Archive for March, 2018

Vaccines and Autism: the Final Word

Friday, March 30th, 2018

OK, let’s do this…
I grieve for the state of public discourse in this nation. It’s seems that way too much of the discussion and debate is focused on endless reiterations of the same tired arguments which, to any sensible person, have been over for a long time. I mean… flat Earth? seriously?!? Yes, there are quite a number of people who are seriously arguing that Earth is flat.

Another “you’ve got to be fucking kidding me” argument is the claim repeated endlessly that vaccines cause autism. They don’t. “Do vaccines cause autism?” has been one of the most studied questions in all of science over the last twenty or so years. And the answer, over and over and over and over, is NO. Studies in all of the most prestigious relevant journals. Big studies, medium sized studies, short-term studies, long term studies. In vitro. In vivo. Mercury in vaccines does not cause autism. Aluminum in vaccines does not cause autism. Formaldehyde in vaccines doesn’t cause autism. Vaccines don’t cause autism when they’re administered several at a time. Vaccines don’t cause autism on a fewer, slower schedule. There’s no known mechanism by which vaccines could possibly cause autism. While the symptoms of autism may show up after vaccines have been administered, the disorder’s causes are genetic and epigenetic, so it’s spectacularly improbable that vaccines administered well after birth could be a cause. The question has been answered. It’s not up for debate. It does not need further study.

Nonetheless, the claims keep coming. There seems to be a cottage industry of professional quacks, cranks, and crackpots endlessly recycling the same drivel and a herd of credulous paranoid laypeople who parrot anything that supposedly supports their worldview.

In a completely unrelated Facebook thread a little while back, I was commanded to refute this “peer reviewed” study from a “.gov website”:
“A two-phase study evaluating the relationship between Thimerosal-containing vaccine administration and the risk for an autism spectrum disorder diagnosis in the United States”
Unfortunately, I have other things to do besides produce detailed rebuttals to pseudoscience sometimes, so it took a while… several months, in fact. But here goes. Let’s look at the study, the journal it was published in, and the authors. Perhaps this will serve as a microcosm of the larger discussion which we really shouldn’t be having any more. Maybe we’re talking about this specific study, maybe we’re talking about the supposed link between vaccines an autism, maybe we’re talking about the state of academic journals today, maybe we’re talking about the sad state of scientific literacy in our country. Or some mix of all those things.

The Study

Yup, the study is on a .gov website, alright. But that site is just a great big database of medical studies. Inclusion does not impart any particular credibility to the study. The study was (electronically) published in an online journal called Translational Neurodegeneration.
More about that journal and scientific journals in general later.

But, on to the study itself. Here’s my detailed, comprehensive analysis of this study: so what?

Seriously. So what? A group of folks re-crunched some carefully selected old numbers and claim to have found a correlation in that particular dataset between vaccines containing thimerosal and autism. That is all. So fucking what. Do I have to say it? Really? OK. Correlation does not prove causation. Anybody who has ever done a high school science project is supposed to know this. One more time: correlation does not prove causation. Repeat as many times as necessary for this to stick in your brain.

This study does not address the issue of why so many other studies with larger, more reliable datasets have found no correlation. It does not propose a mechanism by which vaccines could cause autism. It does not address the fact that diagnoses of autism continued to rise long after thimerosal was removed from childhood vaccines. It does not address the question of why it’s the mercury in vaccines supposedly causing autism when there are other sources of more mercury (and mercury in a form which accumulates in the body) in people’s lives. Plus, it uses the unreliable, unconfirmed data from the VAERS database.

Admittedly, I don’t have the medical and statistical chops to perform my own detailed analysis of the specific methods and statistical techniques used in this paper, but the foundational flaws are so obvious, it doesn’t really seem necessary. I poked around the web anyway, to see if anyone had analyzed this study. I didn’t find much about this particular study. I found lots about the authors, though. More on that later. The closest thing I could find to an analysis of this specific paper was this dismissive tidbit from Matt Carey: “Ever heard of that paper? That’s what happens to mediocre science published by biased authors. No one cares.” In short, so what?

The Journal

Scientists communicate their findings to each other through peer reviewed technical journals. For the elite journals, this peer review process is brutal. Leading experts in the fields relevant to the paper analyze every aspect of the study, from the original question through the study’s design, data gathering, statistical analysis, and conclusion looking for flaws. When a study goes through this process and is accepted for publication, it has earned a degree of credibility. The conclusions are not automatically correct, but there’s real support for them. It should be noted that one study, even a well-designed one, does not instantly overturn established scientific consensus.
Less prestigious journals are a mixed bag. For some, the peer review process is nearly as exhaustive as the most prestigious journals, while others have only a thin veneer of peer review. Still others are predatory journals  which will publish anything the authors pay to have published. Their “peer review” is a bad joke. There are several cases where people have pranked them by sending gibberish papers. As long as the check clears, they sail through “peer review”.

So, what’s an interested lay person to do? How do we tell if a particular journal is really peer reviewed or if it’s “peer reviewed”? There isn’t one absolutely reliable gauge for a journal’s credibility, but it’s easy to check three important factors:
• Impact factor. This is a measure of how many times the articles in this journal are cited by other researchers. In general, better research is cited more, so journals with more highly-cited studies are more reliable.
• Longevity. Shady journals seldom last long. Newcomers are suspect.
• Paper publishing. Printing on real paper costs money, and it’s dependent on subscriptions from real libraries to make it viable economically.

Sure, there are excellent journals which e-publish, haven’t been in business for long, and don’t have an impressive impact factor yet. But those are the exception, rather than the rule. But studies in journals which match those criteria warrant extra scrutiny.

Here’s a good overview of the state of peer reviewed journals today:

So, what about Translational Neurodegeneration? It was established in 2012. It e-publishes. According to Scimago Journal & Country Rank, its H-Index (a measure of impact) is 19, averaging 3-4 citations per document. For comparison, Cell, one of the tippy-top journals, has an H-index of 655, averaging around 28 citations per document. Cell has been publishing on paper since 1974. I don’t see any reason to accuse Translational Neurodegeneration of being a hard-core predatory journal, but it’s not especially prestigious, either. One point in favor of the credibility of this journal is that they did have the editorial integrity to retract a paper by Brian Hooker, one of the authors of the study in question here.

For what it’s worth, the specific study under discussion here has been cited 11 times, which sounds sorta-kinda halfway impressive, except that 7 of the citations are by authors of the study. On the one hand, there’s nothing wrong with self-citations. Statements along the lines of “in my previous study, I determined a-b-c; in this study, I’ll attempt to determine if d-e-f necessarily follows” are perfectly reasonable. On the other hand, self-citations do not indicate that the paper in question has made a significant impact in the field.

The Authors

Hang on, folks, the ride gets wild at this point.
Two of the authors of this study are the father and son team of Mark and David Geier. They have a long history as anti-vaxx researchers. Mark has some legit degrees and was, until recently, a practicing MD. David has no discernible degrees or other qualifications, but that doesn’t seem to stop him from practicing medicine and authoring research papers. One consistent theme of criticism of the Geiers’ anti-vaxx work is that they basically write the same paper over and over again. They use the same unreliable, old data sources, the same dubious statistical analyses, selectively report only the data that supports their pre-determined conclusions, and cling to the same a priori assumptions in study after study.
“…the studies by Geier could not establish a causal relation between MMR and autism because of their methods—such as using statistical measures incorrectly and omitting facts about their research approach. Similar problems were found in six other studies by Geier.”
{Wilson K, Mills E, Ross C, McGowan J, Jadad A (2003). Association of Autistic Spectrum Disorder and the Measles, Mumps, and Rubella Vaccine: A Systematic Review of Current Epidemiological Evidence. Archives of Pediatric and Adolescent Medicine, 157:628-634. quoted in}

Until recently, the Geiers had a substantial following in the anti-vaxx crowd. Mark testified as an “expert” in over 90 legal cases about vaccines. They both spoke at anti-vaxx gatherings around the USA.

But what has really made the Geiers famous (or infamous) is their “Lupron Protocol” for treating autism. Basically, they got an idea in their heads that testosterone binding to mercury was the root cause of autism. Do I need to mention that there was no credible medical or biochemical basis for this? They used human subjects for trials of their treatment which involves chemical castration of children. I do not understand why the Geiers are not in jail. The results of this horrific, unethical study are that the Lupron Protocol doesn’t work (duh!), Mark has lost his license to practice medicine, David has been fined and booted from Maryland’s Autism Commission (why was he there in the first place?), and the Geiers are disgraced, even in the eyes of most of the anti-vaxx crowd.

I was going to go into some detail about the other authors of this study, but this draft has been sitting around for a couple of months, so I’m just going to post it. Sufficeth to say, they’re professional crackpots who’ve made their careers in the field of bad science.

So, in conclusion, the vaccine-autism “debate” for lay people boils down to: who are you going to believe, the overwhelming consensus of top researchers in relevant fields and their dozens of large, carefully-controlled,  brutally peer reviewed studies in all of the most prestigious journals; or a small band of crackpots recycling unverified data and morphing it into obviously bad papers which would be laughed out of a seventh grade science fair but are nonetheless published in dubious journals so that anti-vaxx cuckoo birds can have “peer reviewed” studies to cite in an attempt to look credible?

This site is protected by Comment SPAM Wiper.