Thursday, December 06, 2012

Are All Epidemiological Autism Studies Flawed?

Autism diagnoses have been increasing at incredible rates.  Within the past two decades estimates of the numbers of persons diagnosed with autism has increased from 1 in 500 to 1 in 250 to 1 in 110 to 1 in 88.  Those are startling numbers to this humble, unsophisticated mind.  Yet to many in the autism research and Neurodiversity communities these numbers do not reflect a real increase. These smarter than the average bear types are sure that these incredible numbers are simply due to diagnostic definition changes, increased awareness and the existence of readily available of free treatments and services for people with autism. The  less certain, less Neurodiversity adherent among the autism epidemic deniers will admit that, at best only 50% of these increases can be explained by the diagnostic change, social awareness and access to services factors.   Assuming that these very smart people are right what does it mean for our understanding of what causes autism disorders?

Of what value are studies which purport to determine the possible role of any factor or event in causing autism based on autism rates over time  if those studies lack the ability to determine whether increasing autism rates  are real and if so to what extent?  

Are all epidemiological autism studies flawed that rely on measuring changes in autism diagnoses over time?  Will we have to content ourselves with discovering yet more gene, gene groups, gene expressions etc etc etc that MIGHT have SOME RELATION to autism?  Is this where autism research stands today?

1 comment:

Sue Gerrard said...

Originally, 'autism', was a descriptive label for the withdrawn behaviour seen in schizophrenia. Kanner originally used it descriptively as well, and it was only after seeing several children with similar symptoms that he began to suspect he might be looking at the same disorder in each of the children concerned.

Today a 'diagnosis' of autism is still based solely on behavioural characteristics and everyone whose behavioural characteristics meet the criteria for autism *and* whose symptoms have no known cause, qualifies for this 'diagnosis'.

I've put diagnosis in quotes because a diagnosis of autism is simply matching behavioural characteristics, not giving any indication of what causes those characteristics. A diagnosis couched in terms as vague as autism is, is bound to experience 'diagnosis drift', and to start to include those who would have been considered within the normal range 70, 50 or even 30 years ago.

A far more likely explanation for autism is that the triad of behavioural characteristics can be caused by lots of different medical conditions; in their 1992 book "The biology of the autistic syndromes" Mary Coleman and Christopher Gillberg list many different medical disorders documented as associated with autism. They don't know, of course, whether the autistic characteristics are caused by those other medical disorders, but it's a pretty safe bet some of them are responsible.

So it's quite possible that the so-called autism 'epidemic' might have been caused by several factors including broadening the diagnostic criteria - in practice if not in theory, increased awareness, and increased cultural expectations of conformity, as well as medical problems that are often not even looked at after a diagnosis of autism - because all symptoms are widely attributed to some hypothetical single condition we call 'autism'.