Parents of children with autism don't get much good news: It's still not clear what causes the often devastating disorder, which affects as many as 1 in 150 children and for which there is no cure. As a result, theories abound on potential causes, the most notorious being the 1960s-era notion of "refrigerator mothers."
In recent years, much energy has been expended on arguing whether vaccines could cause autism: Some parents think that the measles-mumps-rubella vaccine or thimerosal, a mercury-based preservative in other vaccines, is the culprit. Scientists, on the other hand, think autism is largely genetic, and have focused on looking for genes that could be at fault. That disconnect has been frustrating to parents and sometimes dangerous; an unproven treatment known as chelation therapy, which leaches heavy metals such as mercury from the body, resulted in the death of a 5-year-old boy in 2006 after he was administered the wrong drug.
The best evidence to date that vaccines are not responsible is published today in the Archives of General Psychiatry. Researchers with the California Department of Public Health found that the number of new cases of autism reported in California has risen consistently for children born from 1989 through 2003, which includes the period when thimerosal was phased out. Studies in other countries, including one from Canada published in 2007, have also exonerated vaccines and thimerosal.
"This study is very important for public health," says Eric Fombonne, an autism researcher who heads the department of psychiatry at Montreal Children's Hospital. He hopes the new evidence will encourage all parents to get their children vaccinated, and persuade parents of children with autism to shun chelation and other untested treatments. "When you have a disorder that is so complex, that affects humans at their core nature, people want explanations," says Fombonne, who is optimistic that research now underway on linking genes with autism will provide them. "I think the situation will change dramatically in the next few years, as we discover more genes."
Such discoveries won't help children who are already struggling with the disorder. But one recent bright spot in treatment is evidence that the earlier children get behavioral interventions designed to teach language skills, thinking, and social interaction, the better they seem to do through life. Scientists are now trying to push the limits of that theory, targeting even tiny babies, as young as 6 months, who are too young to be diagnosed with autism.
"Early intervention leads to more positive outcomes," says Annette Estes, a research assistant professor of psychiatry and behavioral sciences and behavioral sciences at the University of Washington. Unpublished work by Estes and her colleagues has found improvements in children who are given intervention at as young as 18 months. The idea is that since small children's brains grow in response to their environment, babies destined to develop autism can develop new neuronal pathways to work around their brain deficits.
This month, Estes started recruiting 200 families with newborns in the Seattle area who have an older child with autism. Those babies have a much higher risk of developing the disorder than kids in the general population: about 1 in 20, compared with about 1 in 150. Half of the parents will be trained to work with their babies using a program called "Promoting First Relationships," which was designed to improve the social environment for normal children in high-stress situations like homeless shelters. The other 100 babies will be evaluated and referred to early intervention services in the community. All the children will be re-evaluated at 12 months, and those who show signs of autism will be switched to an autism intervention plan. When all the children are 2 years old, an age at which signs of autism can be more reliably diagnosed, the researchers will evaluate them to see if the children who got the more intensive interventions have fewer symptoms. "We don't know how effective this will be," Estes says. "But we owe it to the families to try."
A similar effort is underway at the Kennedy-Krieger Institute in Baltimore, where for the past 18 months researchers have been teaching about 50 babies basic baby skills like reaching out and grabbing a rattle while also interacting with the person who gave it to them. Some of the children have autistic siblings; some are at risk of developmental delays for other reasons, such as preterm birth; and others are developing normally. At 6 months, the babies who end up later being diagnosed with an autism spectrum disorder seem to do just as well as the normal babies when interacting one-on-one with a parent or caregiver. But when those babies are asked to pay attention to other activities at the same time, they start to struggle. "We want the baby to be able to shift attention between objects and people," says Rebecca Landa, director of the institute's Center for Autism and Related Disorders.