Pages

Wednesday, 17 October 2018

Autism as a Hierarchy of Impairments


A French Pyramid, worth visiting

Today’s post is not full of complex science.
I am reminded from time to time that I am supposed to be writing a book about translating autism science into practical therapy. To even partially do justice to all the science, things have to get a little complicated, at which point it will inevitably lose many readers.
What is much easier to achieve is to explain what autism is, and is not, and what, if anything, you might want to do about it.
I think you can consider autism as a hierarchy of impairments that together define a particular person’s “autism”.  For example, epilepsy is not just a comorbidity of someone’s autism, it is an integral part of it, and very much so biologically.
All of this is a simplification, but I think it does actually help represent what is currently diagnosed as autism.




Most people diagnosed today with autism are at the lower end of the pyramid/hierarchy, they have impaired social and communication skills to some degree and some of the issues in the level above, maybe some anxiety or ADD or ADHD.
People with severe autism rise through the levels to the summit, perhaps escaping from some elements.
When you then add prevalence to this hierarchy of impairments, you get the graphic below.
Really severe autism is thankfully rare. This was the old autism defined under the diagnostic regime of DSM3.  DSM is an abbreviation of the Diagnostic and Statistical Manual of Mental Disorders, published by the American Psychiatric Association.
In 1994 DSM version 4 introduced Asperger’s as an extension of autism.
We are currently on DSM5 which dropped the term Asperger’s opting for three levels of severity.  Severe autism is called level 3 and mild autism is level 1.  So, a genuine little professor type of Asperger’s would be level 1. Some people are getting diagnosed at intermediate points like 1.5.
Given the fact that the underlying biology is actually extremely complex, involving many hundreds of affected genes, it is perfectly possible to have a person with impaired social skills, who has a high IQ, no physical impairments, but self-injures.





  

Having identified where a person fits in this autism hierarchy, it is then time to see what are the likely consequences.
Having understood the consequences, you can then make plans to mitigate them.









In the case of the person with Asperger’s (DSM5 level 1) there may be very few issues that need to be addressed; but if you ignore the fact they may spend their school years being bullied and feeling excluded, they may fall victim to the 9 times elevated risk of suicide.
Ignoring what appear as minor quirky issues may have major consequences later.
At the summit of the pyramid the big dangers are seizures, self-injury and early death, but not from suicide.
Aggression and self-injury have to be brought under control during childhood, because in adulthood society does not tolerate it.  In most countries there is a lack of appropriate places to house adults with such behaviours and then bad things will inevitably happen.
Some people’s physical impairments fade away, some people never have any, but for some others such issues remain lifelong.
Cognitive dysfunction is part and parcel of DSM3 autism, what now is called Level 3 autism, under DSM5. As we have seen in this blog, some aspects of cognition can be improved using biology.

Personalized Medicine?
When deciding whether to treat a 2 or 3-year-old with autism using personalized medicine it is very important to understand the consequences. If the young child has severe autism (DSM3, or DSM5 level 3) then you know what the likely outcome will be if the child remains untreated. We know that 10-15% of these cases will dramatically improve without any intervention, but 85% will not. Intensive ABA interventions will accelerate skill acquisition in many cases, but it does not address the biological dysfunctions. The end result is a shortened lifespan (on average 40 years), much of it likely in an institution of one kind or another.  This you compare against the risk and cost of personalized medicine.
If you have a 3-year-old with mild autism (DSM5 level 1), the biological issues are quite mild and you will likely achieve great things with simple steps like teaching social skills and finding the right schools (small class sizes and no bullying). If you have very mild autism you may well find the positives outweigh the impairments associated with autism. Great attention to detail, perseverance, reliability and perhaps a high IQ may not make you cool at school, but are highly valued in the "right" workplace.

Not surprisingly, it is mild autism (DSM5 level 1) that gets most of media attention these days. At some point perhaps they will add DSM5 level 0.5 to include even mildly quirky people, but the next target for diagnosis appears to be adult females who could fit DSM5 level 1, but who slipped through the net.  Expect prevalence to continue to increase.







17 comments:

  1. with my limited knowledge, I once wanted to start a yearbook called Autism Now, which would publish a number of significant autism research every year - the number depending on quality of sources. The idea was to educate doctors around the world in a condensed and prechewed way. there were two reasons that I gave up on the idea: 1. wanting to spend as much time as I can investing in my own family and general disappointment with humanity and lack of interest in helping large swathes of people. 2. my husband pointed out that the current status is actually quite helpful. as long as the ‘system’ is not paying attention to an illness, there are so many venues to explore. when an illness gets all sorts of official procedures and codes and therapies in the system, the system is working not only to provide those but also to stop anyone doing anything else. like I said in one of my previous replies here, we humans have enough knowledge currently to significantly reduce the incidence and severity of autism in children. what we do not have is the necessary qualities as a species. no amount of disseminating information or books or effort of any kind will change that.

    ReplyDelete
    Replies
    1. I think this is a great idea even though doctors here in the United States already have the DSM, medical journals, etc. At least you can show them that as a parent you have an open mind and happen to be one of the more motivated parents that will put in the time to try and understand how to help your child and be unwilling to take the path of least resistance for your child. There is a saying that doctors work to earn, not to learn so there would need to be a condensed version a doctor could be presented with, perhaps maybe an abridged version of Peters book would do the job.

      Delete
    2. I think you need to have low expectations about mainstream doctors treating autism. You usually go to the doctor expecting firm answers and you cannot get them with idiopathic autism. This is why only highly unusual mainstream doctors try and treat autism, leaving the field open to a very mixed bunch of others.

      In the case of genetic autism syndromes the situation is very much better. There really should be clinics for kids with Rett or Fragile X, because there now is some evidence based medicine.

      Delete
  2. Is anyone here using Zotero and can recommend it (or not)?
    It is a free tool for researchers who want to keep track of their sources/quotes in a kind of digital library. I was thinking it might be useful to collect all those useful articles you've read and want to save and perhaps also tag with your own labels.

    My bookmarks folders in my webbrowser are pretty overloaded by now and not supereasy to browse.

    /Ling

    ReplyDelete
  3. I kinda thought it was time to take a peek at language issues again. It's really so central, and even if the answers aren't out there yet it would be nice to have a hunch of what pathways to look at.
    As with ASD, there are probably many contributing things that affect speech; neurological hearing, fine motor skills of the mouth, verbal memory (remembering what to say), social motivation and understanding of grammar. Dysfunctional speech is usual in many disorders, and is interestingly not even necessarily linked to low IQ.

    So, a first attempt at the subject could be the genetic perspective. An important hub in the genetic network seem to be the FOXP2 gene. Wikipedia describes the famous KE family with an inherited mutation in the FOXP2 gene, causing oral apraxia, impaired grammatic skills, dysfunction in Brocas area and some minor learning deficits.
    https://en.wikipedia.org/wiki/KE_family

    This more technical paper (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5826363/) shows that as far known, there are very few promoters of the FOXP2 gene. Either it upregulates itself, or is promoted by FOXP1 (autism gene), FOXP4 (less important) or TBR1+CASK together (both autism/ID genes). How to affect these are a bit beyond the scope of this comment, but at least for TBR1 there might be opportunities.

    In 2014, a first paper (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4176457/) showed that “FOXP2 promotes neuronal differentiation by interacting with the retinoic acid signaling pathway”. There are two notes to the article worth reading, as is this related paper:
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4660430/

    FOXP2 in humans and animals is very much linked to speech or song capabilities, and this in turn is affected by retinoic acid. Unfortunately retinoic acid doesn’t look like a good idea…

    So what is the next step? Maybe we should look at the downstream effects.

    /Ling

    ReplyDelete
    Replies
    1. There could also be a connection between language and the serotonin system. Peter has mentioned SAM-e as anecdotally being beneficial for speech. Also Bacopa (antagonist to 5HT2a and 5ht6, agonist to 5ht3a) and Fisetin (MAO-A inhibitor, agonist to 5HT1A) are mentioned as beneficial for speech and now Tatjana mentions a 5HT3 (subunits unknown?) antagonist, Ondansetron, with positive impact on language.

      The quote "Retinoic acid signaling is affected by photoperiod and could play an important role in circannual timing." (https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/retinoic-acid) caught my eye and makes me think of SAD - seasonal affective disorder, which very likely links to serotonin up- or downstreams.

      /Ling

      Delete
    2. There are many things downstreams of retinoic acid, and it is hard to know which ones are important for speech and which ones are not.

      Here is one potential clue, but there are surely others:

      The reelin protein/RELN gene, which is down in autism and many other neurodisorders, happens to be upregulated by retinoic acid. It is also upregulated by bacopa, fluoxetine (antidepressant), olanzapine, nicotine and even valproic acid. The gene itself is downstreams of TBR1+CASK (see two comments above), but is it downstreams of FOXP2 too?

      Reelin regulates NMDA functionality (among many other important things), and especially the NR2A subunit. NR2A gene mutations are especially indicated in language dysfunction and epilepsy from early years, but not as much in ID as other NMDA subunits).

      /Ling

      Delete
    3. NR2A subunit is encoded by the GRIN2A gene:
      "GRIN2A-related speech disorders and epilepsy are characterized by speech disorders in all affected individuals and a range of epilepsy syndromes present in about 90%. Severe speech disorders observed can include dysarthria and speech dyspraxia, and both receptive and expressive language delay/regression; more mildly affected individuals may display subtly impaired intelligibility of conversational speech.
      [Characteristics:]
      Acquired aphasia
      Auditory agnosia (impaired recognition of sounds)
      Dysarthria
      Speech dyspraxia"
      https://www.ncbi.nlm.nih.gov/books/NBK385627/

      So we have one axis of
      Retinoic acid or TBR1+CASK->FOXP2
      ->
      reelin
      ->
      GRIN2A

      What else?

      Surely there has to be something related to language comprehension and E/I balance. Those loop diuretics seem to be involved in both ototoxicity (https://www.sciencedirect.com/science/article/pii/S1672293016300629) AND hearing improvements (https://www.sciencedirect.com/science/article/pii/S1672293008500067).

      /Ling, totally overposting today

      Delete
    4. FOXP2 shows up as being dysregulated as part of the pathogenesis of several neurodegenerative disorders such as frontotemporal dementia. That research might be a start in looking for more autism specific solutions.

      Delete
    5. You are right Tyler, there are a lot of hits when you google FOXP2 and frontotemporal dementia. Interestingly, I found this article that also connects this type of dementia with dyslexia genes:

      "A high frequency of neurodevelopmental learning disability, including dyslexia, has been reported in FTD patients and their first-degree relatives [..] . Additionally, dyslexic individuals showed structural and functional changes of the left temporal regions, those regions selectively damaged in FTD patients [..]
      Here we explored whether variations within three related-dyslexia genes, namely KIAA0319, DCDC2, and CNTNAP, might affect cortical thickness measures in FTD patients."
      (DCDC2 was not correlated, but the other two were)
      https://www.nature.com/articles/srep30848

      CNTNAP2 is definitely linked to FOXP2:
      "Recently, several studies have pinpointed the involvement of common variants of the Contactin-Associated Protein-Like 2 (CNTNAP2) gene, whose transcription is regulated by the product of FOXP2"
      https://www.ncbi.nlm.nih.gov/pubmed/23277129

      The other dyslexia gene, KIAA0319, was upregulated 2.85 times compared to wild type mice in a TBR1 -/- model.
      "Of 124 TBR1 target genes, 23 were reported to be associated with ASDs. In addition, one gene, Kiaa0319, is a known causative gene for dyslexia, a disorder frequently associated with autism."
      https://www.ncbi.nlm.nih.gov/pubmed/25600067

      Stuttering has its own genes, and those seem not directly connected to FOXP2:
      "This study provides an improved estimate of the contribution of mutations in GNPTAB, GNPTG and NAGPA to persistent stuttering, and suggests that variants in FOXP2 and CNTNAP2 are not involved in the genesis of familial persistent stuttering. This, together with the different brain expression patterns of GNPTAB, GNPTG, and NAGPA compared to that of FOXP2 and CNTNAP2, suggests that the genetic neuropathological origins of stuttering differ from those of verbal dyspraxia and SLI [specific language impairment}."
      https://www.ncbi.nlm.nih.gov/pubmed/24807205

      /Ling

      Delete
    6. Today I stumbled over yet another substance that I haven't heard about before: Diosgenin (DSG)
      "Diosgenin, a yam-derived compound, was found to facilitate the repair of axonal atrophy and synaptic degeneration and improve memory dysfunction in a transgenic mouse model of Alzheimer’s disease (AD). It was also found to enhance neuronal excitation and memory function even in normal mice.
      []
      For this placebo-controlled, randomized, double-blind, crossover study, 28 healthy volunteers (age: 20–81 years) were recruited
      []
      Among the 12 individual standard cognitive subtests, diosgenin-rich yam extract use significantly improved the semantic fluency. No adverse effects were reported."
      https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5691776/

      Interesting, I thought, and dug a little more. Diosgenin seem to work on estrogen receptors, but which ones? Probably both alpha and beta, but I don't know.

      "Diosgenin promotes oligodendrocyte progenitor cell differentiation through estrogen receptor-mediated ERK1/2 activation to accelerate remyelination."
      https://www.ncbi.nlm.nih.gov/pubmed/22461009

      "DSG significantly increased nuclear expression of ERβ. [..] In response to DSG stimulation, ERβ bound with RXRα and dissociated RXRα from PPARγ, leading to the reduction of transcriptional activity of PPARγ."
      https://www.ncbi.nlm.nih.gov/pubmed/26408789

      While this other paper was on metabolic dysfunction and not on brain functionality it mentions something that might be a clue. That RXRα thing in the above quote happens to be retinoid X receptor alpha. Is this something that connects estrogen, retinoic acid pathways, language and PPARgamma?

      Unlike Peter, i don't have any nice conclusion at the the end of all these posts.

      /Ling

      Delete
  4. We haven’t done genetic tests. But in our case first, ABA/VB with picture cards helped to speak single words. Galantamine 4mgx2 per day along with ABA helped to speak in 3 to 4 word sentences.

    ReplyDelete
  5. Hi Ling!

    Hope all is well.

    First of all, you’re definitely not overposting :) You always have great info to share, and you can never overpost great info!

    It’s funny you’re looking into FOXP2 Ling because I had recently looked at trying to find the connections between FOXP2 and my daughter’s mutated gene given her speech issues. I don’t know if FOXP2 dysregulation is the issue in my case, but given the relevance of FOXP2 to speech, I tried to find the connection, but wasn’t able to make the connection (yet).

    Great work on the FOXP2 analysis above by the way.

    If this helps, I did use two really great tools you’re probably already aware of, but these tools come with a well-earned warning – using these tools may lead to lack of sleep as you will find yourself all of a sudden realizing it’s 2am on a worknight and you still have rabbit holes you need to go down – trust me, been there done that (a lot :) )

    The first tool is Harmonizome – I’ve provided the links for SATB2 and FOXP2 but you can input any gene:

    http://amp.pharm.mssm.edu/Harmonizome/gene/SATB2

    http://amp.pharm.mssm.edu/Harmonizome/gene/FOXP2

    Simply scroll down the page and check out the many functional associations by clicking on the “+” signs (and say goodbye to sleep!)

    The second one is STRING:

    https://string-db.org/cgi/input.pl?sessionId=nf0azLSTIBw6&input_page_show_search=on

    Type in your gene of interest, enter Homo Sapien, and you’ll find a graphic representation of the gene of interest with connections to close and relevant genes (you can click on the connections or the genes).

    I hope you find this useful, and don’t blame me if you spend a lot of time on these sites :)

    Let me know what you think if you haven't used them before.

    Have a great night Ling!

    AJ

    ReplyDelete
    Replies
    1. Thank you AJ - and don't you tell me about deprived sleep issues! *lol*

      The result from STRING was actually the first thing I got when it came to SATB2 research, it is what I have been building on ever since. It was a great starting point, but it took a while to find the relevant pathway though.

      I haven't seen Harmonizome before, so I guess I'll have something to do for a couple (ahem..) of evenings.
      ;-D

      Good luck with the science digging AJ!

      /Ling

      Delete
  6. Here is some new research showing the mother's microbiome and the expression of the inflammatory protein interleukin-17a contribute to autism symptoms which is nothing new as poor maternal diet and health as well as interleukin-17a have both in repeated studies been shown to increase the risk of autism more than just about any other type of environmental exposure other than perhaps valproate exposure during pregnancy:

    Press Release:

    https://www.sciencedaily.com/releases/2018/07/180718113343.htm

    Paper:

    http://www.jimmunol.org/content/201/3/845

    What is new is that the researchers show that improving the mother's microbiome through a variety of different methods can significantly reduce autism risk. This may not be useful information to help children already diagnosed with autism, but for mothers who already have a child with autism and who are afraid of having additional children with autism, interventions which improve microbiome health may give some real practical benefits in decreasing autism risks and giving the parents, and especially the mother, some peach of mind.

    Specifically, dietary changes, probiotics, and even fecal transplants are options to help improve the inflammatory state of the microbiome in all mothers, but especially those with increased risk of having a child with autism. The researchers also suggest directly inhibiting IL-17A pharmacologically is an option as well, but are hesitant to employ this as an intervention due to excessive attenuation of the inflammatory response making the mother and baby more susceptible to various types of infections.

    ReplyDelete
  7. IL's are expressed in a multitude of tissue types.

    ReplyDelete

Post a comment