Sunday, 4 May 2014

Weighing up genetics and environment in autism

"Environment as important as genes in autism, study finds" read the press release about the study by Sven Sandin and colleagues [1]. Even the authors seemed to be taken aback by the weighting their results gave to a role for non-genetic factors contributing to autism risk as per their comments: "We were surprised by our findings as we did not expect the importance of environmental factors in autism to be so strong... It's now clear we need much more research to focus on identifying what these environmental factors are".

And so the titanic battle between genes and environment continues in autism research as memories flood back about the paper by Joachim Hallmayer and colleagues [2] talked about in a previous post (see here).

Backing up slightly, there are a few points to make about the Sandin paper:

  • Based on data derived from one of those very useful Scandinavian health registries, this time based in Sweden, the starting point for researchers were the records of over 2 million children born between 1982 and 2006. 
  • Fourteen and a half thousand children diagnosed with an autism spectrum disorder (ASD) were included in those millions of records, 5689 of them diagnosed with autism up until the end of December 2009. One certainly can't say that it was an under-powered study in terms of the participant numbers examined despite the focus being on anonymised data from medical records. 
  • As per another write-up of the study (see here) the authors looked at both heritability, "the proportion of risk in the population that can be attributed to genetic factors" and relative recurrence risk (RRR) which "measures individual risk for people who have a relative with autism".
  • Results: "Autism risk rises 10 fold if first child has disorder" for full siblings reported one newspaper. So, having a full brother or sister with autism places a sibling at some significant risk of developing autism or ASD. Even a half brother or sister is associated with something like a 3-fold risk for autism, and a 2-fold risk if it is a cousin. Monozygotic twins.. well the RRR came in at 153 which I'm pretty sure speaks for itself.
  • But... "The ASD heritability was estimated to be 0.50 (95% CI, 0.45-0.56) and the autistic disorder heritability was estimated to 0.54 (95% CI, 0.44-0.64)". This is the part of the Sandin paper that most media have gone with, suggesting that for autism "heritability is only half the story, with the other 50 percent explained by environmental factors such as birth complications, socio-economic status, or parental health and lifestyle". 

I don't think I need to tell you how important the Sandin paper might be to autism research. Not only because the findings further inform about the potential recurrence risk of autism [3] and the implications for genetic counselling, but also that they bring back to our attention how genes and environment might [variably] interact when it comes to autism risk. Certainly sweeping generalisations that I've heard down the years about 'autism being genetic' might have to be slightly tempered. Regular readers of this blog probably already know how I feel about generalisations when it comes to autism, sorry the autisms, and how aspects of environment (however you wish to describe this) are finding quite a bit of research favour in recent years (see here and see here).

I'd also like to think that the Sandin study might further stimulate some research interest and funding when it comes to looking at the bridges between genes and environment when talking about autism risk. I'm thinking about the recent genes and air pollution work (see here) as a template of where to start; understanding that although the door has been further opened to a role for environment in autism risk, this doesn't mean that genetic contributions are all bunk. Think BCKDK gene mutations as one example (see here) and keep in mind the heterogeneity of autism and its potential overlap with other conditions (see here).

And then there is epigenetics. I know that some people still see this upcoming area as a fad or distraction from work on more traditional structural changes to the genome such as all those SNPs, CNVs and other issues. I don't. I see it as very much complementary and how concepts such as DNA methylation or mention of words like histones might have some really important connections to conditions like autism. Think HERVs - human endogenous retroviruses - as one example (see here) under epigenetic control, and indeed, whether as the paper by Shuvarikov and colleagues [4] implied: "HERV-H elements as a mechanism of [genetic] deletion formation".

I could go on but I won't. Suffice to say the autism research wheels keep turning and science books about autism keep being rewritten...


[1] Sandin S. et al. The Familial Risk of Autism. JAMA 2014; 311: 1770-1777.

[2] Hallmayer J. et al. Genetic heritability and shared environmental factors among twin pairs with autism. Arch Gen Psychiatry. 2011 Nov;68(11):1095-102.

[3] Ozonoff S. et al. Recurrence risk for autism spectrum disorders: a Baby Siblings Research Consortium study. Pediatrics. 2011 Sep;128(3):e488-95.

[4] Shuvarikov A. et al. Recurrent HERV-H-mediated 3q13.2-q13.31 deletions cause a syndrome of hypotonia and motor, language, and cognitive delays. Hum Mutat. 2013 Oct;34(10):1415-23.

----------- Sandin, S., Lichtenstein, P., Kuja-Halkola, R., Larsson, H., Hultman, C., & Reichenberg, A. (2014). The Familial Risk of Autism JAMA, 311 (17) DOI: 10.1001/jama.2014.4144