Last week, I wrote a piece here, On the Corn Syrup Theory of Autism, which took a critical look at a Grist story concerning a scientific paper proposing that high fructose corn syrup consumption was responsible for the rise in autism cases in the United States. My point was that the author, Tom Laskaway, was far too quick to embrace that conclusion, to cheer lead for a paper which was, at best, speculative, and, at worst, not credible.
And I wasn’t the only person to make that point. At his blog, Collide-A-Scape, environmental journalist and former Audubon magazine editor, Keith Kloor, called it “shoddy and irresponsible journalism.” And Grist’s readers were just as tough in their response. To partially quote one comment: By writing an article about this terrible study and giving it such a misleading title you are revealing yourself as either a) too stupid to be trusted with the responsibility of writing about science or b) willing to knowingly mislead the public in the service of your irrational personal hatred of HFCS.
As Grist editor, Scott Rosenburg, wrote in the aftermath, a genuine furor arose. Some furors are healthy, he noted; they encourage people to consider new ideas, to challenge their own assumptions. But some furors are more like, “Guys, you messed up.” I’m afraid that from where I sit this was one of the latter kind.
Rosenburg’s post-mortem is both thoughtful and, I think, wonderfully non-defensive. He zeroes in on the essential fault in Laskaway’s piece – that it was a superficial response to a flawed paper about a complicated issue. And he details Grist’s rapid response to that realization.
First, editors changed the original headline from a statement “New Study Links Autism to High Fructose Corn Syrup” to a question: “Paper Asks: Does High Fructose Corn Syrup Contribute to A Rise Autism?”
And second, Rosenburg asked science blogger and writer Emily Willingham to take a more thorough look at the paper itself. Before I tell you what she wrote, let me take a minute here to say that I think tagging Willingham – who is one of the best science writers working today – was a brilliantly right move. She’s one of those writers – as evidenced, for instance, in this January piece for Slate about GMO food alarmist – who always leaves the reader feeling smarter.
Actually, I think I usually am smarter after reading her work. Willingham’s piece for Grist, which ran Tuesday, is titled Why That Corn Syrup and Autism Study Leaves Such A Sour Taste. And I’ve rarely read a better deconstruction – or maybe evisceration is a better word – of a deeply flawed scientific paper. There’s history here, there’s original statistical analysis, and there’s deep understanding of what science is, how it works, and when it doesn’t.
As Willingham notes, the autism paper (technically a review, rather than a study) waves “several red flags of pseudoscience”.
For instance, the authors’ theory of autism is based on the idea that high-fructose corn syrup is contaminated with mercury which they link to developmental disorders. They base this on a three-year old paper by the lead author, Renee Dufault, which claimed to find such contamination. As Willingham points out, studies attempting to confirm that finding were unable to do so. And she points out, the whole mercury question is dismaying vague. The authors randomly rule out any other sources of mercury (including emissions from coal-fired plants) and they fail to even analyze for or distinguish between different types of mercury exposure. Pure elemental mercury, for instance, is far less of a health risk than such mercury compounds as methylmercury (the type that tends to bioaccumulate in fish) or salts of mercury, such as mercury bichloride).
But my favorite part of Willingham’s debunking of the Dufault paper is a graph she produced, based on her careful analysis, showing that U.S. consumption of high fructose corn syrup doesn’t track actually with the increased rates of autism diagnosis. As it shows (I’ve included it at the top of this story) those rates have clearly been rising even as HFCS consumption has been falling.
It’s such a good breakdown that I wish it could be required reading; certainly my science writing students are destined to study it. So I come here, at last, not to criticize Grist but to praise it. This is one of the best responses to a bad article I’ve ever seen. Absolute redemption, Scott Rosenburg!
— Deborah Blum