Tuesday, 31 March 2015

Reassessing Cognitive Impairment in Autism – Improving the Prognosis

When Monty, now aged 11 with ASD, was diagnosed aged three and a half we were told that he had autism and “this may be indicative of the presence of an associated learning disability, but it is impossible at this stage to give a prognosis as to his future difficulties” and also “he is not yet able to take part in formal assessments of his cognitive ability. When his skills and ability to share interests with adults and to follow direction/instruction develop, it will be possible to formally assess his cognitive skills using standard measures.”

Off the record, we were also told that he might develop epilepsy.

We never measured his IQ and he has never had a seizure.

With hindsight, it is interesting what they said about it being pointless to try and measure his IQ.  Apparently it is not uncommon to do just that.

Improving Cognitive Function

This post is about cognitive improvement, so do not be put off by the introduction to MR/ID.  Several regular readers who are using some of the suggested drugs discussed in this blog are now also commenting on the resulting cognitive improvement, so it really is not just a case of N=1.

Nobody here is measuring the change in their child’s IQ, so these remain anecdotes.

To start on a happy note

Monty, aged 11 and diagnosed with classic autism, has been learning the piano for three years.  At the start he was not very cooperative with his teacher and after a few months the lessons stopped.  

27 months ago he started on bumetanide, the first part of his autism Polypill.  After years of ABA, slow but solid development appeared to have reached a plateau; but then he began to accelerate.  We restarted piano lessons again, with a new teacher.  Having added Atorvastatin, from the very next day he began to practice daily, playing without his teacher.  He has had two 40 minutes lessons most weeks since.  Two years later this is the result:-


 Click the image to play, turn up the volume (video may not work on Apples)

So there is no doubt that Monty got smarter.  When I heard him playing this piece, I thought it was the piano teacher, but she was recording Monty on her phone.

Big brother also did not believe little brother was playing this, until he saw the full video (with moving fingers).

Reading, writing, and numeracy have all improved and are now at a similar level to those of many of his NT classmates (who are 2 to 3 years younger than him).  Rather unexpectedly, he was recently the only one in class who understood how to multiply fractions; this was never taught at home.   

Prior to starting the Polypill drugs, I had spent three years, on and off, trying to teach Monty prepositions, without much progress.  This is almost always a difficult area for those with classic autism.  In the end he figured it all out by himself, with a little help from Bumetanide.

Recently yet another cognitive step forward seems to have have occurred, which appears to be the result of PAK 1 inhibiting propolis and/or the tangeretin flavonoid.  Monty's assistant in school was today proudly showing me his latest school test result, "73% and it was all his own work".  She thinks it is the tangeretin.

In earlier years school was for “socialization”, not learning.  This is fine as long as the learning takes place at home, otherwise inclusion means no education.

Cognitive Function, IQ, MR/ID

I prefer to talk about cognitive function, and its improvement or enhancement.  Drugs that achieve this are usually called Nootropic.

I think that many people remain skeptical about Nootropics.

Psychiatrists, Pediatricians and Psychologists prefer to think about IQ, MR/ID.

People affected do not like the old term of Mental Retardation (MR) and so quite recently, in English speaking countries, it was replaced by the term Intellectual Disability (ID).  The World Health Organization still use the old term, as does almost everybody else.

Somebody is diagnosed with MR or ID if their IQ is below 70.  In a typical group of 100 people, two people (2.2%) would be expected to fall into that category.  The average IQ (mean, median and mode) is 100.

In theory as you progress through childhood and into adulthood your IQ is expected to stay the same.  So the tests used adjust for your age.  There are special non-verbal tests.

If you acquire new skills at a lower rate than typical, your IQ would appear to fall over time.  This does not mean that you have lost skills just that you are acquiring new skills at a slower rate than your peers.  This explains why parents of kids with ASD, who do have their IQ tested, often find their score goes down as they get older.

Measuring IQ in Autism

I think it is generally a bad idea to measure IQ in people with autism.

There is anecdotal evidence to show that the results are often not valid, because the test is based on the assumption of compliance and that the child is actually doing his/her best.  Not surprisingly, the experts have found that children undergoing an ABA program improve their measured IQ by 10s.  After a few months of ABA the previously unfocused child has been trained to sit down, sit still, pay attention and work.  Of course they then get a higher score, but are they now more intelligent?

One reason put forward for not measuring IQ, is that while people will go a long way to help a child with autism to learn, once you add a diagnosis of MR/ID, some people will try much less hard.

Nonetheless people do measure IQ in autism and it is worth a quick look at what is known.

Some people are saying that 50% of people with autism have MR/ID.  I always found that odd, and what exactly do they mean by autism?

Using the previous US DSM definitions, Asperger’s was a part of the autistic spectrum but had the precondition that there was no MR/ID and no language delay.  Then you had the middle group with the odd name of PDD-NOS    (Pervasive Developmental Disorder Not Otherwise Specified).  This groups the people with more issues than Asperger’s, but without many of the problems experienced by those diagnosed with Autism.

So in the old US system there were 3 main categories, plus 2 minor ones:-

1.     Asperger’s
2.     PDD-NOS
3.     Autism
4.     Retts Syndrome
5.     Childhood Disintegrative Disorder (CDD)

Very few people have Retts or CDD.

The Autistic Spectrum was, in effect, also called PDD (Pervasive Developmental Disorder) just to confuse people a little more.  PDD = ASD = the above five conditions.

The latest version DSM5 went several steps backwards.  Everybody affected in the US is now just ASD, all five categories were merged.  

Hopefully nobody else in the world will pay any attention.

Unfortunately being Psychiatrists, they again have to muddy the water and all the future data/statistics.  A portion of people formerly diagnosed with PDD-NOS, will now get diagnosed with SCD (Social Communication Disorder) which is set outside the new definition of ASD.

“Congratulations you are off the Spectrum” 

I really do wonder about the IQ of these Psychiatrists.

Reliable Data on ASD

I do like to have some reliable data.  The quality of data in the field of autism is usually very poor and incompatible (i.e. rubbish).  Most data, like that from the CDC in the US, is unreliable.  It seems that richer “Ethnic European” parents push to get an autism diagnosis much harder than poorer “Hispanic” and “African American” parents.  Perhaps hard to believe as an outsider, but in the US poverty equals low diagnosis of autism and wealth equals high diagnosis.  Incidence does not equal diagnosis.  CDC data is just who got diagnosed; in the US many poorer people do not get diagnosed.  If you live in a country with free socialized healthcare, as in Europe, this will look strange.

I have chosen a highly regarded, and very highly cited, Canadian source for my data.

Éric Fombonne  is a French psychiatrist and epidemiologist based at McGill University in Montreal.  He co-authored a pair of studies in 2001 and 2005 with Suniti Chakrabarti, that examined the entire preschool and early school population of one large area of the United Kingdom (falling under the South Staffordshire Health Authority).

There is a stable population of indigenous British people with a small (1.4%), mostly Asian, immigrant population. The total population living in the area was 320 000 people.

The studies are:-

All children from 2.5 to 6.5 years old were screened, a total of 15,500 children in total 97 children (79.4% male)  were found to have a PDD (i.e. be somewhere on the autistic spectrum)

Of the 97 children, 29 (29.9%) had no functional use of language defined as the daily spontaneous use of 3-word phrases. The proportion of children without functional language was however strongly associated with diagnostic subtype (AD, 69.2%; Asperger syndrome, 0%; PDD-NOS, 16.1%).

Of the 97 children, 37 children underwent Merrill-Palmer testing and 56, Wechsler Preschool and Primary Scale of Intelligence testing. Four children could not be tested for practical reasons. Overall, 24 (25.8%) of 93 children had some degree of mental retardation. The 2 children with childhood disintegrative disorder and Retts syndrome scored in the moderate range of mental retardation.

Side-note about shoddy research

To show those of you still unconvinced that published, and moderately highly cited, autism research can be rubbish, the authoritative sounding paper written by a Professor of Psychology below also reviewed the above research data.

The author commented:-

Recent epidemiological surveys have shown that the prevalence rates of MR in children with autism is between 40% and 55% (e.g., Chakrabarti & Fombonne, 2001), much lower than the typical rates cited in the literature.”

The Chakrabarti & Fombonne 2001 paper clearly shows 69% of people with the narrow diagnosis of “autism” had MR whereas 25.8% of those with the broader diagnosis of ASD/PDD.

More than 100 other papers now cite the author’s incorrect readings of the original research.

I do not know what IQ you need to have to be a Professor of Psychology, but it clearly needs to be increased.

Back to the Fombonne studies

Four years later they repeated the same study on the next cohort of English school children:-

The rate of mental retardation in the autistic disorder group was 66.7%, compared to 12.0% in the group with pervasive developmental disorder not otherwise specified and 0.0% in the Asperger's disorder group

So I will take the average of the two studies

            Autism                       68% with MR/ID
            PD-NOS                     10% with MR/ID
            Asperger’s                    0% with MR/ID

What surprised me was the breakdown of the children by PDD.  Most kids (>50%) were PD-NOS, I did not expect that.

  Autism                        31%
            PD-NOS                     52%
            Asperger’s                  16%
            CDD/Retts                    1%

So the percentage all PDD (i.e. all ASD) with MR was 27%

Combining this as a graphic:-  


So now when people tell me that 50% of kids with autism have MR/ID, at least I know the likely reality.  It is either more, or less, depending on what you mean by “autism”; but is not 50%.


Most people think you cannot change your IQ.

The reality is that testing a young child with severer ASD is highly likely to underestimate their IQ, since the test assumes that the child will comply with the tester.  Most young children with autism do not comply with their parents, let alone an IQ tester.

ABA will improve compliance and hence improve an IQ test result.

Long term ABA use will, in many cases, gradually improve the child’s ability to learn and hence boost cognitive function and by implication an IQ test result.  You will find references to people saying ABA raised their kids IQ score by one or two dozen.

Correcting the biological dysfunctions underlying autism undermines the whole shaky DSM system.

PDD-NOS is, in effect, milder classic autism without the stereotypy .
Take some N-acetyl cysteine pills, you lower oxidative stress and you can stop the stereotypy.  So then your “expert” diagnosis would change from classic autism to PDD-NOS?

Take a few more pills and instead being in the 68%, with an IQ of less than 70, you can move up to 85 and, who knows, maybe much higher.

But it has to be said that the concept of IQ is something many people think they understand.

If one day, I were to make a clinical trial of my autism Polypill, I would definitely include a before and after measurement of IQ, alongside all the usual behavioral measures that most people would not understand.

And then …

“Wonder drug rescues people with autism from mental retardation”

In the meantime, we can continue correcting the remaining biological dysfunctions underlying autism and thus improving cognitive function.

Note the rocket, for those with classic autism doing just this and changing their prognosis.

Friday, 27 March 2015

Antibiotics and Autism(s) – Pass the Bacteroides Fragilis?

Today’s post allows me to cross off several topics from my to-blog list, since I can link them all together.

N = 1 or N > 1

If you are the parent of one of more children with autism, you will have a very specific view of autism, since your kind is the only kind that affects you.  This is natural and so for most readers it is case of N = 1.

When it comes to everyone else, and what they (should) teach medical students, it is the big picture that matters.  So large clusters of people behaving in a certain way is more significant that any outliers.  If you are the outlier, this is not much consolation.

In the world of autism, rather strangely, it is the very rare types that have an established medical therapy.  This ranges from the types caused by rare metabolic disorders to the more common PANDAS/PANS.

The large cluster that is classic autism remains untreated.

Polypill N = 3

To date I am aware of only a handful of people who have implemented the majority of my suggested Polypill for classic autism.  Three parents found major improvements and one found no impact; but the no impact case was not classic autism, it was very late regressive autism, later diagnosed as mitochondrial disease.

Many parents have implemented 1-2 elements of the Polypill with good results; these usually are elements that are the non-prescription drugs.

Three is not many, but it is more significant than one; and three out of four is a pretty good success rate.

As it stands, the Polypill will be a therapy for some children whose parents happen to be doctors, or own a pharmacy.

What does this have to do with antibiotics?

The other day I wrote a post about a recent 6 month clinical trial of Minocycline, an antibiotic.  The hope was that drug would reduce microglial inflammation and improve autism; but it did not.

Then I received a comment from Seth, a regular reader of this blog, to say that in his son tetracycline antibiotics really do improve autism.

I just read about John, another Dad, who found his child’s autism improved greatly while on antibiotics.  He has started his own charity N of One (N = 1) to raise funds for autism research and published an account of what he noticed.

There are many other accounts of certain antibiotics improving certain people’s autism.

In the case of PANDAS/PANS antibiotics are just the initial part of the therapy, but unless you live in the US you are unlikely to get diagnosed with PANDAS/PANS, let alone treated for it.

I will not be able to solve this puzzle today, but I will make my observations, for what they are worth.

First of all, Seth is talking about tetracycline-class antibiotics, one of which is Minocycline, the subject of that six month autism trial.  Now as we saw in a recent post, that trial was deemed a failure, but that was a trial of 10 children with regressive autism.  

Note that what people mean by "regressive autism" varies widely; most autism has some degree of regression.  In classic autism, the person is born different and then gradually becomes more evidently "autistic" during early childhood. Regressive autism, as defined by Chez, is when things are normal for at least the first 12 months.  Language can be normal or abnormal and then lost.  

I should also highlight that are other reports of Minocycline being beneficial in Schizophrenia and other neurological disorders.


Pharmacological interventions to treat psychiatric illness have previously focused on modifying dysfunctional neurotransmitter systems to improve symptoms. However, imperfect understanding of the aetiology of these heterogeneous syndromes has been associated with poor treatment outcomes for many individuals. Growing evidence suggests that oxidative stress, inflammation, changes in glutamatergic pathways and neurotrophins play important roles in many psychiatric illnesses including mood disorders, schizophrenia and addiction. These novel insights into pathophysiology allow new treatment targets to be explored. Minocycline is an antibiotic that can modulate glutamate-induced excitotoxicity, and has antioxidant, anti-inflammatory and neuroprotective effects. Given that these mechanisms overlap with the newly understood pathophysiological pathways, minocycline has potential as an adjunctive treatment in psychiatry. To date there have been promising clinical indications that minocycline may be a useful treatment in psychiatry, albeit from small trials most of which were not placebo controlled. Case reports of individuals with schizophrenia, psychotic symptoms and bipolar depression have shown serendipitous benefits of minocycline treatment on psychiatric symptoms. Minocycline has been trialed in open-label or small randomized controlled trials in psychiatry. Results vary, with findings supporting use in schizophrenia, but showing less benefit for nicotine dependence and obsessive-compulsive disorder. Given the limited data from rigorous clinical trials, further research is required. However, taken together, the current evidence suggests minocycline may be a promising novel therapy in psychiatry.

Minocycline is not just an antibiotic; it has several other known modes of action.

Minocycline is the most lipid-soluble of the tetracycline-class antibiotics, giving it the greatest penetration into the prostate and brain, but also the greatest amount of central nervous system (CNS)-related side effects, such as vertigo.

In various models of neurodegenerative disease, minocycline has demonstrated neurorestorative as well as neuroprotective properties
Minocycline is also known to indirectly inhibit inducible nitric oxide synthase (NOS).

As an anti-inflammatory, minocycline inhibits apoptosis (cell death) via attenuation of TNF-alpha, downregulating pro-inflammatory cytokine output.

Early research has found a tentative benefit from minocycline in schizophrenia


The antibiotic that John (from N=1) found to have magical properties was Amoxicillin, a very common type of penicillin.  Amoxicillin is a standard therapy for a strep throat.

Streptococcal infections are the initial trigger for PANDAS (Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections)

Amoxicillin seems to be have just one mode of action, that of an antibiotic.  This means it is a type of antimicrobial used specifically against bacteria, and usually used in medical treatment of bacterial infections. Antibiotics may either kill or inhibit the growth of bacteria.

          From the US National Institute of mental Health  (NIMH):-

Can penicillin be used to treat PANDAS or prevent future PANDAS symptom exacerbations?

Penicillin and other antibiotics kill streptococcus and other types of bacteria. The antibiotics treat the sore throat or pharyngitis caused by the strep by getting rid of the bacteria. However, in PANDAS, it appears that antibodies produced by the body in response to the strep infection are the cause of the problem, not the bacteria themselves. Therefore one could not expect antibiotics such as penicillin to treat the symptoms of PANDAS. Researchers at the NIMH have been investigating the use of antibiotics as a form of prophylaxis or prevention of future problems. At this time, however, there isn't enough evidence to recommend the long-term use of antibiotics.

However, a quick “google” will show more n=1 cases, of people claiming their child’s autism/PANDAS improving on Penicillin and then regressing again afterwards.


The other antibiotic that has been researched in autism is Vancomycin.  This drug is not absorbed from the intestine, so for systemic therapy it has to be taken by injection.  

When given orally it is used for things like treating bacterial infections of the intestines that cause colitis.  Orally administered vancomycin is recommended as a treatment for intestinal infection with Clostridium difficile, a common side effect of treatment with broad-spectrum antibiotics.

Vancomycin was first isolated in 1953 at Eli Lilly, from a soil sample collected from the interior jungles of Borneo by a missionary.

In most cases symptoms of autism begin in early infancy. However, a subset of children appears to develop normally until a clear deterioration is observed. Many parents of children with "regressive"-onset autism have noted antecedent antibiotic exposure followed by chronic diarrhea. We speculated that, in a subgroup of children, disruption of indigenous gut flora might promote colonization by one or more neurotoxin-producing bacteria, contributing, at least in part, to their autistic symptomatology. To help test this hypothesis, 11 children with regressive-onset autism were recruited for an intervention trial using a minimally absorbed oral antibiotic. Entry criteria included antecedent broad-spectrum antimicrobial exposure followed by chronic persistent diarrhea, deterioration of previously acquired skills, and then autistic features. Short-term improvement was noted using multiple pre- and post-therapy evaluations. These included coded, paired videotapes scored by a clinical psychologist blinded to treatment status; these noted improvement in 8 of 10 children studied. Unfortunately, these gains had largely waned at follow-up. Although the protocol used is not suggested as useful therapy, these results indicate that a possible gut flora-brain connection warrants further investigation, as it might lead to greater pathophysiologic insight and meaningful prevention or treatment in a subset of children with autism

What is going on?

The truth is that nobody knows for sure what is going on.  That also applies to PANDAS & PANS, which is why most of the world does not recognize them as genuine diagnosable conditions.

It would seem to me that various different processes are likely involved.  It would not be so hard to do some detective work, on a case by case basis.

For example, both Seth and John were using broad spectrum antibiotics.  If they gave Vancomycin a quick trial, they would find out if the problem was in the intestines, since that is the only place oral Vancomycin can have an effect.

John has written in his paper all about possible changes to the gut microbiome and how repeated antibiotic use early in life could set the stage for the development of autism in some children.  It is very easy to test this hypothesis, just try some Vancomycin.

We know that ulcerative colitis is comorbid with autism.  We know that this will lead to a permeable gut and the flow of unwanted substances to other parts of the body.  We see that Vancomycin is used for treating bacterial infections of the intestines that cause colitis.

So it is no surprise that in some people with autism, Vancomycin will improve behaviors.  You just need to identify which people.

Once apparent that Vancomycin is indeed effective, at least you know where the problem is.  Then it is a question of finding long term solutions to manage the problem.

We already know much about the so-called “leaky-gut” and the many GI problems in autism.  This is very well covered on the SFARI site and blog, so here are some highlights from there.

The new study is the first to show that maternal infection alters the microbiome in the offspring. The finding is significant for autism, as many children with the disorder are plagued by gastrointestinal problems, including diarrhea, vomiting and stomach discomfort. 

Leaky gut is also reported in children with autism and is associated with several other disorders, such as inflammatory bowel disease and Crohn’s disease, and perhaps with Alzheimer’s and Parkinson’s diseases, says Sarkis Mazmanian, professor of biology at the California Institute of Technology.To diagnose leaky gut in the mouse pups, the researchers fed them a carbohydrate molecule attached to a fluorescent molecule. The molecule later turned up in their blood, showing it had escaped through the gut wall. The mice also showed elevated gut levels of an immune molecule called interleukin-6 (IL-6) — a prime suspect in mediating the effects of maternal infection

The researchers then treated the mice with B. fragilis. This strain of bacteria isn’t commercially available, but exists naturally in about 20 percent of the human population. 
Mice treated with B. fragilis at 3 weeks of age don’t have a leaky gut five weeks later, their levels of blood 4EPS and gut IL-6 plummet, and the assortment of bacterial species in the gut reverts to something closer to that of control mice. And the mice do better behaviorally: They stop obsessively burying marbles in their cages, become as vocal as controls and are less anxious.

Sarkis K. Mazmanian, Ph.D.California Institute of Technology
Most research into autism spectrum disorders has focused on genetic, behavioral and neurological aspects of the illness, but people with autism also show striking alterations in immune status.

What’s more, a significant subset of children with autism spectrum disorders show chronic intestinal abnormalities, such as loose stool and altered bacterial microbiota (the collection of beneficial bacteria within the intestine). Antibacterial treatments are reported to provide behavioral improvements in some cases.

In addition, many children with autism have been diagnosed with food allergies and are on special diets. Societal advances (including 'Western' diets and antibacterial products) may have paradoxically compromised human health by reducing our exposure to health-promoting gut bacteria.

The connection between gut bacteria, intestinal disease and autism is a promising area of investigation. Sarkis Mazmanian and his team at the California Institute of Technology used mouse models that show autism-like features to evaluate the efficacy of probiotics.

They found that specific probiotic bacteria ameliorate autism-like behaviors in both environmental models of ‘induced’ disease (by mimicking viral infection of the mother during gestation), as well is in two genetic models of autism spectrum disorder.

These studies are an important step in furthering research that addresses the connection between the gut microbiome and altered behaviors, a link suggested by studies in humans. Finally, Mazmanian’s findings may help validate the use of probiotics as a safe and effective treatment for autism when it is accompanied by gastrointestinal abnormalities.

What it means?

It certainly appears that some people with ASD and GI problems have a something similar going on to my case of “N=1” (ASD + pollen allergy).  An allergic reaction has caused mast cells to degranualate releasing histamine and  IL-6.  That histamine causes further release of IL-6 elsewhere.  IL-6 is a pro-inflammatory cytokine and “public enemy number one” in the case of autism flare-ups.

It does appear that some people with autism + GI problems improve somewhat with supplemental digestive enzymes, like Creon/Kreon.  This does appear to be the basis of CM-AT, the long awaited therapy from Curemark.

However, based on feedback from this blog, it appears that blocking the calcium channel Cav1.2 with Verapamil may be even better.  It will certainly be much cheaper.

The standard treatment for this type of allergy related GI problem, is Cromolyn Sodium, a mast cell stabilizer.  Verapamil is also a mast cell stabilizer, among other properties.
Interestingly, some people “do grow out” of some allergies.  I myself, as a child, was prescribed Intal (Cromolyn Sodium) for GI problems of unknown origin.
You will find countless reports on the internet of children with “autism” who, on various diets, “recovered”.  You will hear plenty of people saying that young children will “grow out of” their autism.  It is generally accepted that most people’s autism does moderate as they become adults, just like many people’s asthma.
There is some sense in all of this.  Allergies can seriously aggravate autism.  So if you have someone with very mild autism, but a severe allergy, when you control the allergy you will see dramatic behavioral improvement.
Some readers of this blog have found that common allergy treatments like Zyrtec (cetirizine), have a profound behavioral improvement on their child, who was supposedly allergy-free.
In “my” subgroup of classic autism one underlying problem appears to be a channelopathy (Cav1.2); this might be genetic, or it might be an “epigenopathy”.  In either case, you could detect it, with existing technology, if you really wanted to.

The clever people at the NIMH think that PANDAS/PANS is a kind of Rheumatic Disease, where an autoimmune disorder (triggered by strep throat infections) causes the body to produce antibodies against the invading bacteria, and the antibodies help eliminate the bacteria from the body. However in a rheumatic disease, the antibodies mistakenly recognize and may attack the heart valves, joints, and certain parts of the brain.  When they attack the joints it is called Rheumatic Arthritis, when they attack the brain it is now called PANDAS.
The NIMH thinks that PANDAS/PANS is distinct from autism.
If you regularly read the research in this blog, you may disagree with the NIMH and see that PANDAS/PANS is just another autism variant.  Likely many things, other than strep infections, can also trigger this over-active immune system.
Many strange things occur in autism, one being that adults apparently cannot have PANDAS.  Of course they can; it just would have to be called ANDAS.
If an adult with autism wants to check for some rare for metabolic disorders leading to “autism” he/she may need to get referred to a children’s hospital, like Arkansas Children’s Hospital.  All the while, some of their diagnoses/treatments continue to be regarded as quackery by many other clinicians.
Some people with Schizophrenia, who improved on Minocycline, should try Vancomycin.  If the benefit is lost (as I suspect, it will be), then we would know that the effect was elsewhere than in the intestines.
Having established that Minocycline had no benefit in children with regressive autism, perhaps Johns Hopkins and NIMH should trial it in early-onset autism (classic autism).  It is Johns Hopkins after all, who believe that regressive autism is primarily mitochondrial disease.  The research indicates that mitochondrial disease is but one feature of classic autism.
Vancomycin is a useful diagnostic tool, rather than a long term therapy, but if Vancomycin improves behaviour, then you have plenty of choices:-
·        Cromolyn Sodium
·        Verapamil
·        Digestive enzymes like Creon/Kreon and, eventually, CM-AT
·        Probiotics & Prebiotics  (one day even Bacteroides Fragilis)
·        Exclusion diets

So if your child improves after taking antibiotics, or anything else, my suggestion is to investigate it yourself, rather than found yet another autism charity.
There is actually plenty of existing research and clever people, like those at the Simons Foundation, are funding further work on a prolific basis.
Other than readers of the SFARI blog and the Questioning Answers blog, is anyone actually reading (nearly) all this research? (let alone applying it)   Evidently not.
The academic researchers just read narrowly around their very focused area of interest.  The majority of clinicians read almost none of the research.

If you want to solve a complex problem, collect all the available data, look for connections and then think about it.
You should not have to do this for yourself, but with autism you do.

Monday, 23 March 2015

“Epigenopathies” in Autism and Epigenetic Therapy in Current Use - Part 1

Today’s post is about epigenetics, a complex area of science, that has been touched upon in previous posts.

Since none of us are experts in genetics we will focus on the application of epigenetics rather than going into the excruciating details.  Skip over any parts that get too technical. Some of the interesting studies, that are of more academic interest, I will put in a later post.

Epigenetics is just one way in which gene expression (whether genes are turned on or off) can be altered.  There are other ways, which may be equally important. It is evident that epigenetics plays a role in many conditions including autism, schizophrenia, inflammation, asthma, COPD and cancer.

Even based on today’s highly superficial review, there is an immediate, practical, therapeutic prospect, worthy of investigation.  Thanks to Professor Peter Barnes in London and again those irrepressible researchers in Tehran, who were actually trialing theophylline for entirely different reasons.

You do need some basic definitions to understand what is going on in epigenetics, but in essence epigenetic changes are just like bookmarks.


DNA is a molecule that encodes the genetic instructions used in the development and functioning of all known living organisms.

The problem with DNA is that there is a lot of it.  It has to be very tightly packed since it has to fit inside every cell in your body.  In order to tightly fold up all that DNA you need Chromatin.


Chromatin is a complex of macromolecules found in cells, consisting of DNA, protein and RNA. The primary functions of chromatin are to:-

1) Package DNA into a smaller volume to fit in the cell
2) Reinforce the DNA macromolecule to allow mitosis
3) Prevent DNA damage
4) Control gene expression and DNA replication.

The primary protein components of chromatin are histones that compact the DNA.

Epigenetics – book marks on your DNA

Rather like you might stick post-it notes on your cookery book, or science text book, your body has various mechanism to highlight specific genes.  In effect these bookmarks turn on, or turn off that gene.

So in the jargon:-

Epigenetic changes involve non genetic changes in chromatin structure resulting in changes in gene expression

The important thing to note is that we are not talking about genetic defects, mutations, CNVs etc. which are usually what you might think about.

We all have these epigenetic markers and they are subject to change. Some of these markers become fixed and can then be inherited.  So if your ancestor lived/worked in a highly polluted place, you might have inherited some of his/her DNA tags/bookmarks, this would affect how your genes are expressed today.

The problem occurs when these markers get stuck, or are in the wrong place.  Imagine having a bookmark to remind you how long to roast your chicken and instead it takes you to the page with the recipe for pancakes.

In some inflammatory diseases, like COPD, the “good” genes are turned off and the “bad” genes have got stuck turned on.

Epigenetic change is reversible

Whereas genetic defects are irreversible, epigenetic changes are potentially reversible.  You just need to figure out how to rub them out.

Epigenetic Mechanisms

Just as you might use a variety of objects to mark pages in a book, so nature employs multiple methods to tag your DNA.

1.     DNA methylation

In this process the tag is a methyl group (CH3); to silence the bad gene you add more tags (Stimulate methylation).  To reverse a good gene that has been silenced, remove the tags (use a DNA methyltransferase inhibitors e.g. azacytidine)

Applicable to lung cancer & inflammation
Problems of specificity and targeting

I could only find a current methylation epigenetic therapy for schizophrenia:-

Recently, Satta et al. reported that nicotine decreases DNMT1 expression in GABAergic mouse neurons leading to decreased methylation at the GAD67 promoter and increased GAD67 protein expression. This effect was found to occur as a result of nicotinic receptor agonism. These improve cognitive functioning in schizophrenia, and may suggest in part why 80% of schizophrenia patients use tobacco. The specific nicotinic receptors that mediate this improved cognition have yet to be established. However, an alpha7-nicotine receptor agonist has been shown in small studies to improve cognition in schizophrenia subjects.

2. Histone modification 

Histones are the chief protein components of chromatin, acting as spools around which DNA winds.

There are several types of histone modification, that act as tags on your DNA:-

·        Lysine methylation
·        Arginine methylation
·        Lysine acetylation
·        Serine/Threonine/Tyrosine phosphorylation

The most studied variant is acetylation, this involves the addition or removal of acetyl groups (O=C-CH3)


In medicine a group of drugs already exists, called Histone Deacetylase inhibitors (HDAC inhibitors, HDIs).  HDIs are a class of compounds that interfere with the function of histone deacetylase.

HDIs have a long history of use in psychiatry and neurology as mood stabilizers and anti-epileptics. More recently they are being investigated as possible treatments for cancers, parasitic and inflammatory diseases.

The prime example of this is valproic acid, marketed as a drug under the trade names Depakene, Depakote, and Divalproex. In more recent times, HDIs are being studied as a mitigator for neurodegenerative diseases such as Alzheimer's disease and Huntington's disease.  Enhancement of memory formation is increased in mice given the HDIs sodium butyrate or SAHA.  While that may have relevance to Alzheimer's disease, it was shown that some cognitive deficits were restored in actual transgenic mice that have a model of Alzheimer's disease (3xTg-AD) by orally administered nicotinamide, a competitive HDI of Class III sirtuins.

Autism and HDIs 

There is research in mouse models showing that HDIs can improve autism.

Readers of this blog who are using the Supersprouts broccoli powder may not realize that the Sulforaphane produced, is a potent HDI (Histone Deacetylase Inhibitor).

In the autism world, the HDI research is still generally on mice, where social cognition is seen to improve.

Follow up study by Foley:-

In utero exposure of rodents to valproic acid (VPA) has been proposed to induce an adult phenotype with behavioural characteristics reminiscent of those observed in autism spectrum disorder (ASD). Our previous studies have demonstrated the social cognition deficits observed in this model, a major core symptom of ASD, to be ameliorated following chronic administration of histone deacetylase (HDAC) inhibitors. Using this model, we now demonstrate pentyl-4-yn-VPA, an analogue of valproate and HDAC inhibitor, to significantly ameliorate deficits in social cognition as measured using the social approach avoidance paradigm as an indicator of social reciprocity and spatial learning to interrogate dorsal stream cognitive processing. The effects obtained with pentyl-4-yn-VPA were found to be similar to those obtained with SAHA, a pan-specific HDAC inhibitor. Histones isolated from the cerebellar cortex and immunoblotted with antibodies recognising lysine-specific modification revealed SAHA and pentyl-4-yn-VPA to enhance the acetylation status of H4K8. Additionally, the action of pentyl-4-yn-VPA, could be differentiated from that of SAHA by its ability to decrease H3K9 acetylation and enhance H3K14 acetylation. The histone modifications mediated by pentyl-4-yn-VPA are suggested to act cooperatively through differential acetylation of the promoter and transcription regions of active genes.


Histone modification is also implicated in inflammation.  We know that Autism is an inflammatory condition and of course we know which are the much better studied inflammatory conditions.

In terms of the brain, schizophrenia and even sometimes ADHD are better studied.

In the rest of the body arthritis, asthma and COPD are interesting.  Thanks to Peter Barnes at Imperial College, the COPD research is again leading the way.

COPD is like a severe drug resistant form of asthma.  Barnes has almost completely figured out the mechanism and how to best treat it.  One of the findings is to use a common drug called Theophylline in low doses as a HDAC activator.

The usual modes of action of Theophylline are:-

1.     competitive nonselective phosphodiesterase inhibitor, which raises intracellular cAMP, activates PKA, inhibits TNF-alpha  and inhibits leukotriene  synthesis, and reduces inflammation and innate immunity

2.     nonselective adenosine receptor antagonist

The usual dosage involves concentration of 10-20 micrograms/mL blood.  At this level there can be some side effects.

Barnes found that at sub-therapeutic doses (<8 micrograms/mL) , Theophylline actually has a different mode of action; it behaves as a HDAC activator; because the other modes of action were no longer present, no longer were their side effects.  He also showed that as the dose increases, the HDAC activation actually fades.  Another case of less being more.

Once deacetylated, DNA is repackaged so that the promoter regions of inflammatory genes are unavailable for binding of transcription factors such as NF-κB that act to turn on inflammatory activity. It has recently been shown that the oxidative stress associated with cigarette smoke can inhibit the activity of HDAC2, thereby blocking the anti-inflammatory effects of corticosteroids.)

Theophylline is a novel form of adjunct therapy in improving the clinical response to steroids in smoking asthmatics and people with COPD (some of whom do never smoked).

By using a low dose of Theophylline, steroid medication became much more effective allowing lower doses of steroids to be used.

Below is a presentation and one of Barnes’ many papers on this subject:-

Targeting the epigenome in the treatment of asthma and chronicobstructive pulmonary disease.


Epigenetic modification of gene expression by methylation of DNA and various post-translational modifications of histones may affect the expression of multiple inflammatory genes. Acetylation of histones by histone acetyltransferases activates inflammatory genes, whereas histone deacetylation results in inflammatory gene repression. Corticosteroids exert their anti-inflammatory effects partly by inducing acetylation of anti-inflammatory genes, but mainly by recruiting histone deacetylase-2 (HDAC2) to activated inflammatory genes. HDAC2 deacetylates acetylated glucocorticoid receptors so that they can suppress activated inflammatory genes in asthma. In chronic obstructive pulmonary disease (COPD), there is resistance to the anti-inflammatory actions of corticosteroids, which is explained by reduced activity and expression of HDAC2. This can be reversed by a plasmid vector, which restores HDAC2 levels, but may also be achieved by low concentrations of theophylline. Oxidative stress causes corticosteroid resistance by reducing HDAC2 activity and expression by activation of phosphoinositide-3-kinase-delta, resulting in HDAC2 phosphorylation via a cascade of kinases. Theophylline reverses corticosteroid resistance by directly inhibiting oxidant-activated PI3Kdelta and is mimicked by PI3Kdelta knockout or by selective inhibitors. Other treatments may also interact in this pathway, making it possible to reverse corticosteroid resistance in patients with COPD, as well as in smokers with asthma and some patients with severe asthma in whom similar mechanisms operate. Other histone modifications, including methylation, tyrosine nitration, and ubiquitination may also affect histone function and inflammatory gene expression, and better understanding of these epigenetic pathways could led to novel anti-inflammatory therapies, particularly in corticosteroid-resistant inflammation.

COPD and Autism

COPD is not autism, but there are some similarities.  Both conditions are associated with chronic oxidative stress and inflammation.

The antioxidant NAC is effective in both conditions.

The Nrf2 activator Sulforaphane (from broccoli) is being trialed for both conditions and is shown effective in much autism.

Inhaled steroids keep people with COPD alive, and oral steroids are beneficial to many people with autism.  Their use in autism is severely limited by side effects of long term oral steroid use.

Some HDI drugs improve autism and some HDI drugs improve COPD.

It would seem that the Epigenopathies of autism and COPD may well overlap.  Could the COPD epigenetic therapy be effective in some autism?


Theophylline for Neurological Disorders?

You might have realized that epigenetic therapy should be highly focused, since some genes need to be switched on while others need to be switch off.

Nonetheless that natural question to ask is what is the effect of Theophylline on neurological disorders like autism.

I cannot answer that question; but we can see the effect on ADHD (autism-lite).

It should be noted that the below trial was nothing related to epigenetics and the dosage was the more typical high dosage.  The histone modifying (epigenetic) effect would have been greater at a slightly lower dosage.

At these doses Theophylline would act as a mild stimulant;  note that Theophylline is very closely related to caffeine.  Somewhat counter-intuitively, psychiatrists treat hyperactive people with stimulants.

A total of 32 children with ADHD as defined by DSM IV were randomized
to theophylline and methylphenidate dosed on an age and weight-adjusted basis at 4 mg/kg/day (under 12 years) and 3 mg/kg/day theophylline
(over 12 years) (group 1) and 1 mg/kg/day methylphenidate
(group 2) for a 6-week double-blind and randomized clinical trial. The principal measure of the outcome was the Teacher and Parent ADHD Rating Scale. Patients were assessed by a child psychiatrist, at baseline and at 14, 28 and 42 days after start of the medication.

The results suggest that theophylline may be a useful for the treatment of ADHD. In addition, a tolerable side-effect profile is one of the advantages of theophylline in the treatment of ADHD.

In autism it would be nice if somebody made a trial with 2mg/kg

Let us digress a little and see just what is Theophylline:-

Theophylline is naturally found in cocoa beans. Amounts as high as 3.7 mg/g have been reported in Criollo cocoa beans.

Trace amounts of theophylline are also found in brewed tea, although brewed tea provides only about 1 mg/L, which is significantly less than a therapeutic dose.

As a member of the xanthine family, it bears structural and pharmacological similarity to theobromine and caffeine

Derivatives of xanthine (known collectively as xanthines) are a group of alkaloids commonly used for their effects as mild stimulants and as bronchodilators, notably in the treatment of asthma symptoms. In contrast to other, more potent stimulants like sympathomimetic amines, xanthines mainly act to oppose the actions of the sleepiness-inducing adenosine, and increase alertness in the central nervous system. They also stimulate the respiratory centre, and are used for treatment of infantile apnea. Due to widespread effects, the therapeutic range of xanthines is narrow, making them merely a second-line asthma treatment. The therapeutic level is 10-20 micrograms/mL blood; signs of toxicity include tremor, nausea, nervousness, and tachycardia/arrhythmia.

Theophylline degrades to caffeine.

Inhibitor or Activator of HDAC ?

You may be wondering why we would want an HDAC activator for autism, if we know that Sulforaphane (broccoli) does just the opposite; it is an inhibitor.  The reason is that we have made a few simplifications in the science; there are many types of HDAC, and you might need an inhibitor of one type of HDAC and an activator of another.  Worse still, you might need something on one part of your body and something quite different in another part.

The HDACs can be divided into 3 classes based on their structure and sequence homology: class I consists of HDACs 1, 2, 3, 8, and 11; class II includes HDACs 4, 5, 6, 7, 9, and 10; and class III enzymes are HDACs originally found in yeast and include Sir2-related proteins. Increased HDAC activity and expression are common in many cancers and can result in repression of transcription that results in a deregulation of differentiation, cell cycle, and apoptotic mechanisms. Moreover, tumor suppressor genes, such as p21 appear to be targets of HDACs and are “turned off” by deacetylation. Prostate cancer cells also exhibit aberrant acetylation patterns. The use of class I and class II HDAC inhibitors in cancer chemoprevention and therapy has gained substantial interest.


When the epigenetic bookmarks appear in the wrong place, trouble will follow.  Genes that should be “off” are turned on and vice versa.

These events have recently been a new name “Epigenopathies”

Just as we can look at many dysfunctions in autism as Channelopathies; those dysfunctions in ion channels and ion transporters, we will be able to consider others as Epigenopathies.

Who first came up with this terminology is not certain, but it might be a clever Frenchman called Mark Millan who works at the Unit for Research and Discovery in Neuroscience, Institut de Recherches Servier, beside the river Seine.

The good news is that here is a very clever neuroscientist with an interest in autism, but not obsessed by it.

France generally has quite an old fashioned view of autism, you will not find much in the way of ABA in France, and the State is certainly not going to be the one paying for it.

Millan nicely summarizes the implications:-

Neurodevelopmental Disorders (NDDs) are accompanied by aberrant "epigenetic" regulation of processes critical for normal and orderly development of the brain. Epigenetics refers to potentially-heritable (by mitosis and/or meiosis) mechanisms controlling gene expression without changes in DNA sequence. In certain NDDs, prototypical epigenetic processes of DNA methylation and covalent histone marking are impacted. Conversely, others involve anomalies in chromatin-modelling, mRNA splicing/editing, mRNA translation, ribosome biogenesis and/or the regulatory actions of small nucleolar RNAs and micro-RNAs. Since epigenetic mechanisms are modifiable, this raises the hope of novel therapy …

In the next post on epigenetics we will look at the research that is specific to  neurodevelopmental disorders.  It is interesting, but does not really have any obvious therapeutic implications.  One point I will highlight in this current post is the following:-

ASD is not associated with systemic differences in global DNA methylation

What this means is that, as far as one key type of epigenetics is concerned, autism is not characterized by too many or too few epigenetic tags; the problem is that they are not all in the right place.  Many alternative therapies in autism are rather simplistic.  It is not a case of too much methylation, or too little.

In the twin study the ASD Twin and his unaffected sibling has almost the same amount of total DNA methylation.