Studies into the phenotype of rare genetic syndromes largely rely on bivariate analysis. The aim of this study was to describe the phenotype of Cornelia de Lange syndrome (CdLS) in depth by examining a large number of variables with varying measurement levels. Virtually the only suitable multivariate technique for this is categorical principal component analysis. The characteristics of the CdLS phenotype measured were also analysed in relation to parenting stress.
Data for 37 children and adults with CdLS were collected. The type of gene mutation and relevant medical characteristics were measured. Information on adaptive functioning, behavioural problems, the presence of the autistic disorder and parenting stress were obtained through questionnaires and semi-structured interviews with the parents. Chronological age and gender were also included in the analysis.
All characteristics measured, except gender, were highly interrelated and there was much variability in the CdLS phenotype. Parents perceived more stress when their children were older, were lower functioning, had more behavioural problems, and if the autistic disorder was present. A new perspective was acquired on the relation between the gene mutation type and medical and behavioural characteristics. In contrast with earlier research the severity of medical characteristics did not appear a strong prognostic factor for the level of development.
Categorical principal component analysis proved particularly valuable for the description of this small group of participants given the large number of variables with different measurement levels. The success of the technique in the present study suggests that a similar approach to the characterisation of other rare genetic syndromes could prove extremely valuable. Given the high variability and interrelatedness of characteristics in CdLS persons, parents should be informed about this differentiated perspective.