In The Politics of Autism, I explain:
When a pregnancy is under way, doctors can detect certain kinds of disorders, but neither amniocentesis nor any other prenatal test can currently tell us whether a fetus will become autistic. Suppose that such a test did exist. “The best case use of a prenatal test at the moment would be if you could say to a parent, your child has got an 80 percent likelihood of autism and so once the baby's born, we would like to keep a close eye on that child in case they need extra support like speech therapy or social skills training or some sort of behavioral approach,” says leading autism scientist Simon Baron-Cohen. But would the “best case use” be the most common? When amniocentesis indicates Down Syndrome, most mothers choose abortion. A study of autism parents in Taiwan found that just over half would abort if a prenatal test indicated that their next child would be autistic. We cannot be sure what the figures would be if such tests were available in the United States, but it seems likely that a large share of autism pregnancies would end in abortion.
Azeen Ghorayshi at NYT:
Academic research labs across the country are working to find biological markers that can predict whether a child is at risk of developing autism. And companies are rushing to turn the findings into commercial tests, despite limited evidence to back their validity, raising concerns that their results could mislead desperate parents.
They include one test that examines a strand of hair to rule out an autism diagnosis in babies as young as one month old. Two other tests just entered the market. One promises to predict autism risk based on skin cells collected as early as days after birth. Another looks for the presence of certain antibodies in a mother’s blood to determine whether her children, or babies that she might have in the future, are at risk of developing autism.
All the tests are based on autism research by scientists at academic institutions... But the new tests, largely aimed as a screening tool for the general population, are not yet reliable enough to be offered commercially, outside scientists familiar with the tests say, especially in a landscape where families are already inundated with incorrect or unverified information about autism. None of the tests has gone through large experimental trials or had its validity evaluated by a regulatory agency.
...
The new antibody test, which is based on the work of Judy Van de Water, an immunologist at the University of California, Davis, has perhaps the most research behind it of any of the measures.
The test looks for specific antibodies, known as autoantibodies, that mistakenly attack normal cells in a person’s body as if they posed a threat. Dr. Van de Water’s research group has identified eight autoantibodies that cross the placenta during pregnancy and can target proteins involved in fetal brain development.
...
Michael Paul, chief executive of MARAbio, the company selling the new antibody test, said that Dr. Van de Water’s research was based on “a very, very robust data set.”
“I would feel unethical to not bring it forward,” Dr. Paul said, “because we believe this is a test that families should have.”
Dr. Paul said the company hopes to limit the number of false positives by making clear to doctors and patients that the test is not intended as a screening tool for the general population. Instead, the company is marketing the test only to women who already have a child with an autism diagnosis and are deciding whether to have more children, or those who have a child who is exhibiting clinical signs of the disorder. The company is currently validating the test for use in the general population, Dr. Van de Water, who also is the founder of MARAbio, said.
The risk of false positives is also one reason the company is not yet making the test available to pregnant women — a group that Dr. Van de Water said has been asking her for a test like this for years. But the restriction, she added, also allowed the company to sidestep thorny debates about abortion and autism.
“We do not want to go there. And we don’t have to,” Dr. Van de Water added, noting that, unlike common prenatal tests that rely on detecting fetal cells, the antibody test needs only a sample of the mother’s blood on which to base predictions.
