Tuesday, 23 August 2011

Screening for metabolic conditions in autism

I start with a question. It is not a question meant to offend anyone or provide a platform for extolling a particular view, merely a question that people probably come across quite regularly in relation to autism spectrum conditions.

What causes autism?

Your answer to this question will probably depend on lots of different factors: your relationship to autism, your interest in autism, your personal experiences of autism, your scientific/political/religious points of view and persuasions.

Some people might argue that this is not a question that needs answering and that autism, or at least a segment of the autism spectrum, is just neurodiversity. Other people might argue that autism, or again at least specific phenotypes of autism, is either genetically-controlled, a product of environment or lies somewhere in-between these concepts. Other people just sit on the fence.

I've thought long and hard about this question and the various answers given to it down the years. Those answers have really shaped the history of autism, from the dark times of the blame-game, to the hopefully more enlightened times we live in now. Whilst accepting that autism is in fact autisms, a heterogeneous group of conditions/perspectives/ways of thinking which span ability ranges from significant 'disability' to significant ability alongside the adage 'correlation does not imply causation', I am inclined to say that in some selected cases, science might be beginning to have some pretty good clues about how autism might come about. In much the same way that other behavioural, developmental conditions have been subject to such questioning, I reckon the evidence for causation in autism is probably on a par with things like ADHD.

Consider a few lines of inquiry (appreciating that this list may not be exhaustive).

First, the various research on Fragile X syndrome (which manifests autistic behaviours in some cases). Some might say Fragile X syndrome is not autism but rather autistic behaviours. Others might say, so what; autistic behaviours are autism - or at least autism spectrum. I note on the National Fragile X Foundation website, they use the word 'cause' to describe the relationship.

Second the growing evidence on things like the immune system, mitochondrial dysfunction, brain chemistry and architecture seen in some cases of autism: causative or coincidental?

Third are the various 'associations' between autism and environmental factors including that linking viral infections such as rubella and encephalitis. I did cover possible post-malarial autism in this post a while back. Whether there is any connection to other organisms (viral or bacterial) as there might be in some cases of schizophrenia remains unanswered.

Fourth, although taking a hit these past few weeks, the various genetic findings linked to cases of autism. You see them quite regularly in the scientific literature; this gene or that gene manifesting as autistic behaviours (we think).

Finally we have evidence from the various in-born errors of metabolism which forms the basis for this post.

I linked to a connection between PKU and autism when it came to listing inborn errors of metabolism simply because PKU is the archetypal metabolic condition and seems to show more than a passing relationship with autism. I suppose given the reports of dietary factors being also potentially linked to some cases of autism, it is perhaps all the more useful to list PKU as an example. When it comes to screening for inborn errors of metabolism, many parts of the world are actually getting quite good at this. Thanks to Robert Guthrie, who lent his name to the newborn heel prick test, several potentially life-changing conditions are screened for shortly after birth using the dried blood spot method. Importantly for some of these conditions, there are changes that can be made to off-set some of their effects on health and wellbeing.

My interest in metabolic conditions in relation to autism has been piqued recently following the publication of a couple of articles on the subject. This article by Schiff and colleagues reviewed a couple of hundred children with autism who passed through a French centre. They reviewed a variety of test results carried out including testing for disordered purine metabolism which has been discussed previously on this blog, and concluded that aside from 2 patients, there was no autism-specific relationship between indications of metabolic disorders and their cohort. At least one of their cohort who presented with abnormal results, one patient with elevated urinary creatine excretion, is interesting. Interesting because this is not the first time this has been reported either in the scientific literature or via a personal account.

A second article published in 2010 by Wang and colleagues suggested similar things with regards to there being no overall significant difference in the rates of creatine deficiency syndrome in autism vs. controls. That being said the authors did note elevated mean levels of urinary creatine in children with autism and their siblings as a result of several individuals in these groups showing elevated creatine levels. Importantly despite their 'negative' results, the authors do not rule out routine screening for such issues in autism.

There are various other 'associations' dotted around the autism research landscape. Glucose-6-phosphate dehydrogenase (G6PD) deficiency, L-2-hydroxyglutaric aciduria, and disorder of the urea cycle have all been mentioned. Autism and Smith-Lemli-Opitz syndrome is the topic of a blog post scheduled for the not-too-distant future on cholesterol and autism. Such conditions give us some interesting clues into the nature of autism (i.e. not necessarily 'caused' by only one factor) but also providing some interesting possibilities for management and intervention.

So back to the question of what causes autism, do we perhaps know more than we think?