Ben Goldacre
Saturday February 24, 2007
The Guardian

It’s hard to know how you’d react in a crisis, and so if there’s one group of people who can’t be blamed for the weirdness around autism, it’s parents. But academic journal editors have different responsibilities. The current issue of Lancet Neurology has a review of a book on autism: the book is for a lay audience, and it flatters the views of the growing fringe autism movement on speculative biological causes and treatments for the condition. The review is by a man with a long and worrying history of working on that fringe – and it’s certainly very flattering.

Now I’m not advocating censorship for one moment, but I do think that in a situation as extreme as this, it’s only fair to give the reader some background.

The reviewer, Mark Geier, is an American doctor, vaccine activist, and highly criticised expert witness in vaccine cases, who promotes the idea that mercury in vaccines causes autism, and that testosterone is somehow also implicated. He performs expensive, drastic, controversial, experimental and unproven treatments on children with autism, including “chelation” to remove this mercury (a procedure from which a child died in 2005, under the care of another activist doctor) and “chemical castration” of children using leuprorelin, a testosterone blocker.

Research on mercury in vaccines and autism suggests no link. And the evidence on chelation therapy and autism doesn’t support its use, nor does the evidence on “chemical castration”.

If you are going to use such radical treatments, then I would argue that you might want compelling evidence, rather than a theory, and also gold standard ethics committee involvement in every experiment you do.

Mark Geier does research with his son, David Geier, who is president of MedCon, a medical-legal consulting firm that helps vaccine injury claimants get compensation. Dr Geier (Mark) was a recipient in the multimillion pound MMR expert witness payout.

In more than 10 of his legal cases, particularly the more recent ones, Dr Geier’s testimony was either excluded, or accorded little or no weight, after it was found he was testifying beyond his expertise. He had “largely irrelevant” qualifications, and acted as a “professional witness” in areas for which he had “no training, expertise, and experience”. His “speculation” was directly contrary to the conclusions reached in well-respected and numerous epidemiologic and medical studies ranging over two decades”. He was “neither board certified nor [had] training in paediatrics and paediatric neurology”.

He was further criticised [PDF] by judges for his work not being “based on scientific validity, valid methodology, peer review or testing, and more than minimal support within the scientific community”.

My favourite of all was the finding that his testimony was “intellectually dishonest” and his affidavit was “nothing more than an egregious example of blatant, result-oriented testimony”.

A recent academic paper of the Geiers has been retracted by the journal without explanation – so far – and some of their published “scientific” work is so laughable you can explain the flaws in a jokey national newspaper column.

In 2003, they published a paper in the journal of a Miami hospital using data from a voluntary reporting system for adverse events. They compared the number of reports of autism after use of the vaccine MMR and the vaccine DTP, and found the MMR jab was more likely to be associated with autism.

But the DTP vaccine was a very odd choice for comparison as it is given to babies at two, four, and six months of age, whereas MMR is given to toddlers at 18 months on. You aren’t very likely to find autism in a four-month-old baby, since it is a developmental disorder affecting things like language – and not a lot of four-month-old babies can speak.

That didn’t stop the Daily Mail from giving the study a glowing write up, and the headline “MMR raises risk of brain disorders say researchers”.

As I say, I’m not hostile to people like Geier having a voice. And the idea that there might be a biological cause, or treatment, for autism is a seductive and interesting one. All I ask is that when you take someone as far out as Geier, and bung him in an academic journal reviewing a slightly maverick book, you owe your readers a tiny bit of a warning.

· Please send your bad science to science@guardian.co.uk

Alice wins an as yet unspecified prize for being the first person to send in a photo of herself wearing a stylish yet subversive badscience.net bib. More badscience t-shirt nonsense here.