Saturday, March 04, 2006

Below Junk Science

The junk science team of Geier and Geier (father and son) has outdone themselves this time.

Once, they were content to mine the contaminated VAERS database (also see here) to find the data that they or their clients had planted there. It was a simple thing, really. They knew that groups like “Defeat Autism Now!” had been encouraging their members to report to the VAERS database that their autistic children (and their family members’ children and their neighbors’ children and…) had been injured by vaccines.

It should be no surprise that when they looked at the VAERS database, the Geiers found those reports. This “data” does not support the “mercury-causes-autism hypothesis” any more than finding Easter eggs at an Easter egg hunt would prove the existence of the Easter Bunny

But now, they have begun to step away from bad “epidemiology” and move into a new realm. In 2005, they announced their new business – Fairy Tale Science - in the "journal" Medical Hypotheses (the National Inquirer of medical journals - see here). In their introduction, Geier and Geier trot out all of the “usual suspects” in the autism-mercury world – including the article by Redwood, Bernard and Brown that predicts autistic children will have higher hair mercury levels than non-autistic peers and the article by Holmes, Blaxill and Haley that shows that autistic children have lower hair mercury than non-autistic controls.

[Note: the autistic children in the Holmes et al article had hair mercury levels about twice the average for their age range found in the NHANES study of over eight hundred children – the “normal” controls had hair mercury levels over fifteen times higher than the NHANES average. This highlights the danger of ignoring data that doesn’t fit your preferred hypothesis - and the dangers of using a lab like "Doctor's Data".]

After this attempt to muddy the waters by bringing in contradictions, irrelevancies and poor research – including “research” that involves absolutely no data – the Geiers spin their “Just So” story about how mercury raises testosterone levels and thereby causes autism. Drawing on a single 1976 publication showing that the enzyme that converts DHEA to DHEA-S is inhibited by mercury (actually, any enzyme that transfers a sulfur would probably be adversely affected by mercury), they assert that this is the piece of the puzzle that explains how most of the autism “therapies” work. Of course, they omit the possibility that the apparent effectiveness of these autism “therapies” may be an artifact of poor research (“no research” being a subset of “poor research”).

Their article provides a number of poorly designed flow charts that purport to show how inhibition of this one enzyme greatly increases testosterone in both autistic boys and girls – never explaining how a “buildup” of DHEA would not lead to elevated estrogen in girls. Actually, this is just a small part of the unexplained features of this “hypothesis”. Here are some others:

[1] Studies are cited that show elevated testosterone in autistic subjects, yet no data is given to show how these elevated levels could cause autism. Certainly, elevated testosterone is linked with aggression, but not (yet) to autism. Causation is not shown, only an association.

[2] No data is given to suggest that DHEA is elevated in autism – this would have been a relatively simple test to do.

[3] No data is given to suggest that the enzyme converting DHEA to DHEA-S is impaired in autistic subjects, or that inhibiting this enzyme would raise testosterone levels. This is especially true for females, who it seems would be as likely to develop elevated estrogen.

[4] No data is given to suggest that the enzyme converting DHEA to DHEA-S is significantly impaired at mercury levels seen in autistic children. While you can argue that retained mercury could cause continued brain impairment, it is hard to imagine that there is no turnover of this enzyme in the years since these children were born. And it also hard to imagine that – given the extensive homeostatic regulation in larger eukaryotes (like humans) – that no increase in enzyme synthesis would occur as a result of a buildup of DHEA.

So, starting with two pieces of real data ([a] some autistic children have elevated testosterone levels and [b] mercury can impair the function of one enzyme related to testosterone production), the Geiers have spun a tale of how mercury can cause autism, how various unrelated autism “therapies” are working through this common pathway, and how giving autistic children drugs to reduce testosterone can make them “all better”.

Is anyone else seeing the holes that I do? There is no data to support what they are saying, beyond the rather thin threads mentioned above. Now, if they were advocating the use of a water-soluble vitamins and absorption-limited “minerals” (ala Rimland), that would be one thing. The risks of that sort of “therapy” are low and the costs are likewise low. However, Lupron (the drug they are advocating on their “snake-oil circuit”, if not the only drug mentioned in their article) is potentially very dangerous and it is very expensive.

A few minutes on the phone with the local wholesale pharmacy revealed that the cost of an average dose of Lupron is about $1200 (US). After several more minutes on the phone with Blue Cross (a health insurance company), I discovered that the “patient contribution” for this drug – if the drug was indicated - would be about $240 (they were rather vague about the number) a dose.

Now, from the chatter on the Internet, it appears that Geier and Geier have been saying that reducing testosterone will help remove mercury from autistic children. Apparently G&G have been saying that mercury combines with testosterone to form “sheets” and that reducing testosterone levels “frees” the mercury.

This – whether it comes from G&G or not – is pure fantasy. There is nothing about testosterone that would make it have a greater affinity for mercury than its parent molecule, cholesterol. And there is whole lot more cholesterol in the body – even a pubescent male body – than testosterone. So, if mercury could form “sheets” with testosterone, it would be even more likely to do so with cholesterol, which is so much more prevalent. Let’s not even mention estrogen, progesterone, DHEA, cortisol and all the other steroid hormones – all of which are equally likely (or, more precisely, equally unlikely) to bind with mercury.

Geier and Geier, in their Medical Hypotheses article and their advocacy of Lupron “therapy” for autism are setting new lows for their “research” standards. This stuff isn’t even up to the standard of “Junk Science”.



Blogger Big Lebowski Store said...

Well said, Prometheus,

I looked into this at Kathleen Seidel's request. I found only Geier's on this and no one else.

It strikes me that even families with children diagnosed with autism that are suffering terrible stress wouldn't get nutty enough to buy this story. It's way too weird.


04 March, 2006 18:22  
Blogger Interverbal said...

Nicely done Prometheus.

The Geiers flubbed the IDEA data as well; when I get a moment I am going to do a post how they they managed to mangle that data base.

04 March, 2006 19:48  
Blogger Bartholomew Cubbins said...

They're full of sheet, that's for sure.

I'm too lazy to check on daddy G's CV, but it would be interesting to know what his former coworkers and advisors (if they're still alive) have to say about his antics.

Maybe his freshman Chem and sophmore Biochem instructors could send daddy G a copy of the relevant textbooks to remind him of the basics of biochemistry.

I've got my fingers crossed that I never go insane and make such scientifically idiotic statements in public (happy hour aside).

I really like the way you took these guys down, I'm going to link to it.

04 March, 2006 20:32  
Anonymous Anonymous said...

As far as I can see, the article was published not in Medical Hypotheses but in the Journal of American Pysicians and Surgeons (Vol II Numer 1). This journal apparently is peer-reviewed and with some sort of quality control.

05 March, 2006 00:40  
Blogger Big Lebowski Store said...


The Geiers have indeed published in JAPS, but not the article we're talking about, re: testosterone.



05 March, 2006 04:01  
Blogger Prometheus said...

The Journal of American Physicians and Surgeons (JoAPS) is technically a peer-reviewed journal, but their peer-review process often appears to be subordinate to their political aims. If you read JoAPS (which I have, on occasion), you will see that it has a distinct anti-government and anti-establishment tone. This is not necessarily a bad thing, but allowing its political goals to trump good peer-review is probably why it is not indexed on MedLine (even Medical Hypotheses is indexed on MedLine).

Besides the Geiers, JoAPS has provided a platform for none other than Jeff Bradstreet, MD. In fact, JoAPS is the only journal that has published Bradstreet's highly questionable conclusions on autism therapy.

In my career, I have come across a few journals that so consistently publish bad science that I would be embarrassed to have my name in them. Medical Hypotheses is one and JoAPS is another.

BTW, the Geier and Geier article in question is in Medical Hypotheses - just follow the link to its abstract.


05 March, 2006 08:31  
Anonymous Anonymous said...

Truthfully, the "Journal of American Physicians and Surgeons" is really more of a political publication than a medical one. It is NOT indexed at PubMed.

It is more of a joke than "Medical Hypotheses"!

05 March, 2006 11:36  
Anonymous Anonymous said...

I stand informed. Thanks.

05 March, 2006 12:48  
Blogger The Probe said...

There can be no question that the Geiers produced this fish-wrap to substantiate their testimony fees. According to Quackfiles,, Geier's testimony has not washed well 11 times. Wikipedia also mentions this number. I would not be surprised if it were considerably higher.

07 March, 2006 09:56  

<< Home