Oct 07 2009

Autism Prevalence

Two recent studies concerning the prevalence of autism in the US have been getting a lot of attention, because they indicate that autism prevalence may be higher than previously estimated. This, of course, fuels the debate over whether or not there are environmental triggers of autism.

One study was conducted by the CDC but has yet to be published. The results were announced ahead of publication by the US Health and Human Services Secretary Kathleen Sebelius to the autism community. She reports that the new prevalence of autism spectrum disorder (ASD) is now estimated at 1% or 100 in 10,000 children. This is an increase over the last few years. In 2002 the prevalence was estimated to be 66 per 10,000.

The second study was published in the journal Pediatrics and is a phone survey of 78,037 parents. They asked if they had any children who had ever been diagnosed with an ASD. Here are the results:

The weighted current ASD point-prevalence was 110 per 10,000. We estimate that 673,000 US children have ASD. Odds of having ASD were 4 times as large for boys than girls. Non-Hispanic (NH) black and multiracial children had lower odds of ASD than NH white children. Nearly 40% of those ever diagnosed with ASD did not currently have the condition; NH black children were more likely than NH white children to not have current ASD. Children in both ASD groups were less likely than children without ASD to receive care within a medical home.

These are slightly higher numbers than the CDC data. This is to be expected, however, since the Pediatrics study is a phone survey. Diagnoses were not confirmed by a clinician.

There is no question that the number of ASD diagnoses has been steadily increasing for the last two decades. The burning question has been – is this a real increase in the disorder or an artifact of increased surveillance and an expanded diagnosis. I have reviewed the evidence in depth previously – the evidence strongly supports the conclusion that the increasing autism prevalence is due to increased efforts to make the diagnosis and a broadening of the definition of autism. The evidence is not sufficient to conclude that there is not also a real increase in ASD incidence, but nor is there data to support this conclusion.

Further, since I wrote my last summary, there has been more data to support the conclusion that autism rates are really flat. The National Health Service also has recently published a survey of autism prevalence. While the US studies looked at children from 3 to 17, the NHS study looked at all age groups. Their question was this – is the prevalence of ASD the same or different among various age groups? If the incidence of ASD is truly increasing, then younger age groups should have more ASD than older age groups.

They found a consistent prevalence of 1% in all age groups they surveyed. This is remarkable for two reasons – first, they found the exact same 1% figure as the CDC US survey (assuming the CDC data is more accurate than the phone survey published in Pediatrics). This supports the conclusion that the 1% figure may be close to the true prevalence of ASD in the population.

Second, the NHS study found that the prevalence of autism was the same in all age groups, strongly suggesting that true ASD incidence has not been increasing over recent decades and supporting the increased surveillance and definition hypothesis.

Of course, the anti-vaccine community is exploiting these two new studies (the US studies – they have already found excuses and conspiracies to dismiss the implications of the NHS study). David Kirby, writing for Age of Autism, a notorious anti-vaccination site, wonders aloud why there is no alarm about these new ASD prevalence numbers. He is still playing his coy “conspiracy of silence” card.

It is also interesting that Kirby and others in the anti-vaccine community were predicting that autism rates should decrease following the near complete removal of thimerosal form the childhood vaccine schedule by 2002 in the US. Autism rates continued to rise at a steady rate, essentially killing the hypothesis that thimerosal is a significant contributor to autism. Now Kirby is citing the same evidence (continued rising autism rates) that punctured his previous hypothesis as support for his new hypothesis – the HepB vaccine.

That’s right – the anti-vaccine community has moved from the MMR vaccine to thimerosal to other “toxins” in vaccines and are now focusing their sites on the HepB vaccine. As each claim fails to be supported by scientific evidence, they simply migrate over to another claim – all with the core feature that vaccines are always to blame.

Meanwhile, the strong consensus of evidence is that ASD rates are not truly increasing at all.

Note: This entry was cross-posted over at Science-Based Medicine.

25 responses so far

25 thoughts on “Autism Prevalence”

  1. superdave says:

    The problem at the AoA is worse than moving goalposts, there are multiple sometimes mutually exclusive goalposts. This is because they never truly move on to another argument. People on that website will still bring up mercury and thimerisol. I’ve even asked people who comment on that site what percentage of Autism they feel is triggered by environmental causes and I have received anywhere from a small minority of genetically susceptible children to all of them. The only common thread is that “They” know what the cause is and “We” either are ignorant or in on the conspiracy. When you take into account the pseudoscience peddled on their site concerning the cause of autism with the pseudoscience concerning the treatment of autism, the site is such a mess false information it’s a self contained ecosystem of woo.

  2. Draal says:

    Dr. N, you mentioned potential environmental factor(s) that may contribute to an increase in autism. In all the years of study of autism, is there any knowledge of how autism arises (i.e. mechanism)? Pertaining to this, when does autism develop (within zygotes, embryoes, fetuses, infants, toddlers)? What physically differentiates an autistic brain from a normal brain (mutations in proteins, expression, neuron stucture/function, chemical imbalance)? And are there any confirmed environmental factor(s)?

    And has there been a population density map been compiled of the occurrences of autism? There are a number of famous legal cases where a local community has an elevated number of health problems (ex. Erin Brockovich).

  3. capheind says:

    What about the Dr. Mom/Dad factor? Parents seem to think that knocking boots makes them instant MD’s, they latch onto a theory to explain why their child fails to meet their ridiculous expectations based on their extensive Google university education, and then they go to the doctor expecting immediate confirmation and a pat on the back. When parents were pumping their kids with every form of sugar known to man ADHD and Bipolar disorder became the big ticket item, now in the era of the geek its HFA and Asperger’s. All these conditions run in my family, genuinely on one side, due to a rich Dr. Mom on the other.

  4. BA says:

    The study in Pediatrics may well overestimate ASD prevalence but the NHS study in the UK probably underestimates ASD prevalence. They excluded individuals in assisted care. Though I believe evidence points strongly to the best hypothesis on increasing autism prevalence being diagnostic substitution, prevelance is probably going to be more frequent than 1 in 100.

    That said, the NHS study is very important in that its survey of prevalence in adults is based on what seems a pretty sound method that involves a follow-up diagnostic work-up (ADOS) for a portion of the sample identified through a less accurate screening tool (AQ).

  5. s says:

    there are a couple of studies suggesting a possible connection between environmental factors and autism. Some examples follow:

    1) “Autism Spectrum Disorders in Relation to Distribution of Hazardous Air Pollutants in the San Francisco Bay Area” (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1570060/) suggests “a potential association between autism and estimated metal concentrations, and possibly solvents, in ambient air around the birth residence”.

    2) Then there is a small but pretty interesting work done by Baumgardner (Geographic Analysis of ADHD in Children: Milwaukee County study, http://www.fammed.wisc.edu/our-department/newsletter/fall-2007/geographic-analysis-adhd-children-milwaukee-county). In it autism cases are clustered especially by the airport and main roads to it. The authors have not suggested any possible causes.

    3) And finally there is Palmer et al studies which have been, uhm…, critiziced on SBM and “Respectful insolence”: “Proximity to point sources of environmental mercury release as a predictor of autism prevalence” (http://www.ncbi.nlm.nih.gov/pubmed/18353703) and “Environmental mercury release, special education rates, and autism disorder: an ecological study of Texas” (http://www.ncbi.nlm.nih.gov/pubmed/16338635).

    A more general discussion and overview can be found in “Epidemiologic evidence of relationships between reproductive and child health outcomes and environmental chemical contaminants” (http://www.ncbi.nlm.nih.gov/pubmed/18470797).

    There is also a hypothesis suggesting vitamin D deficiency (as insolation varies with latitude and there is a possible correlation between latitude and autism–there is also the acse of autims in somali immigrants in eg the US and Sweden, (http://www.ncbi.nlm.nih.gov/pubmed/18754897)). “Autism and vitamin D” (http://www.ncbi.nlm.nih.gov/pubmed/17920208).

    Another more general hypothesis put forth in MedHyp combine the vitamin D, heavy metal and general pollution hypotheses. “Environmental risk factors for autism: Do they help cause de novo genetic mutations that contribute to the disorder?” (http://www.ncbi.nlm.nih.gov/pubmed/19699591). (Rather see Epidemiologic evidence etc.)

  6. sbg says:

    Thank you for reviewing this. I was pulling my hair out when I noticed the blurb about autism rates rising in the NY Times – Science Times this week — esp. that the author gave the ‘rising’ story without any allusion to the NHS Survey. In fact, I haven’t seen it addressed by the NY Times at all (not that this should be a beacon of scientific understanding, but it does seem newsworthy for the Science Times segment). While the survey must be replicated, I am thrilled to finally see someone address the big question.

  7. Draal says:

    Thanks S for the minireview. I took a look at the references that you provided and I quickly saw that the literature is just littered with garbage when it comes to the causes of autism. The literature is just all over the map.

    Here’s another one:
    Ockham’s Razor and autism: The case for developmental neurotoxins contributing to a disease of neurodevelopment
    Author(s): DeSoto MCSource: NEUROTOXICOLOGY Volume: 30 Issue: 3 Pages: 331-337 Published: MAY 2009

    A long held standard of evaluating competing theories is parsimony. A single theory that explains multiple findings is superior to circuitous explanations that require untestable assumptions. It is hard to imagine how the set of findings could be more easily explained than including an etiological role for environmental toxins.

    Obviously, her logic is infallible ’cause she says so.

  8. HCN says:

    Also, anything published in Medical Hypotheses is pretty much worthless. That publication has come under fire for not even reading much of the papers coming in for publication. It has included papers with some “hypotheses” such as masturbation helping with sinus congestion, and where children with Down Syndrome are called “mongoloids.”

    For more information see:

    And let us not forget it is where the truly horrible paper claiming autism was a unique form of mercury poisoning written by people without any qualifications (from a financial analyst to a chemical engineer, woo hoo!). Mercury poisoning has completely different symptoms. Plus, as noted many times that thimerosal was removed from vaccines almost nine years ago (and even the HepB vaccine!), and there have been no effects on autism rates (cue Kirby on his plumes from China hypothesis!).

  9. springle says:

    I wish everyone would stop bickering over whether Autism numbers are higher due to changes diagnostic tools and criteria or is becoming more rampant, it detracts from the point which is 1 in 100 children have a serious condition that needs our full attention. We will all pay dearly to take care of these people and it would do us good to stop fighting and try to find a cause and then a cure!

    When I say to someone “did you know 1 in 100 children have Autism” I really don’t want to hear their opinion on where those statistics came from, I want to hear what they think we should do about it.

  10. Andrew Bienert says:

    I found the following point interesting:
    “Nearly 40% of those ever diagnosed with ASD did not currently have the condition”

    I wonder what criteria were used to determine that? Since I don’t have access to the full text I’m assuming it means something like, that for 40% of those ever diagnosed they no long have a clinical diagnosis of ASD? The methods in the abstract don’t seem to outline it great detail, so I’m not sure if I’m reading it right. But as a parent of a child recently diagnosed with HFA/Asperger’s my rather lame pattern seeking monkey brain saw that figure straight away. I just hope the peddlers of woo don’t latch onto that and somehow misrepresent it.

    And as we broaden the definition and find more and more individuals in the population with autism can we correctly define the entire spectrum as a disorder anymore? I could imagine there are many HFA/Aspies out there who might just as easily look at the rest of the population, albeit the other 99% (and seemingly still falling), and maybe think of them as having Neurotypical Spectrum Disorder.

  11. eiskrystal says:

    I was also surprised that the 1% is across all age groups as it has been mentioned before that people do improve over time and therefore some would likely end up no longer meeting the diagnosis of autism. How was this accounted for? or am I missing something?

  12. tmac57 says:

    springle-“I wish everyone would stop bickering over whether Autism numbers are higher due to changes diagnostic tools and criteria or is becoming more rampant, it detracts from the point which is 1 in 100 children have a serious condition that needs our full attention.”
    I don’t agree that this is a petty argument (bickering). Whether or not autism rates are rising is an important question that needs to be addressed. If in fact the rates are rising, that could be an important clue that could be linked to some other changing factor. Also , if rates are not rising, that could be a red herring that has sent researchers down a blind alley, so yes, this is something that needs to be answered.

  13. springle says:


    I agree with you that it is an important question which could help focus scientific research and that it needs to be answered. My point is that I feel like the argument is used to stop the conversation. The implication seems to be that there have always been this many people walking around with this disorder and we just didn’t call it Autism. The problem is that’s like saying there is nothing to be alarmed about, when this should be very alarming. I often find that people who bring up the arugument when confronted with the numbers use that point to stop further inquiry into possible environmental causes and/or new treatment theories. I appreciate your comment because I was trying to find a way to say what you said too, while still making my point.

  14. sbg says:


    I had the opposite desire from these statistics, that we shouldn’t be alarmed but comforted that rates are likely not rising. This supports that there might not be some toxin or other agent claiming more and more children each year .

    The ‘autism rates rising’ creates alarm and crazy behavior and also has obscured the fact that most children with autism benefit tremendously from behavioral therapies (not from wacko diets, non-vaccinations and chelation). These same claims obscure our understanding of what autism really is.

    We should not be alarmed. We should be informed and we should help these children get safe, efficacious treatments in lieu of identifying all the causes of autistic behavior. And finally, we should have done these studies ages ago to actually assess rates of autism across ages.

  15. Michael Kingsford Gray says:

    # springle:

    Your comment misses the point almost entirely.
    The post is about the mortal criminal negligence that is being actively encouraged and supported by anti-vax liars, and the delusional followers of their religion.

    Discussion of rates of autism are being forwarded to reveal one of their outrageous lies about vaccines.

  16. s says:

    Well, there’s a lot of bad articles in science journals of every field. ASD research seems just even more fraught with it (maybe not unsurprising due to the emotional dimension).

    PS Daedalus,
    this paper on fever in autism may interest you in case you haven’t already read it: Autism, fever, epigenetics and the locus coeruleus. (http://www.ncbi.nlm.nih.gov/pubmed/19059284)

  17. daedalus2u says:

    s, I looked at that paper, I find it quite annoying they way they talk about the epigenetic programming of the pathways involved in autism as “dysregulation”, and then in their hubris talk about restoring normal regulation.

    A final common pathway in essentially all conditions of “stress” is low nitric oxide. Stress in utero does program the brains of experimental animals in ways that would be suggestive of autism-like programming were they to happen in humans. Prenatal stress is strongly associated with autism. Postnatal social deprivation does decrease the number of NO producing neurons in the brains of experimental animals. Exposure to xenobiotic chemicals decreases NO levels. Even exposure to sugar reduces NO levels.

    Fever does raise NO levels. I have several blog posts about fever, autism, mitochondrial dysfunction and immune system activation. Nitric oxide physiology is a common pathway that links them all together. Low NO exacerbates autism-like symptoms, high NO ameliorates autism-like symptoms. That is not due to “dysregulation” per se, it is due to good regulation around a bad setpoint. The setpoint is bad because the basal NO level is too low and all NO regulated pathways require the right amount of basal NO to function properly. The NO level can be raised acutely. This has the effect of reducing the severity of the stress responses mediated by low NO (which is a great many of them). This allows neurodevelopment to continue on a higher NO trajectory, on a lower stress trajectory which results in a more neurotypical outcome.

  18. Joseph says:

    @s: You missed a not so well known study of EPA Superfund sites in Minnesota.

    The problem with most of those studies is that they fail to control for urbanicity, while relying on administrative autism counts. It’s well known that autism is most often diagnosed in urban areas. But it’s probably not the case that autism is actually more common in urban or wealthy areas. For example, a recent prevalence study in a semi-urban area of Sri Lanka found the prevalence of ASD to be over 1%.

    Palmer was told about this confound after his first study. In his second study, he did control for urbanicity, and the effect size dropped considerably. Unfortunately, he did not control for urbanicity very well. He treated it as a discrete variable, when in reality, it’s a continuous variable. I’ve suggested these types of studies need to control for (log of) population density.

  19. s says:

    I sure did, thanks for pointing that out. The article is: CM DeSoto “Ockham’s Razor and autism: the case for developmental neurotoxins contributing to a disease of neurodevelopment. (http://www.ncbi.nlm.nih.gov/pubmed/19442816)

  20. tom says:

    Autism is a common end-result that develops from a hugely diverse set of genetic, environmental, and developmental factors. It is not mysterious, and the rational behind it is straight forward. In fact, it is stupidly obvious once you know what to look for. Keep your ears open on this issue.

  21. Hyperion says:

    The 1% figure actually makes some sense in light of comorbidity of ASD with other disorders. One thing that has always made me wonder about whether the older 1/166 number might have been on the low side was the comorbidity rates with ASD and ADHD.

    I’ve seen a few studies that have implied that around 30% of patients with ADHD have comorbid ASD. While searching for those studies, I also came across a study looking at whether there were similar genetic causes between the two disorders, using twins:


    I mention this partly because obviously finding genetic correlation between ASD and ADHD, which is already known to have a high genetic correlation, is one more blow against the anti-vax idiots, but more for what they mention about comorbidity, finding that 22% of the children with ADHD also met the criteria for ASD.

    How does this relate to that 1% number? Well, multiple reviews have put the worldwide prevalence of ADHD as being around 5% of the population. If 20-30% of those have comorbid ASD, then that 1% statistic starts to make a whole hell of a lot more sense.

    I’ve mentioned this before on Dr. Gorski’s other blog a while back, I think, and postulated at the time that 1/166 would be below what one would expect based on comorbidity data with other conditions whose prevalence was better known.

    My guess is that eventually (and this will not be anywhere in the near future), some better understanding of the development of dopamine pathways will make all of this make much more sense. Right now all anyone can do is make diagnoses based on symptom presentation. Functional neuroimaging may eventually show that just as people with ADHD have impaired mesocortical dopamine transmission in specific regions of the frontal lobes, there might be far more pervasive issues in people with ASD, maybe looking at mesolimbic dopaminergic activity as well, throughout many areas of the brain.

    So where small, localized problems of dopamine transmission might be implicated in ADHD, Tourette’s, OCD, and a few other neurodevelopmental conditions, it wouldn’t surprise me if ASD simply involved more widespread problems in dopamine transmission everywhere.

    Of course, armchair speculation about pathophysiology is less important than my original point, which is that this 1% prevalence number is actually somewhat unsurprising if you were to take the data and extrapolate backwards from comorbidities with other conditions whose prevalence is better known. Obviously directly measuring the prevalence of ASD is better than trying to extrapolate from the prevalence of other conditions, but it’s certainly good when the numbers seem to match up between the two.

  22. SensoryDad says:

    Read this great front page article from the Atlantic last week, VERY interesting:


Leave a Reply