Autism-like findings in relatives of autistic children and the evolutionary biology of autism

I came to this reference via medlogs. It's quite fascinating. The more we learn about congenital structural and organizational disorders of the brain and mind, like schizophrenia and autism, the more oddity we see in how they're expressed. Here we learn that some structural aspects of "autism", a "disorder" that seems to be strongly inherited, may manifest in high functioning "non-autistic" adults ...

Brain deficits found in relatives of autism sufferers
Unaffected family members show characteristic abnormalities.
Jim Giles

People can have physical brain abnormalities similar to those found in autistic individuals without having the disorder themselves. These results come from two studies, which were presented at a conference over the weekend. Brain scans show striking similarities between the brains of autistic patients and those of their non-autistic parents and siblings.

... In one study, Eric Peterson of the University of Colorado at Boulder and his colleagues scanned the brains of 40 parents of autistic children and compared the results with functional magnetic imaging (MRI) scans from 40 [jf: normal, non-related] controls. The data look much like those obtained for comparisons between autistic and non-autistic brains, says Peterson. The results were discussed on 13 November at the annual meeting of the Society for Neuroscience in Washington.

Some areas of the brain region known as the prefrontal cortex were smaller than normal in the parents of autistic children, for example. This part of the brain is involved in understanding other peoples' motivations, something that autistic people find difficult and is thought to lie behind the problems they face in interacting socially.

Another typical symptom of autism is the tendency to avoid making eye contact. This behaviour was studied by Brendon Macewicz and colleagues at the University of North Carolina, Chapel Hill. He gave nine families with an autistic child and unaffected brother a digital camera and told them to take pictures of friends and family. Macewicz then mixed up the shots with images of strangers and tracked the childrens' eye movements while asking them to say whether the people they saw in the pictures were familiar or not.

Most people rely heavily on looking at the eyes when asked to complete this task. But autistic children are known to avoid the eyes and focus on other areas of the face. To Macewicz's surprise, the non-autistic siblings did almost exactly the same.

"This piqued our curiosity," he says. The team then ran MRI scans on the brothers, focussing on the part of the brain known as the amygdala. This area is involved in fear and is typically smaller in autistic people. "It was very interesting," says Macewicz. "The children showed a similar decrease in amygdala size to their autistic siblings." The difference was around 5-10%.

The results are intriguing, say the researchers, because the parents and siblings had not been diagnosed with autism. Macewicz says it is likely that in the unaffected siblings other brain areas, perhaps in the frontal lobes, are helping to regulate the amygdala and compensate for its smaller volume.

It may be that a core set of brain abnormalities has to be present for autism to occur, adds Peterson, and that the parents he studied do not have them all. He points out that some autism-related behavioural traits have previously been seen in the relatives of people with the condition, but that these current studies are among the first to show similarities in brain anatomy.

The clinical concept of "autism" is very vague. It's a"diagnosis" made by school systems, social services, psychiatricsts, psychologists, and researchers. There are children than all would label "autistic", but there's no doubt the concept is itself ill-defined. The group studied here is probably more homogenous than the general "autistic" population.

The results are fascinating. I do wonder how many of the parents would have been labeled as "autistic", were they children today.

This study lends credence to the "Silicon Valley nerds" theory of the increasing prevalence of autism -- that many high IQ "autistic" children are the result of increased rates of marriage between persons with autistic traits, who congregate in the tech indutries. It also strengthens the long suspected link between pre-autistic traits and "geekiness".

Classic autism is not a very adaptive condition in most human environments. Autistic children would probably die quickly in a harsh environment. So why is autism a relatively common disorder? We know from many, many examples in human evolution that a serious genetic disease (ex. sickle cell anemia) will persist when some of its component traits have adaptive advantage. It's very likely that some pre-autistic "traits" or genetic components have adaptive advantages.

I would like to know what the correlation of autism is in identical twins ...

Update 2/24/07: Correlation in identical twins can be very high, probably depending on the subtype of "autism":

... different studies have shown that if one identical twin has autism then there is a 63-98% chance that the other twin will have it. For non-identical twins (also called fraternal or dizygotic twins), the chance is between 0-10% that both twins will develop autism. The chance that siblings will be affected by autism is about 3%.

The population risk is supposedly about .7%, so siblings have about a 500% relative risk. The large spread in co-occurrence for twins is very compatible with diverse genetic causes; again we see that the word "autism" is used for a wide variety of distinct but unnamed disorders.