The problem with articles on autism risks and how to evaluate studies. Or why circumcision is unlikely to increase rates of autism.

It seems like every month, articles pop up all over the internet blaming something else for causing autism. To the average person, it looks as though scientists have no idea what they are doing and are just throwing out darts to see what sticks. It’s hard for the average person to have faith in science when it seems to change more easily than the weather here in North Carolina where you can wear shorts and snow pants in the same week. It’s not just a problem with studies on the cause of autism. In scientific circles, it is well-known that popular science articles frequently misinterpret and misunderstand scientific studies, to the point that it’s become a joke. However, because of the ‘popularity’ of autism and the fear-mongering done in the media, it seems as though studies on autism catch fire rather quickly.

Because of my background, I’m often asked by friends to interpret articles posted in popular press, or to cut through the overwhelming information on certain subjects and deliver what most researchers in that field actually believe. There are many excellent guides on how to find and evaluate online studies that go into the subject in more detail, such as Understanding Scientific Studies, A Guide to Looking Smart on the Internet: How to Find and Evaluate Online Information, How To Read and Understand A Scientific Paper: A Guide For Non-scientists (start with this one if you are looking for a quick overview), and for those interest in articles on birth, Understanding Research, The Birth Professional’s Guide. This article is not meant to replace those excellent guides, merely give an overview of how a study on autism can be misinterpreted, and what I do to evaluate popular press science articles.

Here is how a recent study is presented in popular press:

The Mirror reports:

Circumcision in young boys can DOUBLE the risk of developing autism

and the article goes on to state, “Scientists believe the finding may be linked to stress caused by the pain of the procedure.

The study of more than 340,000 boys in Denmark found that circumcision raised the overall chances of an autism spectrum disorder (ASD) before the age of 10 by 46%.”

The Mirror is a British tabloid newspaper which is quite controversial, so the first thing I do when I encounter something like this is try to find more details in a more respectable publication, such as Science Daily. The headline on in Science Daily,

Ritual circumcision linked to increased risk of autism in young boys

is already a much more accurate portrayal of the research. The Mirror implies casual relationship, while Science Daily reports just a correlation. The Mirror implies that multiple or even the majority of scientists believe in the veracity of this study, while Science Daily notes that it is simply the two authors of this piece who hold this belief. More importantly, Science Daily links to the actual study.

When looking at studies, the first thing I do is examine where the study is published and who conducted the study. Not all journals are created equal and there are some journals that are pay to post, where anyone can get a study in provided that they pay a fee. These studies have been duped into publishing outrageous studies, including one written by Maggie Simpson. This study was posted in The Journal of the Royal Society of Medicine. From the wikipedia page on the journal, we can learn that “[i]t claims to act as ‘a forum for debate, education, and entertainment for clinicians interested in UK medicine and relevant international developments and research. The aim of the journal is to influence clinical practice and policy making across the whole range of medicine’ and that it has “open peer review, a system in which authors and reviewers know each other’s identities on the assumption that this improves openness in scientific discourse. This made it one of the few medical journals in the world with open peer review”. Of course, even if a study makes it into a prestigious journal, that doesn’t mean that study is 100% accurate as was scene in Wakefield’s retracted Lancet study, but a study published in a prominent journal is more likely to be reviewed by professionals before and after publication.

A good journal will also note conflicts of interest, which can be seen in this study. Under competing interests, they note:

MF has been an author of articles on health-related and sexual consequences of male circumcision and has taken part in national and international debates on the ethics of male and female circumcision.

Even if the journal had been one that did not declare conflict of interest, a google search of Morten Frisch, the primary author, reveals that he is an anti-circumcision activist with claims such as male circumcision leads to a bad sex life. While researchers who have personal interests in subjects can still be 100% correct in their assessment, even good scientists are at risk of personal biases, and a critical eye needs to be turned to all studies that support the beliefs of the scientists conducting them. I try to be more critical of studies that match what I wish to be true because I know I am naturally more inclined to believe ones that match my opinions, which is known as confirmation bias. In this particular circumstance, I am against RIC and didn’t cut my boys, so I have no emotional biases against this study. I only mention these facts so as not to be accused of being in denial about the results of the study.

In the study, the authors state, “[p]ainful experiences in neonates have been shown in animal26 and human studies4,27,28 to be associated with long-term alterations in pain perception, a characteristic often encountered among children with ASD..29” The citation for this statement is this study. When I looked at the cited study, I found this abstract:

There is a lack of knowledge about pain reactions in children with autism spectrum disorders (ASD), who have often been considered as insensitive to pain. The objective of this study was to describe the facial, behavioral and physiological reactions of children with ASD during venipuncture and to compare them to the reactions of children with an intellectual disability and nonimpaired control children. We also examined the relation between developmental age and pain reactions. The sample included 35 children with ASD, 32 children with an intellectual disability, and 36 nonimpaired children. The children were video-taped during venipuncture and their heart rate was recorded. Facial reactions were assessed using the Child Facial Coding System (CFCS) and behavioral reactions were scored using the Noncommunicating Children’s Pain Checklist (NCCPC). A linear mixed-effects model showed that children’s reactions increased between baseline and venipuncture and decreased between the end of venipuncture and the recovery period. There was no significant difference between groups regarding the amount of facial, behavioral and physiological reactions. However, behavioral reactions seemed to remain high in children with ASD after the end of the venipuncture, in contrast with children in the 2 other groups. Moreover, we observed a significant decrease in pain expression with age in nonimpaired children, but no such effect was found regarding children with ASD. The data reveal that children with ASD displayed a significant pain reaction in this situation and tend to recover more slowly after the painful experience.

Bolding mine. Displaying a significant pain reaction is not the same thing as experiencing a significant pain reaction. We can’t assume that the display is an accurate representation of perception. Even if autistic children do have a higher perception of pain compared to neurotypical children in the same circumstance, it is quite a stretch to say that “lifelong deficits in stress response” show that early pain causes autism. Is it that children with autism have deficits in their stress response or is it that children with autism find the world more inherently stressful due to other mechanisms? 

The study also states

Specifically, using data from nine countries, Bauer and Kriebel31 observed that with each 10% increase in a population’s neonatal circumcision rate, the estimated prevalence of ASD increased significantly by 2.01 per 1000 boys. The present study was carried out to address the hypothesis that ASD might be a rare adverse outcome in boys undergoing ritual circumcision during a vulnerable period of life.

But when you examine the study in question, the authors were examining a possible link between acetaminophen/paracetamol (Tylenol) usage and autism, not circumcision itself. Under the category ‘Weaknesses’ within the Danish study, the authors note, “Unfortunately, we had no data available on analgesics or possible local anaesthetics used during ritual circumcisions in our cohort, so we were unable to address the paracetamol hypothesis directly,” but it seems too big a factor to not seriously consider or at least attempt to measure. (And if you are concerned about the autism/ADHD link with acetaminophen/paracetamol, please see this excellent write up.)

Returning to the study, we see the authors note:

In Denmark, ritual circumcision is an uncommon procedure, except among Muslim and Jewish families…Overall, 3347 boys (0.98%) in the cohort were ritually circumcised in a hospital department or a doctor’s clinic before their 10th birthday, including 2903 circumcisions among 26,664 boys (10.9%) with a likely Muslim cultural background and 444 circumcisions among 316,213 other Danish boys (0.14%)

Since cultural difference can throw off studies, the authors attempt to correct for this.

To address any such influences, we created for all cohort members a cultural background variable based on information about the country of birth for their parents and grandparents. Using the criterion of having at least one parent or grandparent born in one of the following 17 predominantly Muslim countries (Turkey, Iraq, Pakistan, Iran, Somalia, Lebanon, Afghanistan, Morocco, Egypt, Syria, Indonesia, Algeria, Jordan, Bangladesh, Kuwait, Tunisia and Kosovo; no other predominantly Muslim country in the world accounted for more than 0.1% of all non-Danish born citizens), this variable categorised 7.8% of both boys and girls in the cohort as belonging to a family with a likely Muslim cultural background.

Yet, there are many potential problems with this as well as the fact that many religious circumcisions are likely performed outside of medical centers. The authors do note that themselves later on

Recently, Muslim authorities in Denmark explained to the National Board of Health that Muslim circumcisions are often made by private practitioners in their clinics or in the boys’ homes. One possible explanation for the apparent low overall circumcision rate in boys with a Muslim family background is that when payment occurs directly between parents and doctor, no public record will be available. The extent to which such arrangements take place, or similar ones involving non-Danish doctors or non-medical circumcisers, will determine the degree to which our 10.9% figure is an underestimate.

Based on the huge differences in cultural background between these two groups, I am suspicious of the results of this study, as I don’t believe the authors did enough to control the cultural differences.

The authors also speak of “infantile autism” versus “all other types of ASD”, where classic autism is considered separate from a diagnoses of Asperger Syndrome or Pervasive Developmental Disorder- Not Otherwise specified. It is worth noting that individuals with “infantile” or classic autism are typically diagnosed at a much younger age and the authors noted no risk factor with all other types. All of these diagnosis are including under the autism spectrum umbrella, so it seems odd that only classic autism shows a greater risk factor if circumcision is the mechanism by which it is caused.

Another part that raises suspicion is:

In a study from the United Kingdom where infant circumcision is uncommon the estimated mean age at ASD diagnosis was 66 months.43 This is considerably older than the corresponding average age of 39 months in Israel where most boys are circumcised on the eighth day according to Jewish tradition.44

It is worth noting that the comparison is age of diagnosis rather than overall rates of diagnosis. The UK study also includes non-classic autism, while is it not clear if the Israel study did as well. Children in Israel also undergo more thorough assessments for autism, which is possibly catching classic autism earlier. Cultural studies of autism reveal that diagnosis rates change based on how countries view autism, so I find it highly unlikely that the difference is do to circumcision rate.

The most important takeaway from this piece is, “Possible mechanisms linking early life pain and stress to an increased risk of neurodevelopmental, behavioral or psychological problems later in life remain incompletely conceptualised,” which of course, seems to be all but ignored in popular press stories on this study.

Is this study accurate? It is impossible to say given the limited data presented. Other studies will examine this piece and lend credit to it or disprove it. One study is not enough to state that circumcision can double the risk of autism, especially a flawed study like this. Science is a process, not a set of beliefs. When an article says “scientists say…” the average reader tends to interpret it as a consensus, when in reality, it sometimes means that more than one scientist holds a position. Unfortunately, most popular press reporting on autism tends to be along these lines. The next time you see a headline proclaiming what can cause or ‘cure’ autism, take it with a huge grain of salt.

Again, if you are interested in learning more about this subject, please check out these links: Understanding Scientific Studies, A Guide to Looking Smart on the Internet: How to Find and Evaluate Online Information, and especially How To Read and Understand A Scientific Paper: A Guide For Non-scientists. If you are a busy person and don’t have time to research every topic in detail (I sure don’t), then pay attention to what studies the prominent researchers are sharing on social media or backing. If you have questions about studies, then ask in places frequented by researchers, such as r/askscience.


  1. Great analysis. One thing probably worth mentioning is the problem of reporting relative vs absolute risks.

    • It is appalling that the NHS has gone back on the deiiocsn of the Gardner report where it was announced that non-therapeutic genital cutting would not be paid for! It should not even be possible to elect to have a child’s genitals modified at a whim. Another example of politicians not following a deiiocsn to its logical conclusion, namely a ban. Belatedly girls were protected, so why not boys?

Speak Your Mind