In the past year, there have been numerous reports linking autism and exposure to, among other things, pesticides¹, power plants² and pyrethrin-based pet shampoos.

Statistical associations, however, are just the first step toward implicating risk. Connecting exposure to an autism-relevant biological pathway bolsters the evidence that environmental risk factors play a role in this highly heritable disorder.

For example, one hypothesis that is gaining credibility suggests that heavy metals can cause oxidative stress that, in some genetically susceptible children, leads to neural deficits associated with the attention and cognition deficits seen in autism³.

“Genes and environment ― thatʼs where the action is going to be,” says Ian Lipkin, epidemiologist and director of the Center for Infection and Immunity at Columbia University in New York.

Lipkin is among those researchers who say that studying the environment is not only worthwhile, it is prudent. “If we waited to make genetic linkage between smoking and cancer,” he notes, “we would have missed an opportunity to intervene.”

The genetics of autism has been tricky to unravel. Beyond the few single genes that explain a small percentage of autism cases, de novo mutations and epigenetic control of gene expression also factor into autism incidence ― leaving room for the possibility that environmental factors play a role in their occurrence.

“Could these genes or pathways be involved in environmental exposures―and mistakenly counted as genetic cases?” asks Craig Newschaffer, an epidemiologist at Drexel University. It is well-documented that chemicals in the environment, such as lead or mercury poisoning can cause neurological injury, Newschaffer notes.

Clues from other complex disorders, notably diabetes, suggest the environment can indeed play such a role.

“We know what you eat has an effect on obesity and diabetes risk,” says Margit Burmeister, professor of psychiatry and genetics at the University of Michiganʼs Molecular and Behavioral Neuroscience Institute.

For example, Pima Indians who follow their traditional diet maintain a healthy weight. But on a Western diet, they become significantly overweight and develop a high risk of type 2 diabetes compared with Europeans on the same diet⁴.

“You can call it purely genetic or environmental depending on point of view,” Burmeister says.

Nature and nurture:

A role for the environment has also been observed in the psychiatric realm. For instance, there is evidence that maltreatment of children or stress in adulthood is linked to depression⁵.

But, she notes, researchers were only able to detect a genetic effect in people with depression while looking for the interaction between genes and the environment. “You canʼt start with genes and do environment later. The researchers wouldnʼt have found the gene without also looking at the environmental insults,” Burmeister says.

The ideal approach, researchers say, is to use animal models of chemical toxicity during gestation to help identify plausible hypotheses for a specific factorʼs effect on brain development.

For instance, biomarkers at birth ― ranging from nutritional deficiencies to antibody responses against viruses ― may offer clues to environmental risk factors. Prenatal exposure seems to be important, so research should also take into account maternal health.

Reliable observations can then direct studies to investigate the interactions between genes, the environmental factors and the disorder.

“Once you combine information on genes and how they are related to one another, we can add a layer of information about chemicals targeting those specific neural pathways disrupted in autism,” says Isaac Pessah, a toxicologist at the M.I.N.D. Institute at the University of California at Davis.

Pessah was part of the team who found that women who use pet shampoos containing pyrethroids, a synthetic version of a chemical insecticide produced by flowers such as chrysanthemums, double their risk of having a child with an autism spectrum disorder.

The study was retrospective, and relied on participantsʼ recollection of pet shampoo use. Pessahʼs UC Davis colleague Irva Hertz-Picciotto reported the preliminary findings at the International Meeting for Autism Research in May, where they were promptly picked up and publicized by the media.

Pessah says the findings merit follow-up in part because pyrethroids are known to block GABA receptors, a class of nerve receptors that maintain normal nerve activity by balancing excitation and inhibition of the central nervous system ― a pathway disrupted in individuals with autism.

Interestingly, another California-based study also found an association between autism incidence and the use of two pesticide chemicals known to be GABA receptor blockers.

“If a toxicologist were to put a bet on any one compound that might be associated with autism, it would be a GABA blocker,” says Pessah.

Tough task:

Proving this, or any other gene-environment interaction, is a thorny task, however.

Epidemiological studies on the environmentʼs effects need to be long and comprehensive ― including genetics and neurological responses ― which makes them forbiddingly expensive.

Experts say the ideal study would follow newborns over a long period of time to avoid biased reporting of past events^6,7,8. But the sheer number of potential environmental factors to track is daunting.

“The environment ― and all the factors that influence brain development ― is far more complex than the genome,” says Pat Levitt, director of the Kennedy Center for Research on Human Development at Vanderbilt University.

Fewer than three percent of the more than 75,000 chemicals listed on the US Environmental Protection Agencyʼs Toxic Substances Control Act Chemical Substance Inventory in the environment have been tested in any way in a biological system, Levitt notes. And, thatʼs just one subset under the label of environment. Maternal antibodies⁹, nutrition¹⁰, and the fatherʼs age¹¹ are among the other factors that have been implicated in autism.

There are also massive gaps in scientistsʼ understanding of environmental data, says Danielle Fallin, an epidemiologist at Johns Hopkins University.

“Itʼs like the puzzle is right there in front of you, and we could figure it out if we focus long enough,” she says. “But we donʼt have enough information on the timing of environmental exposures ― particularly in utero ― or even of parent genomes, to piece together what is happening.”

To gather information on timing, Newschaffer will soon begin enrolling pregnant women who already have one child with autism and are therefore genetically predisposed to having children with the disorder. The researchers plan to collect samples from 1,200 women throughout their pregnancy and after, including newborn core blood and meconium, an infantʼs first bowel movement, to assess chemical exposure burden in utero.

Mathematical models may also help make sense of the massive amounts of environmental and genetic data available. But Levitt says epidemiological studies must first be done to reveal population-wide patterns of exposure that merit investigation.

“Talking about dog shampoo is no different than talking about immunizations or any other single environmental component,” says Levitt. “Whatʼs missing is the integration of the biology that connects the genetics and environment.”

References: