As a science writer, I see lots of press releases about researchers having dug up a gene for this or that disease. They make my eyes glaze. I mean, researchers can’t even predict your height based on the genetic variation they’ve identified for that trait. The situation with complex disease isn’t much different. That’s why I’m more interested in environmental risk factors for disease. See my story on risk factors for schizophrenia, inspired by this exchange with Brandon Keim of Wired Science.
So naturally I perked up when Seth Roberts blogged about a possible link between autism, which has defied easy genetic explanation, and fetal ultrasound. Science writer Caroline Rodgers has done the leg work. From her 2006 article in Midwifery Today (an old favorite from my days at the NYU interlibrary loan department):
Early studies showed that subtle effects of neurological damage linked to ultrasound were implicated by an increased incidence in left-handedness in boys (a marker for brain problems when not hereditary) and speech delays.(5) Then in August 2006, Pasko Rakic, chair of Yale School of Medicine’s Department of Neurobiology, announced the results of a study in which pregnant mice underwent various durations of ultrasound.(6) The brains of the offspring showed damage consistent with that found in the brains of people with autism. The research, funded by the National Institute of Neurological Disorders and Stroke, also implicated ultrasound in neurodevelopmental problems in children, such as dyslexia, epilepsy, mental retardation and schizophrenia, and showed that damage to brain cells increased with longer exposures.(7)
Rodgers has since pointed out that autism rates are lower in Black and Hispanic women in the U.S., who receive maternal care later in pregnancy and less often than White women. She also turned up a curious dip in the autism rate for Hispanics in Alabama between 2002 and 2006, following a reduction in Medicaid coverage for prenatal care in that state.
The biggest hitch in Rodgers’ argument may be biological plausibility:
Dr. Rakic’s research team, cited earlier in this article for its recent study on mouse brains and ultrasound, pointed out that “the probe was held stationary for up to 35 minutes, meaning that essentially the entire fetal mouse brain would have been continually exposed to the ultrasound for 35 minutes…in sharp contrast to the duration and volume of the human fetal brain exposed by ultrasound which will typically not linger on a given tissue volume for greater than one minute.”(42)
Let me be clear: I have no idea if her argument holds water. It sounds plausible, but I’m not enough of an autism geek to have a considered opinion. I do think it’s at least as worthy of consideration as the claim that sleeping on a metal bed frame causes cancer.
Anyway, I am posting on this subject because I wrote something tangentially related for the physics web site Physical Review Focus, about simulations showing how ultrasound punches molecule-size holes in the membranes around cells — an effect called sonoporation. I have no idea how sonoporation conditions compare to what’s used to image a fetus in the doctor’s office — I’m guessing sonoporation involves much more intense ultrasound — but if I were writing a story on ultrasound and autism, I’d be sure to get that sorted out.
Related fun fact: fetuses may be able to hear ultrasound exams.