Autism is a soul piercing word. Just the sound of it stops the heart. In pediatrics our lexicon is usually generously sprinkled with words like cute or cuddly, and autism is not part of the speech center that we like to use. We deal mostly with happy mothers and smiling babies, but that seems to be changing. Sadly, autism has become more a part of our vocabulary as we all become better acquainted with a neighbor's boy with the diagnosis or a cousin's child you knew wasn't quite right.

There is a very important debate about autism. Is it more common today than before? This is critical, because if autism is growing in numbers, it implies a cause or causes are also multiplying. The thought that something is out there affecting our children and our children's children's brains is pretty frightening. It also implies a strategy for prevention. If a medical condition is static, this year there will be, let's say, 100, and 100 years ago there were also 100 in the same population. It is fixed, and the environment that has changed over time seems not to have any influence on whether a person gets it or not. Then the etiology, or what causes the disease, is also something stable, probably in our genes. But if a century ago there were 100 and today there are 100,000 in the same group, then it is easier to see that this something is not in our genes but perhaps our water or air or food or the way we treat children, or who knows what. Therefore, it is critical to understand if autism is more common or not.

Establishing how often a diseases attacks its victims is not as easy as it seems. It is no simple matter to know what is autism and who has it and is that number greater today than five or 10 or 100 years ago.

The first problem is just knowing what is autism. It has a list of three domains where the brain seems to be broken. They are communications, socialization and use of objects and symbols. Yet unlike the absolute counting of dead or alive, there is a huge range of severity. What to count involves severity as recorded by different people, timing of diagnosis in light of normal developmental changes, interventions along the way, definitions of normal variations etc. It is never just a simple 1, 2, 3.

Another challenge to know how often a disease is showing up in the community is that the definitions may change. In times past, the behaviors seen in a child with autism would be called mental retardation. So are there more children with autism or fewer kids with mental disabilities? Clinical skills in applying the criteria also make a big impact. A doctor, like anyone else, sees what they know. If I only know about childhood schizophrenia, I might see certain characteristics of that in autistic children, or if I don't know something, I may not give it any name at all.