Wednesday, 25 April 2012

Rich kids, older dads and autism? Or poor kids, older mums and autism?

I'm confused.

I must admit that it is fairly easy to confuse me as family members and work colleagues will readily attest. Not to dwell too much on the point but I'm the sort of chap who easily gets lost even with a map and car GPS, to the cries of "we've been past that building already" and "just ask someone". Given that I am a man, asking someone for directions is just not in my vocabulary.

The source of my current confusion is contained in one edition of the Journal of the American Academy of Child & Adolescent Psychiatry (JAACAP) containing two papers published fairly recently. The first by Rai and colleagues* reports on parental socio-economic status (SES) and risk of offspring autism in Sweden. The second is by Sandin and colleagues** looking at advancing maternal age and risk of offspring autism. The results: kids with a lower SES in Sweden have a greater risk of autism and older mums conceiving later might increase the risk of their children being diagnosed with autism.

A short summary of the Rai paper first:

  • Rai et al were intrigued about how the data on SES and autism (high family SES = increased risk of offspring autism) seemed to fly in the face of many other conditions where lower SES increased risk. 
  • Based on a total population of just over half a million children aged 0-17 years living in Stockholm county between 2001-2007, they wanted to look at the 5000 or so cases of autism included in the population, matched 10-1 with age and gender controlled asymptomatic children, to see if SES factors were relevant. 
  • Importantly, they controlled for various potential sources of bias including things like parental age, migration status and various birth characteristics.
  • Results: lower SES seemed to be associated with a risk of autism in offspring.

And for the Sandin paper:

  • A slightly different approach based on a meta-analysis of epidemiological studies reporting on autism up to January 2012.
  • Sixteen publications met their criteria for study inclusion resulting in a total sample population of just over 25,000 people with autism - autism spectrum conditions - compared with 8.5 million controls.
  • Results: based on various analyses of age and age-brackets, a dose response effect of maternal age on risk of offspring autism was apparent in most studies. So older mums with a greater 'age dose' were at greater risk of having a child with autism. Mums who were 35 years old or older at conception/birth were about 1.5 times more likely to have a child with autism compared to those mums falling into the 25-29 years age bracket. Interestingly, young mums (below 20 years of age) showed a significant decrease in risk of offspring autism.

Why then my confusion?

Well as has been pointed out by a few people in the Twitter-sphere, quite a lot of data has already been produced and published suggesting that SES might have a link to autism risk; the findings generally reporting that kids from families with a higher SES (rich kids) might be at greater risk than their poorer counterparts. Added to the fact that quite a lot of attention has also been directed to older dads, older sperm and more possibility of some kind of genetic-environmental effect being linked to an increased risk of autism in offspring - remembering the Nature papers on de novo mutation and dads recently - you can perhaps see why I might be a little confused with these results which seem to suggest something slightly different.

I hasten to add that I am not questioning the studies, current or previous, how they were carried out or anything like that, it's just that with all the fuss being made about factors like higher SES and paternal age, one can't help but wonder if such issues are as definitive as they are often portrayed. Of course there are caveats to such generalisations. Who was looked at and where do they live alongside the myriad of interfering variables which might not be controlled for. Autism being an extremely heterogeneous condition with aetiology most likely linked to lots and lots of different factors variably depending on genes and environment should also not be forgotten. I hark back to other studies on SES and autism which to some degree substantiate the current findings from Rai and colleagues; to quote from King & Bearman*** (full-text) "We find that the socioeconomic gradient for autism has begun to reverse".

Still, what these latest figures serve to show is that autism is still very much of an enigma and one should always be wary of sweeping generalisations when it comes to risk. The added assumption also when asking questions like "what causes autism" or rather "what factors are associated with the increase in cases of autism", as well as being mindful of heterogeneity and comorbidity, we should also be responsive to the fact that the causes may be fluidic and fluctuating. What might drive increase in one cohort at one time period, may not necessarily carry the same weight in others at other times. I will talk more about this in subsequent posts on the back of some interesting suggestions regarding environmental pollutants (see here for a sneak preview of the research).

To finish, I'm minded to suggest something a little bit lively to relieve me of my confusion. How about a spot of B52s?

* Rai D. et al. Parental socioeconomic status and risk of offspring autism spectrum disorders in a Swedish population-based study. JAACAP. 2012; 51: 467-476.

** Sandin S. et al. Advancing maternal age is associated with increasing risk for autism: a review and meta-analysis. JAACAP. 2012; 51: 477-486

*** King MD. & Bearman PS. Socioeconomic status and the increased prevalence of autism in California. American Sociological Review. 2011; 76: 320-346.


  1. I'll have to add "poor white trash" to the growing list of my son's risk factors (heehee).

  2. Thanks Mrs Ed.

    Appreciating that we should all be grateful there is research on-going about 'risk' and autism over a purely 'autism is genetic' view (prevalent not so long ago), I can perhaps see how it does seem to be getting a little bit silly with all the associations being made. I'm pretty sure that for example, the authors of the obese mums - autism risk study did not make many friends after that particular publication came out.

    Part of this I think, has to do with how the issue of 'risk' is being communicated and part of it to do with the current obsession with linking things, everything to autism - autism as if it were some homogeneous condition with little or no variation or comorbidity.

    There aren't any easy answers I'm afraid to getting round these factors; although one would perhaps suspect that as the endophenotype view starts to become more widely observed in autism research, such generalisations might become a little more focused on the parts rather than the whole.