Monday, 2 April 2012

Six developmental trajectories, autism and bloomers

"If you've met one person with autism, you've met one person with autism" is how the saying goes. Heterogeneity is a keyword accompanying most descriptions of autism spectrum conditions; that and the fact that comorbidity is a mighty complex thing when it comes to presentation and interaction with core autism features.

Enter then a study by Christine Fountain and colleagues* (full-text) and the suggestion that from diagnosis in early infancy to aged 14 years old, children with autism spectrum conditions are likely to fall into one of six developmental trajectories related to their symptom presentation and progression. Perhaps even more surprising is the suggestion that within one of these trajectory groups, about 10% of children experience "rapid gains", blooming and outgrowing their severe disabilities by adolescence according to some of the media reports. Interesting isn't it?

The paper is open-access but here is a short summary of the findings:

  • Group based trajectory modelling was carried out based on the core triad of symptoms for 6975 children with autism identified through the California Department of Developmental Services (DDS) records.
  • Core symptoms of autism were assessed based on a Client Development Evaluation Report (CDER) which whilst not diagnostic for autism, has items covering the domains of communication, social interaction and repetitive behaviours. Several CDERs for each participant provided a longitudinal measure of how symptoms developed over time.
  • Six groups were modelled from the cumulative data and grouped high to low based on functioning within the triad of domains.
  • The results: based on the CDER scores for the communication domain, all groups showed an improvement in communicative function at 14 years compared with 2-3 year old baseline scores. Where you started in terms of 'functioning' seemed to determine where you ended up: more able children finished with better scores at study end. This was true aside from one group, termed 'bloomers', who showed some quite astonishing improvements in communicative function crossing over other group profiles and peaking at the time puberty would be expected before plateauing slightly. Same scenario for the social area trajectory. Group scores for repetitive behaviours were less uniform, with several groups actually showing a greater frequency of such behaviours over time and maturation as per other studies (here and here).
  • There was some noticeable heterogeneity between participants and groups and intellectual / learning disability also seemed to play a hand in outcome.
  • What were the characteristics of 'bloomers'? Well, generally-speaking, aside from presenting with quite severe problems in communicative and social areas at an early age, they were more likely not to have intellectual disability comorbid and perhaps have mothers who were white and quite well-educated. I'm not totally impressed by these last socio-economic factors and their involvement, but can perhaps see why they might be included in the analysis.

As the authors note, this study is a bit of a first in that it collected lots of consistently-derived data over quite a long period of time charting childhood for quite a lot of children diagnosed with autism. Following on from another quite recent paper asking about whether it is possible to grow out of autism, the answer from this study is that at least for a proportion of children who start with quite severe autism, the prospects for improvement are potentially really quite good over the period of childhood. Puberty however seems to offer its own little challenges as many parents of children with autism can testify.

There are however a few gaps not covered in this latest paper. So for example, how many children used some kind of intervention, be it educational, behavioural or biomedical and what effect this might have had on core symptom presentation; whether other important comorbidity such as epilepsy or seizure-type disorders had any bearing on symptom progression; and more practically whether behavioural scores relating to the triad of autism symptoms reflected real-world progression in terms of educational attainment and use or loss of the diagnostic label of autism.

I have to say that I am really interested in this paper. Interested because of the close developmental trajectory across both language and social domains noted in this paper as conceivably offering support for the proposed dyadisation (is that a real word?) of the diagnosis of autism. But mainly interested in the 'bloomer' category of children across the different domains.

All participants were diagnosed with autism, but I wonder about this bloomer group and whether their 'autism' was somehow different in terms of something mimicking a more classical presentation of the condition. Without trying to sound too contentious, the first thing that passed my mind was infection as being potentially related. Encephalitis, meningitis and other related conditions affecting children early and leading to the kind of symptoms noted in the current paper. As time goes on, the infection passes, the brain taps into its fantastic plasticity properties and hey presto, symptoms start to abate - at least to some degree.

Post-encephalitis cases of autism have been reported down the years (here and here) and whilst correlation does not imply causation, autism cases overlapping with the seasonality of viral meningitis have also been recorded. Of course there are lots and lots of other potential infections which could also show some involvement as per the suggestion 40 years ago now linking autism and rubella. I made a suggestion a post back about looking at T.gondii antibodies in relation to autism given the quite startling links being made with other conditions. Who's to say that they might also show some involvement? (Dare I even mention PANDAS or is it called PANS now?)

Those that read this blog will know that I am a fan of looking at endophenotypes in relation to autism and other conditions. The way I see it is that the categories reported in this study are ripe for this 'phenotype' kind of approach and in particular a detailed inspection of those bloomers in relation to the immune system, mitochondrial function, overlapping autoimmune conditions, even bowel symptoms, alongside lots and lots of other parameters just to see if there are any potentially important differences.

I don't want to lose sight of the 'individuality' that accompanies autism and its presentation despite the evidence presented by Fountain and colleagues. The move however to recognising that a group of children on the autism spectrum may possess a slightly different developmental trajectory to others is something that really should start to make people sit up and ask 'who' and 'why'?

* Fountain C. et al. Six developmental trajectories characterize children with autism. Pediatrics. April 2012.
DOI: 10.1542/peds.2011-1601