Ideas on Autism Diagnosis

Autism Inflation
Jerome Kagan and Robert Pozen, 06.06.05


Doctors should use biology, not only behavior, to evaluate children.
The bad news is that the reported incidence of autism has tripled over the last decade. The good news is that this figure is probably exaggerated.

Autism's sharp rise is, in large part, a matter of definitions. Is a child with severe learning problems autistic? What about a child who is insensitive in social situations? What about children who have trouble communicating or seem to retreat into their own shells? These days a large number of children who fit any of those descriptions are likely to be tagged with the autism label, or their parents will be told that they have a disorder (like Asperger's syndrome) that falls somewhere in the autism spectrum.

This looseness of definition is getting in the way of medical progress. We will not find effective cures for autism until we add biological markers to behavioral symptoms in diagnosing children.

Three different types of biological features can help to distinguish among the children currently labeled as autistic by behavioral symptoms.

A larger brain, as revealed in MRI scans, is one sign. The larger brains of some autistic children could be due to an excess of myelin, the sheath of white matter around the nerve cells. This could mean that the neurons are overly active, since neuron activity stimulates the production of myelin. Other autistic children have abnormally small neurons in the prefrontal and temporal cortex. These parts of the brain play a major role in language development and behavioral control--key challenges for many autistic children.

A second set of biological markers involves the responses of neurons to external stimuli as measured by the timing and magnitude of brain waves. Some autistics do not show the expected wave form when they hear a change in a spoken syllable (e.g., from "pa" to "ba"). This suggests that there is something deficient in language perception.

The biochemistry of the brain supplies a third differentiating feature. A set of the relevant molecules can be measured by examining the cerebrospinal fluid, which bathes the spinal cord and extends upward to the brain. Some autistic children show abnormally high concentrations of molecules called gangliosides. Other biochemical features, which can be measured by brain-scanning techniques, show that certain autistic children have abnormally high amounts of serotonin, a critical molecule for regulating mood.

If children with autistic behaviors were evaluated for these biological features, physicians would likely find several distinct profiles among the children, despite the similarities of their behavior. In other words, what we call autism today is likely to constitute two, three or even more quite different disorders.

The currently preferred treatment for children exhibiting early signs of autism is enrollment in intensive behavioral programs. These programs involve one-on-one interactions between the child and the therapist from 20 to 40 hours per week. Scientific studies of the most popular of these programs have shown modest results. Although intensive behavioral therapy can reduce the frequency of certain autistic behaviors for certain children, it does not cure the basic language and emotional deficiencies of most autistic children.

To find new cures, we need more research into the biological bases of the various types of autism. Over the last few years the National Institutes of Health has greatly increased funding for such research, but it must continue to coordinate the studies at various laboratories to ensure systematic conclusions on these biological markers. Once these conclusions are established, we'll be well on our way to prescribing new therapies that target each child's particular biological disorder, rather than general therapies aimed at all children labeled as autistic.

Perhaps in a few decades, when a young child begins to exhibit autistic symptoms, a physician will order blood tests and brain scans. These will lead to a targeted set of therapies that will be far better than the uniform response we offer today. Autism Inflation


Jerome Kagan, professor of psychology at Harvard, and Robert Pozen, chairman, MFS Investment Management.