In their October 23 opinion piece “Why Does Autism Impact Boys More Often Than Girls?” Renee Joy Dufault and Steven G. Gilbert attempt to argue that autism diagnoses are on the increase because of inorganic mercury content in processed foods. Going a step further, they try to construct a rationale for blaming mercury for the perceived bias in autism rates among boys compared to girls. Using the example of one observational study reporting that mercury affects chemical tagging of a single gene in one cell type differently in boys and girls, the pair constructs a fragile chain of putative links between this single study and their claim that “inorganic mercury has been rising for many years in American blood.”
The claims are problematic on many levels, but let’s just take a trip to the ground floor: evidence.

First, mercury levels in “American blood” and urine are decreasing, not increasing. The latest analysis of values of inorganic mercury in urine and total blood mercury, published online September 6 in  Environmental Toxicology and Pharmacology, finds that from 2005 to 2012 among all age groups, urinary inorganic mercury decreased. Total blood mercury, which includes organic (carbon-bound) and inorganic forms, also decreased in all age groups during that time. These conclusions are based on data from the U.S. Centers for Disease Control and Prevention (CDC) National Health and Nutrition Examination Survey (NHANES). Meanwhile, other CDC data indicate that autism prevalence has increased. The trends for autism prevalence and mercury levels in people living in the United States are in opposite directions.

The study that Dufault and Gilbert use to claim increased inorganic blood mercury levels also relied on NHANES data from an earlier period. The study’s sole author, Dan Laks, concluded that from 1999 to 2006, the proportion of people with detectable levels of inorganic mercury had increased tenfold. But the analysis involved some important caveats that Laks details. One is that as sample storage time increases, mercury detection results with the methods applied become less reliable, and inorganic mercury measurements are “particularly susceptible” to such chemical caprice. Second, the “overwhelming majority” of the population in this study—70–95 percent—had undetectable mercury levels. Laks attempted to resolve this issue by assigning the same mercury value to every sample that was below the limit of detection. He discusses this research in a video that is part of the bonus material for Evidence of Harm, a film claiming that mercury amalgam fillings cause devastating health effects.

The bulk of our mercury exposure is the organic form, which we ingest largely through eating certain fish. Another source of mercury in our bodies is mercury vapor from coal burning emissions. This can accumulate in our food chains by contributing to organic mercury, or people can inhale it. In other words, processed foods with potential trace inorganic mercury contamination are the least of our exposure worries when it comes to mercury, especially given that Dafault and Gilbert provide no evidence that we ingest or accumulate it at significant levels that way.

Second, abundant evidence from studies all over the world forms a solid foundation to explain why autism diagnoses seem to be rising: increased awareness along with diagnostic shifts and expansions. As an example, a study using Swedish registry data for almost 20,000 twins and more than 1 million children in total found that documentation of autism features changed little over the years even as autism diagnoses increased steadily. The authors concluded that autistic children were always present at the same prevalence but going unidentified. They describe any influence of environmental changes as “marginal.”

Journalist Jessica Wright’s breakdown of the huge effect of awareness and diagnostic changes lays out the evidence in clear detail. As she rightly notes, the best evidence-based candidates for environmental factors include the age of the father. Most findings point to genes as the dominant actors in autism, suggesting limited openings for any environmental contenders. PON1, the gene that Dufault and Gilbert cite as a link in the chain of events from maternal inorganic mercury ingestion to boys with autism, is a rather lukewarm selection in the catalogue of autism-related gene variants.

And that takes us to point three. Right now, the CDC data put the ratio of males to females with autism at about five to one. Many researchers have interpreted this skewed ratio to indicate some enhanced autism susceptibility for boys. But there’s another plausible explanation with increasing evidence to back it up: That diagnostic shift that captures more people with autism than ever before may be far more effective at identifying boys compared to girls. It wouldn’t be the first time, certainly, that a focus on males over females has yielded skewed impressions.

Although Dufault and Gilbert are careful to dismiss seemingly deathless claims of mercury as a link between vaccines and autism, they are still dragging mercury back onto the autism research stage largely based on speculation and single-study argumentation. From the assumption in the headline to the claims of diagnostic increases and sex biases, substantial evidence exists to weaken or break every link in their logic chain. Short of some startling new, overwhelming evidence to crush years of research involving more than a million children, the science shows that mercury in any form and autism have nothing to do with each other and never have.