Pediatrics has a now famous study in which it was found that kids exposed to certain pesticides had significantly increased risk for ADHD.

There is universal praise for this piece of garbage, bow to the only true god: Evidence!

Michael L. Goldstein, MD, who was not involved in the study, said the study results are "very interesting findings from a very well-done study from a good database." The report, he said, "certainly got my attention when I read it; I was really impressed by it. I think it is a groundbreaking study..."

There are no dissenters, the worst that is said is "well, correlation isn't causation."

Sorry, there is one dissenter-- a pesticide rep:

Garry Hamlin of Dow AgroSciences, which manufactures an organophosphate known as chlorpyrifos, said he had not had time to read the report closely. But, he added, "the results reported in the paper don't establish any association specific to our product chlorpyrifos."

Ha! And the fact that he objects must mean it's true.  (Never mind chlorpyrifos was specifically not associated with an increased risk in this study.)

But the problem isn't the data-- in psychiatry, the problem is almost never the data.




The study purports to show some relationship between pesticides and ADHD.  But this is silly; a priori, pesticides at high enough doses disrupt cognitive function.  Duh? So what's the point of the study?  Have they found that kids are getting exposed to much higher levels to pesticides than we ever thought?  Or that the pesticides are even more toxic than previously known?  No.  So?

The purpose isn't to show that pesticides can cause ADHD; the purpose is to solidify in your mind that ADHD is real.

While you're busy teasing apart statistical or methodological issues, such as whether a single static urine sample is sufficient to infer pesticide exposure in general, or the lack of control for diet, or how did they calculate the odds ratio?  you are not questioning whether ADHD is a real disorder, and that what they say it is here is what others say it is elsewhere.

But it's a trick.  You cannot use a correlation between two variables to prove the validity of one of the variables.

You cannot use a statistical correlation between pesticides and ADHD to reinforce the validity of ADHD, but that is what they did.  And everyone fell for it.

II.

Note that the rate of ADHD in this study (12%) is double the national average.  You don't have to agree with me on this, but:  that national average is likely inflated, i.e. kids were diagnosed ADHD that should have been bipolar, or anxious, or normal.  So whatever happened in this diagnostic sniper attack (telephone interviews of parents, and did not exclude symptoms that could/did occur in another disorder) is suspect.   

Regardless, it is indisputable that under the best circumstances-- a two hour interview with PET scan and genetic testing-- the diagnosis "ADHD" is not biologically homogeneous.  Whatever Billy has, even if it appears identical to what Jamal has, may be biochemically completely distinct.  Early HIV and the flu both appear identical in the beginning, but I'm not having sex with either.

This doesn't mean that "it doesn't exist"-- the behaviors are real. Nor does it mean the same medicine (e.g. Adderall) can't help it/them.

So now to the study, which finds a correlation between pesticides and an increased risk for a report of a cluster of behaviors.  No argument from me at all.  But the rigorous explanation of this study is that pesticides increase the risk for these behaviors, not "ADHD."

III.

If I've lost you, think of it like this: if this study found not just correlation but causation-- if the study proved that pesticides actually caused all of the symptoms that are detected by the ADHD inventory, then that isn't ADHD, any more than psychosis caused by cocaine is the same as schizophrenia.  This would be pesticide poisoning.  And if you're treating it with Adderall, you're an idiot.

IV.

Stated explicitly: the purpose of these studies is not to establish a link between variables, but to gain popular support from the public for one of the variables so that it becomes insanity to ask whether it's real.

Here's an example.  If the study determined a link between pesticide exposure and IQ, you'd say, "oh, I guess that makes sense that pesticides would lower intelligence."  If you do enough of these studies, say, 10,000 of them, you wouldn't have learned anything more about pesticides and intelligence-- because what else would there be to say? but you would have drowned in the passive assertions that IQ measures intelligence.  So it would become insane to question whether IQ is a good measure of intelligence ("oh come on, you're not asking rigorous questions, we've been doing it this way for decades").  At best-- another trick-- the debate may turn towards whether the Wechsler test is a valid measure of IQ. 

Maybe it is, maybe it isn't; but there are other kinds of studies that would tell us that.  You can't use these correlation studies to reinforce the validity of IQ.  But that's what would happen,10,000 garbage studies later.  "Oh, yeah, IQ.  Everyone knows about that."


V.

"You're just silly.  You haven't discussed construct validity..."

If you had this thought, then you're the problem.  I have discussed construct validity, without saying the words "construct validity," because the words distract from the discussion.  Same with "statistical significance."  But that's the move, drown the public in jargon and they'll cling tightly to whatever they do understand, which in this case is the title: "Pesticides linked to ADHD."

Every second you spend talking about construct validity is a second you aren't talking about the fact that the kid in front of you has something far more serious than ADHD.

If you are a talking to a scientician and they say "construct validity" or sling around a chi-squared, grab their tie and slam their face into the desk, and then run, run, run, you are being lied to.

VI.

"Why didn't you publish this post earlier, when the story first came out????"

You mean because the further away we get from the hot news cycle, the less you're interested in the truth? And so apathy allows it to stand as true, because it all seems plausible (face validity, bitches.)  That's social science for you.

But the answer to your question is simple: I don't get paid to criticize studies.   I have a job and a life and an alcohol problem that doesn't leave much room for blog posts.

If you don't like something going on in politics, there are ways to voice that.  But looking deep into a "scientific" concept requires a level of sophistication which, if you possess it, means you're probably busy with a day job.  So you have neither the time nor energy to dissect articles in Pediatrics-- and, anyway, no forum.  I, at least have a semi-popular blog;  but even so I'm subject to the same time and monetary constraints.  If I need money for rum, do I spend an extra hour at work, or write a post?

Had it not been for the little money I do get from this blog, maybe I would not have written this post at all, because... why bother?  And so all you are left with is their dogma.  Search the internet, the Letters To The Editor.  Anyone else saying what I just said in this post?  You may not agree with my post, but you are nevertheless better for having read and considered it.

I'm not saying I'm the only one who thought it; I'm saying that those who thought it, or could have thought it, are busy with other things, and the people who should have thought of it physically can't.

What do you think would happen if the NIMH paid astrophysicists to look into psychiatry? Ain't never gonna find out, they have their own problems.  Specialization doesn't mean I'm the best, it means stay off my lawn.  

Since I/us/astrophysicists don't have the resources to publish at all/as fast as Pediatrics or CNN, they get to decide what's true.
 
The people who get paid to look at ADHD are all unanimously on the same side, that's what conveys the illusion of objectivity.  Pretending to be divided pro- or anti-Pharma, or "dopamine hypothesis" vs. "noradrenergic hypothesis" makes it look like there's a lot of vigorous debate, but debating what color is a ki-ran just means ki-rans are real.

Please don't misunderstand me, I am not blogging for money.  I'm showing you that so much of what you think you know, what you think is fact, is established not by the force of evidence but by the absence of resources for the opposition.   The ones with the money are all unanimously in agreement. 

So yes, I would have liked to have posted day one; but I would have liked it better if someone else on the inside said, "umm, wait a minute?  Really?  We're going to CNN with this?"


-------
http://twitter.com/thelastpsych

---

More on pesticides here.

Note: I added another section to the earlier Love Means Not Letting The Other Person Be Himself.