Viewpoint on the brain disorder in autism (2000)

Working papers Viewpoint on the brain disorder(2003) (View in 2000)
and notes: The auditory system The inferior colliculus Hemoglobin & the brain
Concepts of autism Autism spectrum Social responsibility


Site Map Home
Conrad Simon Memorial Research Initiative /inc/docinfo.shtml
Page posted: April 24, 2000 08:02 PM
Updated (email address): January 14, 2002 04:35 AM
Website links updated (& subheaders added):  January 20, 2023 06:44 PM

My purpose in putting together this website is to tell Conrad's story, express my views on autism, and solicit feedback and discussion. Please contact me at with any ideas or information on cases like Conrad's. Thanks!

For starters

  1. Conrad's traumatic birth was an outrage. His older brother had been injured at birth, but my pleas for extra attention to prevent trauma the second time were ignored. The medical record of my pregnancy with Conrad describes me as "over anxious". Birth may be one of life's most traumatic experiences for many. Recent increases in the number children with autism may be due to (a) increased survival of children after traumatic birth, (b) increased use of over-the-counter and prescription drugs during pregnancy, and (c) failure to detect use of alcohol and other drugs of abuse during pregnancy.
  2. Conrad's death and its cover-up are outrageous beyond belief! State departments of mental retardation should be required to report all such cases to the Food and Drug Administration and for publication in journals read by clinicians who care for disabled people. No one in authority in Massachusetts planned to do this, so I did. I submitted a case report to the American Journal of Psychiatry because this is the most widely read journal by doctors who prescribe chlorpromazine (Thorazine). They returned my submission and told me to shorten it to a 500-word letter-to-the-editor, which I did and which they rejected. The contents of this letter are included in my description here of Conrad's death. How many other such deaths have taken place without being reported?
  3. following are excerpts from replies to a letter of mine that appeared in the January 11, 2000 issue of the journal Neurology. My comments follow each excerpt. My letter was written in response to four articles on autism published in the March 23, 1999 issue. I am pleased that that the authors took the time to reply to my letter. This is the kind of discussion I hope to generate:

Scroll or click for [papers on autism]   [my letter]   [replies & my comments]      [top]

Papers on autism in the March 23, 1999 issue of Neurology

  • Two of the papers were research reports:
  • The other two were a new hypothesis (by DeLong) and an editorial (by Rapin):
    • DeLong GR, Autism: New data suggest a new hypothesis.

      DeLong proposed that there are two distinct forms of autism, the first characterized by bilateral damage of the temporal lobes, and the second with a genetic etiology rather than damage to the brain. He proposed that the genetic predisposition might be reduced serotoninergic modulation of corticogenesis, especially in the left hemisphere.

    • Rapin I, Autism in search of a home in the brain.

      Rapin emphasized the need to determine the neural basis of autism. She discussed the five subtypes of pervasive developmental disorder (PDD) described in the DSM-IV manual of mental illness, one of which is classic autism for which DeLong proposes a defect in serotonergic modulation of left temporal lobe (language area) development. Another PDD subtype is disintegrative disorder, the form in which bilateral hypometabolism within the temporal lobes has been observed (Chugani et al., 1996, Annals Of Neurology 39:643-649). Rapin also pointed out that in about 20% of individuals with autism a currently definable biological cause has been determined. These include prenatal rubella infection, prenatal exposure to thalidomide or valproate, phenylketonuria, tuberous sclerosis, and fragile-X syndrome.


In my letter I proposed that perinatal disruption of energy metabolism might explain dysgenesis of temporal and frontal lobes, defects in serotonergic modulation of development, abnormalities observed in the cerebellum, and oculomotor signs - a much abbreviated version of the working paper I present here. Letters to the editor are limited to 500 words.

In my cover letter I listed the following points that I wanted to discuss except for the 500-word constraint; the authors did in fact mention these points in their replies:

  1. Beyond knowing that temporal, frontal, or parietal lobe and cerebellar functions are impaired, the etiologies of these impairments need to be sought. A Wernicke-type encephalopathy could be caused by many of the etiologic factors associated with autism, prenatal exposure to alcohol, viral infection, phenylketonuria, lactic acidosis, etc.
  2. There is evidence that prenatal exposure to alcohol may be a factor in autism (Nanson, Alcohol Clin Exp Res 1992;16:558). The case described by Rodier et al. (J Comp Neurol 1996;370:247) may well represent autism caused by maternal alcohol use.
  3. The essential lesion that produces the core syndrome of autism appears to be part of the more widespread brain damage caused by disorders such as prenatal exposure to alcohol, phenylketonuria, tuberous sclerosis, and prenatal rubella infection.
  4. Piven et al. found decreased size of the corpus callosum in autistic subjects (Am J Psychiatry 1997;154:1051), which has also been reported as part of the neuropathology in fetal alcohol syndrome (Roebuck et al., Alcohol Clin Exp Res 1998;22:339) and maternal phenylketonuria (Levy et al., J Pediatr 1996;128:770).
  5. Leigh syndrome is an example of a neurological disorder associated with Wernicke-type pattern of brain damage, a progressive degenerative disorder due to mitochondrial DNA mutations in some cases. The core syndrome of autism is not progressive, yet there may be some marginal dysfunction of mitochondrial energy production that affects primarily the auditory system with its high metabolic demands in some autistic children.

Replies and my comments:

Scroll or click for replies from  [DeLong]  [Minshew]  [Müller &  Courchesne]    [back]

From Isabelle Rapin, MD:

  • Perinatal factors are not among frequent nongenetic causes of autism. Experimental asphyxia in neonatal monkeys damaged brainstem nuclei with the highest metabolic rate, but whether these animals would have had autistic symptoms is unknown and relevance to Leigh's disease, let alone autism, is tenuous.

    My comment: First, I believe that perinatal factors are downplayed as predispositions to autism. Second, monkeys do not develop language but asphyxiated monkeys had developmental deficits in motor function and perception as discussed by Windle in the October 1969 issue of Scientific American. Denny-Brown (Proc Royal Soc Med, 1962, 55:527-538) reported disturbed responsiveness in monkeys with ablations of the superior and inferior colliculi - the area of the brain damaged by experimental asphyxia. I discussed Denny-Brown's findings in my paper on echolalic speech (Arch Gen Psychiatry, 1975, 32:1439-1446).

  • In my experience unstigmatized autistic children have no clinical signs of cranial nerve involvement. Involvement of brainstem nuclei in autism was raised by three observations: the autopsy study by Rodier et al. (1996) of an autistic woman exposed to alcohol in utero, cranial nerve involvement and autism in 4 of 100 cases of thalidomide embryopathy, and reports of an occasional child with Moebius syndrome and autism.

    My comment: Children with autism do not have stigmata usually associated with fetal alcohol syndrome and other forms of severe mental retardation. Autistic children are often remarkably attractive. On the other hand individuals exposed to toxic substances in utero show characteristic physical anomalies involving facial features and may display autistic behaviors. Moebius syndrome is a congenital condition in which facial nerves are paralyzed bilaterally and some of these children display autistic behaviors. If there is cranial nerve involvement in unstigmatized autistic children, it is minimal but I believe may be evident in decreased facial expression. Wernicke's encephalopathy in its severest (and purest) form has been shown to affect the auditory system more prominently than the cranial nerve system (see Vortmeyer et al., 1992, Journal of Neurology, Neurosurgery and Psychiatry 55:826-829 and Chen et al., 1997, Pathology International 47:748-756 ).

  • Preliminary data do not point to a subcortical cause for the language disorders of verbal autistic children. Verbal auditory agnosia (VAA) is not the language deficit of autism. It is the most frequent language deficit in acquired epileptic aphasia (Landau-Kleffner syndrome, LKS), but in LKS the evidence points to pathology in temporal auditory cortices, not the brainstem.

    My comment: Language depends upon intactness of the temporal lobes. Development of the temporal lobes may be dependent upon integrity of the brainstem auditory pathway. Landau and Kleffner (Landau et al., 1960, Neurology 10:915-921) also reported neuropathology in a case of congenital aphasia in which the most prominent damage was in the medial geniculate bodies of the thalamus.


From G Robert DeLong, MD:

  • Wernicke-type encephalopathy from hypoxic-ischemic injury is common in the perinatal period, but no significant correlation has been established between such perinatal hypoxic injury and autism.

    My comment: A significant correlation of hypoxic-ischemic injury with autism has not been established. I believe this should be a more vigorous focus of research. Matsuishi et al. (J Autism Dev Disord ,1999, 29:161-166) reported a significant correlation between meconium aspiration and autism in survivors of a neonatal intensive care unit.


From Nancy Minshew, MD:

  • The theory that perinatal brainstem injury that interferes with development of the forbrain is reminiscent of the "Whisper of the Bang" theory proposed by Tanguay and Edwards (J Autism Dev Disord, 1982, 12:177-184) to explain abnormalities in brainstem-evoked potentials, but there is no clinical or empiric neurologic evidence to support such a pathophysiologic model of autism.

    My comment: No clinical or empiric neurologic evidence supports a model of disturbed caudal-to-rostral development in autism. Brainstem auditory evoked potential (BAEP) data indicate subcortical anomalies in the auditory system of some children with autism. This should stimulate more research on auditory function in autism.

    Not all BAEP data supports auditory impairment in autism, but this may be a function of the type of auditory stimulus used. Damage at the level of the inferior colliculus should interfere with orientation to direction of sound sources. Sounds presented out of phase at each ear might be one way of investigating the integrity of the inferior colliculi.

  • Disruption of aerobic metabolism in the brainstem is a common consequence of neonatal hypoxic-ischemic injury, but empiric evidence does not support a connection to development of autism.

    My comment: Many reports of brainstem damage in infants who died at birth have been published. Who knows what developmental disabilities would have been evident in these children if they survived? Neuropsychological tests might be devised that could determine which groups of developmentally disabled children survived with less severe injury to these brainstem centers.

  • In my experience children with autism very rarely have clinical evidence of cranial nerve involvement. Rather, the paucity of facial expression is of the type seen in PD (Parkinson disorder?) - facial expression is masked during social communication but becomes animated when the autistic individual thinks of something funny or experiences emotion internally.

    My comment: Children with autism do not exhibit signs of neurologic impairment, which is why the syndrome is defined by behavioral characteristics rather than in terms of any neurological system. Again, neuropsychological tests might be developed that could detect impairments of subcortical sensory pathways and cranial nerves.

    Facial expression in individuals with autism may look distorted, especially as they grow older. For example, a smile may look more like a grimace. This is even true of some of the forensic patients where I now work - an attempted smile may look like a snarl. I believe many of these men who have come to the attention of the police may represent cases of Asperger's syndrome. Some are permanently inebriated due to their own alcoholism, which accounts for classic oculomotor and cranial nerve signs.

  • The study we reported on saccadic eye movements demonstrates that there is no evidence of abnormalities in the function of the oculomotor nuclei in the brain stem.

    My comment: Other testing methods might reveal oculomotor anomalies in autistic children. This was one of the points in my letter. Results of the procedure used are informative but do not irrefutably demonstrate that there are no abnormalities in the function of oculomotor nuclei.

    First, a technician gave all instructions over an intercom. Even high functioning autistic subjects may still have trouble with auditory communication; written instructions would be preferable. Further, some method might be needed to confirm comprehension of the instructions. High functioning autistic subjects tend to respond in "sound bites" (similar to earlier echolalic speech). Thus a technique for obtaining a fully re-worded explanation of the instructions is needed.

    Second, subjects were to remember the location of a peripheral target but to continue looking at a central target, and to look at the peripheral target only after the central light was extinguished. Peripheral vision is a function of the superior colliculi. What if there is a disturbance with this function? Most people, including autistic children, learn to compensate for deficits of any kind; these compensations may be practiced to the point where they become reflexive. At least one alternate way of testing frontal control should be used.

  • The mechanisms that underlie the development of neocortical systems are exciting areas of ongoing research. The importance of defining these mechanisms and their genetic basis is the reason the disorders with a strong genetic component like autism have become of such great interest to neuroscience.

    My comment: Serotonin, GABA, endorphin, and other yet to be discovered modulators of neocortical development are exciting areas of research. As a computer scientist I find construction of systems that respond to multiple external signals to be equally exciting. I do believe a systems approach to understanding neural development and evolution will be more informative than starting with its genetic basis.

    What survival purpose did evolution of lateral line systems, the vestibular system, vision, and hearing provide? How have neocortical elaborations enhanced these further? At what stages of evolution or maturation do serotonergic or endorphin neurons appear? These questions can best be explored beginning with the most primitive single-cell to multi-celled species.

    It is not proven that autism is genetic in all cases. There are many other pre- and perinatal hazards that can interfere with development of the nervous system. Widespread use of over-the-counter and prescription drugs in addition to drugs of abuse may account for the increasing numbers of autistic children, plus the fact that more infants at risk can now be kept alive.


From Ralph-Axel Müller, PhD and Eric Courchesne, PhD:

  • We agree that the effects of intrauterine neurodevelopmental disturbances, such as hypoxia or neurotoxins, may share certain features with autism... However, we do not believe that autism can be explained in terms of a unique etiology.

    My comment: Disruption of aerobic metabolism can be thought of as a unique etiology with many causes, including anoxia, infections, or exposure to toxic substances. Such perinatal factors have been associated with autism. Interference with aerobic metabolism is probably more likely than genetic or environmental factors that directly affect enzyme systems for neurotransmitter synthesis or nerve propagation.

  • One of the established etiologic findings in autism is the importance of genetic factors.

    My comment: Genetic factors are not clearly established in the etiology of autism. Concordance in identical twins is not 100% and concordance in fraternal twins is higher than among singleton siblings. Even one pair of identical twins discordant for autism provides a counterexample that refutes a single gene etiology. Higher concordance in fraternal twins implies an intrauterine factor in some cases. Polygenetic factors such as infants with large heads born to mothers with small pelvic dimensions may be significant.

  • It should be noted that auditory agnosia or oculomotor abnormalities, although often observed in autistic individuals are by no means necessarily linked to autism. Most individuals with auditory agnosia or oculomotor impairments are not autistic.

    My comment: Clear-cut auditory agnosia or oculomotor impairments would be the basis for a neurological diagnosis, even if autistic features were observed.

    Oculomotor impairments are not at the root of autistic disorder. If a Wernicke-type pattern of neuropathology is, then minor oculomotor deficits might be an accompanying sign, depending upon how extensive the involvement of periventricular tissue.

    Verbal auditory agnosia should be a focus of research in autism. Agnosia for stressed syllables or other sounds that signal separate words in streams of speech may be at the root of the language disorder characteristic of children with autism, especially echolalic children.

  • It appears unlikely that autism can be explained as a single disorder resulting from a single etiologic path and affecting a single brain structure or region.

    My comment: I believe that impairment of function within the inferior colliculus could cause verbal auditory agnosia, and the defect of general awareness characteristic of children with autism. The inferior colliculus is clearly vulnerable to many injurious factors, but other brainstem nuclei of high metabolic rate are also likely to be affected by factors that impair functions of the inferior colliculus, which may account for differences from case to case.