“There are in fact two things, science and opinion; the former begets knowledge, the latter ignorance.” –Hippocrates

May 11, 2012

There’s a certain logical disconnect that seems to pop up among everyone in the anti-vaccination crowd.  They dismissively ignore all actual evidence that doesn’t support their firmly held belief, but they will crow from the mountaintops any time they think they’ve found scientific evidence that backs them.  While this sort of thing is far from exclusive to the anti-vacc contingent, they are among the most obnoxious about it, and they have enough of a profile (thanks to noted biomedical scientist Jenny McCarthy) that you can’t help but hear a lot of what they spout, whether you want to or not.

Which brings us to Mark Moore of the Arkansas Watch blog.  Mark is, it seems, part of the anti-vacc group that is just absolutely sure that vaccines cause autism.   Once you know this, it becomes much less of a surprise that he would entitle a post, “How do you give a Monkey Autism?  Administer them the Vaccines we give to our Children!

The kind of tests that should have been done a long time ago have finally been done on lab monkeys. The result is that young monkeys given a vaccine schedule from the 1990’s tend to develop autistic symptoms while a control group did not.

Then he provides a link to the study. If you’ve followed the anti-vacc movement at all over the last few years, and if you ignore that Mark makes it sound like this is a new study, you might guess that he’s going to link to the study by Dr. Laura Hewitson from 2010.  You’d be more or less correct; he links to an article on LewRockwell.com.  That article, while also making it sound like we’re dealing with a new study, links to an article at Age of Autism written by (noted shaken-baby-denialist) Catherine Frompovich.  Frompovich is talking about the Hewitson study, and she’s hitting all the same notes that you’d expect: (1) the study showed that “biological changes and altered behaviors did occur in vaccinated monkeys, which resembled and were similar to those observed in ASD diagnosed children;” (2) the same changes and behaviors did not occur in the non-vaccinated monkeys; (3) that people who stand to profit from vaccines don’t want this study to be replicated; and (4) that this study, once replicated, will prove ex-Dr. Andrew Wakefield correct.

Wow.  That sounds like quite a study, no?  To someone who was unfamiliar with the study (or who was simply looking for something to confirm what they already “knew” to be true), that’s tantamount to a smoking gun.

Of course, once you look at the study — which I would bet most people touting the results have not done — you realize it’s less a smoking gun and more a smoldering pile.  First of all, there’s some important information in the Methods section of the study.  See if you can spot the problem:

For this pilot study, a total of sixteen male infant rhesus macaques (Macaca mulatta) were randomly allocated to either the exposed or unexposed study group in order to complete peer groups (Ruppenthal and Sackett 1992) such that each peer group contained animals from either the unexposed or exposed study group.


Four infants were assigned to the unexposed study group and received saline injections according to the schedule in Table I (although one control had to be withdrawn due to a scheduling an error). The remaining 12 infants were allocated to the exposed study group and received the complete, age-adjusted childhood vaccine regimen (Table I).


A complete set of MRI data at both T1 and T2 were obtained from 9 exposed and 2 unexposed animals.


A complete set of PET data at both T1 and T2 were obtained from 9 exposed and 2 unexposed animals.

So, the entire group was only 16 monkeys, and the control group was only four? Then, of those four, only TWO were actually part of the data that was analyzed. It is mind-blowingly difficult, if not nearly impossible, to achieve a difference in results that would be statistically significant when you have a two-monkey control group.  Short of the vaccinated monkeys all spontaneously combusting as soon as they received their injections, while the control monkey did not, I can’t imagine how you could draw conclusions here.

Oh, plus there’s this little gem when they explain why they split the group 12-4 instead of 8-8:

We purposefully assigned a larger number of animals to the exposed group in order to optimize the chances of observing what we anticipated to be an uncommon or idiosyncratic effect.

As anyone who has spent more than about five minutes doing scientific experiments can tell you, if you’re looking for an uncommon effect and trying to attribute it to something, you have to have a large control group so that you can demonstrate that the uncommon effect did not happen there. When you structure the experiment the other way, as the authors of this study did, you’re not controlling for the cause of the uncommon effect — you’re just trying to increase your odds of seeing that effect in the experimental group so that you can point to the outcome as “proof” of something that it is not in fact proof of.

But it gets worse. The study specified that it started with 16, split 12-4, but only analyzed the data from 11.  They explained that one monkey from the control was removed “due to a scheduling an [sic] error,” but there’s not even a footnote, let alone an actual explanation, as to why the other 4 monkeys (3 vaccinated, 1 control) were not analyzed.  That’s just sloppy science.  And, when the earlier versions of the study included a prominent mention of Dr. Andrew Wakefield, what appears to be sloppy science gives rise to a strong implication of shenanigans.

Also suggestive that not everything is on the up-and-up here?  When the author of a study inexplicably changes the number of animals in the data set and the results of analyzing that data.  In 2008, this study was presented with the following in the abstract (emphases mine):

Methods: Macaques were administered the recommended infant vaccines, adjusted for age and thimerosal dose (exposed; N=13), or saline (unexposed; N=3). Primate development, cognition and social behavior were assessed for both vaccinated and unvaccinated infants using standardized tests developed at the Washington National Primate Research Center. Amygdala growth and binding were measured serially by MRI and by the binding of the non-selective opioid antagonist [11C]diprenorphine, measured by PET, respectively, before (T1) and after (T2) the administration of the measles-mumps-rubella vaccine (MMR).

Results: Compared with unexposed animals, significant neurodevelopmental deficits were evident for exposed animals in survival reflexes, tests of color discrimination and reversal, and learning sets. Differences in behaviors were observed between exposed and unexposed animals and within the exposed group before and after MMR vaccination. Compared with unexposed animals, exposed animals showed attenuation of amygdala growth and differences in the amygdala binding of [11C]diprenorphine. Interaction models identified significant associations between specific aberrant social and non-social behaviors, isotope binding, and vaccine exposure.

So, that’s data crunching on 13 vaccinated monkeys and 3 non-vaccinated. If you’ve been paying attention, you know that the latter paper — again, based on the same experiment, not some later, separate experiment — referred to 12 vaccinated monkeys and 4 non-vaccinated, then only looked at 9 and 2, respectively. And what did the 2010 paper claim to find?

For the exposed group there was a nonstatistically significant increase in right amygdala volume over time (P=0.16; Table IIa). For the unexposed group there was a significant drop in right amygdala volume over time.


For the exposed animals there was an increase in left amygdala volume over time, although this was not statistically significant (P=1.0; Table IIa). In contrast, for the unexposed animals there was a significant
decrease in left amygdala volume from T1 to T2 (P<0.0001; Table IIa).

With our 13-3 group, we had reduced amygdala growth when compared with the control; with the 9-2 group (who, again, were a sub-group of the 13-3 group), we get a slight increase in amygdala growth over time, while the control group showed a precipitous shrinkage in the same.  What does this mean?  Well, at a micro level, it shows the inherent flaws in having a two-monkey control group.  The amygdalas of these monkeys generally do not shrink over time, so having it happen in the two control monkeys is the kind of statistical noise that a larger control group would have minimized.

On a macro level, it just means that this study, like others before it, has failed to show anything; the amygdala growth in the vaccinated monkeys tracks perfectly with what you would expect in the brain growth of normal monkeys, meaning that the null hypothesis in this experiment was not proven false.  Though, to be fair, I suppose that the experiment did show that anti-vacc believers will tout anything that they think supports their argument, even when that “evidence” contradicts previously proffered evidence and even when it the evidence contradicts itself.

Oh, but back to Mark Moore.  Regarding the study, he notes:

This result contains some variables not present today. Thimerosal has been largely removed from vaccines today, so if that was the factor which contributed to the increase in autism associated with the vaccines then it could mean that today’s vaccines are safer. Regular readers of this space will know that my view is that it was NOT the Thimerosal which contributed to the autism, but rather the presence of rouge human proteins from live virus vaccines cultured in aborted fetal tissue. The MMR and Chicken Pox and some whooping cough vaccines fit into this category.

Skipping over the weirdness of Mark’s touting a study as proof when he doesn’t even believe that the subject being studied was the cause of autism, there is still some facepalm-worthy silliness going on here.  This “aborted fetal tissue” causation is pushed by Dr. Theresa Deisher (among others), and the claim generally goes something like this:

Today, more than 23 vaccines are contaminated by the use of aborted fetal cells. There is no law that requires that consumers be informed that some vaccines are made using aborted fetal cells and contain residual aborted fetal DNA.


How could the contaminating aborted fetal DNA create problems? It creates the potential for autoimmune responses and/or inappropriate insertion into our own genomes through a process called recombination. There are groups researching the potential link between this DNA and autoimmune diseases such as juvenile (type I) diabetes, multiple sclerosis and lupus. Our organization, Sound Choice Pharmaceutical Institute, is focused on studying the quantity, characteristics and genomic recombination of the aborted fetal DNA found in many of our vaccines.

Preliminary bioinformatics research conducted at SCPI indicates that “hot spots” for DNA recombination are found in nine autism-associated genes present on the X chromosome. These nine genes are involved in nerve-cell synapse formation, central nervous system development and mitochondrial function.

Wow, that sounds all science-jargon-y! It’s also wrong.

First off, the viruses used to make some vaccines come from human cells, which were derived from aborted fetuses . . . back in the 1960s and 1970s. Since then, they’ve been propagated in cell cultures. If the DNA from those cells is the cause of autism as Deisher, et al., claim, then why was the increase in autism diagnoses not seen until the early to mid-90s?  Secondly, at least regarding the way Deisher is trying to explain it (we’ll come back to Mark), the whole thing doesn’t even make sense: she never explains how the “recombination of aborted fetal DNA” would trigger autism.  Were the aborted fetuses carriers of an autism gene?  (If so, that really undermines the idea that vaccines cause autism, considering the fetuses had not been vaccinated.)  Or is it simply that introducing DNA from other people causes a person to develop autism?  (If so, we should expect a similar incidence of autism among children who received a blood transfusion or a bone-marrow transplant or any of a host of other procedures that introduce foreign DNA in much larger doses than a vaccination does.)

Or — and this question is really big one, so I’m setting it apart in its own paragraph — if causing a permanent genetic change in a human is as easy as injecting a minute amount of some other DNA into someone’s arm, why has gene therapy taken over thirty years to develop reasonably long-lasting treatments?  (Or, since the anti-vacc argument here suggests that it doesn’t have to be autism genes that are injected and that the introduction of any foreign DNA can trigger autism, are anti-vacc advocates willing to suggest that gene therapy should never, ever be performed on children?)

Related to the previous hypothetical, do the opponents of these vaccinations also oppose the use of stem cells to treat disease?  (I know Dr. Deisher doesn’t; she touts herself as an “internationally renowned” stem-cell expert and is part of a group designing stem-cell-derived vaccines for children, because, apparently, that DNA . . . uh . . . won’t cause . . . things?  Conflicts of interest: Dr. Deisher has them.)

Also, riddle me this: the fetal-tissue-derived vaccine at issue here is the MMR vaccine, yet no one is suggesting that children who contract measles, mumps, or rubella “naturally” show an increased incidence of autism, right?  Except, since those diseases are human diseases, when little Timmy catches measles from little Sally, the measles virus in Timmy’s body contains traces of Sally’s DNA, and there’s a LOT more of that DNA in the contagious disease that he contracts than there would be in the vaccine against the disease.  Though, I suppose that saying that contracting the disease can also cause autism — which it should if there is any scientific validity to their argument — undermines the idea that vaccinating against the disease is a bad idea.

Of course, the answers to the above questions show that the argument is without merit, but so does the mere fact that the logistics involved in this argument make no sense.  Vaccines are injected intra-muscularly.  The amount of DNA present in them is measured in nanograms.  That DNA, to the extent it is absorbed anywhere, is taken up by the surrounding muscle tissue.  Yet this argument would suggest that such a microscopic amount of DNA, once taken up by a few muscle cells in the arm or hip area, somehow continues to spread and replicate such that it winds up affecting the brain while exposures to larger amounts of DNA do not?

Well, forgive my reliance on an old Carl Sagan cliche, but extraordinary claims require extraordinary evidence.  And the proponents of this theory have offered exactly zero.  Rather, they’ve made broad, sweeping generalizations designed to pique the interest of the anti-vacc crowd:

“Now when we vaccinate our children, some vaccines also deliver contaminating aborted human fetal DNA. The safety of this has never been tested,” says Dr. Theresa Deisher, President of the Sound Choice Pharmaceutical Institute (SCPI).

SCPI, a group that educates the public on the use of aborted human fetal material for drug production, warns that the MMR (measles, mumps and rubella) vaccines introduced to the US and UK in 1979 and 1988 respectively, were produced using aborted fetal cells, while previous versions were made using only animal cells. This switch coincides with what SCPI says are “dramatic” increases in the rates of regressive autism in children, in which the child’s social and verbal development halts.


Which, again, brings us full-circle to Mark’s comment about “rouge human proteins.”  To the extent that he’s referring to DNA generally, everything we’ve just covered explains why that’s incorrect.  To the extent that he might be talking about a specific human protein, I note that every human protein contains DNA, so the argument would have to be that the specific protein (rather than the DNA in it) causes autism.  Not surprisingly, there’s less support for that argument than there is for the other anti-vacc/autism arguments; at least the latter have some incorrect, highly flawed “data” to support them.  The protein-as-protein argument can’t even claim that.

I understand the innate desire of people like Mark to know why a loved one has autism.  I really do.  If it were my child, I would also be searching for an answer.  But there’s a difference between searching for a real answer and looking for stuff that supports a pet theory.  And that’s all the anti-vacc movement is: a pet theory for people who want to understand something and who ultimately want somewhere to lay the blame.  My hope is that this mindset will be changed as we see more studies like this one, which, using diffusion tensor imaging MRI, found evidence of brain abnormalities as early six months of age in children who later manifested autistic symptoms.  Such studies greatly undermine the idea that the 18- or 24-month vaccinations were the root cause of autistic symptoms.  As the study’s authors explain:

It is clear that the neurodevelopmental story of ASDs neither begins at 6 months of age nor ends at 24 [months].

This is far from the only study to find visible evidence of ASD in the brains of children around six months of age, which fits with the studies that suggest ASD-type behavior is also visible as early as six months.  Of course, suggesting that autism is genetic puts the “blame” (as it were) on the parents, and that’s simply too uncomfortable for many people to accept.  However, it’s only uncomfortable because (a) autism doesn’t present signs that the average person can identify until well after birth and (b) people like the anti-vacc types scream loud and long about how “normal” children were “turned autistic” because of [insert trendy-but-scientifically-incorrect assertion].  Education can address part (a) by teaching new parents what to look for in the first six to twelve months of their child’s life, and maybe educating parents as to part (a) will lead fewer people to listen to the group in part (b), fixing that part of the problem as well.