This week on Science-Based Medicine I wrote an article about a new study looking at the onset of autism symptoms, showing that most children who will later be diagnosed with autism will show clear signs of autism at 12 months of age, but not 6 months. This is an interesting study that sheds light on the natural course of autism. I also discussed the implications of this study for the claim that autism is caused by vaccines.

Unfortunately, I made a statement that is simply wrong. I wrote:

Many children are diagnosed between the age of 2 and 3, during the height of the childhood vaccine schedule.

First, this was a vague statement – not quantitative, and was sloppily written, giving a different impression from the one I intended. I make these kinds of errors from time to time – that is one of the perils of daily blogging about technical topics, and posting blogs without editorial or peer-review. Most blog readers understand this, and typically I will simply clarify my prose or correct mistakes when they are pointed out.

However, since I often write about topics that interest dedicated ideologues who seek to sow anti-science and confusion, sometimes these errors open the door for the flame warriors. That is what happened in this case.

J.B. Handley, writing at Age of Autism, saw my error as a way to demonize me before his enthralled mobs – and he dives into his task with gusto – although without much care or attention to detail himself, as we will see. Handley also is clearly not interested in what the science actually says, only in grabbing a propaganda opportunity.

First, to clarify the facts, here is the childhood vaccine schedule from the CDC. As we can see, the majority of vaccines are given prior to the age of 2, many in fact at or before 12 months. While “height” is a vague term, it is certainly inaccurate in this case. Before I explore this issue further, however, let me address the other factual claim made in that statement.

I wrote that “many children are diagnosed between the ages of 2 and 3. About this statement Handley writes:

Firstly, the last time I checked, the average age of diagnosis for a child with autism was somewhere between 3-4 years of age, not 2-3.

In fact, the average age of diagnosis is 3.1 years (although to be fair other studies give the average at 3.6  – there is some regional variation).  This is “somewhere between 3-4 years of age” but Handley’s point is still incorrect. I did not write that the average age was between 2 and 3 but that many children are diagnosed between 2 and 3, which is certainly true if the average age at diagnosis is 3.1. In his exuberance, Handley simply got this wrong.

I don’t expect him to make a correction, however. The last time he attacked me, he make a rather amateurish mistake, confusing incidence and prevalence, and used his error as the basis of his criticism. He never admitted or commented in any way on his gross error.

But onto the substance of his latest attack. The point that I was trying to make, which I did in fact clarify later in my post, is that when parents attempt to date the onset of their child’s autism they typically will date the onset later than the true onset. As we now know, from multiple studies, true clinical onset (biological onset is likely earlier) is between 6 and 12 months. Parents may not notice this onset until much later, and formal diagnosis is later still. This diagnosis happens within the childhood vaccination schedule, so it is likely that parents will have some recent vaccine to point to when looking for factors that seem to correlate with the onset of autism.

That was my point – a point that I and many others have made previously. By pushing earlier the true clinical onset of autism this study adds to evidence against the involvement of later vaccines. It is true that many vaccine are given in the first year of life, and of course this study by itself does not let all vaccines off the hook – nor did I say that it does.

But I did discuss one vaccine in particular – the MMR vaccine. MMR has received more attention than any other vaccine as a potential cause of autism, thanks to the now-discredited work of Andrew Wakefield. The first dose of MMR is given at a minimum of 12 months of age, with the second dose being given between 4 and 6 years. Certainly this study is relevant to the claim that MMR is a significant cause of autism

There is much evidence to support the conclusion that there is no correlation between MMR and autism. It should be obvious that this study is further evidence against a correlation. If most children with autism show signs by 12 months, then a vaccine which is not given prior to 12 months cannot be to blame. This was not obvious to Handley, however. He wrote:

In fact, between the ages of 2 and 3, children receive all of 2 vaccines, accounting for 5.5% of the vaccines they receive, while a full 70%, including MMR, come in their first 12 months of life, perfectly matching the time when this new study reported the beginning of a regression into autism!

Wrong, Handley. The MMR does not come “in their first 12 months of life.” At the very earliest it comes AT 12 months, which obviously cannot be responsible for symptoms that are present by 12 months. Maybe Handley was just sloppy in his choice of words (but it is odd that he went out of his way to mention MMR by name). Or maybe he is guilty of all the things of which he falsely accused me. This is certainly the same kind of error I made, although more specific. My error was inadvertent and I have readily admitted my mistake and am taking great pains to correct it. Let’s see how Handley responds when this error is pointed out to him.

The bottom line is that this study does in fact add to the body of evidence against an association between MMR and autism, because of the timing of the onset of autism.

Handley also takes exception to my point that parents may not accurately observe and remember when the onset of their child’s symptoms were. He makes one obvious point – that onset is not a moment in time but a process. I agree – but never implied otherwise. But even for diseases and disorders of insidious onset, there usually is a relatively brief period of time when the patient or family member really notices it – and that is when they date the onset.

I have the experience of actually seeing and diagnosing patients, reviewing their histories and comparing them to documented evidence in some cases. So I, like other experienced clinicians, understand this phenomenon well. For example, I see many patients with dementia, like Alzheimer’s disease, and they or their family will often date the onset of symptoms at a point in time some months prior to presentation, often anchored to a specific event (a phenomenon actually known as “anchoring”). But when I probe for specific details, it is apparent that there were signs of dementia for 1-2 years prior to the family’s dating of the onset. Or, I may have the benefit of a documented exam or history, clearly showing onset prior to the memory of those giving the history.

Handley seems naive to all of this, and rather he is content to grossly mischaracterize my point as calling parents “dumb,” which, of course, I never did. This is because Handley’s purpose, in my opinion, is not to meaningfully explore the evidence, but to demonize scientists and physicians with whom he disagrees. Read the comments to his blog and you will see that his attempts at demonizing me and others are quite successful in the echochamber of his followers. He wrote:

Further, this notion by Novella that we parents are “telescoping” is simply the ridiculous introduction of a new and confusing term to try and explain away the chorus of tens of thousands of parents all screaming the same thing about what happened to their kids.

The authors of this study itself applied the concept of “telescoping” or dating autism onset as more recent that it really was. I simply used their term, which they in turn took from the literature. It is a well-described concept, not invented by me or this study’s authors – people remember events in the past as being more recent than they actually were. For Handley’s purposes, however, he wants to characterize a well-known and scientifically established psychological phenomenon as being equivalent to calling people stupid, or spinning reality.

I also showed that this phenomenon is absolutely relevant to the question of vaccines and autism, using the Cedillo case as an example. The Cedillos believed that their child acquired autism after receiving the MMR vaccine. However, home movies reviewed in the court case brought before the Autism Omnibus shows signs of autism in the first year – prior to the first MMR vaccine.

Further, Handley is now trying to argue that this new study supports a correlation between vaccines and autism. In fact, it does nothing of the sort. It does all but eliminate MMR, varicella, and Hep A as having any potential role in autism, as these vaccines all come after the onset of autism in most cases. Handley, if he were being intellectually honest, should admit this, but he doesn’t, and very dishonestly implies that MMR specifically is still a potential cause.

Average age of diagnosis of autism is about 3.1 years of age, and when that was the best information we had to go on the anti-vax movement argued that this showed a correlation with vaccines. Then clinical studies showed that the diagnosis could be reliably made between age 2-3 years of age, but that’s OK, that still correlates with vaccines. Now we know the age of clinical onset is between 6 and 12 months, and Handley is saying this still correlates with the vaccine schedule.

Since the HepB is given at birth, and other vaccines at 1-3 months, moving the diagnosis of autism up even further would still correlate with some vaccines. Since Handley is willing to blame any vaccines, regardless of type (live virus or otherwise) and ingredients (thimerosal or not) no age of onset would disprove his cherished vaccine hypothesis.

There is also nothing about the vaccine hypothesis that led anyone in the anti-vaccine camp to predict that the true age of onset of autism is earlier than it is being diagnosed, and certainly not to within 6-12 months of age. So Handley is just retrofitting – declaring whatever evidence there is as supporting his position.

Finally, Handley pulls the “pharma shill” gambit on me as part of his smear campaign. He trots out an accusation he has made before, portraying my association with the ACSH as being sinister. As I have already explained, my association with the ACSH is limited to me agreeing to advise them on areas of my expertise. That’s it. I have, in fact, performed zero work for them. I have had no contact with them, other than them sending me their public material, and I have never received any kind of remuneration from ACSH. I have corrected Handley on this before, so I know that he knows what he is writing is wrong, and he has failed to correct his errors.

Further my associations with the pharmaceutical industry are minimal – a couple lectures and consultations for nominal fees years ago (and nothing ever to do with vaccines). I have never received a dime from any company to express an opinion on a scientific topic, or write a particular blog or article. I suppose that Handley believes my many hours of work producing podcasts and blogs and promoting science and skepticism is all an elaborate smokescreen for shilling for industry, all without receiving a dime for my efforts, which makes me the worst shill ever.

Actually, I don’t suppose that at all. Rather I think that Handley knows that what he is writing is pure BS, especially since I have called him on it before, and openly challenged him to produce a shred of evidence to support his false accusations. But Handley knows the narrative of the anti-vaccination movement – everyone who denies that vaccines are evil are themselves evil shills for even more evil industry.


In the end, all that matters is the science, which clearly shows that there is no association between vaccines and autism. This one study has minimal implications for an alleged connection, except that it clears the most often implicated vaccine – MMR. It also supports other evidence that the onset of autism is earlier than many parents observe and much earlier than formal diagnosis, which calls into question any casual observations about the timing of onset to any potential triggers.

In my first article on this topic I was sloppy in that one sentence about the vaccine schedule – an error I have now corrected. But that error did not affect the relevant points I made in the rest of the article, which I have also amplified here.

J.B. Handley thinks he has scored some points for his side by jumping on my error, but he has only shown himself, once again, to be a propagandist with no regard for science, accuracy, or even common decency.

Posted by Steven Novella

Founder and currently Executive Editor of Science-Based Medicine Steven Novella, MD is an academic clinical neurologist at the Yale University School of Medicine. He is also the president and co-founder of the New England Skeptical Society, the host and producer of the popular weekly science podcast, The Skeptics’ Guide to the Universe, and the author of the NeuroLogicaBlog, a daily blog that covers news and issues in neuroscience, but also general science, scientific skepticism, philosophy of science, critical thinking, and the intersection of science with the media and society. Dr. Novella also contributes every Sunday to The Rogues Gallery, the official blog of the SGU.