Send in the Clowns

There’s far more causality between paying attention to the news, or worse social media, these days and becoming deranged, than there is between using Tylenol and either 'getting autism', or if you are a pregnant woman, using Tylenol, causing your child to 'get autism'.

The joint announcement by the current President and HHS Secretary RF Kennedy that Tylenol, or its primary functional ingredient, acetaminophen, causes autism is another example of the erosion of trust and perhaps the down right endangerment to the public, being caused by our current federal guv’mint. I would say the latest, but as is my wont, I am slow in getting this post to post, so I am confident, equally or even stupider things may have been put forth since the Tylenol scare.  

Anyway, a few factoids. The largest study done on possible links between acetaminophen and autism comes from a large (>2,000,000 participants) study in Sweden looking at a whole host of health impacts, one of them being linkages to autism. This study reported:  ‘that around 1.42% of children exposed to paracetamol (another name for acetaminophen) during pregnancy were autistic, compared with 1.33% of children who were not exposed — a “very small” (<0.1%) difference, according to a lead author.’ In fact, when accounting for uncertainty in such a large, and by its nature, uncontrolled study, really no difference at all. Furthermore, these researchers followed up with a smaller study of identical twins and could find no link between ‘tylenol’ use and autism.

It seems the primary ‘evidence’ for this association cited by Kennedy et al, is a recent meta-analysis indicating a relationship between acetaminophen use and neurological disease or symptoms. A meta-analysis is essentially an analysis of analyses. In this case the authors used scientific publication engines to find studies that had words like acetaminophen and autism in the text, and then used something called a Navigation Guide to identify studies that met their criteria of being legitimate studies aimed at finding if there was an association between acetaminophen usage and autism. This resulted in identification of 46 studies, 27 of which showed a potentially positive association between acetaminophen usage and a range of neurological issues in new borns.  This 42 page paper (with 108 references) includes 26 pages of tables that are qualitative (not quantitative) assessments of the scientific literature, and some other stuff. If there was any hardcore statistical analysis in this paper it was hidden somewhere in the background. I should know, since my statistical acumen is embarrassingly weak (in my Ph.D days we were real men and women, and statistics was only something for losers), so the hardcore statistics part of the paper would have been mostly incomprehensible to me.  

Meanwhile another paper published this year also undertook a meta-analysis looking at what different research studies had shown on the connection between acetaminophen and autism. This paper was shorter, 9 pages total and 75 references, but found only 9 studies deemed worthy of assessing the causality between acetaminophen and autism, many of them used in the other paper. The conclusion from this paper was that there was no connection. This paper also did not seem to have rigorous statistics.

The fact these two meta-analysis could only find a comparatively small number studies investigating linkages of autism with something like acetaminophen is itself telling. The root causes of autism, or even really what it is, are bedevilingly complex, really akin to understanding the causes of cancers. Profound autism (that does seem to have a fairly strong genetic component) has increased relatively little, while mild autism cases account for most of the ‘epidemic’ cited by Kennedy et al (perhaps if I had been born 30 yrs later, and my parents had foolishly had me ‘tested’ (perish the thought) I could be part of the epidemic myself). Mild autism is especially challenging to diagnose, or perhaps more rightly to identify specific criteria for what it is, and how to unambiguously test for it. The issue is that understanding ill-defined conditions like autism do require many well-designed studies that use powerful statistical approaches to sort through myriad possibilities. Meta-analyses like the ones presented in these two papers are important approaches to sorting through the different possibilities, for example genetic versus environmental factors, and helping to design better clinical studies. Nonetheless any undergraduate at ‘Whatsamatter U’ would get a well-deserved F for a paper that analyzed the currently weak scientific literature on the subject, and categorically concluded: ‘Don’t take Tylenol because it causes autism!’

This gets to the larger point that this administration is mis-using science in the worst ways imaginable to make a pre-ordained judgement sound like it has ‘scientific backing’.

It’s hard to fully figure out what the backroom game is here. Obviously, Kennedy has primarily made a name for himself by A) being a Kennedy, and B) questioning authority(?!?) by pushing the link between vaccines and autism among other wacky health ideas (well not all wacky, he’s supposed to be opposed to the most widely used, legal, and dangerous substance in the world: high fructose corn syrup, but hasn’t said a peep about it since he came into and under the power of the food industry lobbyists). Still where acetaminophen comes in seems a mystery. Kennedy has at least reasonable synaptic qualifications, he attended Harvard in the early 1970s,  given the times and his family background, no surprise, but also proof he was not the ‘the stupid one’ (i.e. perhaps profoundly autistic). So perhaps, it’s that he’s found himself in a conundrum of his own making, and is so desperate not to be hoisted by his own petard that he’s grasping at any evidence that has the thinnest varnish of authenticity to make some points about autism. Obviously, these guys have egos plenty large enough to not worry about babies having anencephaly or congenital heart defects because their moms didn’t take acetaminophen to reduce a high fever during pregnancy.

Of course, the cynic in me just assumes there’s a financial angle either directly to the President and/or those in his orbit, perhaps preparation for a monetary shake down of company that produces Tylenol, perhaps undisclosed financial ties to some alternative cure for autism that can now be touted, perhaps some skimming on the lawsuits that could well result from this Presidential announcement, or perhaps some unassociated financial shenigans that are either not being covered by the press, or are now buried underneath an avalanche of ‘reporting’ on this latest crazy talk, especially since it coincided with yet another flip-flop on the War in Ukraine. “Keep the sucka’s eyes focused over there, while we pick his pocket over here”

I’ll close by saying I remain confident that real science and the real United States of America will prevail, and this current class of clowns will be revealed for the charlatans they most assuredly are, but the damage being done is very sad.

Fun side riff: the paper from the Baccarelli Lab does not provide any citations for the Navigation Guide, so at face value one would assume it is something they invented (out of whole cloth?), when it was developed in the early 2010’s by others. The fact original papers aren’t cited seems like shoddy scholarship (and weak peer review). Then there is this conflict of interest statement at the end of the article :

"Dr. Baccarelli served as an expert witness for the plaintiff’s legal team on matters of general causation involving acetaminophen use during pregnancy and its potential links to neurodevelopmental disorders. This involvement may be perceived as a conflict of interest regarding the information presented in this paper on acetaminophen and neurodevelopmental outcomes. Dr. Baccarelli has made every effort to ensure that this current work—like his past work as an expert witness on this matter—was conducted with the highest standards of scientific integrity and objectivity."

What’s not said is that he received at least $150,000 for his expert witness testimony (also not sure how ‘highest standards’ and neglecting to cite primary work of others jibe). Finally, Baccarelli is a professor at Harvard, the #1 academic target of this administration, and the journal where this study was published, Environmental Health, largely publishes studies that the current EPA administration would defund and ignore, or given current trends perhaps even prosecute the authors.

Hypocrisy for this crowd isn’t a word, it’s a world view.