Question about the delimitation of 'autism'

[Azul]

I'm aware that most people here consider the "autism is in a continuum in humanity" false. But could someone explain for me what is the proof that a brain either is or isn't autistic?

Because a few points confuse me:

First, as a condition with genetic bases whose most common gene shows up in ~1% of the population*, how can we define such group with so much genetic diversity with one specific condition? That is, there isn't a necessary nor sufficient set of genes for defining it. It seems too vague.

* For what I remember. The exact number might be off.

Second, if there are things like classifications of subthreshold autism or parents of autistics with some more noticeable traits but not enough*, then why can't be autistic traits be distributed in the population?

* There's a classification different from PDD that I can't remember for the life of me.

That is, what is an well delineated, objective, definition (neurological, genetic, or otherwise) of autism that can prove an either-or situation, even if such definition is impractical or impossible to obtain with the current technologies?

I realize that the current behaviour based diagnosis isn't satisfactory. This is about a strict definition of what is, biologically, autism, and not so much how we can identify it today. So the DSMV isn't really relevant.

 

[Crossbreed]

This is about a strict definition of what is, biologically, autism, and not so much how we can identify it today.

I am not sure that it has been nailed down to a single biological cause. Some here who are better versed in medicine have proposed that there are multiple causes. It appears to me that the current model of autism is largely based on "black box" analysis.

 

[Au Naturel]

"Autism is in a continuum in humanity" is almost certainly true, ranging from those with barely perceptible symptoms to those who would not survive without support.

Every nonbinary trait in the human condition is expressed as a Bell curve. If you think of height, for example, it's like trying to define what "short" is. Any definition one might devise is an artificial line drawn across a smooth distribution curve. Even though "short" is an entirely subjective concept, "shorter than" is still an objective measure.

There are vanishingly few binary traits; that is a single gene locus with only 2 variations controls everything. If autism were one of them it would be easy to identify. (There are a few rare cases where that has happened.) Depending on what source you look at, autism is between 40 and 90% genetic. That leaves a lot of room for epigenetic and environmental influences to create a Bell curve even in a binary condition.

However, the CDC believes autism is almost entirely genetic. Given that it runs in families and two autistic parents are more likely to produce children who are more autistic than either parent, it is difficult to discount that belief.

One of the reasons for the wide range of estimates is that autism is not easily defined other than its more extreme forms. Many people would be diagnosed positively by one doctor and negatively by another. Some conditions mimic autistic symptoms. (Alternately, you could just as easily say that autism mimics their symptoms.) Consider that:

• autism spectrum disorder (ASD)
• Asperger's syndrome.
• childhood disintegrative disorder.
• pervasive developmental disorder-not otherwise specified.

all prevent very similar symptoms, and there is argument as to whether they are one condition or several related conditions.

Over 100 genes have been found to increase the risk of autism. Some increase the probability more than others. Very very few (4%) autism cases (syndromic autism) are shown to be the result of a specific genetic condition. Fragile X and tubular sclerosis are the most prominent ones.

 

[Tom]

Crossbreed and Gerald are very up on the genetics and perhaps they'll check in. I have only a hazy understanding even though I do read most of the studies published on autism causes/genetics/etc.

That said, there are over 100 gene variations associated with autism. No one has all, just some, and the more you have the more likely you will be symptomatic. Even people considered NT may have a few. Which variations you have will contribute to the way it presents itself in the individual. But with that many variations the ammount of combinations possible is huge (therefore the wide variety between us). But though the variations are different, they tend to occur in specific areas of the genome (2 I believe) that affect our development. So although we are different in detail we also are similar in an overall sense.

Once you have them you can pass them down to children, but how people get them in the first place is I think very debated and not yet understood well. It's been hard to find the cause, and there seem to be more then one.

So really I see it as a wide variety of specific situations but similar enough to be grouped together. For now. It may come to pass some distinctions become more apparent in time as science learns more about it.

 

[Au Naturel]

I'm aware that most people here consider the "autism is in a continuum in humanity" false. But could someone explain for me what is the proof that a brain either is or isn't autistic?

Because a few points confuse me:

First, as a condition with genetic bases whose most common gene shows up in ~1% of the population*, how can we define such group with so much genetic diversity with one specific condition? That is, there isn't a necessary nor sufficient set of genes for defining it. It seems too vague.

* For what I remember. The exact number might be off.

Second, if there are things like classifications of subthreshold autism or parents of autistics with some more noticeable traits but not enough*, then why can't be autistic traits be distributed in the population?

* There's a classification different from PDD that I can't remember for the life of me.

That is, what is an well delineated, objective, definition (neurological, genetic, or otherwise) of autism that can prove an either-or situation, even if such definition is impractical or impossible to obtain with the current technologies?

I realize that the current behaviour based diagnosis isn't satisfactory. This is about a strict definition of what is, biologically, autism, and not so much how we can identify it today. So the DSMV isn't really relevant.

"Either-or" doesn't objectively exist.

Consider that autism might be a case of balanced polymorphism. In early man, people with a few "autistic" genes might have a superior chance of survival and reproduction. Too many autistic genes and you don't. The survival of autistic genes is guaranteed even though it means a few offspring don't survive. Imagine a very complex version of what happens with sickle cell anemia.

Maybe some people have 10 "autistic genes" and others have 100. Nobody has zero. Do we draw our diagnostic line at 50 genes or 51? Surely the environment affects the expression of at least some of those genes. And there are also epigenetic influences to consider.

If you look at that one percent as fading imperceptibly into the rest of the population and not a separate creature, you see that the one percent line has no physical basis. It's just where we decided the symptoms are significant enough to diagnose, an arbitrary and subjective line.

No, modern psychiatry doesn't have a clue about most disorders. In some cases, they are no better off than ancients who grouped stars into constellations to make them easy to study. The stars in the area of the Centaur really aren't related, but when I mention Alpha Centauri, an astronomer knows where to look and which star it is.

 

[Neonatal RRT]

I'm aware that most people here consider the "autism is in a continuum in humanity" false. But could someone explain for me what is the proof that a brain either is or isn't autistic?

Because a few points confuse me:

First, as a condition with genetic bases whose most common gene shows up in ~1% of the population*, how can we define such group with so much genetic diversity with one specific condition? That is, there isn't a necessary nor sufficient set of genes for defining it. It seems too vague.

* For what I remember. The exact number might be off.

Second, if there are things like classifications of subthreshold autism or parents of autistics with some more noticeable traits but not enough*, then why can't be autistic traits be distributed in the population?

* There's a classification different from PDD that I can't remember for the life of me.

That is, what is an well delineated, objective, definition (neurological, genetic, or otherwise) of autism that can prove an either-or situation, even if such definition is impractical or impossible to obtain with the current technologies?

I realize that the current behaviour based diagnosis isn't satisfactory. This is about a strict definition of what is, biologically, autism, and not so much how we can identify it today. So the DSMV isn't really relevant.

1. What is the proof that a brain is/isn't autistic? There are several scientific lectures you can find on YouTube on the topic. There are several research articles on Google Scholar and PubMed that you can access that have examined autistic brains upon autopsy, CTs, MRIs, fMRI, PET scans, and other imagining techniques. Conductivity (voltages) and connectivity (wire thickness) can vary significantly in autistic brains as compared to normal controls. The migrational patterns of the neurons and layering are different than normal controls. The number of synaptic connections/dendrites may have too few or too many in specific areas as compared to controls. The cerebellum of autistics is often associated with too few Purkinge fibers. The concentration and distribution of glial cells and astrocytes are also different in autistics as compared to normal controls. Some variants of autism are missing some of the genes responsible for the normal neuronal "pruning" that occurs between 1-3 years of age, resulting in enlarged brains, but also in the geriatric brain where the brain does not loose mass as compared to controls. We know that there are even differences between female and male autistics on certain imaging studies. There are at least 3 different morphological autistic "brain types" identified. There are differences between an "Asperger's type" brain and other types of autistic brains. You can spend months going through all the literature, so I would limit it to the past 5-10 years on your searches.

2. Genetics. There have been about 200 familial genes, some 1000 other "de novo" genes, and another 4000 or so other genes associated with autism conditions. It is not a simple thing to do genetic testing and interpret it, as autism is one of those conditions that is a result of "genetic loading" and epigenetics. One can do prenatal genetic tests for these genes, but what it doesn't tell us is which of these many genes have been activated,...so just because a fetus has these genes, it doesn't mean that the resultant baby will be born with autism,...he/she may simply be a carrier. Genetic loading of activated genes will determine if and to what extent the condition will present. Furthermore, we know that females require more genetic loading than males in order to present with autistic traits. It also explains, in part, some of the pre-threshold/subclinical traits found in some individuals,...also why every autistic will present a bit different from one another,...I may have sensory issues, but they may not be the same as others, for example.

I think we are reaching a cross-roads in medicine where the diagnosis of autism will be a combination of psychological, neurological, and genetics,...as there is strong evidence for all three to play a role. To add the neurological and genetic components will solidify the idea that autism is a prenatal, genetic, developmental medical condition.

Broad transcriptomic dysregulation occurs across the cerebral cortex in ASD - Nature

RNA sequencing reveals widespread transcriptomic changes across the cerebral cortex in autism spectrum disorder, including primary sensory regions, in addition to association regions, as well as an attenuation of regional identity.

www.nature.com