We read with interest the review article entitled “Is there a link between autism and glyphosate-formulated herbicides?” authored by James E. Beecham, a retired University of Florida pathologist, and Stephanie Seneff, a computer scientist with the Massachusetts Institute of Technology. Our interest was piqued because of our study as toxicologists of glyphosate’s toxicity over a number of years with the University of Ottawa, the U.S. Environmental Protection Agency, or the nonprofit group Toxicology Excellence for Risk Assessment.
The authors, Beecham and Seneff review information and state a hypothesis that a pregnant woman’s exposure to glyphosate-formulated herbicides may produce changes in her unborn child’s brain that are remarkably similar to those found in the brains of humans with autism. These changes could be caused by glyphosate’s known ability to bind manganese, an essential element, inducing deficiency. One result of this deficiency would be a reduction in maternal serum levels of Thyroid-Stimulating Hormone (TSH). Such a hypothesis challenges conventional wisdom on the toxicology of this well-studied herbicide and warrants a careful listing of supporting data. It should be noted that the ability of glyphosate to cross the placenta from the human mother to fetus is limited; only 15% was found in the fetal circulation (Mose et al 2008). In essence even though the mother may ingest this pesticide, over 85% of glyphosate does not reach the fetus and the amount of chemical required for the mother to produce any fetal changes would likely result in maternal toxicity, which means the mother would be affected. However, these authors did not note any maternal toxicity.
Further, obvious supporting data would be found in a review of the existing experimental animal literature on glyphosate’s toxicity for information pertaining to levels of TSH, especially since changes in TSH are readily discernable, and have often been monitored, in such studies. Beecham and Seneff did not do this. Another approach to supporting their hypothesis would be to carefully compare the dose levels of glyphosate in their cited studies to human biomonitoring data garnered in humans who work directly with the herbicide. These biomonitoring data demonstrated extremely low human exposures as a result of normal application practices, 500-fold or more lower than the safe dose for glyphosate set by the U.S. Environmental Protection Agency (EPA 1993) based on developmental toxicity. Beecham and Seneff have not done this either. In addition, the authors might include more contrary citations. For example, extensive studies in Columbia clearly demonstrated that glyphosate did not markedly affect time to pregnancy and maternal health (Sanin et al, 2009).
What the authors have done is to speculate on a number of items that may relate to glyphosate exposure. And while scientific speculation is nearly always the first step in any scientific investigation, without reference to the many well-conducted evaluations of glyphosate’s toxicity in the past by government agencies, including the EPA and Organization of American States (OAS), their current speculation appears premature.
In fact, the authors themselves “acknowledge that we have not yet proven that glyphosate…exposure from common food and beverages causes human autism.” We would certainly agree with this statement, especially since a well-structured hypothesis has not yet been supported, nor tested. These authors also recommend that “mothers are advised to consume and feed their family an organic diet as a precautionary measure,” not understanding perhaps that, over 99% of the pesticides that we eat daily in tiny amounts are naturally occurring and that washing food prior to ingestion removes a significant amount of these pesticides.