Professor Tony Charman's research brief (3)


In this research briefing Professor Charman discusses two current research papers that have raised much debate and discussion across the autism and academic community - the studies of Dawson G et al and Fein D et al.

Progress and protest in autism outcome studies

A couple of papers emanating from leading autism clinical research centres have hit the headlines in the past few months. To some commentators these studies signalled hope for improving outcomes for people with autism; to others they raised more concerning issues and debates.

First, towards the end of last year Geraldine Dawson and colleagues published a follow-up report on the already high profile Early Start Denver Model (ESDM) randomised controlled trial (RCT). In the original paper the authors reported significant improvements in developmental abilities, in particular language and communication abilities in the ESDM group (who had received 15 hours a week of this developmental-behavioural therapist delivered intervention over 2 years, augmented by parent training) compared to a group randomised to regular community services1. This follow-up was conducted several months later and involved around two-thirds of the sample undergoing a benign electrophysiological experimental assessment called EEG (electroencephalogram)2. The experiment involved the children (now aged between 4 to 6 years) sitting watching videos of faces or objects. The authors compared the pattern of brain activity to those seen in typically developing children. Although the groups were similar on several of the measures, on a couple of measures the ESDM group had brain activity more similar to the typical group and different from that in the community services controls. As the authors indicated in their title, they concluded that the ESDM treatment had ‘normalised’ the brain activity of the young children with autism. The use of the term ‘normalised’ provoked considerable reaction both from the autism advocacy and rights community and in discussions amongst other academics that I discussed the paper with. In contrast, much of the media coverage reflected the ‘brain bias’ that often accompanies reporting of neuroscience3. Although the original ESDM study 2 years ago demonstrated clear behavioural/developmental advances due to the treatment, now that there was evidence of a ‘brain difference’ then this really must be true! I take the view that the findings are both intriguing and provocative (not necessarily in a bad way in spite of the unfortunate language used in the title) and need further investigation and follow-up. There are now a number of groups worldwide including ours who are combining neuroscientific methods with developmental/behavioural interventions and this will be the first of several to emerge in the next few years.

The second study was by Deborah Fein’s group and was published in February this year (2013)4. The authors reported on a group of children with autism who had what the authors describe as ‘optimal outcome’. They recruited a sample of children and young adults who had gone on to show little or no autism symptomatology, despite having had a clear diagnosis and some evidence of language delay in early childhood. On a battery of measures this group scored the same as typically developing individuals, with one or two exceptions on language measures. The media preferred the phrase ‘recovery’ than ‘optimal outcome’ to describe the phenomenon (and if you were a headline writer I suspect that you would too). Although the authors avoided using the word ‘recovery’ in their article; they did speculate when speaking to the press about the possible effects of intervention in creating this outcome. Those of a more cautious disposition, including this author, were concerned that by definition such selectively sampled, retrospective and uncontrolled designs are least able to reliably answer questions about treatment effectiveness. However, the more positive message is to remind us that outcome in autism is highly variable, and that we know much less than is desirable about what factors influence outcomes.

Despite some concerns about how these two studies were covered by the media, they are important as they highlight how research on interventions and outcomes has taken centre stage in the autism research field. After decades where little (and almost no high quality) intervention studies were being done, in the past 10 years there has been a surge in intervention research and some has been of increasingly high quality and rigous5. Perhaps it is inevitable that intervention research garners headlines and moreover its fair share of misleading ones. However, it is important for the autism community that the challenge of understanding which supports and interventions can promote positive outcomes is taken on and tested using the most rigorous scientific methods available.


1. Dawson G et al. (2012). Early behavioral intervention is associated with normalized brain activity in young children with autism. Journal of the American Academy of Child and Adolescent Psychiatry, 51, 1150-1165.

2. Dawson G et al. (2010). Randomized, controlled trial of an intervention for toddlers with autism: the Early Start Denver Model. Pediatrics, 125, e17-23.

3. Bishop DV. (20123). Research Review: Emanuel Miller Memorial Lecture 2012 - Neuroscientific studies of intervention for language impairment in children: interpretive and methodological problems. Journal of Child Psychology and Psychiatry, 54, 247-259.

4. Fein D et al. (2013). Optimal outcome in individuals with a history of autism. Journal of Child Psychology and Psychiatry, 54, 195-205.

5. Charman T. (2011). Glass half full or half empty? Testing social communication interventions for young children with autism (Invited Commentary). Journal of Child Psychology and Psychiatry, 52, 22-23.

Professor Tony Charman

Chair in Clinical Child Psychology
King's College London, Institute of Psychiatry
Department of Psychology
Box PO77, Henry Wellcome Building
De Crespigny Park
Denmark Hill
London SE5 8AF


Want to discuss this article with others? Why not set up a discussion here

Download a PDF copy of the breifing here

Date added: 22nd March 2013