Autism and Environmental Chemicals: A Call for Caution

Emily Willingham

Pardon me for a moment while I get a bit sciency on you. In a former life, I was a scientist who conducted research in the field of endocrine-disrupting compounds. We focused on compounds that accumulate in body fat. The list of these compounds is long … almost endless … and many of these chemicals occur in what we consume, wear, sit on, wash with, and eat from.

The term “endocrine disruptor” doesn’t even encompass the physiological systems that some of these compounds affect, and one system that interacts and overlaps with the endocrine system — the two cannot be separated, frankly, and I dare anyone teaching physiology to try — is the neurological system. Our neurology and our endocrinology are integrated, and compounds that influence or disrupt one often will do the same to the other.

It all started with what we used to call environmental estrogens. Then, it expanded as we realized that the effects of these compounds weren’t always estrogenic — some were anti-estrogenic, anti-androgenic, androgenic, thyroid inhibiting, thyroid boosting, adrenal affecting, and, yes, neuroeffective. While my focus was the influence of these compounds during embryonic development on sex development — gonadal development and penile development, specifically — there are hundreds of other endpoints that these chemicals can affect.

The key factor that these compounds — and pretty much any chemical that has an influence on a developing organism — share is that they may seem to have little in the way of negative effects on an adult, but they can have permanent disruptive effects if the exposure is embryonic. That makes sense if you think about alcohol — ethanol. You can go out and get kneewalking drunk and suffer the acute effects the next day, but that single episode of exposure likely won’t do you lasting harm. But do that to a fetus, and the processes in motion at the time of exposure may be disrupted in ways that do not allow recovery.

Another feature of the chemicals is that the doses required to cause an effect in an adult organism can be many, many orders of magnitude greater than the doses that disrupt normal development. In my research models, it would take as little as a drop in a trillion drops of some compounds simply to shift the sex development of the embryo from male to female. One. Drop. In. A. Trillion.

For several years now, I’ve had these compounds on my mind as I consider human neuroendocrine development and my own scientific work. It’s hard to formulate any firm hypotheses while the jury remains out on whether or not autism rates are genuinely on the increase because of a real increase in autism or whether or not the increase is the result of better diagnostic criteria and recognition. Until that question is decided, it will be difficult to identify any real correlation that would lead to a hypothesis of causation of any specific compound.

And as anyone in endocrine-disruption research will tell you, a single compound can be difficult to tease out of the thousands to which we’re exposed every day. However, there are some prime candidates for human exposure and developmental disruption. These include phthalates (used to soften plastic toys), bisphenol A (BPA), PBDEs (polybrominated diphenyl ethers, otherwise known as flame retardants), and PCBs (polychlorinated biphenyls). These have all been identified as having endocrine-disruption capabilities. But results vary about which tissues they effect, what the effects are, what doses have these effects, what the timing of the dose is to have an effect, and more. The PCBs alone constitute a class of hundreds of separate chemicals. In other words, endocrine-disruption research is extraordinarily complex.

The few solid human-based studies have identified some of these at high levels in women — in breast milk, in particular — and also have identified them in umbilical cord blood of infants, meaning that they passed from mother to infant. So, yes, we’re exposed. And for most of these, there is a great attraction to fat, which is mobilized at some of the worst times for exposure for children, as when we breastfeed.

Thus, it comes as no surprise to me that mainstream science is paying attention to this potential link ( between these exposures and autism. A handful of chemicals — not of the kind we’re passively exposed to but of the kind we take therapeutically — have already been linked to autism. These include valproic acid (an anti-convulsant sold under the name Depakote), which also is an endocrine disruptor.

What we need to be careful about is talking about any links as established before the work has even been done. Nicholas Kristof in The New York Times ( tries to make this argument, but I’d call it a big fail from the get-go, as the headline itself is a screaming warning of “Do Toxins Cause Autism”?

Kristof states, and he’s right, that “these are difficult issues for journalists to write about. Evidence is technical, fragmentary, and conflicting, and there’s a danger of sensationalizing risks. It’s quite true. The studies of the effects of these compounds in humans are mindboggingly complicated, with endpoints that may be under the influence of a host of confounding factors. And once again, we can’t hypothesize an influence of any environmental factor as being an actor in the rise in autism rates unless we’ve established that the rise is genuinely an autism increase, rather than an increase in diagnostic accuracy. And the jury is still very much out on that, although most evidence points to the latter explanation as valid.

Where does that leave us? I advocate for simply doing the best we can to remove these compounds from the environment or at least to stop contributing them. Whether they are a factor in autism or not, they’re patently not safe for developing vertebrates, and we should be addressing that. Period.

And as someone who has seen the power of these chemicals to alter vertebrate development, I can only tell you what I do now. I do not use cosmetics, and I do not use shampoos or soaps on my children that contain phthalates, tea tree oil, or lavender oil. We do not use plastics with bisphenol A. I am careful about my purchases of fish oils and other fat-related items, checking to see if the persistent organic pollutants have been removed. I almost never microwave in plastic.

That said, I have to note that after all of my work with endocrine-disrupting compounds, including the study of congenital urological malformations in boys, I have two children who have … urological malformations and at least one child who has autism. The thing is, I can’t determine whether these outcomes are a result of genetics — the urological problems run in the family, and I can see clearly where my children get some of their spectrum-associated manifestations — or whether they’re the result of what I, the mother, have been exposed to simply through living in modern times and through my laboratory work. We already know genetics contributes strongly to autism. The question is, What else, if anything, does?

And that’s not an easy question to answer — if it can be answered at all. Certainly, there won’t be an answer that a single, sensational headline can encapsulate.

This essay was originally published at A Life Less Ordinary,