Mothers with high levels of the pesticide DDT in their blood during pregnancy are more likely to bear children who develop autism, according to a study of blood samples from more than one million pregnant women in Finland.
The World Health Organization estimates that globally, one in 160 children has autism. Any case of autism is likely due to a number of factors, including genetics and other environmental exposures.
Although the authors stress that the findings do not prove that autism is caused by DDT — whose use has been banned in many countries for decades over concerns about its effects on wildlife— it is the first such association using a direct measure of exposure to the pesticide. Researchers who investigate links between environment and disease say that further studies are needed to determine the mechanism, if any, by which DDT exposure could trigger autism.
The study, published on 16 August in the American Journal of Psychiatry1, also examined mothers’ exposure to another set of chemicals known as polychlorinated biphenyls (PCBs), and found no association between these substances and autism. That finding deepens questions about whether or how DDT might be linked to autism.
DDT — which is still sometimes used in Africa to control mosquito populations — lingers in soil and water for decades, accumulating in plants and the animals that consume them. PCBs, which used to be common in building materials and electronics, tend to accumulate to high concentrations in certain fish.
Looking for links
Previous research has linked both DDT and PCBs to cancer3, and has suggested that the chemicals might affect brain development and cognition in early childhood2. Most of these studies, however, have assumed exposure to these chemicals based on the participants' proximity to a contaminated site; they did not directly measure levels of the chemical in pregnant women's blood during pregnancy.
To get a better sense of direct exposure, Alan Brown, a psychiatrist and epidemiologist at Columbia University in New York City, turned to a biological database in Finland, which has collected and stored blood-serum samples from pregnant women since 1983.
Brown and a group of researchers in Finland compared childrens' health records against a cohort study that had collected blood serum samples from more than a million women who gave birth between 1987 and 2005. They found about 1,300 children who had been diagnosed with autism and compared 778 of them — and their mothers — to 778 child-mother pairings without an autism diagnosis, each carefully matched for place and date of birth, sex and residence.
The researchers analysed serum samples from the mothers and from 778 women whose children had not developed autism, looking at the levels of certain chemicals produced when the body breaks down DDT or PCBs.
Brown's team found no correlation between the PCB by-product and autism. But when they measured DDT by-product levels in the blood samples, they found that mothers with high concentrations of this chemical — those in the top quartile — were 32% more likely than women with lower DDT levels to give birth to children who developed autism. The likelihood that a child with autism accompanied by intellectual disability was twice as high in mothers with elevated DDT levels compared to those with lower levels.
The study “is really amazing”, says Tracey Woodruff, who studies reproductive health and the environment at University of California, San Francisco. She is impressed by the number and quality of the samples in the Finnish database, and finds the association between DDT and autism striking. “This just confirms that banning it was a good idea.”
Brown is surprised that his study did not find a link between PCB exposure and an increase in autism rates, as other epidemiological studies have suggested5. “What it tells me is you really can't assume that if one toxin is related that every toxin will be related,” Brown says.
It’s unclear how exactly DDT could raise the risk of autism, but Brown proposes two hypotheses. DDT has been shown6 to cause low birth weight and premature birth, which are known risk factors for autism. DDT is also known4 to bind to proteins in the body called androgen receptors, which allow cells to respond to testosterone and other hormones. (PCBs do not bind to androgen receptors.)
Research in rodents has shown that some chemicals that bind androgen receptors can disrupt fetal brain development, particularly in boys, who are more likely than girls to be diagnosed with autism. Brown says his group is beginning rodent studies to test these hypotheses.
Bruce Lanphear, an epidemiologist at Simon Fraser University in Burnaby, Canada, also praises the study and says that the correlation seems to be at least as strong as those between autism and several genetic mutations. “We’ve learned some of the most important things from these types of observational studies,” he says. “Those are what led to the greatest improvements in health.”
Brown cautions that although there seems to be a link between autism and DDT exposure, the overall risk of having a child with the disorder is low — even among women with high DDT levels. His group plans to look at other organic chemicals in the Finnish database to determine whether they might affect fetuses by interacting with DDT.
Jonathan Chevrier, an epidemiologist at McGill University in Montreal, Canada, is interested in knowing whether DDT levels are linked to intellectual disability in children who do not have autism. He is currently following more than 700 children in South Africa — where DDT is still used — which could provide hints as to the mechanism by which the pesticide might affect the brain. It’s an important question, he says, given how much DDT persists in the environment, even in places that have banned its use. “At this point, essentially the entire planet is contaminated with DDT,” he says.