2016;7(4):17–21. Commentary on the tension between Bossio et al.'s data and their stated conclusion.
- Sexual function follow-up within the African randomised trials — see Kigozi G, et al. (Rakai; BJU International, 2008;101(1):65–70) and Krieger JN, et al. (Kisumu). Both report no adverse effect on sexual satisfaction or function.
- Physician fee data for newborn circumcision. Commonly published ranges put the professional (physician) fee at roughly $250–600 for a hospital newborn circumcision, with facility fees billed separately and Medicaid reimbursement often substantially lower and varying by state. These are illustrative market figures rather than a single authoritative schedule; confirm current ranges before citing specific numbers.
- On work-RVU (wRVU) productivity-based physician compensation generally. Under these widely used models, physicians are paid a dollar conversion factor multiplied by the total work RVUs they generate, so total compensation scales with procedure volume. CPT 54150 (circumcision using clamp or other device with dorsal penile or ring block) carries on the order of 2.8 work RVUs. See the CMS Physician Fee Schedule and standard RVU references.
- American Academy of Pediatrics. "Revisions made to circumcision code, valuation." AAP News, 2007;28(2):26. Reports the Academy's successful advocacy to obtain increased physician work RVUs for the circumcision code (CPT 54150) and to fold the nerve block into the code descriptor. Cited as evidence of an institutional-level financial interest; note that in the same revision total RVUs for the code fell relative to the prior year, so the effect is not a simple pay increase — represent it accurately.
- Van Howe RS. "Incidence of meatal stenosis following neonatal circumcision in a primary care setting." Clinical Pediatrics, 2006. PMID 16429216. Prospective genital exams in a primary-care pediatric practice: meatal stenosis in 24 of 329 circumcised boys older than 3 (7.29%, 95% CI 4.48–10.10%); essentially all cases were in neonatally circumcised boys, and most needed meatotomy. Measures stenosis found on active examination, including cases that had not yet prompted a complaint — hence higher than clinic-diagnosed rates. Also the key source for the observation that the condition is specific to circumcised males.
- Acimi S, et al. "Prevalence and causes of meatal stenosis in circumcised boys." Journal of Pediatric Urology, 2021. DOI via S1477-5131(21)00472-1. Prospective cross-sectional screening of 1,031 circumcised boys aged 5–8: meatal stenosis in 185 (17.9%, 95% CI 15.6–20.3). Found boys circumcised in the first week of life about twice as likely to develop MS (OR 2.08, 95% CI 1.10–3.92). Measures stenosis on universal screening of a school cohort — the high end of the range, capturing asymptomatic narrowing.
- Centers for Disease Control and Prevention (CDC). Sexually Transmitted Infections Surveillance, 2023. Published Nov 2024. US chlamydia rate 492.2 per 100,000 (2023); combined chlamydia + gonorrhea + syphilis approximately 734 per 100,000; over 2.4 million reported cases. A 2024 provisional report shows a subsequent ~9% decline. Reported cases undercount true prevalence (many infections are asymptomatic) and reflect US testing practices, which differ from European ones — so US/EU comparisons above are approximate.
- European Centre for Disease Prevention and Control (ECDC). Surveillance Atlas of Infectious Diseases; Annual Epidemiological Reports on STIs (2022 & 2023 data). 2022 chlamydia notification rates per 100,000: Denmark 708.9, Norway 539.5, Iceland 492.5, Sweden 313.8, Finland 303.9. EU/EEA gonorrhoea up ~300% since 2014. The essential caveat, stated by ECDC itself: a small number of countries (Denmark, Norway, Sweden, UK) account for the large majority of reported chlamydia because they run national screening and test far more heavily — so these rates index testing intensity heavily, not just infection. This is the basis for the "Denmark paradox" above.
- Country circumcision-prevalence figures: Denmark's circumcision rate is approximately 0.4%; the US figure is approximately 55–60% of the adult male population; global prevalence is approximately 37–39%. Country-level circumcision-prevalence estimates come from varied surveys and methods; treat as approximate and confirm the specific figure and its year before citing a precise number.
Wording preserved from the original bibliography — For advocacy organizations and the strongest opposing sources, see Other Organizations.