Sunday, 7 April 2013

Conceptual Map of Behavioural Homeostasis in Autism

In the research there are various scales to measure how autistic a child is, for example the Childhood Autism Rating Scale (CARS).  They are very subjective, but clearly better than nothing at all. I read a study on older children with ASD that was highlighting that as the children get older, they become less autistic.  In the CARS scale there are 14 behavioural areas to grade and then there is number 15, which is the general impression of the clinician.  In effect, number 15 is how autistic the clinicians feel the subject to be;  you would expect that number 15 would be consistant with the findings in the first 14 areas.  In the test the older children all showed a big improvement in areas 1 to 14, but not in number 15, which is the one that really matters.  This really means that either the use of CARS was inappropriate or CARS is flawed.

As children get older the concept of "normal" changes.  So there is not much point comparing a 15 year old with ASD to a normal/typical 3 year old.  So I decided that kids with ASD deserve a better framework, so that clinicians and parents have a better way to understand many important issues, ranging from prognosis, to just what is autism and where its boundaries really are.

In the Peter Autism Scale (PAS), typical kids start at birth with a score of zero. After 12 months, typical kids show rapid development; they acquire an increasingly negative PAS score. After puberty has ended, the PAS score stops declining and gradually heads back towards minus 20.  Much later in life, after starting their own family, the PAS drifts back up towards minus 10.

Autistic children are likely born with a positive PAS score; as they show either regression, or just the emergence of ASD-type behaviours, the score will increase. At a score of 60, self injurious behaviour is exhibited; at a score of 80 there is serious aggression and violence. The scale stops at 100, where the child needs physical restraint and is likely to be sent to an institution (depending on which country they live in).

There has been an explosion in the number of children diagnosed with ASD, most notably in the US.  Undoubtedly a great deal of this is explained by a bizarre, ever-widening of the definition of autism.  At one end of the scale, what were formerly nerds/geeks are now "on the spectrum" and at the other end of the scale, cases of mental retardation (MR) are now labeled as severe autism.  This ever changing, ever stretching, diagnosis is extremely harmful to the search for genuine therapies.

Mild social difficulties and obsession with computers or music is nothing new; it may indeed be worthy of a diagnostic term, but I do not think it is autism, or ASD.  On the PAS framework, you will note a blue cloud.

Asperger's syndrome already has a nice name; in the PAS conceptual map it follows in parallel the general path of typical development but has a score 12 higher.  It passes through the lower edge of the blue cloud.

Now to the serious case of what I shall term "disabling autism".  These are children with a PAS score of greater than 20.  These children need immediate therapy and their parents need external training and support.  If this help is received, there is good chance that the PAS score will never reach 60, let alone 80.  If help is not received and the PAS reaches 100, there will not be a happy ending.

In a perfect world, the prompt intervention (both behavioural and pharmacological) will give the child a fighting chance of gradually heading south to that big blue cloud.

Monty no BU no NAC
This is my estimation of Monty's development with no drugs
Monty + NAC
This is my estimated outcome using NAC alone
Monty + BU + NAC
This is my curreent prognosis using Bumetanide and NAC
Monty + BU + NAC + Agent X
This is my optimal prognosis achieved by adding the unkown Agent X

Monty himself is the first subject to have a PAS assessment.  Clearly this is just conceptual and so is purely illustrative;  nonetheless I think it has value.

Monty was not born a typical baby; he was always a bit different.  These differences grew and it was in his fourth year that he really began to acquire and demonstrate lots of new skills.  Then, with a great deal of behavioral support, it was plain sailing until he was about 8 and a half years old, when almost overnight, his world fell apart.  His longtime assistant had to leave him and his cosy world was overturned.  All the behaviours that he had avoided earlier, such as head banging and aggression to others, emerged with a vengeance.  Nine months later things had finally returned to "normal", but at a higher plateau of autism than 12 months before.

Three months later I trialed Bumetanide (BU in the chart) and three months after that plateau I trialed NAC.  In the PAS framework, Bumetanide appeared to provide a one-off 40% reduction.  Then with NAC, I took a wild guess at a 25% one-off reduction in PAS score.

Now we are in sight of the blue cloud, but a lot of work remains to be done.

Within 12 months, I would like to have found Agent X, which would give another 25% reduction in PAS score.  I think that will be as much help as can be reasonably expected from drugs.     

If it was not for puberty, it would be plain sailing into the future.  In the prediction, I have forecast another behavioral meltdown (loss of behavioural homeostatis), but since we now have experience of such a shock, I hope to manage it better.



  1. Interesting reading Peter. I dont' think there really can be a good testing scale. We are dealing with subjective humans - both the testers and the children with ASD. Keep up the good work Peter. I do hope you find the missing link.

    1. Thanks for the comment.
      I have a feeling that the "missing link" may actually be just a "lost link" hidden away in old research.

  2. Hi Peter, I think this is a really interesting idea and would very much like to try and 'map' my own son's behaviour vs time and compare behavioural homeostasis with interventions. I think this shows a really clear picture of your son's development, far more than any of the standard assessment scales could.

    I would say, however, that it also really highlights the heterogenity of ASDs. For example the points you mark on the PAS (self harm, aggressive behaviour) have little relevance in my sons's case, and to map his behaviour there would downplay the seriousness of his impairments. Instead, I would need to mark "Total zombie state" at the extreme end of the scale, with other points on the scale being "loss of/emergence of speech" and "oblivous to other human beings"; or possibly "extreme hyperactivity".
    I would also like to add a second y-axis, perhaps with medical issues that are also addressed through intervention. Again, in my sons' case this would be "severity of bowel issues". I imagine we would see a link between behaviour and comorbid medical conditions, punctuated by relevant interventions.
    As the previous comment points said, perhaps there cannot be a good testing scale, purely because to compare our children is almost impossible, but I think your mapping idea is great for looking at one child's progress, vis-a-vis their key issues and specific interventions.

    Looking forward to future posts.

    1. I am glad you liked this idea. I enjoyed coming up with it and it helped me to rationalise my own situation. As you point out, all autistic children are unique and deserving of their own custom-built framework.

    2. Stumbled across this today, an interesting approach indeed. Although my son has an Asperger's diagnosis, this model/scale shows his case is rather severe - and he is not alone. He had severe sudden self-injurious behavior at 6 years of age, and after a crisis-ridden year where all the best supports and environmental modifications were increasingly inadequate, he how attends a specialized school with extensive support for kids with severe aggression and physical interactions. If we lived in the other 99% of the US, he would have to be in a residential boarding program. For better or for worse, Asperger's is viewed as mild, "the geek syndrome". However, by definition, the social impairment is severe - not mild - and other impairments can be severe as well; only the language development is relatively intact. I actually prefer the new ASD definition, placing a severity level based on the amount of support required - not on whether and when language developed. Perhaps people's first assumption based on my son's diagnosis will no longer be "oh, what a gift, he must be very smart".

    3. Self-injurious behavior is indeed the worst aspect of any ASD. Until you have experienced it yourself, you cannot imagine how horrible it is. Now that I have found drug interventions that put my son back in control of his body, I am much less concerned about any such possible behavior in future. At the time we only had ABA to help us, which did work, but took almost a year.

  3. My daughter had a similar experience to Monty's. Following a transitional period that included her beginning preschool, a major change at her ABA therapy center, and my return to work, she abruptly began exhibiting symptoms such as extreme auditory sensitivities and anxiety that had been virtually non-existent. Her social interaction skills worsened somewhat as well. It took us several months to help her through this period, but when her symptoms finally leveled off, they remained more severe than they had been and still have not returned to pre-"crisis" levels. I have not known whether to consider this period a late regression of sorts (it also coincided with several back-to-back illnesses, which seemed to lend credence to the regression idea), but your story supports the more logical notion that her symptoms worsened in response to increased stress. Do you have an explanation for why stress would generate autism symptoms that a child had never presented before? Also, do you have a theory about why overall symptoms would be worse than before even long after the stressful event was over and the child returned to "normal"?

    1. Emotional stress triggers biological stress, which is actually called oxidative stress.

      Oxidative stress is a key part of autism and indeed many other conditions. An increase in oxidative stress can then trigger other changes, some of which appear to be epigenetic. Epigenetics is another key element of autism; it sounds complex and is. It is a reversible change in gene expression, so it could be just turning "on" a gene that should be "off".

      Prior to my son's stressful experience he never demonstrated any aggression or self-injury. Now he does. It is like a learned behavior, that cannot be "un-learned". The real cause is very complex.

      Something similar happens with seizures. It is actually very hard to have your first seizure, but that first seizure makes a change (I assume epigenetic) that lowers the threshold at which further seizures can take place.

      So the key is to avoid these stressful events. This can be done by having wide exposure to many therapists, relatives, favourite places, even pets etc. Then change becomes normal and nobody/nothing becomes irreplaceable.


Post a comment