In a recent paper purporting to link autism and vaccines, the author, Gayle DeLong, chose to lump autism and speech or language impairments together to create an autism “prevalence” from special education data:
To determine autism prevalence by U.S. state, the number of 8-year old students classified with either (1) autism or (2) speech or language impairments (speech disorders) was divided by the total number of 8-year-olds in the state.
I won’t respond to the personal attacks, but I will correct one error. The Herbert reference in the paper is incorrect. The correct citation is: Herbert and Kenet (2007) Brain abnormalities in language disorders and in autism. Pediatr. Clin. North Am. 54:563-583 (abstract: http://www.ncbi.nlm.nih.gov/pubmed/17543910). The paper shows that brain injury of people with autism is similar to brain injury of people with speech and language disorders. Another paper that makes much the same point is Herbert et al. (2002) Abnormal asymmetry in language association cortex in autism. Ann. Neurol. 52:588-596 (abstract: http://www.ncbi.nlm.nih.gov/pubmed/12402256).
Speech impairment is such a fundamental symptom of autism that the two conditions cannot be separated, especially when the child has a speech/language impairment that is strong enough to be classified as a learning disability.
First, kudos to Ms. DeLong for taking her statement to Orac’s blog. The participants are clearly not supportive of her statements—such as:
“Speech impairment is such a fundamental symptom of autism that the two conditions cannot be separated, especially when the child has a speech/language impairment that is strong enough to be classified as a learning disability.”
All I can say is that I disagree. Strongly.
Here is the definition that California uses for Autism as a special education category:
56846.2. (a) For purposes of this chapter, a “pupil with autism” is a pupil who exhibits autistic-like behaviors, including, but not
limited to, any of the following behaviors, or any combination thereof
(1) An inability to use oral language for appropriate communication.
(2) A history of extreme withdrawal or of relating to people inappropriately, and continued impairment in social interaction from
infancy through early childhood.
(3) An obsession to maintain sameness.
(4) Extreme preoccupation with objects, inappropriate use of objects, or both.
(5) Extreme resistance to controls.
(6) A display of peculiar motoric mannerisms and motility patterns.
(7) Self-stimulating, ritualistic behavior.
(b) The definition of “pupil with autism” in subdivision (a) shall not apply for purposes of the determination of eligibility for
services pursuant to the Lanterman Developmental Disabilities Services Act (Division 4.5 (commencing with Section 4500) of the
Welfare and Institutions Code).
One can not say that SLI and autism are directly linked.
The data don’t support it either. Her own paper gives administrative prevalence numbers for autism+SLI which are as high as 10%, about a factor of 4-10 than reported for autism.
Also, consider this: if you look at the administrative prevalence of SLI with age in a given year, it is sharply peaked at around age 6 or seven. Here are the data for SLI and autism from the most recent year data are available in California. ( click to enlarge )
(note—when this was first published, the graph was not in color).
No, that isn’t a “tidal wave” of SLI. This is what SLI looks like every year. It is a category mostly for younger children. I’ve often wondered if many of these children end up in the specific learning disability category in later years as this category is largely made up of older kids. Either way, it is clear that children tend to leave the SLI category, and they certainly aren’t being reclassified as autistic. Clearly, SLI is not something which can not be separated from autism.
I don’t know why Ms. DeLong chose to lump autism and SLI together for her study. I do feel quite strongly that the idea that they are basically the same is incorrect.
Herbert and colleagues reported that there was some similarity in the structural abnormalities of one cortical region (Broca's area) in kids with SLI and kids with autism who have language difficulties (but not in autistic kids with age-appropriate language). It's an interesting result, because the region in question is known to be very important for language. it's feasible that abnormalities of Broca's area cause language impairment in both groups (although it seems just as likely that the reverse is true). But it's one small region of the brain. To say that the nature of "brain damage" is similar in autism and SLI is to ignore the whole of the rest of the brain!
PS. I think the reference she was looking for was De Fosse et al 2004 (also in Annals of Neurology) - the Herbert et al 2002 study doesn't have any kids with SLI in it!