To understand autism, talk to the grandmothers

Commentary from the Spring 2017 newsletter

With Dr. Temple Grandin, backstage after an autism conference in Northern California.

Last weekend I delivered a talk at an autism conference, among the openers for keynote speaker Temple Grandin (sitting with me at left). After the conference we got to chatting and I mentioned my "time bomb" hypothesis of autism and the idea that grandmaternal smoking, for example, increased the risk of autism in grandoffspring via glitches induced in vulnerable fetal germline. "Well," said the legendary author and new inductee to the Women's Hall of Fame, "my mother's mother smoked like a chimney!"

I have lost count of the number of times I've heard this sort of information about autism grandmothers. When I first started interviewing autism parents about their own prenatal exposures about five years ago, my sole focus was the synthetic steroid hormone "anti-miscarriage" treatments like those to which I had been exposed in utero. Yes I found those hormone stories, but most of the responses went something like this: "No, I doubt my mom took any drugs like that, but she was a heavy smoker," or "My mom smoked a pack a day back when she was pregnant with me," or "Both my mom and mother-in-law were smokers," or "My mother-in-law smoked like crazy," or, of course, "My mother smoked like a chimney."

At first I thought nothing of the information, I considered cigarette smoke a mundane and uninteresting fetal exposure compared to the fake hormone protocols my mother was given. But after hearing the smoking story for what seemed like the bazillionth time I did some research to try to connect the dots. And what I found hit me with a wallop: cigarette smoke harbored a long list of toxicants such as benzo[a]pyrene and nicotine, induced mutagenesis and somatic mosaicism, damaged germ cells on many levels, had adverse epigenomic effects on fetal cells, and, from a phenotypic point of view, did direct damage to the fetus, including low birth weight and some evidence of behavioral differences such as ADHD. And there was new evidence from animal models of intergenerational adverse effects of gestational tobacco exposure, including what I would call "mousie ADHD" resulting from nicotine given to the gestating grandma mouse. When paired with the temporality of the autism increase (decades after the parental in utero exposures), the strange socio-demographics of autism, recurrence of ASDs and related pathologies among siblings, and the much-replicated findings of heterogenous de novo mutations, not to mention my somewhat casual surveying of autism families, grandmaternal smoking struck me as a rather formidable hypothesis.

When we ponder tobacco toxicity we tend to think of lung cancer, heart disease and stroke, among other somatic pathologies. And sometimes we even think of the millions of fetuses who were heavily exposed in those decades when pregnancy smoking was not only common, but at times prescribed for the purposes of appetite control or anxiety relief. But what about the vulnerable fetal germ cells simultaneously exposed back then during this critical window of germline synthesis? Had anyone looked?

To my amazement, the answer was pretty much no. And no autism studies were considering the question either. So I decided to make this question a priority even though it was absent from radar when I began my research quest in 2012. I'm happy to report that some of our projects are now looking at neurodevelopmental outcomes in grandoffspring of women who smoked during pregnancy. Knowing this could be a question of national, or international, importance, I have also contacted about a half dozen PIs of various autism cohorts asking (well, let's be honest, begging) them to inquire about grandmaternal smoking as they do their work evaluating either genomics or proximal fetal exposures, and though I don't yet have any takers, I detect growing openness to the question.

One more thing for autism research to consider—the so-called "broader autism phenotype." Are we seeing the BAP in sibs and parents because of natural genetic variation, or at least in part because of acute but forgotten prenatal exposures suffered by us autism parents born during the Mad Men era of maternal medicine? Born in 1965, I was a young subject in a landmark 1977 study finding what we today would call "Aspie traits" in offspring exposed to heavy doses of synthetic progestins. (See the study here: Reinisch and Karow (1977) Prenatal Exposure to Synthetic Progestins and Estrogens: Effects on Human Development). Could the same induced variation hold true of autism parents exposed as fetuses to developmental toxicants such as cigarette smoke (and/or alcohol or drugs or meds)? Well, again, no one has asked. The BAP has so far been presumed to be all natural, a bit of dogma and assumption-making perhaps rooted in ignorance of biological history.

In sum, might there be at least one "smoking gun" behind a subset of the autism increase? Perhaps in autism research we should start not with the autistic kids, or even their parents, but with the grandmothers.

Jill Escher
February 27, 2017

---