Friday, 29 December 2017

Childhood Disintegrative Disorder (CDD) – Not a Useful Diagnosis?

Today’s post is about CDD (Childhood Disintegrative Disorder) also known as Heller’s Syndrome, which sounds rather nicer. It was first identified in 1908 by an Austrian, Theodore Heller. Later on came Hans Asperger, another Austrian and Leo Kanner who was born in what was the Austro-Hungarian Empire.

Why were Austrians so interested in Autism?

I started this post expecting that I would naturally be a supporter of the continued use of CDD as a diagnosis, I do firmly support calling an Aspie an Aspie after all. 
CDD is a diagnosis used for late onset severe regressive autism, which has fast onset, making it scary for all concerned.
Unlike many syndromes, Rett for example, Heller’s syndrome is not a defined genetic condition, it is just another observational diagnosis. This probably explains why it has been folded into the ASD diagnosis in the current DSM5, which sadly was also the case with Asperger’s.
Since it is not really a syndrome I will call it by its other name CDD (Childhood Disintegrative Disorder). I have not previously given much mention to CDD in my blog because I had assumed it was much worse than “regular” severe autism and that the disorder was well defined, so that it would be a clear case of CDD or autism.  It turns out this was a mistake. 

Back in 1994
In 1994 when Yale researcher, Fred Volkmar, was writing about CDD and severe autism he noted
More boys than girls appear to be affected. Childhood disintegrative disorder is perhaps 10 times less common than more strictly defined autism and is estimated to occur in between 1-2 children per 100,000. 

He was thinking strictly defined autism (SDA) was present in 0.015% of the population (15 children per 100,000), whereas I think today it is 0.3% (300 children per 100,000). I have got by zeroes in the right place. That is a massive twenty-fold increase in severe autism in 20 years, something that is never seriously investigated because DSM5 autism now includes 1% (1,000 children per 100,000) who have mild autism (mainly Asperger’s), so no statistics are directly comparable. 

Present Day
In today’s post we will see that CDD is just another broad umbrella term for a large number of different, often genetic, disorders. The case of CDD presented by a Yale researcher below appears to show a young lady less severely affected than many people with a diagnosis of severe autism. 

Well in case of Gina in the above article, it looks no worse than “regular” severe autism, certainly not terrifying. She rides a bicycle and helps around the home.

The Yale researcher in the above article is rather indignant that his disorder has “disappeared” into autism in DSM5 and so now he struggles to get funding.  What is telling is his comment that if you had a CDD diagnosis, clinicians would then naturally look for its biological origin, but now with a diagnosis of autism with ID (Intellectual Disability), no clinician will investigate further. Why is that??
The distinction has clinical implications. With a CDD diagnosis, the initial push is to hunt for a reversible cause. If the patient is diagnosed with autism and intellectual disability instead, that hunt never happens. "This means we may be missing a whole world of possible treatments for kids on the low-functioning end," says Westphal. 

Now this raises a question of why people with CDD are more worthwhile investigating, than anyone else. Surely all cases of severe autism should be investigated? There are no more wonder cures for CDD than there are for any other autism. As we have seen previously, there are numerous rare inborn errors of the metabolism that cause autism/ID that are treatable. 
Westphal wrongly assumes that there are no possible therapies for severe idiopathic autism. 

CDD as a sub-type of regressive autism
It would seem best to consider CDD for what it is, late onset regressive autism, with or without a “prodrome”, which is a distressing period of great anxiety lasting a month or two as the regression from normal to severe autism takes place.
In the young lady profiled in the above article, there was no prodrome and her regression took place at 3 years old and left her in a condition better than some I know with an autism diagnosis.
So here I actually agree with DSM5 that she would be better off with an autism + ID diagnosis; she should then have been taken to Johns Hopkins to see Dr Kelley, to check for mitochondrial disease. She looks no different to one of his cases of severe autism secondary to mitochondrial disease (AMD).
It looks like genuinely late regression, say 5 to 10 years old, points to some rare disorders like leukodystrophy, which is itself a family of genetic myelination disorders. These can be late onset and are degenerative.
Very likely within what has been diagnosed as CDD are hundreds of rare disorders, some degenerative but most not. It is the degenerative potential that makes CDD scary, but most people do not have this.
Some of these non-degenerative disorders will overlap with people diagnosed with severe autism, which itself likely includes many hundreds of rare biological disorders.
It certainly is important to get as precise diagnosis/description of each type of autism as possible.
So why not keep CDD along with Asperger’s as sub-categories of autism? All these rather vague observational diagnosis carry risks. You can have mitochondrial disease at age 5 that causes a regression to severe autism, I think the Yale team might well (mis)diagnose that as CDD, which is much more of an arbitrary diagnosis than I had realized.
An important issue is whether the disfunction is degenerative or not. It turns out that in a small number of CDD cases, the disintegration continues to an early death, but in most cases the condition becomes stable and in a small number of cases there is partial recovery, as with Dr Kelley’s AMD. This tells us that the observational diagnosis is pretty pointless. As always, what matters is a biological diagnosis.
The CDD researchers think some people with severe autism have CDD. I think people with CDD either have a rare genetic dysfunction, which may or may not be degenerative, they have mitochondrial disease, or it is “just” another case of idiopathic autism (the “I don’t know” category). So I do not really see the point of the CDD diagnosis.
The diagnosis should be ASD, its severity, speech delay or not, cognitive level, nature and time of onset, and then possible biological origin based on genetic testing. 

I think people diagnosed with CDD, who have no identifiable genetic disorder, really need to get tested for mitochondrial disease.
I wish some more intelligent people were in charge of autism research and collecting and interpreting data on prevalence. Is it possible that the prevalence of Strictly Defined Autism (SDA) was just 0.015% in 1994, versus 0.3% today? That means going from very rare to just rare in two decades. Maybe clinicians back then diagnosed a much larger group simply with MR/ID, that today would get diagnosed with autism + MR/ID. Pre Dr Wakefield and the MMR vaccine scandal, autism was rarely spoken about, even among doctors; today it has become quite a fashionable diagnosis. That is the best explanation I can think of; we do know that the diagnosis of MR/ID has fallen substantially in the last decades. 


  1. I do agree that an autism diagnosis is both a diagnostic dead end,and an excuse by lazy doctors not to look for deeper,underlying causes.I do not see this situation changing any time too soon.Too often what we diagnose as autism,especially regressive autism,and autism with MR/ID,is merely a feature of an underlying genetic disorder.I believe every person with regressive autism,or autism with MR/ID needs to have as complete a workup as possible.Metabolic,immune,and genetic,in the form of whole exome or whole genome sequencing.It is up to families to push doctors for more extensive workups.Doctors are not going to do it on their own.I firmly believe if we were to to both aggressively test for underlying conditions,and automatically rule out those with confirmed genetic disorders from being counted in the autism numbers,the incidence would fall dramatically.

    While not commonly known,regression is often not a one time-thing ,as I know all too well.You can have a severe regression,then if you are lucky,not suffer another acute illness for years.The autism and brain function can gradually improve,but then you suffer another severe illness,and you can have another regression,that leaves you in a worse position than you were before.

    Those who study CDD have a lot to learn from more garden variety regressive autism.Whoever wrote the Wikipedia page on CDD has a better grasp of the condition than most doctors do.CDD usually ha an underlying medical cause.While the diagnosis may have made sense when Heller first coined the term in 1908,it has no reason to exist in,what will soon be 2018.

  2. Hey Peter I was doing some reading on PV interneurons and came across a paper on schizophrenia in which one paragraph in particular caught my attention that discussed how mu opioid receptors can affect disrupt gamma waves in the hippocampus via inhibiting GABA release from PV interneurons directly. Prior to this, some other interesting info you probably know intimately already dealing with NKCC1/KCC2 is mentioned as well in terms of the context of schizophrenia.


    Nothing specific to autism (other than how maybe mu opioid receptor agonism can suppress GABA), just thought you might find the paper interesting.

    1. Thanks Tyler, and a Happy New Year to you and all our readers.


Post a comment