Friday, 4 March 2011

Sex ratio and autism

An article delivered as part of my almost daily BMC journals email update caught my eye this morning, reporting on the sex-biased brain in birds (the feathered variety). It got me thinking again about the sex / gender gap in cases of autism spectrum conditions and all the various debates contained in this issue - the sex ratio, the genetics of it, the psychological theories about it and the diagnostic issues around it (girls perhaps less likely to be diagnosed, variable effect of symptom severity & learning disability, etc). My interest in this issue has been stirred over the last few years; not least because we published a paper on it and also within psychological circles there has been some debate on the nature of studies and theories put forward to explain it. The more recent publication of results related to RORA add to my interest.

It has probably been one of the only major consistencies in autism: males are more likely to be diagnosed with an autism spectrum condition than females. It is almost written in tablets of stone that the gender ratio in autism is roughly 4 males to every 1 female (not including Rett syndrome) right from the earliest report by Kanner.

More recently however there have been a few studies critically questioning whether the 4:1 ratio is indeed the best estimate we have of the gender gap or whether perhaps the rates are changing.  Our own data (which I won't talk about too much for fear of being accused of being a self-publicist) suggested that the overall rate of all autism spectrum conditions in the UK/Ireland (autism, Asperger syndrome, ASD) between 1986-2007 was 7.38:1. When we looked at the individual diagnoses there were also differences: autism = 6.54:1, AS = 12.07:1, ASD = 6.84:1). Similar data from Cambridge University suggested that again in the UK, the overall rate was around 4:1 (for children born between 1988-1993) but (and it is an important but), the sex ratio fluctuated between 3:1 and 8:1 depending on schooling-type (determined largely by ability). A final dataset for the UK derived from the ALSPAC study concluded that in this prospective study, the gender ratio for autism cases between 1991-1992 was 6.8:1.

So there we have it (at least in the UK). Various data derived from various methodologies plotting quite a different gender ratio from that originally established. Of course there are other studies that confirm the 4:1 ratio and so I could be just cherry-picking here. There are a few methodological caveats to all of this of course relating to how and what we diagnose, how autism is viewed and the important issue of co-morbid learning disability.

I will perhaps end with one final question: where is the gender ratio in autism spectrum conditions heading? Again back to our data (sorry!), plotting year-on-year over the course of our study, there appeared to be an increasing gulf opening between the sexes in terms of diagnosis. Why this is happening has not been established. Is it genetic (remember our CNVs)? Is it something environmental? What will happen if the DSM V changes are implemented and Asperger syndrome disappears from future diagnostic literature?

There are a lot of interfering variables in this debate.

Anyway back to my songbirds.