A Case for Neurobiological Work-up in Autism

Glenn Vatter
3124 Henneberry Road
Jamesville, NY 13078-9640
(315) 677-3844
  Reprinted with the permission of the author
© 1998 Glenn Vatter
All rights reserved

I have been noticing some parents taking the shotgun approach to seeking interventions for their autistic child. They try every known intervention until they find the one that appears to work. Perhaps a more rigorous diagnostic procedure would better identify the nature of the autistic behaviors. Then an intervention based on the diagnosis could be applied. Such a regimen is suggested in The Biology of the Autistic Syndromes:
The likelihood that a medical diagnosis can be made for a child with autism is no longer remote. In a recent study by Steffenburg (1991) 37 percent of a population-based series of carefully defined autism cases - conforming to DSM-III-R (American Psychiatric Association 1987), and Rutter (1978) criteria - were diagnosed after intensive neurobiological investigation as having at least one of the medical conditions described in this book. (Moreover, this percentage does not include the epilepsy/infantile spasms cases in this group.) One important finding of the Steffenburg study was that in at least half of these cases the diagnosis would not have been made had the medical investigation not been exhaustive.

In addition to the 37 percent of children who had a definite medical diagnosis, 50 percent had clear signs of major brain dysfunction not seen in normal children. Another 6 percent had a close relative with Asperger syndrome/autism, providing the only clue to a possible aetiology. A few more had a positive family history combined with clear evidence of brain damage. Of the total study group, in only about 5 percent was there no clue at all as to the underlying aetiology of their autistic symptoms.

These figures could be taken to mean that at least 35 percent of all autism cases are causally related to an underlying known medical condition, 5 to 10 percent are clearly related to other hereditary factors (autism/Asperger syndrome or some other autistic-like condition), 50 per cent have unspecified major signs of brain dysfunction which could be brought about by medical conditions (hitherto undiscovered), hereditary factors or a combination of the two, and only 5 percent have no strong or clear-cut clue as to the underlying aetiology.

Two professionals I contacted on this issue reported different percentages in number of autism cases for which the cause is found. One friend has reported 50% causal, 50% unknown and another friend has reported 8% causal and 92% unknown. The Steffenburg study shows 95% causal and 5% unknown. This is probably due to a more extensive work-up.

Inasmuch as The Steffenburg study gives hope that a complete neurobiological assessment may give a clue as to the underlying cause of the autism in a great many cases, it seems that may be a wise move by parents who could afford such an assessment. In some cases the early intervention program, the school district, or family insurance may cover the costs. Enclosed is table 1, which expands a table found in the Biology of The Autistic Syndromes. A discussion of the assessment is given in the book. This chart also appears in my book Billy, A Family's Experience With Multiply Disabled Child. Additional traits such as hand wringing, allergies, sleep problems and so forth could be added to the assessment checklist.

My hope is that, given such an assessment, and discovering something about the probable etiology, then perhaps the more likely to succeed intervention could be tried first. For instance, if the etiology appears to point to brain anomolies, then the Applied Behavior Analysis method may be called for. This method may work by finding new pathways in the brain, i.e. connecting up dendrites to accomplish pathways which are deficient. Much data is now emerging about the way a young child's development is helped by the environment, i.e. lots of oral and visual stimulation. If, on the other hand, a metabolic disorder is found to be the underlying cause of the autism, then a nutrition-based intervention would be in order. If a metabolic disorder is the problem and not brain anomolies, then one could theorize that the dendritic connections are already there and what is needed is better regulated chemistry, i.e. neurotransmitters, to carry the messages. Nutritional-based interventions ultimately serve to better regulate the chemistry. (This is my theory, do any professionals out there agree?)

One of my professional friends commented that her take on this is that the more severely impaired kids (often those with organic brain damage) respond better to ABA not because of neuro pathways but rather because of cognitive abilities (low) and methods such as ABA need very little level of functioning.

My suggestion that Applied Behavior Analysis applies best in cases where there are brain anomolies rather than metabolic problems correlates with the Lovaas studies done in the 1980's. In those studies 47% of the autistic children were vastly improved by the ABA protocol. Were these the ones with brain anomolies? Remember the Steffenburg study above reported 50% of subjects with clear signs of brain anomolies.

Autism diagnosis is typically done using DSM-IV, ICD-10, or other protocols based on observations of behavior, and not routinely followed up with a complete neurobiological, neuropsychiatric assessment. Such an assessment is quite involved and may be beyond the resources of many parents. Also, one does not run blindly through all of the tests. The developmental specialist who is organizing the tests should make an assessment as to which tests are appropriate for each child. However, I have observed some parents going to no ends to find appropriate interventions. These parents may be interested in pursuing the complete work-up if they were aware of it. The professionals they work with may select only certain portions of the work-up depending on their observations of the child and his history. After completion, more than one therapy may be found to be appropriate.

If a metabolic disorder is suspected, The DAN protocol may be used to pinpoint it. Tests for possible metabolic deficits, along with suggested treatment, are outlined in the report entitled Clinical Assessment Options For Children With Autism And Related Disorders. This report emanated from a conference of Autism professionals (the Defeat Autism Now! conference ) in January 1995. This report is written as a guide for doctors and parents seeking to narrow down the specific metabolic errors in their autistic children. This report is available from the Autism Research Institute in San Diego, California. Tests for the following possible deficits are contained in this report:

There seems to be a strange quirk in autism diagnosis not seen in very many diseases. The diagnosis is done by observation/reports of behaviors. Given a fit to DSM, ICD, or other diagnostic checklists, the diagnostician declares his job done. He tells the parents " Your child is diagnosed with autism (or whatever subgroup applies). There are many theories as to causes and treatments but nothing definite. I will steer you to some sources of information. Good luck". This is like going to the doctor and he says "Yes, you sure are sick. Good luck". This is a fallacy in the autism diagnostic process. In most diseases the diagnostician does a complete work-up of tests to find out the cause of the illness. From the cause, a treatment protocol is prescribed. In most cases of autism, the parents are left to the discovery process by joining a parent support group, going to seminars and conventions, buying books, joining the ASA, and blindly trying suggested interventions until maybe one is effective. The Steffenburg study and the information contained herein suggests that there is hope for a more classical approach to finding and treating the cause of the behaviors.

Table 1
Neuropsychiatric assessment checklist for autism

Materials Needed

  1. Autism assessment questionnaire: use validated questionnaire such as HBS (Wing 1980) or CARS (Schopler 1980) {or Rimland E2 questionaire, plus other non-verbal psych-educational tests as needed to assess intellectual/developmental abilities}
  2. Medical records of patient (gestational, birth, neonatal, pediatric)
  3. Cranial circumference tape
  4. Small ruler to measure auricle length and interpupillary distance
  5. Woods lamp (to look for diagnostic skin changes)
  6. Standard equipment for neurological examination, [opthalmoscope, relex hammer, etc.]
  7. (optional) Video equipment for making a permanent recording of child's behavior for your records.

Special space requirements

  1. Room for child to run [with low visual and auditory stimulation]
  2. Stairs for child to climb
  3. Table with chairs having foot and arm rests.

History and clinical assessment

Before seeing the patient and family
Review of medical records: Family present
  1. Detailed structured assessment in respect to autism using validated questionnaire
  2. Detailed psychiatric history:

General physical examination

Non-intrusive observation throughout the examination. Observe posture, gait, and what the child is grasping in his/her hand. Is child rocking or whirling? Or catatonic? Observe any adventitious movements, such as self-biting, hands over ears, hand flapping, clasping, wringing, clapping. Fingers or other objects in the mouth? Is there facial grimacing? Is child performing visual self-stimulation on a pattern in your office? Note and record any myoclonic jerks. If possible, observe spontaneous handedness. Record spoken language by child, if any. Observe how the child is processing information. Is he getting overstimulated or distracted by visual/auditory information? Is there tactile defensiveness? Delay in response?



Age-appropriate neurodevelopmental/neurological examination

Cranial nerves
Assess by inspection if child uncooperative
Check standard deep tendon reflexes; also plantar flexion and clonus. Test for snout reflex 1 and visual rooting reflex2.
Sensory examination
Usually impossible; make a try at praxis testing at least. If possible, observe play skills in a large area and on suspended equipment.
Extrapyramidal examination
Assess muscle tone, observe child on stairs. If child runs, look for dystonic, choriec and/or athetoid posturing. Check milkmaid sign3.

Neuropsychological evaluation

Should be performed by a clinical psychologist experienced at testing children with autism. The psychologist must know how to do cognitive testing of children with autism, which includes tests appropriate to age, developmental level, amount of language and degree of cooperation (WISC-R, Leiter, Rave, and Griffiths may all be appropriate).

Laboratory examinations

  1. Blood tests:
  2. 24 hour urine:
  3. CSF: protein electrophoresis (to rule out progressive encephalopathy)
  4. Other essential laboratory tests:

  1. Snout reflex - primitive reflex in infants, an involuntary pouting elicited by tapping the center of the closed lips.
  2. Visual routing reflex - another primitive reflex, an involuntary rooting in response to the sight of food.
  3. Milkmaid sign - a sign of underlying basal ganglia disease; the examiner "milks" the finger of the patient as if milking a cow; the sign is positive if the finger tremors.