Monday, 15 August 2011

Sibling risk of autism

It has been known for quite some time that autism spectrum conditions, whilst not being a wholly genetically based set of condition, does seem to occur more often in siblings of affected children. The research base in this area is fairly large and the 'amount' of risk potentially present varies according to study sample sizes and diagnostic methods and criteria used.

One of the more comprehensive studies carried out fairly recently (2010) hinted that the sibling re-occurrence rate for autism spectrum condition could be anything up to 10%; indeed even higher if one were to start looking at the broader phenotype presentation to include things like speech and language problems. You are probably thinking 10%.. that sounds pretty high. Well a recent study suggests that in some cases it might actually be much higher.

The study by Ozonoff and colleagues* published in the journal Pediatrics estimated that the risk of sibling re-occurrence of an autism spectrum condition might be as high as 18%. There are quite a few descriptions of the study on the web. Geraldine Dawson from Autism Speaks provides a brief overview here and the BBC have also covered the study. I have not yet seen the full-text so can't provide too detailed an overview yet. From these collected and other sources, there are however a few details which can be reported from the data:
[Update: the full-text of the paper can be viewed here]

  • This was a prospective study whereby 664 infants who had a sibling with an autism spectrum condition were followed for the first 3 years of their lives.
  • Children were assessed for autism using the ADOS (one of the gold-standards for assessing autism) and by expert clinical diagnosis.
  • Of the original 664 infants, 132 children met the cut-off criteria for an autism spectrum condition. Of these, 54 exceeded the cut-offs from autism and 78 infants for the wider autism spectrum. 
  • Male siblings were more likely to be diagnosed with autism in all its forms (26%) compared with females (9%). This translates as 1 in 4 younger brothers were diagnosed with an autism spectrum condition compared with 1 in 9-10 younger sisters.
  • Where there was more than one older sibling with autism, the risk to younger siblings presenting with an autism spectrum condition was also increased (to an average of 32%).

Whilst appreciating that risk is not necessarily eventuality, these are some quite startling figures from a research point of view as once again some of our commonly held views on autism are starting to be overturned. It started with the sex ratio, then it was the rates in MZ-DZ siblings and now the recurrence rates.

The current study is a strong one methodologically. Given that it was part of the High Risk Baby Siblings Research Consortium and hence relied on data from more than one location, the generalisability of the data outside of just one geographical area is strong. The use of ADOS and clinical opinion similarly means that diagnostic verification is high.

Looking at this new data and the previous studies 15-20 years ago I am struck by one question: did they really get it so wrong back then or is the autism of today, the same as it was in yesteryear?

* Ozonoff S. et al. Recurrence risk for autism spectrum disorders: a baby siblings research consortium study. Pediatrics. August 2011.