Verified by 
The classification of autism has been a mishmash. The very real disorder started out in the 1920s as children who refused to play with others and were indifferent to parental affection. All this was described much earlier than Leo Kanner’s work at Johns Hopkins University in the 1940s.
Then autism became steadily more convoluted as symptom after symptom was piled into the category, symptoms that themselves were very real but that did not necessarily belong to classical autism: “black-and-white thinking,” and so forth. Rather, they seemed part of other as yet poorly specified developmental disorders. Autism was becoming like the former “hysteria”: an undifferentiated wastebasket into which one could throw anything troublesome or problematic.
DSM-5 in 2013 attempted to clarify matters with the concept of “Autism Spectrum Disorder,” as though there were a single spectrum, like temperature, along which all cases of autism could be arrayed on the basis of severity, or whatever. This spectrum abolished such favored diagnostic categories as “Aspergers,” which said that children who were library rats were basically dissimilar from children who were unable to dress themselves.
The idea of a spectrum has met with general disbelief: The children who are playing the piano at Carnegie Hall have basically the same disorder as the children with an IQ of 50? Really?
Now Jennifer Foss-Feig at Yale University has devised a fairly sensible effort to make sense of the chaos of “autism” symptoms. Going on the model of schizophrenia, she has classified the symptoms of autism into (1) “positive” symptoms, things the kids do but shouldn’t, like pronoun reversal; (2) “negative” symptoms, things the children are supposed to do but don't, like making eye contact; (3) “cognitive features” (once called thought disorder in schizophrenia), meaning difficulties in thinking, as for example “difficulties with theory of mind.”
This tripartite classification of symptoms in schizophrenia has held up fairly well, given that some of the symptoms (the “positive” ones) are responsive to medication and others, namely the negative symptoms, are not. When something responds to medication, it’s always good to give it a separate listing in the disease classification as a signal to clinicians that, hey, here is a symptom we can do something about.
And this tripartite classification of autism symptoms seems vaguely feasible, although nothing in it is particularly responsive to medication.
But there is a huge problem with both the DSM-5 and the Foss-Feig classifications, and that is that they heavily include symptoms of another illness: catatonia.
This is one of the scandals of pediatric psychiatry: that for decades they have assigned such catatonic symptoms as repetitive movements to a separate illness that they just made up: “Stereotypic Movement Disorder.” DSM-5 accepts catatonia in adults quite readily, but that it might be found in children, hey, no way, they have SMD.
The failure of pediatric psychiatry to accept catatonia is a sign of a field facing intellectual bankruptcy. This failure is compounded when a really serious symptom in autism and intellectual disabilities, namely Self-Injurious Behavior, is not recognized as catatonic. The work of Dr Lee Wachtel at Johns Hopkins University and the Kennedy Krieger Institute, showing that SIB is a form of catatonia and that, as such, it is responsive to the standard anticatatonic remedies, such as benzodiazepines and convulsive therapy, has met with only partial acceptance by the Stereotypic Movement Disorder gang.
Bottom line: Let’s not get carried away with any of these new classifications. The DSM-5 approach manages to aggregate symptoms into artifactual bins (social communication deficits, restrictive behaviors), and the Foss-Feig approach has not been able to dodge the apples-and-oranges problem.
The tragedy here is that, among that many symptoms of autism that are unresponsive to treatment, catatonic symptoms are highly responsive. Instead, many clinicians are still administering neuroleptics, which may make the patients worse rather than better. Medicine has the power to relieve illness, but switch a train onto the wrong track and it may just ignore the red flags and keep going.
