Factor analysis of behavioural symptoms in Huntington’s disease
- Hugh Rickards1,
- Jenny De Souza1,
- Marleen van Walsem2,
- Erik van Duijn3,
- Sheila A Simpson4,
- Ferdinando Squitieri5,
- Bernhard Landwehrmeyer6,
- The European Huntington’s Disease Network
- 1Department of Neuropsychiatry, Birmingham and Solihull Mental Health Foundation Trust, Edgbaston, Birmingham, UK
- 2Rikshospitalet, Centre for Rare Disorders, Oslo, Norway
- 3Department of Psychiatry, Leiden University Medical Centre, Leiden, Netherlands
- 4Clinical Genetics Centre, Aberdeen, UK
- 5IRCCS-Neuromed, Pozilli, Italy
- 6Ulm University, Ulm, Germany
- Correspondence to Dr Hugh Rickards, Department of Neuropsychiatry, Birmingham and Solihull Mental Health Foundation Trust, Barberry Building, 25 Vincent Drive, Edgbaston, Birmingham B15 2FG, UK;
Contributors All authors contributed to the design, writing and editing of the manuscript.
- Received 20 April 2009
- Revised 16 October 2009
- Accepted 19 October 2009
- Published Online First 14 April 2010
A principal-components factor analysis was performed on behavioural data obtained from the European Huntington’s Disease Network REGISTRY study. 1690 valid assessments using the United Huntington’s Disease Rating Scale Behavioural Rating Scale were included in the analysis. This large data set confirmed previous reports of distinct behavioural patterns within Huntington’s disease comprising a depressive factor, a dysexecutive factor, an irritability factor and a psychosis factor.
Huntington’s disease (HD) is a progressive neuropsychiatric disorder consisting of motor, cognitive, behavioural, affective and perceptual changes. Psychiatric problems in HD are common and lead to significant impairment in the Quality of Life of people with HD.1 2 The Behaviour section of the Unified Huntington’s Disease Rating Scale (UHDRS) was designed to record a range of psychiatric phenomena in people with HD.3 There are few studies analysing the relationship between different psychiatric phenomena in HD. Marder et al4 performed a factor analysis on 960 UHDRS assessments as part of the Huntington’s Study Group (HSG) project. They isolated 15 factors, including ‘depression/anxiety,’ ‘suicidal thoughts,’ ‘aggressivity’ and ‘obsessive/compulsive.’ There were a number of smaller factors including hallucinations in different modalities, something not routinely enquired about in current UHDRS assessment. Craufurd et al5 used factor analysis to examine data on 78 patients with HD who had undergone the Problem Behaviours Assessment in HD (PBA-HD). Principal-components factor analysis revealed three clusters of symptoms, reflecting apathy, depression and irritability. The apathy factor was closely related to the duration of illness, being a stronger feature later in the illness. A further paper by Kingma et al6 using the PBA on 152 HD mutation carriers also showed a three-factor solution: apathy, depression and irritability.
The European Huntington’s Disease Network (EHDN) REGISTRY study (http://www.euro-hd.net/html/registry/) has provided a unique opportunity to re-examine psychiatric symptoms in the setting of HD. This study has recruited around 4000 people with HD, and participants are followed up at yearly intervals.
The aim of our study was to perform a factor analysis on all the completed UHDRS behavioural assessments performed across Europe since the inception of the REGISTRY study. The total number of fully completed assessments at the time of writing was 1690. Elucidation of specific behavioural or psychiatric factors may lead to the identification of specific phenotypes within HD, which in turn could lead to a greater insight into aetiology and aid with treatment.
Data were made available from the EHDN REGISTRY database following approval from the Scientific and Bioethical Advisory committee and the Executive committee of the EHDN. UHDRS behaviour scores were obtained for 1803 assessments on individual participants. The UHDRS behaviour section is made up of 11 individual items, each scored on the dimensions of frequency and severity. The frequency (score 0–4 points) and severity (score 0–4 points) of each item were multiplied to give a single score for each item. The item scores were then added together to give the total behaviour score.
A principal-components analysis was employed, with the ‘frequency multiplied by severity’ score for each item used as a factor. Data on all the factors were analysed using SPSS V.16 (SPSS, Chicago, Illinois). A Varimax rotation was utilised with Kaiser normalisation. Because of missing data, 113 cases were excluded to give a final N of 1690. Full ethical approval was obtained for this study.
As shown in table 1, four factors achieved a loading of over 0.4. These factors accounted for 63.8% of variance in the total score. The four factors were;
depression (depressed mood, low self-esteem, suicidal ideation, anxiety);
drive/executive function (perseveration, compulsion, apathy);
psychosis (delusions and hallucinations).
There was a good delineation between factors, with each item clearly loading to one of the four factors. All factors achieving the threshold of 0.4 were at least 0.6. Eigenvalues, after rotation, were as follows: depression (2.340), drive/executive function (1.714), aggression/irritability (1.580), psychosis (1.385).
This factor analysis builds on the findings of Craufurd et al5 and Kingma et al,6 who both used the PBA, and Marder et al,4 who used an old version of the UHDRS Behaviour assessment. All the studies have shown distinct factors for depression, executive functions and irritability. Both studies using the UHDRS behaviour section had separate factors for psychosis.
These data further support the existence of specific clusters of psychiatric symptoms in HD, which may be related discretely to affect, executive function, aggression/irritability and perception. More research is needed to assess the validity of these categories, as well as their stability over time, including in the premanifest period. Further work is needed to develop valid rating scales that measure changes in these factors as a result of treatment.
Distinct clusters of behavioural symptoms are emerging in HD. These relate in general to depression, executive function, irritability and psychosis. Further studies should aim to tie in these factors with specific endophenotypes in order to establish clear a clinicopathological correlation.
The authors would like to acknowledge the behavioural phenotype working group of the EHDN (http://www.euro-hd.net/html/network/groups/behavioural/) and CHDI Foundation. The list of contributors to the data set is given in online appendix 1.
Funding European Huntington’s Disease Network.
Competing interests This study received no sponsorship and was led by the investigators. HR has sat on an advisory board for MEDA pharmaceuticals. He has been paid honoraria for work for Cambridge Laboratories. SAS received an honorarium from Cambridge Pharmaceuticals for consultancy.
Ethics approval Ethics approval was provided by the UK MREC.
Provenance and peer review Not commissioned; externally peer reviewed.