Vaccines and autism: Same as it ever was

If there’s one thing I’ve learned during the last seven years about the antivaccine crowd invested in the idea that vaccines cause autism, it’s that it reacts with extreme hostility to any sort of studies that cast doubt upon their pet idea that vaccines cause autism. That’s somewhat understandable, given how much of their identity so many members of the antivaccine movement have invested in their idea, but not all studies that fail to support the central dogma of the antivaccine movement (i.e., that vaccines cause autism and are in general evil) are created equal, at least not with respect to the reaction they provoke from the antivaccine movement. Some provoke much more hostility than others.

The sorts of studies that provoke the most hostility of all seem to be the ones that suggest that autism has a large genetic component. I haven’t quite figured out why, other than that if autism is largely genetic in origin then it can’t be due to vaccines. Alternatively, some have speculated that the reason for this particular hostility is that they don’t like the concept that there is a large genetic component to autism because it implies something about the parents. Be that as it may, I saw a perfect example of this sort of reaction over at the antivaccine crank blog Age of Autism yesterday when the media editor for the whole crank operation, Anne Dachel, wrote a post entitled Autism Diagnosis Age. Ann is really, really not happy about a new study and the publicity it got. Originally, I was going to let this study pass because, for whatever reason, I missed what I considered the window of opportunity to blog about it and Steve Novella had already covered it quite well. However, Dachel gave me the perfect excuse to revisit that oversight and take a look, this time in the context of examining the antivaccine movement’s reaction to the study.

Before I get into the study itself, I’ll just do the really brief CliffsNotes version of it. Basically, a group at the Univesity of North Carolina, Chapel Hill that found evidence of brain abnormalities on MRI as early as six months in children who later manifested autistic symptoms. In other words, this study is consistent with an emerging body of evidence over the last five years that suggests that brain abnormalities in autism are detectable far earlier in development than had previously been thought. Scientists are naturally excited by these results, because they suggest a biological basis for autism and because they could help us shed light on the genetic factors that contribute to autism. Antivaccinationists like Anne Dachel, however, are not:

Sometimes it seems that no one is really interested in finding out anything significant about autism. Experts are only obligated to come up with some new findings every few months to make it look like someone somewhere is doing something. That’s what’s happening with this latest research. If scientists can find evidence of autism in babies at six months, it would be proof that the parents who claim that their child suddenly regressed into autism following their 18 or 24 month vaccinations are wrong. It was all just a big coincidence. The signs of autism were really there much earlier.

Yes, that does appear to be the implication of this study. No, what’s happening is not that experts are “obligated to come up with new findings every few months.” Rather, what’s happening is that an increasingly robust body of evidence is being built up over time that suggests that, whatever the cause, the biology of autism manifests itself very, very early in development. The implications, of course, are exactly as Dachel states. If signs of autism are detectable in babies at six months, such observations alone cast huge amounts of doubt on their fanatical belief that vaccines cause autism. Couple this emerging set of evidence that suggests the biology of autism takes hold very early with the now numerous epidemiological studies that have failed to find a wisp of a whiff of a hint of a correlation between vaccines and autism, and it’s pretty obvious that the hypothesis that vaccines cause autism has about as much plausibility and support these days as, well, homeopathy. Ironically, strip Dachel’s little rant of its inherent sarcasm, and she actually got it right. The existing evidence does suggest that autistic regression observed after vaccination is not associated with vaccination any more than one would expect by random chance alone, and studies like this one do strongly suggest that detectable signs of autism are detectable at a very young age, far younger than the age at which symptoms are generally first recognized.

So what did this study show?

Basically, researchers studied high risk children, specifically siblings of children already diagnosed with autism, at six months using MRI. Actually, it was a specific kind of MRI called diffusion tensor imaging. The authors rationale for the study is described in the introduction to the paper:

Autism spectrum disorders (ASDs) are complex disorders of neurodevelopment defined by impaired social communication and restricted, repetitive behaviors. ASDs represent a significant public health concern, affecting upward of one in 110 children, with a recurrence rate among at-risk families of nearly one in five (1, 2). Findings from prospective studies of infant siblings of children with ASDs, who are at higher than average risk for the disorder, indicate that a number of the defining behavioral features of ASDs first emerge around 12 months of age after a period of relatively typical postnatal development (3-5).


Taken together, findings from the behavioral studies and the studies of brain and head size growth in high-risk infant siblings suggest that the latter half of the first year of life is a pivotal time for both brain changes and symptom onset in infants later diagnosed with an ASD. The concurrent timing of these phenomena suggests that brain changes during this pe- riod may have an important role in the pathogenesis of autistic behavior.

Imaging studies were carried out in these high risk children, and the children were then also evaluated clinically at 24 months to see which children met criteria for autism spectrum disorder. The imaging studies were used to characterize the development of the brain’s white matter, which are the parts of the brain that contain the axons rather than the actual neurons and thus carry the nerve signals. Basically, the investigators used techniques known as fractional anisotropy, which allowed them to characterize the white matter. Basically, what they found was that, among the study participants, the 28 infants who later went on to be diagnosed with ASD had significantly different fractional anisotropy trajectories for 12 of the 15 fiber tracts studied than the infants who did not receive a diagnosis of ASD. Specifically, the infants who went on to develop ASD had significantly less white matter development. A clear implication of this study is that, whatever causes ASDs, it must begin well before six months of age. As the authors put it:

It is clear that the neurodevelopmental story of ASDs neither begins at 6 months of age nor ends at 24. Extend- ing neuroimaging downward to infants younger than 6 months would help clarify the temporal origin of diverging trajectories, while extending neuroimaging to later ages would capture increasingly stable neurobehavioral out- comes. Additional data points would likewise refine the calibration of trajectories beyond linear models. Because the present results are limited to at-risk infants, the inclu- sion of a low-risk comparison group is necessary to frame group trajectories in the context of typical development. The addition of a multimodal approach to neuroimaging could sharpen our understanding of early brain changes in ASDs, allowing for the investigation of functional and structural covariance. Linking the structural integrity of specific fiber tracts to behaviors relevant to autism during infancy would similarly clarify mutually dependent de- velopmental processes. To capture the mechanisms that engender and maintain the trajectories evidenced here, future researchers might also consider the role of genetic and epigenetic variables in the development of structural neural circuitry.

Of course, no study is perfect, and this study is not without its shortcomings. Perhaps the most important one is that there wasn’t a low risk control group. In addition, fractional anisotropy is an imperfect measurement of white matter organization. Even so, the results of this study are fairly robust-appearing. Basically, this study is one line among several converging lines of evidence that suggests that the earliest changes in the brain associated with ASDs are detectable around six months, which converges with other lines of evidence suggesting that the first detectable behavioral signs of autism appear around six months too. And, yes, converging lines of evidence that suggest that the pathophysiology of autism is detectable at a very early age are inconsistent with the claim, for instance, that the MMR vaccine causes autism, given that the MMR isn’t given that early.

None of this deters Anne Dachel, however:

Jason Wolff attributed autism to genes and ‘a child’s experiences with the world,’ I’d like to point out to him that a big part of that experience includes an arsenal of vaccinations, starting at birth.

And, if vaccines don’t cause autism when given to the baby, then maybe they do when given to the mother, at least if you believe Dachel:

(And even earlier if a pregnant woman receives the flu vaccine, recommended at any stage of pregnancy. The majority of the flu vaccine available come with a huge level of mercury, a known neurotoxin. This mercury easily passes the placental barrier.)

And if trying to claim that mom’s vaccines cause autism in baby doesn’t fly, there’s always conspiracy mongering:

None of the doctors cited in these stories is worried about the effects of the toxic chemicals regularly found in vaccines. No one seems to care that there isn’t a single study on the cumulative effects of the ever-expanding vaccination schedule. Nor are they concerned that every study disproving a link between vaccines and autism is tied to the vaccine industry.

Pharma shill gambit, anyone?

And if the pharma shill gambit doesn’t work, then there are over-the-top attacks on scientists from commenters:

These people have blood on their hands, and they know it. They will do anything they can to keep that fact from becoming common knowledge.

If that’s the case, scientists sure have a funny way of showing it, given how many studies they’ve done of vaccines and autism, even long after it was clear that the hypothesis was not standing up to scientific scrutiny.

In fact, over the last 10-15 years, there has been a relentless drumbeat of studies that have failed to find a link between vaccines and autism. In the normal course of scientific studies, such a string of completely negative studies would lead to the abandonment of the hypothesis under test, and it has, among scientists anyway. However, to antivaccine zealots like Anne Dachel and her merry band of antivaccine activists at AoA, it’s got to be the vaccines. It is, first and foremost, all about the vaccines. It has always been about the vaccines, and it will always be about the vaccines. Studies such as this are thus immediately attacked and dismissed as either being the result of scientists’ “dogma” that autism must have a significant genetic component rather than scientists following the evidence where it leads. Alternatively, to antivaccinationists, they must be the result of scientists circling the wagon to protect their pharma paymasters from having to admit that vaccines are The One True Cause of Autism. The reasonable conclusion from so many scientific studies that have failed to find even a hint that vaccines are correlated with autism is that vaccines do not cause autism. Unfortunately, when it comes to vccines antivaccinationists are not reasonable people, which is why they will never accept where the evidence goes.