Thursday, 18 December 2014

Activated Microglia and Inflammation in Autism

There have been yet more autism studies recently, highlighting neuroinflammation and the role of cells called microglia.  The result is this rather long post; but there is film to watch, if it gets heavy going.

Glia derives from a Greek word for glue. The original thought was that the glial cells “glued” the neurons together.

It turned out that glial cells do very much more and might be better thought of as “resident immune cells”.  They have other functions including synaptic pruning, which appears to have gone awry in autism.  They also form myelin, and when this goes wrong, big problems follow.

Microglia are inside the blood brain barrier and one of their jobs is to swallow up any foreign bodies that should not be there, before they can do damage.  It appears that this process is mainly modulated via potassium channels.  The majority of research focuses on the calcium-activated K+ channels, particularly KCNN4/KCa2 and 3.1, and ATP-sensitive K+ channels (KATP).  Administration of diazoxide, a classic KATP channel activator, is shown to reduce microglial activation and is neuroprotective in a variety of models involving neuroinflammation. 

However, Kv 1.3 and Kv 1.5 are also involved in activated glia.  We have seen in earlier posts, that blocking Kv 1.3 can be effective in autism (remember those TSO worms).

For the scientists among you:-

Synaptic pruning

A very small Acer Palmatum

Synaptic pruning could itself be the subject of an entire blog.  I will just use the analogy of a different kind of pruning.

With ornamental trees, to obtain the perfect form, pruning is very important.  You have to clear away the dead wood and encourage growth in particular areas to achieve the optimal shape.  You need to know when to cut, where to cut and how much to cut.

The human brain develops with far too many synapses and they too need pruning.  The weak ones need to give way for the strong ones to prosper.  Too many synapses lead to poor brain function.  This process is going on from childhood to early adulthood.  Microglia are heavily involved in this pruning process, as you will see in the video shortly.

We know that synaptic pruning is implicated in autism and very likely in its big brother, schizophrenia.

Activation of Microglia

Microglia can be in either a resting or activated state. In the activated state, for no good reason, they can do damage.  They can also react with mast cells to produce more inflammation.

(here is a link for the mast cell followers of Theoharides; they know who they are)

The subject is very complex.  For those with an hour to spare there is an excellent presentation by Beth Stevens from Harvard.  Click on the link below to go to the SFARI website and the video.

By a bizarre coincidence, there is another B Stevens researching glial cells and autism.  This time it is Bruce Stevens, in Florida.

His paper is interesting because he is using a known anti-oxidant (alpha lipoic acid, ALA) to affect brain glial cells.

One of the odd things is that we know in autism there is both oxidative stress and neuro-inflammation; they are a self-perpetuation combination.  There are numerous effective anti-oxidants; almost too many.  There is, however, a paucity of effective, safe, anti-inflammatory drugs.  In fact the best anti-inflammatory drug is probably an anti-oxidant.  So called Reactive Oxygen Species (ROS) are among the biggest causes of neuroinflammation.  With anti-oxidants you can neutralize the ROS, and thereby you take a big bite out of the neuroinflammation.

Double-stranded RNAs (dsRNA) serve as viral ligands that trigger innate immunity in astrocytes and microglial, as mediated through Toll-like receptor 3 (TLR3) and dsRNA-dependent protein kinase (PKR). Beneficial transient TLR3 and PKR anti-viral signaling can become deleterious when events devolve into inflammation and cytotoxicity. Viral products in the brain cause glial cell dysfunction, and are a putative etiologic factor in neuropsychiatric disorders, notably schizophrenia, bipolar disorder, Parkinson's, and autism spectrum. Alpha-lipoic acid (LA) has been proposed as a possible therapeutic neuroprotectant. The objective of this study was to test our hypothesis that LA can control untoward antiviral mechanisms associated with neural dysfunction. Utilizing rat brain glial cultures (91% astrocytes:9% microglia) treated with PKR- and TLR3-ligand/viral mimetic dsRNA, polyinosinic-polycytidylic acid (polyI:C), we report in vitro glial antiviral signaling and LA reduction of the effects of this signaling. LA blunted the dsRNA-stimulated expression of IFNα/β-inducible genes Mx1, PKR, and TLR3. And in polyI:C treated cells, LA promoted gene expression of rate-limiting steps that benefit healthy neural redox status in glutamateric systems. To this end, LA decreased dsRNA-induced inflammatory signaling by downregulating IL-1β, IL-6, TNFα, iNOS, and CAT2 transcripts. In the presence of polyI:C, LA prevented cultured glial cytotoxicity which was correlated with increased expression of factors known to cooperatively control glutamate/cysteine/glutathione redox cycling, namely glutamate uptake transporter GLAST/EAAT1, γ-glutamyl cysteine ligase catalytic and regulatory subunits, and IL-10. Glutamate exporting transporter subunits 4F2hc and xCT were downregulated by LA in dsRNA-stimulated glia. l-Glutamate net uptake was inhibited by dsRNA, and this was relieved by LA. Glutathione synthetase mRNA levels were unchanged by dsRNA or LA. This study demonstrates the protective effects of LA in astroglial/microglial cultures, and suggests the potential for LA efficacy in virus-induced CNS pathologies, with the caveat that antiviral benefits are concomitantly blunted. It is concluded that LA averts key aspects of TLR3- and PKR-provoked glial dysfunction, and provides rationale for exploring LA in whole animal and human clinical studies to blunt or avert neuropsychiatric disorders

The obvious question is whether other antioxidants have the same effect.  Most likely nobody knows.  I did ask both B Stevens #1 and B Stevens #2 for their thoughts on this – so far no answer.

Brain inflammation a hallmark of autism, according to large-scale analysis

Finally to the subject of this post, the recent Johns Hopkins study that shows inflammation in the autistic brain.

This is the press release from Johns Hopkins so it is quite readable.

While many different combinations of genetic traits can cause autism, brains affected by autism share a pattern of ramped-up immune responses, an analysis of data from autopsied human brains reveals. The study, a collaborative effort between Johns Hopkins and the University of Alabama at Birmingham, included data from 72 autism and control brains. It was published online today in the journal Nature Communications.

There are many different ways of getting autism, but we found that they all have the same downstream effect,” says
Dan Arking, Ph.D., an associate professor in the McKusick-Nathans Institute for Genetic Medicine at the Johns Hopkins University School of Medicine. “What we don’t know is whether this immune response is making things better in the short term and worse in the long term.”

The causes of autism, also known as autistic spectrum disorder, remain largely unknown and are a frequent research topic for geneticists and neuroscientists. But Arking had noticed that for autism, studies of whether and how much genes were being used — known as gene expression — had thus far involved too little data to draw many useful conclusions. That’s because unlike a genetic test, which can be done using nearly any cells in the body, gene expression testing has to be performed on the specific tissue of interest — in this case, brains that could only be obtained through autopsies.

To combat this problem, Arking and his colleagues analyzed gene expression in samples from two different tissue banks, comparing gene expression in people with autism to that in controls without the condition. All told, they analyzed data from 104 brain samples from 72 individuals — the largest data set so far for a study of gene expression in autism.

Previous studies had identified autism-associated abnormalities in cells that support neurons in the brain and spinal cord. In this study, Arking says, the research team was able to narrow in on a specific type of support cell known as a microglial cell, which polices the brain for pathogens and other threats. In the autism brains, the microglia appeared to be perpetually activated, with their genes for inflammation responses turned on. “This type of inflammation is not well understood, but it highlights the lack of current understanding about how innate immunity controls neural circuits,” says Andrew West, Ph.D., an associate professor of neurology at the University of Alabama at Birmingham who was involved in the study.

Arking notes that, given the known genetic contributors to autism, inflammation is unlikely to be its root cause. Rather, he says, “This is a downstream consequence of upstream gene mutation.” The next step, he says, would be to find out whether treating the inflammation could ameliorate symptoms of autism.

The full study is here:-

What I liked about the study was the comment made by Arking, a specialist in genetics, that it did not seem to matter what the genetic cause was, all the brain samples exhibited the same inflammation.  So it does not matter which of millions of possible combinations of genetic dysfunction is present, one key physiological result is shared neuroinflammation.

Take home message:  Treat the neuroinflammation in people with Autism.

The question of course is how.

Since it seems easy to treat oxidative stress, a leading cause of neuroinflammation, we should go to extreme lengths to finish that job. 

I started it with NAC and recently added Sulforaphane/broccoli.  I suspect there are more “low hanging fruit” to be gathered here. Perhaps just an additional supplemental (exogenous) antioxidants, or perhaps something clever like increasing the amount DJ-1, which is needed to support Nrf2 which turns on the anti-oxidant genes. Early 2015 will see my oxidative stress therapy optimized.

Treating Neuroinflammation in Autism

There are lots of possible ways to treat neuroinflammation, some of which we have already covered in this blog.  Sometimes it gets called immunomodulatory therapy.

There are some natural options like quercetin and turmeric.  Turmeric is also possibly chemo-protective:-

“Currently there is no research evidence to show that turmeric or curcumin can prevent or treat cancer but early trials have shown some promising results.”

Cancer Research UK

Interestingly, people who eat a lot of curry (Indians) have a very low incidence of cancer.

1.     Steroids, like Prednisone

These are already used, particularly in regressive autism.  They are potent, but have side effects.

2.     Blockers of Potassium channel Kv1.3

This is a clever approach, since it appears that this potassium channel is involved in mediating the inflammatory response. By blocking these channels the response we have seen that the immune response can be moderated and in some people, there autism moderated.

3.     Activators of Potassium channel KATP

We learned earlier in this post about diazoxide

4.     Other Microglial Ion Channels

The various other potassium, calcium and sodium channels need to be considered.

5.     Ibuprofen

This common painkiller reduces inflammation and is used to reduce inflammation associated with autism secondary to mitochondrial disease.

Do not use acetaminophen/paracetamol/Tylenol.  These will increase oxidative stress, since it depletes GSH and also affect mitochondria.

6.     Leukotriene receptor inhibitors (i.e. montelukast, zafirlukast)

These are interesting because they are used to treat asthma and so are very widely used. They are not steroids and so do not have their side effects.  They are proved to have anti-inflammatory effects.

Montelukast/Zafirlukast is used to reduce inflammation associated with autism secondary to mitochondrial disease.

7.     Pregnenolone

I wrote a post a while back on Pregnenolone, which is interesting, since you do not need a prescription.  But does it work?

Well, after I wrote the post below, the results from a clinical trial in adults with autism was finally published.

The objective of this study was to assess the tolerability and efficacy of pregnenolone in reducing irritability in adults with autism spectrum disorder (ASD). This was a pilot, open-label, 12-week trial that included twelve subjects with a mean age of 22.5 ± 5.8 years. Two participants dropped out of the study due to reasons unrelated to adverse effects. Pregnenolone yielded a statistically significant improvement in the primary measure, Aberrant Behavior Checklist (ABC)-Irritability [from 17.4 ± 7.4 at baseline to 11.2 ± 7.0 at 12 weeks (p = 0.028)]. Secondary measures were not statistically significant with the exception of ABC-lethargy (p = 0.046) and total Short Sensory Profile score (p = 0.009). No significant vital sign changes occurred during this study. Pregnenolone was not associated with any severe side effects. Single episodes of tiredness, diarrhea and depressive affect that could be related to pregnenolone were reported. Overall, pregnenolone was modestly effective and well-tolerated in individuals with ASD.

Trial doses were:-

Days 1-14: 100 mg
Week 1 and 2: 200 mg
Week 3 and 4: 350 mg
Week 5 and 6: 400 mg
Week 7 -12: 500 mg

So it was modestly effective, but the doses were huge.  It is a hormone and our endocrinologist did not much approve of the idea.

I will give this idea a miss.

8.     Statins

The current treatment for neuroinflammation in my Polypill is Atorvastatin.

I have already written a great deal about why statins may be effective in some people with autism; just make sure you do not have low cholesterol or mitochondrial disease.

Arthritis is another disease mediated by inflammation:-

To me it is no surprise that statins have therapeutic value in rheumatoid arthritis.

9.     NF-κB inhibitors

Because NF-κB controls many genes involved in inflammation, it is not surprising that NF-κB is found to be chronically active in many inflammatory diseases, such as inflammatory bowel disease, arthritis, sepsis, gastritis, asthma, atherosclerosis and others.

So perhaps NF-κB is for inflammation ,what Nrf2 is for oxidative stress, a force multiplier?

There are very many other inflammatory diseases like rheumatoid arthritis and so it is quite a well-trod path looking for inhibitors of NF-κB.

Before we get into that, a quick check on what we already know from research to schizophrenia (adult-onset autism).

Many reports suggest that schizophrenia is associated with the inflammatory response mediated by cytokines, and nuclear factor-kappa B (NF-kappaB) regulates the expression of cytokines. However, it remains unclear whether the interaction between NF-kappaB and cytokines is implicated in schizophrenia and whether the effect of neuroleptics treatment for 4 weeks is associated with the alteration of cytokines.
Sixty-five healthy subjects and 83 first-episode schizophrenic patients who met DSM-IV criteria and who were never treated with neuroleptics previously were included. Serum levels of cytokines such as interleukin-1beta (IL-1beta) and tumor necrosis factor-alpha (TNF-alpha) were examined by using sandwich enzyme immunoassay (EIA). Peripheral blood mononuclear cell (PBMC) mRNA expressions of cytokines (IL-1beta, TNF-alpha) and NF-kappaB were detected by using semiquantitative reverse transcription polymerase chain reaction (RT-PCR). NF-kappaB activation was examined by using transcription factor assay kits.
Schizophrenic patients showed significantly higher serum levels and PBMC mRNA expressions of IL-1beta and TNF-alpha compared with healthy subjects. However, treatment with the neuroleptic risperidone for 4 weeks significantly decreased serum levels and PBMC mRNA expressions of IL-1beta in schizophrenic patients. NF-kappaB activation and PBMC mRNA expression in patients were significantly higher than those in healthy subjects. Furthermore, PBMC mRNA expressions of IL-1beta and TNF-alpha were positively correlated to NF-kappaB activation in both schizophrenic patients and healthy control subjects.
Schizophrenic patients showed activation of the cytokine system and immune disturbance. NF-kappaB activation may play a pivotal role in schizophrenia through interaction with cytokines.

It seems fair to conclude that NF-κB inhibitors are well worth investigating.

Interestingly, one of my new “pet” compounds, alpha lipoic acid appears to have another role here:-

Evidence that α-lipoic acid inhibitsNF-κB activation independent of its antioxidant function.



α-Lipoic acid (LA) exerts beneficial effects in cardiovascular diseases though its antioxidant and/or anti-inflammatory functions. It is postulated that the anti-inflammatory function of LA results from its antioxidant function. In this study we tested whether inhibition of NF-κB by LA is dependent on its antioxidant function.


Human umbilical vein endothelial cells (HUVECs) were treated with tumor necrosis factor-α (TNFα) in the presence of various antioxidants, including LA, tiron, apocynin, and tempol. The activation of the nuclear factor-κB (NF-κB) signaling pathway was then analyzed.


LA, but not other tested antioxidants, inhibited TNFα-induced inhibitor-kappaB-α (IκBα) degradation and VCAM-1 and COX2 expression in HUVECs. Although LA activated the phosphatidylinositol-3-kinase (PI3-kinase)/Akt pathway in HUVECs, inhibition of Akt by LY294002 did not affect inhibition of TNFα-induced IκBα degradation by LA. In transient co-transfection assays of a constitutively active mutant of IκB kinase-2 (IKK2), IKK2(EE), and a NF-κB luciferase reporter construct, LA dose-dependently inhibited IKK2(EE)-induced NF-κB activation in addition to inhibiting IKK activity in in vitro assays. Consistent with the effect on luciferase expression, LA inhibited IKK2(EE)-induced cyclo-oxygenase-2 (COX2) expression, suggesting that IKK2 inhibition by LA may be a relevant mechanism that explains its anti-inflammatory effects.


LA inhibits NF-κB activation through antioxidant-independent and probably IKK-dependent mechanisms.


This really makes ALA look very interesting.  It is cheap, widely available and well tolerated.

10.       Low Dose Naltrextone                       

Your local doctor will probably tell you that Low Dose Naltrexone (LDN) is a load of quack nonsense, partly because it is claimed to help so many unrelated disorders.

I would not have questioned that opinion, before I had started by investigation into the biology of the brain and seen how many apparently unrelated conditions are actually interrelated.  This can be established by science, not quackery.

First to note is that tiny doses of some substances do indeed sometimes have effects quite different to large doses.

We saw earlier how a tiny stimulation of the body’s nicotinic receptors produces a different effect to a large dose.

My own experience showed that a tiny, but specific, dose of Clonazepam has a marked effect, whereas conventional medical wisdom would say such a small dose would do absolutely nothing.  In this case, I was just following the clever idea of Professor Catterall, from the University of Washington.

I also found that tiny doses of a TRH analog had a positive effect and quite different to the “regular” dose.

The advocates of LDN suggest it for conditions including Crohn's disease, fibromyalgia and multiple sclerosis (MS).  As I mentioned earlier in this blog, some Fibromyalgia appears to be a condition that was almost autism; perhaps the final hit, in a multiple-hit process failed to occur.  Crohn’s is an immune disease and is a type of inflammatory bowel disease (IBD).  MS is an inflammatory disease in which the insulating covers of nerve cells in the brain and spinal cord are damaged.

Preliminary research suggests LDN may have an effect on inflammation. Naltrexone has an antagonistic effect on Toll-like receptor 4 (TLR4), which are found on microglia, which can modulate the body's response to inflammation. It has been hypothesized that LDN may have anti-inflammatory effects through this pathway.



The immediate conclusion is that there are plenty of ways, already existing, that might very well help reduce neuroinflammation in autism.  They just requires a little further thought and investigation.

The broader conclusion here is about the merit of genetic testing.

Undoubtedly, if you could analyze the entire genome in a person with autism and also measure the expression of those suspect genes in the brain, you would gain a great deal of information.  In a few cases, where there is a single gene causing the “autism”, you might well be able to figure out a therapy.

You cannot take brain biopsies from living people.  We did come across that clever Ricardo Dolmetsch, growing brain samples from skin cells.  He has now moved over to the private sector.

So for the moment genetic testing will just generate a vast amount of data, that in many cases will not be of any immediate clinical relevance.

The good news, as pointed out by Dan Arking, from Johns Hopkins, is that many of these numerous, unrelated, genetic dysfunctions end up with the same biological manifestations.

There may be thousands, or even millions of combinations, of genetic dysfunctions that lead to autism with neuro-inflammation.

You can go ahead and treat the neuro-inflammation, without any knowledge of exactly which gene has which SNP (single nucleotide polymorphisms)  or who had what CNV (copy number variant).

For me, the identification of so-called autism genes like PTEN and BCL2 is interesting, as are the single gene causes of autism.  We can then see that a reduced expression of that gene might contribute to autism, caused by multiple gene dysfunction (multiple-hits).  For the great majority of people with ASD, they have had multiple-hits.

I read Ricardo Dolmetsch’s Stanford research into Timothy syndrome, which is caused just by one gene, albeit a very important one.  I considered that perhaps a partial dysfunction might occur, leading to disturbance in the protein expressed by this gene.  I had no idea whether in my son this dysfunction existed, whether it might be caused by a SNP (there are several known ones) or if a dysfunction was caused as a consequence of a metabolic disruption caused by autism, such as oxidative stress or neuroinflammation,  affecting the function of an undamaged gene.

It did not matter; I just carried on and did a little practical test.  This led me to include Verapamil in my Polypill.  No genetic testing was required.

It was suggested to me that genetic testing might help point me in the right direction.  I think it would likely point me in all directions.  We all carry many genetic errors, and most of us thrive regardless, so most genetic errors are irrelevant.

The clever future diagnostic tool is proteomics.


From now, I will consider autism in terms of a manageable group of clusters.  Once you know, based on symptoms and some measurable biomarkers, which cluster you are in, you would have a good chance of predicting which drugs would be effective.

The underlying genetic causes may, or may not, overlap with other people in that cluster.

Some clusters may overlap. Note the case of siblings with autism, when one is early onset and the other is regressive.  Was the regressive one really symptom free early one? Or, was it just a second hit nudged him “over the edge” and then people noticed?

This would be a practical approach that could be used.  I think when people talk of phenotypes and autisms, they are thinking about very precise biological causes and then it just becomes too complicated to expect your local doctor to ever figure out.

90+% of people quite probably fit into a handful of clusters.  Then you just need a diagnostic flowchart leading to the relevant cluster and then a specific drug toolkit.

My Polypill is the drug toolkit for one cluster; and it is not a rare one.

Wednesday, 10 December 2014

Biotin/Biotinidase Deficiency in Autism and perhaps Autistic Partial Biotin Deficiency (APBD)?

Crete, as seen from the International Space Station
By ISS Expedition 28 crew (NASA Earth Observatory) [Public domain], via Wikimedia Commons

In this blog there is a tab at the top called “Disorders leading to Autism”.  This includes a long list of, supposedly rare, known conditions that lead to the development of autism.

In that list is Biotin deficiency and I even put the name of the gene that is thought to be dysfunctional.  The BTD gene encodes an enzyme called Biotinidase, that in turn allows the body to use and recycle biotin.

Biotin deficiency is a known cause of autism, but it seems that the assumption is made that the cause is Biotinidase deficiency.  The usual test done is for Biotinidase deficiency.

In good hospitals they routinely test for many of these dysfunctions when a child is originally diagnosed with autism.  When I say good hospitals, I mean big US hospitals attached to a university.  In other countries such testing rarely takes place, nor is it even mentioned.

We will see later that even these good hospitals may be getting the result wrong.  They are likely testing for the wrong defect, and so getting a "false negative" in some cases.

The take home message is that Biotin Deficiency may not be rare in autism, only Biotinidase Deficiency is rare.  Both are treatable.

How rare is Biotin Deficiency?

Biotin deficiency is supposed to be extremely rare.

One of this blog’s readers made reference to a recent Greek study.  They checked 187 children in Crete, diagnosed with autism, for various metabolic dysfunctions.

Evidence for treatable inborn errors of metabolism in a cohort of 187 Greek patients with autism spectrum disorder (ASD)

As the reader pointed out, the results are very odd.

The researchers identified 13 children whose results suggested something strange was going on with biotin.  When they did the further tests for biotin deficiency, which is usually caused by deficiency in  biotinidase, they could find nothing unusual.

Nonetheless, they implemented the standard therapy for biotin/biotinidase deficiency.  This involved large doses of oral biotin, which is very cheap and seemingly harmless.

The researchers found that 7 of the 13 made clear advances.  This indicates that they suffered from a biotin deficiency, but not a biotinidase deficiency. Biotinidase is used by the body to recycle its biotin.

Biochemical abnormalities suggestive of IEM

For 12/187 (7%) of patients, urinary 3-hydroxyisovaleric acid (3-OH-IVA) was elevated and sera methylcitrate and lactate levels were also elevated in two of these patients. Despite these biochemical abnormalities, defects in biotinidase, or holocarboxylase synthetase could not be demonstrated in either sera or fibroblasts. Of interest, none of these 12 patients was undergoing valproate intervention, the latter a potential source of 3-OH-IVA elevation in urine. Despite an absence of confirmatory enzyme deficiencies in these 12 patients, we nonetheless opted to treat empirically with biotin for 3 weeks, 2 × 10 mg and then for 6 months at 2 × 5 mg, which led to a clear therapeutic benefit in 7/13 consisting of improvement in the Childhood Autism Rating Scale (CARS; Table Table2).2). For those benefiting from biotin intervention, the most impressive outcome centered on a 42 month-old boy whose severe ASD was completely ameliorated following biotin intervention. This patient was subsequently followed for 5 years, and cessation of biotin intervention (or placebo replacement) resulted in the rapid return of ASD-like symptomatology. This patient currently attends public school without any clinical sequelae and remains on biotin at 20 mg/d.

In the following table are the results showing the effect on the CARS rating scale, before and after treatment with biotin.

Patient #1

Just look at what happened to the first patient in the above table.

For those benefiting from biotin intervention, the most impressive outcome centered on a 42 month-old boy whose severe ASD was completely ameliorated following biotin intervention. This patient was subsequently followed for 5 years, and cessation of biotin intervention (or placebo replacement) resulted in the rapid return of ASD-like symptomatology. This patient currently attends public school without any clinical sequelae and remains on biotin at 20 mg/d.

He went from severe autism to no autism.  (and back, when he stops the biotin)

Yet, if he was tested for the standard biotin(idase) disorder, even at the best center for autism in the world, nothing would show up

Biotin Deficiency

Genetic disorders such as Biotinidase deficiency, Multiple carboxylase deficiency, and Holocarboxylase synthetase deficiency can also lead to inborn or late-onset forms of biotin deficiency. In all cases – dietary, genetic, or otherwise – supplementation with biotin is the primary method of treatment.


Of 187 children, 13 were identified for biotin treatment and 7 responded .  None of these children would have been noticed by the normal diagnostic procedures of even the best laboratory, which look for biotinidase deficiency.

Also of interest is the effect of partial biotin deficiency.

·        profound biotinidase deficiency (<10% of mean normal serum activity)
·        partial biotinidase deficiency (10%–30% of mean normal serum activity).

Children with partial biotinidase deficiency and who are not treated with biotin do not usually exhibit symptoms unless they are stressed (i.e., prolonged infection)

Partial biotinidase deficiency isusually due to the D444H mutation in the biotinidase gene

Profound biotin deficiency would hopefully be noticed

Mild symptoms linked to biotin deficiency:-

  •        Loss of hair colour
  •         Loss of hair
  •         Fine and brittle hair


The results of clinical studies have provided evidence that marginal biotin deficiency is more common than was previously thought. A previous study of 10 subjects showed that the urinary excretion of biotin and 3-hydroxyisovaleric acid (3HIA) are early and sensitive indicators of marginal biotin deficiency.

It does seem that biotin deficiency is usually caused by things that lead to biotinidase deficiency, so let’s look at the data on frequency (Epidemiology)

Biotin Deficiency – Epidemiology
Based on the results of worldwide screening of biotinidase deficiency in 1991, the incidence of the disorder is: 5 in 137,401 for profound biotinidase deficiency

·         One in 109,921 for partial biotinidase deficiency
·         One in 61,067 for the combined incidence of profound and partial biotinidase deficiency
·         Carrier frequency in the general population is approximately one in 120.

Both parents need to carry the genetic defect, for a child to inherit it.

So something odd is going on (in Greece).

In 61,067 people we would expect 600 people with autism.

It seems that in 600 Greek children with autism there may be 22 with a biotin dysfunction.  This is vastly higher than we would expect.

Not everyone with biotin dysfunction has autism and even if they did, in Greece there would be 22x greater incidence than elsewhere.


I think we (and the Greeks) have likely discovered some new phenomenon “autistic partial biotin deficiency”, APBD, which is not caused by the usual lack of biotinidase.  Somehow the dietary biotin is insufficient in these people, even though biotinidase is present.

APBD does not seem to cause all the severe symptoms of biotin deficiency, just the neurological ones and so remains undiagnosed.

Perhaps one of the other odd metabolic disorders in autism is affecting the biotin metabolism?  Remember that Harvard study suggesting the oxidative stress in the autistic brain reduces the activity of a key enzyme D2, that is needed to convert the thyroid pro-hormone T4 into the active hormone T3.  This would mean that despite a “normal” set of thyroid lab results from your doctor, you might well be hypothyroid inside the brain (low on T3).

Those with access to a good laboratory might consider sending a urine sample to measure 3-hydroxyisovaleric acid (3HIA).

Those without these options might have to settle with the option of trying 10-20 mg of Biotin for a short period and see if it has any effect.

Biotin appears to be one of those vitamins, like B12, where even huge doses may have no ill effect; they are just excreted.  The supplement companies are selling 10 mg pills of biotin;  the RDA for a 10 year old is 0.03 mg which is 333 times less.

Based on the Greek study, you would expect about 4% of autistic people to show a clear benefit, without first doing the 3HIA urine test.

A small chance of success per child, but a chance nonetheless.

Note on the study
I have referred to this Greek study once before. On that occasion I was talking about the ketogenic diet and modified Atkins diet.

It is widely accepted that the ketogenic diet can greatly reduce epileptic seizures, so it is not really surprising that it can also help some people with autism (but which ones?).

In the Greek study, via laboratory tests, they identified 9 % children who might benefit from this diet.  Just over a third of these identified children did indeed improve on the diet.

16/30 patients manifested increased sera beta hydroxybutyrate (b-OH-b) production and 18/30 had a paradoxical increase of sera lactate. Six patients with elevated b-OH-b in sera showed improved autistic features following implementation of a ketogenic diet (KD).

This remarkable study was published one year ago.

It has been cited just one time in subsequent literature (although twice now in this blog); this really tells us a lot. (nobody is interested)

Changing diet can require a great deal of effort and, if a fussy eater is involved, it can be even more difficult.  If biomarkers exist to narrow down who would benefit from a modified diet, this is really very significant.

You can easily try biotin pills for a couple of weeks, trying a ketogenic diet just on the "off chance", requires much more bother.