Article Details

A Letter to Dr. Lynn Waterhouse

Categories: | Author: Matt O'Leary | Posted: 2/24/2009 | Views: 56698

This is a copy of the letter I sent to Dr. Lynn Waterhouse, an autism research pioneer, in response to an article she wrote called 'Autism Overflows: Increasing Incidence and Proliferating Theories. My letter attempts to integrate much of the content of her paper with my theories. I will let you know if she responds, which I suspect is unlikely, but we will see.

Dr. Lynn Waterhouse
College of New Jersey
            Re: A Non-Ad Hoc Theory of Autism
Dear Dr. Waterhouse,
I ran across your paper Autism Overflows: Increasing Incidence and Proliferating Theories a month or two ago. It is by far the best summary of many of the divergent avenues of scientific investigation into autism that I have seen. And, I found your description of the ad hockery of this investigation compelling. I also share your apparent concern about the lack of scientific progress in integrating the results of this vast investigation into autism into a useful model that can be used to help autistic individuals. Quoting you back to you:
Despite the overwhelming flood of causal theories of autism, the field has not made progress in creating a synthesized, standard predictive causal theory of autism…
I am writing to you to offer such a theory that hopefully avoids the problems with ad hockery you describe so well in your paper. I believe my theory can offer an avenue to unify some or many of the theories of autism you discuss in your paper. I will try to keep this relatively brief, so that I might increase the chance you will read about my theory. I can offer more detail upon request.
I became interested in autism about 4 years ago after I (at the age of 36) was diagnosed by a pediatrician friend of mine as sensory defensive, and my brother shortly thereafter received an adult Asperger’s Syndrome diagnosis. Upon some investigation, I determined that I (and my father) shared characteristics of the autism spectrum with my brother, though I am not diagnosable as disordered. Being on the broad autism phenotype has provided me some insight into autism that many others likely do not possess. It also provides me some core competencies (i.e. strong problem solving and a tendency to obsess on areas of intellectual interest – autism in my case at this time) that I think suit me well for this type of intellectual pursuit.
Let me start by mentioning a problem I generally have with the various theories of autism you have mentioned, and others you have not cited: they don’t explain the complex array of behaviors, deficits, and abnormalities seen in autism. They tend to focus on one or a few characteristics typically on the autism triad (i.e. lack of emotional reciprocity or speech delay) and generally ignore all of the other stuff, from sensory processing abnormalities, to control seeking behaviors, to a tendency to melt down, to frequently deranged cortisol levels.
Let me also note that I don’t think that autism is ultimately a disease / disorder unto itself. I believe that one reason that science has not identified a common biological marker for autism is that there is not one common to all that can be easily measured. I think autism is ultimately a set of behaviors engaged in and deficits / abnormalities experienced by those who have a series of homeostatic imbalances that are caused by a diverse array of genetic, epigenetic, and environmental factors. That is why we only have behavioral / deficit criteria by which to diagnose autism.
My Theory of Autism
My theory is ultimately rooted in the interplay of two systems, the nervous and endocrine systems, which are linked through a series of positive feedback loops. I believe that autism results from the unbalancing of these systems: in the case of the nervous system, it primarily involves abnormal neurological excitation, and in the case of the endocrine system, it involves an overly reactive stress response. Science’s inability to establish this dual contribution to autism (if my theory is right) has to do with the nature of these systems.
For instance, abnormally elevated neural excitation can result from too much excitation or too much inhibition, or both; it can be caused by genes or by the environment; it can result from abnormalities in the hind brain as well as midbrain and forebrain structures; it can involve problems with interneurons as well as long distance signaling neurons, or problems with how these neurons are myelinated; it can involve some or all of these factors in widely varying combinations. If convergent causation is occurring (as I theorize) such that autism can develop through multiple overlapping pathways, it is no surprise that science is having trouble establishing statistical significance in single experiments that measure one aspect of excitatory balance – like GABA levels in the cerebellum in a cohort of kids.
Also, the unique nature of the stress response tends to confound scientific measurement. This is for many reasons, including the following. First, the stress response is inherently non-specific, meaning that a multitude of environmental exposures (from extreme cold to loud noise to an emotional challenge) result in the activation of the same pathways (i.e. sympathetic nervous system, HPA axis). Second, as described in detail in Why Zebras Don’t Get Ulcers (Sapolsky), chronic activation of the stress response results in different outcomes in different people – for instance, in response to chronic stress, cortisol levels may be generally elevated, may be generally suppressed, may show altered daily variability, and/or may fail to show marginal increases to new stressors. Third, measuring cortisol levels once or a few times is largely useless – to gain meaningful data, you have to measure multiple times a day over a significant period, to establish a trend and any change in that trend depending upon environmental exposures. These factors, and others (like how the simple act of testing a child may result in stress response activation or the fact that levels of stress hormones normally vary widely from individual to individual), confound the ability of science to demonstrate abnormalities in stress response activation.
Excess Neural Excitation
I am convinced that the core initiating factor in most cases of autism is an abnormally reactive nervous system. I am not alone in believing that neural excitation is a key factor in autism. There are many researchers who have postulated that neural excitation, and its balance with inhibition, may play a crucial role in autism (Walsh et al, Markram et al,  Belmonte et al, Rubenstein and Merzenich, Rippon et al, and Hussman).
There is substantial evidence to support the centrality of neural imbalance in autism. There is increasing evidence that the genetic mutations that result in an increased likelihood of autism, including fragile x syndrome, tuberous sclerosis and Rett Syndrome, may cause their effects due to their impact upon neural excitation, particularly through the production of proteins that affect synapse development. Intracellular proteins (with associated genes) implicated include FMRP, MeCP2, mTOR, TSC1/2, and NF1. An excellent paper on this topic was done by Kelleher and Bear. There is also evidence of abnormalities in synaptic binding proteins that appear to regulate the balance of excitatory and inhibitory synapses, including neuroligins (i.e. NLGN4), their binding partner neurexins (i.e. NRXN1), and other related genes / proteins (i.e. SHANK3, CNTNAP2). Furthermore, evidence is accumulating that several genes associated with autism play roles in the development of cortical inhibitory interneurons including c-Met, DLX, and ARX, where rare, partially inactivating mutations cause mental retardation or autism.
There is also much evidence that abnormalities in GABA inhibition are generally associated with autism, which is too voluminous to describe fully here, such as the fact that one of the best mouse models for autism is the GABRB3 knockout mouse. (See GABA in Autism, Dhossche et al.). However, one particularly important factor, and one of the most consistent findings in autism, is a reduction in the size of several structures of the cerebellum, including the Purkinje cell layer and the deep cerebellar nuclei. Both of these cerebellar structures are highly GABAergic. Several of the environmental causes of increased autism incidence potentially work through impacts upon the cerebellum, including prenatal exposure to valproic acid and thalidomide during neural tube closure, as well as fetal alcohol syndrome.
Also, abnormalities in both subtypes of glutamic acid decarboxlyase (“GAD”) (the enzyme that converts glutamate to GABA), GAD65 and GAD67, have been reported in autism, implying a deficit in GABA (or excess of glutamate, or both) in some cases of autism. Autistic parietal and cerebellar cortices are reported to have an approximately 50% reduction in GAD levels.
There is also evidence, though not as extensive as that for GABA, that increased abnormalities in the glutamate system (particularly related to NMDA and mGlu receptors) are also involved in some cases of autism (these are both involved in second messenger pathways that affect synapse stabilization).
            Abnormal Activation of the Stress Response
In my theory, the key partner in autism etiology with elevated neural excitation is abnormal reactivity of the stress response. There is substantial evidence of unusual stress response activation in autism, despite the measurement problems I describe above.
                        Biological Markers
Cortisol levels are typically abnormal in autism, with elevated cortisol being the most frequent finding. General arousal state, mediated substantially by noradrenaline, is consistently elevated in autism. And, several studies have revealed that levels of ACTH were significantly higher in autistic individuals than in normal controls. Also, one study revealed that levels of beta endorphin were significantly higher in the autistic individuals than in normal controls.
Researchers have also shown very elevated levels of various substances known to be substantially increased by stress. Researchers have found highly increased levels of certain neurotrophins and also certain neuropeptides – BDNF, NT4/5, VIP, and CGRP - in substantial numbers of children with autism compared to controls. Another study showed that mean levels of the neuropeptides Substance P, PACAP, and neurotrophin nerve growth factor were elevated in the cord blood of children who later received a diagnosis of autism or mental retardation.
            Co-Morbid Conditions
Researchers have shown that many disorders that are directly associated with chronic stress are substantially elevated in autism, including gastrointestinal and autoimmune conditions, anxiety disorders, and sleep disorders.
                        Behavioral Indicators
Many of the behaviors shown in many if not most cases of autism are indicative of individuals under stress. These behaviors include control seeking, self stimulation and mutilation, eye contact avoidance, social withdrawal, abnormal arousal state, hyperactivity, perseveration, and others.
            Positive Feedback Loops
A very important connector between excessive neural excitation and abnormal stress response activation involves positive feedback loops. According to my theory, upregulation of neural excitation leads to greater levels of stress response activation through various pathways – for instance, the sensory defensiveness typically seen in autism results in a brain that perceives a greater number of threats in the external world, setting off the stress response much more often than is normal. Chronic stress response activation results in upregulation of neural excitation – the most important pathway for this is through cortisol upregulation. Cortisol impacts neural excitation through various mechanisms, including upregulation of glutamate function, inhibition of GABA function, augmentation of amygdala function, and degradation of hippocampus function. This feeds back to higher levels of neural activity, which results in greater stress response activation, and so on.
Explaining Autistic Behaviors
One advantage of my theory of autism is that it offers fairly reasonable explanations for most of the behaviors / deficits / abnormalities seen in autism, including the ones that are not part of the ‘triad’ of impairments and therefore often ignored, but are clearly part of the autism experience. A few examples:
-          Sensory processing abnormalities: these are caused by alterations in the balance between excitation and inhibition in the brain. This imbalance results in a brain that has problems filtering sensory information, experiences problems separating signal from noise, and has difficulties with discrimination, habituation, and sensitization;
-          Communication difficulties: if your brain is constantly overreacting to the external world as threatening, communication other humans (particularly those you can’t read because of your sensory issues) is fraught with danger and likely to result in overstimulation. The best strategy to avoid this is to avoid communication, unless you really need something;
-          Obsession with patterns, routines and ritual: control seeking behaviors that improve cognitive appraisal, reducing the perceptual impact of external stressors;
-          A tendency to withdraw from the world: withdrawal away from stressors that are ever present in the life of a person with autism;
-          The melt down: the likely result of a physiological and psychological inability to deal with additional transient stressors;
-          Seizures: The result of an electrical imbalance in the brain, with the balance shifting to too much excitation.
Integrating My Theory with Those You Cite
Another advantage of my theory is that it can be reconciled with most of the other theories that you discuss in your paper. It is thus less ad hoc (though it certainly may suffer from other flaws) than the other theories you describe.
Reduction in Oxytocin (Domes)
A deficiency in oxytocin in autism is consistent with my theory in several ways. First, exposure to acute and chronic stress reduces oxytocin levels in the brain. Second, oxytocin generally down regulates stress response activation. One way it does this is it facilitates GABA inhibition (GABA in Autism, Dhossche et al.). Increased oxytocin can help restore excitatory balance.
Reduced Region to Region Brain Connectivity (Just and Takarae)
I suspect that the under-connectivity postulated by Just and Takarae is real in many cases. In autism, the balance between neural integration and specialization seems to be thrown off in favor of specialization. This is seen in altered ratios between grey and white matter (more grey), as well as degradation of white matter in many of autistic individuals studied (which may be part of the same phenomenon). I would not be surprised if this phenomena underlay savantism (more processing / less communication). My theory would suggest underconnectivity is a neural reaction to abnormally high levels of excitation in the brain – similar to the way receptors may downregulate in response to high levels of neurotransmitters. I tend to think the immune system is the main actor in this phenomenon, based upon findings of abnormally high levels in autism of antibodies to basic brain tissues such as myelin basic protein (I note a recent interesting article on how the complement system is largely responsible for neural pruning in children).
Weak Coherence and Strong Local Processing (Happe and Frith)
My reconciliation would be similar to that above.
Neuroinflammatory Processes Leading to an Enlarged Head (Courchesne and Pierce)
I have difficulty reconciling to this theory. My theory predicts abnormalities in inflammation generally in autism, largely because of the impacts of the stress response (cortisol levels in particular) on inflammation. And, it suggests that inflammation is involved in the positive feedback loops that develop – such as how IL-6 and related cytokines (elevated in autism) sensitize the nervous system, stoking the flames of excitation. However, I don’t see how my theory would be predictive of this leading to enlarged heads, which are present only in a portion of autism cases. Also, my theory offers alternative explanations for early abnormal head growth that might account for the variation in findings that involve the mTOR / ERK pathways - because of excess neural excitation (see mGluR theory of fragile x) and/or genetic mutations in signaling proteins (as summarized by Kelleher and Bear) - which result in elevated levels of protein synthesis and abnormal growth of neurons.
Disrupted Task Negative Network (“TNN”) (Kennedy and Courchesne)
I suspect these researchers are correct about abnormalities in this functional system. However, I think they are describing a symptom, not a cause. I think I can offer a simpler and more direct explanation for these difficulties related to the default network in many individuals that results from the fact that autistic children are constantly in fight or flight mode, scanning the external world for threats due to the underlying processes I describe above – rarely is the brain in a restful waking state. Neurologically speaking, this involves, among other factors, 1) imbalances between sympathetic and parasympathetic nervous system activity because of perceived chronic stress, 2) problems with arousal state as regulated by hindbrain structures that process sensory information, and 3) abnormal processing of sensory information, typically problems with filtering out irrelevant sensory information. Ultimately, these problems go back to the relationship between neural excitation and stress response activation, and the positive feedback loops between these two factors.
Reduction of Mirror Neuron Function (Martineau)
According to my theory, reduction in mirror neuron function relates directly to the problems with sensory processing seen in many with autism, which ultimately result in eye contact avoidance, failure to recognize body language, over-reactivity to emotional tone in voices, and other problems. This again is a problem ultimately rooted in the imbalances in excitation and stress response function. To the autistic brain, threats that need to be responded to are everywhere, including the people in the autistic individual’s world. The easiest way to relieve the overstimulation resulting from constantly being exposed to threats is to ignore them, withdraw from them, and fail to notice them. Eye contact is neurologically stimulating – thus, it is avoided. Emotional tone of voice is often overstimulating. It can be avoided in part by various techniques autistics use, like vocalization or self-stimulation (controlling their sensory environment). Thus, reduction of mirror neuron function results from the autistic individual, through various mechanisms both conscious and sub-conscious, decreasing their perception of stimuli that would trigger mirror neuron activity.
Loss of Purkinje cells (Rout and Dhossche)
Purkinje cells are one of the strongest GABAergic cells in the brain, and the cerebellum is a strongly inhibitory structure - 4 of 5 of the cerebellum cell types are GABAergic. The authors suggest a mechanism of action that I don’t think is accurate – that cerebellar cell loss results in autism because of the cerebellum’s involvement in motor coordination, working memory, and learning. Instead, I think the reduction in cerebellar volume (and volume of inhibitory cells) results in a general loss in inhibitory control over brain pathways, in particular the cerebello-thalamo-cortical circuit.
Absence of Local Inhibitory Interneurons (Wilson)
This is another potential pathway for reduced inhibitory control of brain signaling to result in an overly excitatory brain.
Increased Gamma Band Oscillations (Orekhova)
According to one paper, gamma waves and fast rhythms likely play a role in neural communication, reflecting integration of information from the external world into the brain. Gamma waves are observed especially during alertness and after sensory stimulation. They are associated with selective attention and conscious perception of the external world. These waves are associated with epileptiform activity, fast rhythms that are seen in the majority of seizures (increased in autism). Abnormally elevated gamma wave activity in autism fits in very well with my theory. Gamma wave activity should be increased in autism because these waves are indicative of heightened arousal and abnormally active sensory processing – both features of autism.
Ultimately, I don’t think the ideas in my theory are really all that new or unique. The basic concepts are rooted in the General Adaptation Syndrome developed by Hans Selye in the 1930s. Selye discovered that that there are common human physiological responses (i.e. adrenal gland enlargement and duodenal ulceration) to a wide variety of external stressors, from injected toxins to sleep deprivation. Using Selye’s language, the body exhibits a series of specific reactions to non-specific stressors of sufficient severity and duration. I think autistic behaviors are ultimately examples of specific responses to the ‘perception’ of unending and severe non-specific stressors that are present from birth and before – ‘perception’ is the key to this sentence. The reality of the stressors in the world may be no different than that for most infants. However, one’s experience of these stressors is shaped by one’s sensitivity to them, which to some large degree is determined by the balance of excitation and inhibition in the brain. An overly reactive nervous system can interact with a normal human environment in a way that the brain perceives a continually and massively stressful existence. And, the body and mind respond to the perception of stress, not the reality of it.
Selye, in his work (or at least that I have read) did not account for specific psychological and cognitive reactions to non-specific external stressors, focusing instead of pure physiology (in part because he worked largely with rats). As a result, he did not delve deeply into that which makes certain individuals vulnerable to chronic stress, and what happens to these vulnerable individuals as they develop in the world. I would bet this is the pathway to understanding how autism works.
Print Bookmark and Share


Leave a Comment
Email: (not displayed)

Enter the code shown above:

Return to previous page