Map of life expectancy at birth from Global Education Project.

Wednesday, February 17, 2010

Well, something is going on

The disorder -- or far more likely multiple disorders -- labeled as autism are much in the news and of blogospheric concern mostly because of the increasingly pathological cult that blames it all on vaccination; and also because of the epidemic of fraudulent treatments. Rather than reviewing all this crap or giving you a collection of links I will simply refer you to Orac who has been all over this like hot sauce on a chicken wing for as long as I've been cyberspacial.

However, just because the issue is infested with destructive bullshit doesn't mean it isn't a real problem for a lot of children and families. A serious underlying difficulty that has helped create room for the snake oil salesman has been a lack of sound epidemiological data. Indeed, as with all collections of symptoms for which we have no etiological explanation, even defining autism and applying the label is problematic. The presence of autism in an individual child, and by extension the prevalence in the population, essentially comes down to whether someone with the appropriate credentials has applied the label. It is this ontological problem which has allowed people to claim there is a growing epidemic of autism, and allowed self-promoting serial liars like RFK Jr. to claim it is temporally associated with whatever causal factor they want to blame.

Since 2000, CDC has been trying to bring at least a semblance of order to the epidemiology of autism through a project called the Autism and Developmental Disabilities Monitoring Network. Here is their latest report. Unfortunately, it doesn't really succeed in reifying the nature of autism. What they do is to collect all the developmental evaluations of children they can get their hands on in certain states or areas, and then have credentialed people review them according to standard criteria. They don't depend on children actually getting a diagnosis of autism in the community -- they can undiagnose an (anonymous) kid who has the label, or diagnose one who doesn't. In some places they have access to medical records only, in others to school-based assessments as well.

Because a lot depends on whether a child ever undergoes such an assessment in the first place, they can't come close to eliminating the problems of ascertainment bias that affect both changes over time, and disparities among demographic sectors. What they can do, however, is put a floor under the prevalence, and the bottom line is it's quite a bit higher than we used to think. It ranges widely from place to place, probably although not necessarily entirely due to differences in the likelihood of ascertainment. Nevertheless, it's worth announcing that they find the overall prevalence to be 9/1,000; it is much higher for boys -- 1 in 70, compared to 1 out of 315 girls. This is most certainly an understatement, since the ascertainment problem, given their consistent criteria, can only result in an underestimate. In other words, there is a screen in that children have to undergo a professional assessment that results in an interpretable record. Not surprisingly, the measured prevalence is lower where only medical, and not educational records are available. Still, the prevalence as ascertained through the network continues to increase over time.

Contrary to the prevailing stereotypes, which are based largely on dramatically interesting accounts of very high functioning individuals with the Asperger's label, the majority of children found to have autism spectrum disorders through the CDC network have low measured IQs. But early diagnosis and appropriate intervention really matters -- as this study shows. Unfortunately, in the CDC surveillance network, most children are identified late, after four years of age.

The fact is, we have absolutely no good idea of what causes the array of developmental delays which get the catch-all label of autistic spectrum disorder. It undoubtedly has to do with genetic predisposition interacting with unknown environmental factors -- which undoubtedly do not include vaccination or "refrigerator" mothers, or exposure to mercury. We just do not know. It is understandable that parents will seize on whatever explanation is offered, but they really, really need to resist that temptation. We do not have the explanation. But we can do something about it, particularly in the less severe cases.

But these prevalence numbers tell us we have a huge loss of human potential, and it truly does mandate urgency in figuring out what's going on. I want affected families to know that the reason we attack the anti-vax cultists and other quacks is because we care. We really do.


kathy a. said...

this is really, really interesting stuff. great to see the final de-bunking of the false link between autism and vaccinations.

also very good to see that early intervention really helps kids, even if we do not yet know what causes these kinds of problems. as you point out, it could be a variety of causes, and possibly varying combinations of genetic endowment and environmental factors. [my own suspicion is that if the family trees of persons on the autism spectrum were investigated -- going back a couple of generations or more, and following branches from there -- there would be evidence that others in the family have or had similar signs and symptoms, even if they were not diagnosed.]

it would be useful to better identify risk factors. but it is also very useful to know that early intervention helps -- which totally makes sense from a developmental standpoint. the human brain is developing rapidly in the early years, and i understand that there has been a lot of work in recent years about factors affecting neurological development, which may have lifelong consequences for the good or otherwise.

Cervantes said...

There is no doubt but that there is a strong genetic component. But to what extent it's familial as opposed to sporadic is unclear. In the past, at least, people with autism probably did not leave many descendants.

roger said...

"In the past, at least, people with autism probably did not leave many descendants."

it can't be good for the race as a whole to enable people with autism to reproduce can it? i know that may be a sensitive subject.

C. Corax said...

Are the rates of occurrence the same in other countries that keep track of such data? And if equivalent surveillance were possible in impoverished countries (I realize it's not possible), it would be interesting to see how prevalent ASD is in those societies.