When not to write about autism

Comment: When not to write about autism
18:26 04 November 2008
NewScientist.com news service
Ewen Callaway


USA Today heads its story: Study: Counties with more rainfall have higher autism rates. The BBC has Rainfall autism theory suggested, while The Daily Telegraph opts for Heavy rainfall could be linked to autism, scientists claim.

These were some of the headlines in stories reporting a paper from scientists at Cornell University showing that between 1987 and 1999, counties in Washington, Oregon and California that got more rain had more cases of autism.

There is no claim that rain causes autism, and the authors are exceedingly conservative in making the connection. They argue that some unknown environmental factor related to precipitation or accompanying behaviour might contribute to autism.

To be fair, the sources mentioned above managed to keep sensationalism to a minimum, but others were not so restrained.

The Palm Beach Register, for example, had a Milli Vanilli-inspired headline as lame as it is misleading: "Autism: Blame it on the rain". Their opening line goes as far as to mention the discredited link between autism and a mercury preservative used in some childhood vaccines.

Misleading the public?
But should the story have been reported in the mainstream media at all? It offers nothing useful for the general public, parents, and even physicians. And press reports, blogs and other accounts of the study could even mislead the public.

I spoke with the study's lead author Sean Nicholson of Cornell last week. He is an insightful economist, and he carefully went over his team's data and its limitations with me.

His team accounted for some variables that might confound an association between autism and precipitation, and they back their claims up with impressive statistics.

They also underscore the speculative nature of the connection between autism rates and rainfall throughout the manuscript, which took a year and a half to get published.

Unclear links
But, Noel Weiss an epidemiologist at the University of Washington in Seattle who wrote an editorial on the paper – peculiarly titled Do These Results Warrant Publication – makes an important point in answering "yes" to his rhetorical question:

"The authors' analysis and the editor's decision to publish it are to be lauded, despite the uncertain ultimate contribution of this work and the possibility (likelihood?) that non-professionals are going to misinterpret and misuse it."

The paper, he writes, is best digested by other epidemiologists, who may use the results to design trials that address potential links to autism that could be related to precipitation and related behaviour, such as lack of vitamin D exposure.

"Hopefully someone will be inspired to do a study that will be more direct and ultimately more persuasive for clinicians," Nicholson agrees.

Most media reports mention vitamin D and other "theories" – TV, household chemicals, and computer games – and the need for follow-up research to confirm these links.

Yet few reports say that these studies likely won't pan out under the microscope of well-designed clinical trials. Nor do they mention the importance of studying individuals with autism, rather than county-wide records, to gain true insight into a potential cause. This is epidemiology 101.

Skip the story
It's unrealistic to expect reporters to always wait for such confirmation before reporting on a study, especially one so tantalising.

An inevitable part of science and medical writing is covering research that others will eventually discredit. It's nearly impossible to determine what will stand the test of time – especially under deadline.

However, irresponsible reporting of the unproven link between autism and vaccines probably led some parents to pursue unproven and potentially dangerous chelation therapies for their kids, or to miss out on protective vaccines.

I can see worried parents hearing about the rain association, second- or third-hand, and keeping their kids in on showery days, or forcing them to play in the rain, or whatever "news you can use" suggestion gets tagged on to these stories.

For this reason, and it might be idealistic, but I think reporters and editors should have taken a pass on this story. There must have been a story on the benefits of broccoli out there.

Journal reference: Archives of Pediatrics and Adolescent Medicine (vol 11, p 1026)