I'd like to think that the review article by Annik Simons and colleagues  (open-access) highlights some pretty strong evidence to suggest there being at least some connection between some autism and some of the collected inborn errors of metabolism. Indeed, when people generally talk about 'not knowing what causes autism' if we perhaps consider a more plural view of 'the autisms', there is a case to be made to say we might know what causes 'some' autism and some of it might lie in this area...
Inborn errors of metabolism (IEM) cover a whole host of different conditions "in which there is an accumulation of toxic and/or complex compounds or energy problems within the cells due to enzymatic defects or other protein dysfunction." The absolutely magnificent work of people like Robert Guthrie who's name is synonymous with the neonatal heel prick test offered to newborns to screen for various IEMs and a jobbing physician called Ivar Asbjørn Følling who lent his name to a condition that was eventually called phenylketonuria (PKU), have proved to be some of the real successes of modern medicine.
For quite a few years, peer-reviewed science has suggested some potentially important 'associations' between various behavioural and psychiatric labels manifesting in both treated and untreated IEMs (see here and see here for examples). Simons and colleagues decided to look through the collected research on this topic to provide "child and adolescent psychiatrists with an overview of metabolic disorders associated with child psychiatric symptoms, their main characteristics and recommendations for further investigations."
So after boiling down the available peer-reviewed literature to some 71 articles (and in so doing making an important distinction between an inborn error of metabolism and the 'metabolic syndrome'!) authors summarise some of the key IEM associated with labels such as autism, attention deficit hyperactivity disorder (ADHD), learning disability, psychosis and eating disorders. Given that (a) the paper is open-access and (b) this blog tends to favour autism research, I'm gonna focus in on some of the details pertinent to the autism spectrum. I do however recognise that when it comes to the term 'over-represented comorbidity' in autism some of the other diagnostic labels covered by Simons et al might also come into the frame.
Long quote coming up: "Known metabolic disorders in autism are phenylketonuria, disorders in purine metabolism (such as adenosine deaminase deficiency, adenylosuccinate lyase deficiency, dihydropyrimidine dehydrogenase and dihydropyrimidinase deficiencies), organic acidurias (such as propionic academia, 3-methylcrotonyl-CoA carboxylase deficiency and pyridoxine dependency), disorders of branched-chain amino acids creatine deficiency, biotinidase deficiency, cerebral folate deficiency, succinic semialdehyde dehydrogenase deficiency, Smith–Lemli–Opitz syndrome (SLOS), late infantile ceroid lipofuscinosis, histidinemia, Sanfilippo disease, glucose 6-phosphate dehydrogenase deficiency, urea cycle disorders, X-linked ichthyosis, and mitochondrial disorders." I've popped in a few links to other occasions where a specific IEM has been associated with autism and covered on this blog.
Simons and colleagues also cover some of the important research findings where specific amino acids have been analysed and found in unusual levels in cases of autism as potentially being important too. This is relevant because disordered amino acid levels as noted in the case of phenylketonuria (PKU) and the aromatic amino acids phenylalanine (and tyrosine) can be an important finding in relation to some IEM. That they specifically focus on some of the research looking at homocysteine levels and autism is rather interesting (see here) and something that I am going to be discussing in future posts.
What's more to say? Well, I think it is also important to highlight how Simons and colleagues talk about 'other signs and symptoms of the metabolic disease' [IEM] alongside the presentation of autism. This is important in the context that science is starting to more fully understand how a diagnosis of autism rarely exists in some sort of diagnostic vacuum (see here) and quite a lot of different types of comorbidity seem to be 'over-represented'. I'd be inclined to suggest that this detail provides even stronger evidence for how IEM and at least some autism represent an important partnership.
Finally, I refer back to one of the statements made by the authors on "recommendations for further investigations." They suggest that those presenting with: "A positive family history of metabolic disease... Symptoms or signs are triggered by food intake (esp high protein content foods), fever, fasting, surgery (catabolism)... Feeding difficulties, food refusal, failure to thrive, eating disorders combined with symptoms of myopathy or fatigue... Mental retardation and/or regression... Epilepsy, episodes of lethargy or confusion... Dysmorphic feature" should be considered for further investigations. Yes, some of the language is not what I would use and yes, that covers quite a bit of clinical ground but screening is the first part to ruling out such a potential organic correlate of some autism and may in some cases, yield potentially important insights (see here)...
 Simons A. et al. Can psychiatric childhood disorders be due to inborn errors of metabolism? European Child & Adolescent Psychiatry. 2016. Sept 30.
Simons A, Eyskens F, Glazemakers I, & van West D (2016). Can psychiatric childhood disorders be due to inborn errors of metabolism? European child & adolescent psychiatry PMID: 27695954