Seminar 3 –
Characteristics and Comorbidity
Elizabeth Hill:
University of London
Modelling comorbidity: implications for DCD
Two striking aspects of the study of developmental coordination disorder (DCD) are: (i) the heterogeneity of behaviour observed within those diagnosed with the disorder, and (ii) the existence of DCD-like motor impairments in those with other neurodevelopmental disorders. If we are to understand fully the aetiology of this common, but little understood, disorder, we will need to have a clearer awareness of the subtypes and comorbidities that exist. Furthermore, we need to have a clearer conception of the framework within which we evaluate the aetiology of the disorder and within which we link aetiology with observed behaviour. Here, I advocate the use of a causal modelling approach to this end (cf. Morton, 2004).
In this presentation, a range of approaches were considered in order to provide an overview of the state of our knowledge in these areas, to evaluate useful research methodologies and to identify a range of questions for future work. To understand the nature of comorbidity we can consider research on motor skill in individuals diagnosed with other neurodevelopmental disorders. A number of these disorders have been associated with DCD-like motor difficulties. Most notable among these are autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), specific language impairment (SLI) and developmental dyslexia. Overall, it is clear that the prevalence of DCD-like symptoms is widespread in these (and other) disorders. However, motor impairment is not central to these disorders per se in that motor difficulties are not seen in all individuals diagnosed with these disorders, nor are they a necessary condition for their diagnosis. This suggests that it is possible to have ASD+DCD, for example, and that these are two ‘separate’ disorders.
A further way of understanding comorbidity is to evaluate the performance
of an individual with a diagnosed neurodevelopmental disorder across more
than one cognitive domain. In this approach, performance in the particular
cognitive domain seen as core in that disorder (e.g. reading in dyslexia)
and in domains that are not believed to be part of that disorder (e.g. movement
or attention in dyslexia). Rigorous studies using this approach are limited
and although there are advocates for dropping the labels for individual disorders
and adopting instead the label atypical brain disorder, many argue that co-occurring
difficulties arise from different aetiologies and are therefore separate.
Certainly studies comparing autism and ADHD suggest a very different aetiology
to account for co-occurring disorders. Explanations of comorbidity that allow
for co-occurring (but separate) disorders include: (i) one impairment (e.g.
language) plays a specific role in the deficits seen in another cognitive
domain (e.g. movement), (ii) a neuromaturational delay that gives rise to
deficits in different domains that do not, individually, cause each other,
and (iii) a neural cause for comorbidity between motor and cognitive development.
This relates to the tight relationship between the cerebellum and frontal
cortex in terms of neuronal connections and the process of brain maturation.
Cluster analysis has been applied as a method of understanding the heterogeneity
observed within the DCD population. This has been a valuable technique since
clear clusters have been identified.
However, the outcome of a cluster analysis will be affected by the variables
measured, a priori predictions and the sample included. Therefore, to enable
such a method to be truly informative, a clear theoretical framework, representative
sample and detailed description of participants and methods are required.
This will allow replication and theoretical advance, which in turn will lead
to higher chances of intervention success. It is important to note, however,
that the identification of DCD subtypes does not tell us directly about the
aetiology of this disorder. Therefore, clear studies are needed to investigate
the different mechanisms underlying the clusters identified. Crucial will
be to understand if/why comorbidity is associated with a particular subtype(s).
This necessitates a research programme that focuses on understanding subtypes
in direct relationship to comorbidity.
Many methodological issues are raised with respect to understanding heterogeneity within DCD and homogeneity of motor difficulties across neurodevelopmental disorders. Of greatest benefit will be the use of multiple methods of research and analysis to build a broader picture of the strengths and weaknesses observed and the clustering of these difficulties. At present, clear theoretical frameworks are few and far between within the study of DCD and this poses a significant challenge to researchers. It is necessary to link the aetiology of any disorder with the behaviours observed within that disorder. Here, I advocate the use of a causal modelling approach to this end (cf. Morton, 2004). This is a way of thinking about any neurodevelopmental disorder and provides us with three levels of description – behaviour, cognition and biology. In this framework, cognition is the link between behaviour and biology and is thus a crucial component of the puzzle. The influence of environmental effects (e.g. prenatal environment, parental bonding failure) is also considered within this framework. I suggest that this causal modelling approach will allow those working with individuals with DCD to come to a clearer understanding of the within-group heterogeneity (subtypes within DCD) and the characteristics that are homogeneous across disorders (the similarity of motor impairment between DCD and some individuals with autism, for example). This, in turn, will lead to an understanding of the aetiology of different subtypes. This information can then be used to have a positive impact on the daily lives, educational and emotional outcomes for individuals with DCD and their families.
In conclusion, research has identified a number of subtypes of DCD and an overlap in symptoms between those diagnosed with DCD and other developmental disorders such as autism, specific language impairment and developmental dyslexia. A clear theoretical framework is needed in which to evaluate and develop these findings further. The importance of understanding the links between biology, behaviour and cognition are central in this endeavour.
Elisabeth Hill
Goldsmiths College, University of London
Morton, J. (2004). Understanding developmental disorders: a causal modelling approach. Oxford: Blackwells.