Beyond that, we don't even know that the prevalence of ASD has increased. All we know is that the prevalence of children getting the diagnosis has increased. A 2009 study in England found that 1 percent of adults have ASD—the same prevalence as among today's children. If true, there hasn't been a change in prevalence over the past 50 years, just an increased tendency to identify a disorder that was there all along.
But because the ASD diagnosis is so common now, parents and health officials are clamoring to find its cause. Genetic factors account for the lion's share of known causes: A constellation of neuropsychiatric disorders (depression, anxiety, alcoholism, obsessive-compulsive disorder) runs in the families of children newly diagnosed with ASD. Specific genetic disorders (e.g., Down syndrome and fragile X syndrome) give rise to ASD, as do certain genetic disorders of metabolism.
Here's what we do know: Confirmed causes of autism with an environmental element are prenatal exposure to rubella, thalidomide (an immunoregulatory drug), and valproic acid (an anticonvulsant)—a surprisingly short list. So far, there are no confirmed postnatal causes.
Suspected causes that researchers are looking into include prenatal exposure to misoprostol (an abortion-inducing agent), organophosphide insecticides, and phthalates (chemicals in plastic and vinyl, found in hundreds of household items from makeup to toys to flooring).
Finally, there is no credible evidence that immunizations, gluten, or yeast cause ASD, despite the play they've gotten in the press.
There's an extremely strong association between people carrying umbrellas to work and storms later in the day, but umbrellas obviously don't make rain. Two new autism studies highlight the difficulty in bridging the gap between associations and causes. Both investigations are based on the geographic distribution of kids in California diagnosed with ASD (which, remember, is not the same thing as the distribution of ASD itself).
The first study found that living close to a major highway is associated with an increased chance of having a child with an ASD diagnosis. By the researchers' estimates, children born to mothers living within about 300 meters of a freeway are twice as likely to be diagnosed with autism.
The second noted that the prevalence of children with ASD diagnoses is higher in certain zip codes. The "hottest" zone was West Hollywood, where four times as many children per capita get an ASD diagnosis than anywhere else in the state (the rate is one in 69 in West Hollywood, vs. one in 275 in California). Conversely, the area of lowest risk (in the range of one in 500) was in the desert communities outside of San Diego. It's possible that shared exposure to an environmental factor—say, something in the swimming pools—is to blame. Alternatively (and in the view of some scientists, more likely), parents living in West Hollywood simply have greater resources to secure diagnoses for their kids, compared to parents in poorer neighborhoods.
The authors of the highway study conclude that researchers need to measure air pollutants near highways, in an attempt to identify a culprit. The authors of the "hot zone" paper state that local environmental or social dynamics need to be investigated to clarify the zip code discrepancy. We're inching forward, but we haven't circled in on a definitive cause of ASD in either of these cases.
Because ASD manifests behaviorally and responds to behavioral treatment, it's particularly difficult to understand scientifically. The fact that it covers a wide range of symptoms and capabilities further complicates the picture.
The best way to get from associations to cause-and-effect is to show that manipulating the putative risk factor alters the rate of disease occurrence: Dogs were trained to smoke cigarettes via breathing tubes; they got lung cancer. It would be unethical to manipulate risk factors in an effort to give children ASD, but researchers can use animal models. Thalidomide and valproate have achieved the status of proven ASDcauses because they've been shown to cause brain abnormalities and "autistic-like" behaviors in rodents.
It's also possible to look for "experiments of nature"—instances in which luck (good or bad) has manipulated variables. Comparisons of the incidence of ASD between identical and fraternal twins, for example, have shed light on the relative impact of genetic vs. environmental factors. (The risk of having "autistic-like" features, such as language disorder or anxiety, approaches 100 percent for an autistic individual's identical twin; among fraternal twins, it may be as high as 25 percent).
