Examining dimensional models of psychopathology experienced by adults with intellectual disabilities

Melville, Craig (2010) Examining dimensional models of psychopathology experienced by adults with intellectual disabilities. MD thesis, University of Glasgow.

Abstract

Background:
Classification systems for use in the diagnosis of mental disorders have been developed based on a categorical model of psychopathology. Although current
categorical diagnostic classification systems have been found to have good utility and reliability, studies have questioned whether these systems have adequate
validity. Dimensional models of psychopathology have been examined as an alternative to categorical diagnostic classification systems and found to be more strongly related to clinical parameters, such as the severity and outcome of mental disorders.

A literature review found a small evidence base on dimensional models of psychopathology experienced by adults with intellectual disabilities. However, the findings were limited by small sample sizes, biased samples and inclusion of only a limited range of items of psychopathology. Furthermore, the methods of exploratory
factor analysis used do not meet established best practice guidelines.

Informed by the existing literature, this thesis aimed to;
1. identify a dimensional model of psychopathology experienced by adults with intellectual disabilities
2. examine the associations of a dimensional model of psychopathology with measures of the severity and outcome of mental disorders
3. compare the predictive validity of dimensional and categorical models of psychopathology.

Methods:
The Psychiatric Present State- Learning Disabilities (PPS-LD) was used as a structured instrument to collect psychopathology data. Exploratory factor analysis
(EFA) following best practice guidelines was used to identify dimensions of psychopathology. Continuous measures representing the dimensions of psychopathology were calculated. Meeting criteria for the diagnosis of a mental
disorder from the Diagnostic Criteria for Psychiatric Disorders for use with Adults with Learning Disabilities (DC-LD) was used as the variable representing the
categorical model of psychopathology.
Baseline data was collected on four measures of severity; the Health of the Nation Outcome Scales- Learning Disabilities (HoNOS-LD), Global Assessment of
Functioning (GAF), Clinical Global Impression (CGI), and the Camberwell Assessment of Needs for Adults with Developmental and Intellectual Disabilities-
Research version (CANDID-R) unmet needs. These measures were completed again at follow up 4-5 years later and change over time used as a measure of longitudinal
outcome.

Bivariate statistics and multivariate linear regression were used to examine the associations of the dimensions of psychopathology, and DC-LD diagnosis, with the measures of the severity of and longitudinal outcome of mental disorders. Relevant socio-clinical variables, associated with psychopathology in previous populationbased
intellectual disabilities studies were included in the analyses: gender, age, living circumstances, level of intellectual disabilities, autism, Down syndrome,
epilepsy, sensory impairments, mobility problems and incontinence.

Key results:
A model of psychopathology with four dimensions was extracted from the EFA. This model was stable in two additional EFA using random samples. There were no
significant correlations between the four dimensions which were labeled depressive,organic, behaviour-affective and anxiety.

Only the anxiety dimension of psychopathology was not associated with any of the measures of severity of mental disorders. The depression dimension was independently associated with severity on the HoNOS-LD (β=.413, p<.001), GAF (β=-.402, p<.001) and the CGI (β=.457, p<.001). The organic dimension was independently associated with severity on the HoNOS-LD (β=.205, p=.004), GAF(β=-.326 p<.001) and CGI (β=.266, p<.001). The behaviour-affective dimension was independently associated with severity on the HoNOS-LD (β=.332, p<.001), GAF (β=-.286, p<.001), CGI (β=.253, p<.001) and CANDID-R unmet needs (β=.178,
p=.018). Level of intellectual disabilities was independently associated with severity on the HoNOS-LD and CANDID-R unmet needs. Finally, younger age (β=-.208,
p=.010), living independently (β=-.599, p<.001) and not having a visual impairment (β=-.191, p=.009) were associated with greater CANDID-R unmet needs.
None of the baseline measures of psychopathology were associated with longitudinal outcome on the CANDID-R unmet needs. Baseline scores on the depressive dimension were significantly associated with longitudinal outcome on the HoNOS-LD(β=.297, p=.034), GAF (β=.342, p=.002) and CGI (β=.373, p=.001). Similarly, the behaviour-affective dimension was significantly associated with longitudinal
outcome on the HoNOS-LD (β=.292, p=.033), GAF (β=.244, p=.036) and CGI(β=.298, p=.009). The organic dimension was only associated with longitudinal outcome on the HoNOS-LD (β=-.382, p=.006). Individuals with mild intellectual
disabilities had poorer outcomes on all four measures of longitudinal outcome.Hearing impairment was associated with poorer outcome on the GAF (β=-.483, p=.000) and CGI (β=-.331, p=.004), and poorly controlled seizures with poorer
outcome on the CGI (β=-1.638, p=.004).The variable representing the categorical model of psychopathology was only independently associated with severity on the HoNOS-LD (β=.178, p=.026), and longitudinal outcome on the GAF (β=.259, p=.045) and CGI (β=.257, p=.044).
However, when categorical and dimensional models were both included in the regression analyses only the dimensional model of psychopathology was retained as independently associated with these measures of severity and outcome.

Conclusions:
The description of a stable dimensional model demonstrates the value of using multivariate statistical methods to examine psychopathology experienced by adults
with intellectual disabilities. Since the findings suggest that dimensional models have better validity than categorical models of psychopathology, the use of EFA, and other multivariate methods, could contribute to the development of valid diagnostic classification systems.

The presence of affective items of psychopathology across the depressive, behaviour-affective and anxiety dimensions highlights the possible relevance of a global affective model of psychopathology. Findings reported in this thesis support the potential relevance of models of affect regulation and affective arousal to developing
an understanding of psychopathology experienced by persons with intellectual disabilities.

There are similarities between the dimensional model in this thesis and the tripartite model of depression and anxiety psychopathology, described in the literature- which
has depressive, anxiety and general distress dimensions. Overlaps between the behaviour-affective dimension, and general distress dimension within the tripartite model, suggest that there may be an association between affective psychopathology and problem behaviours. However, it could be that this association is with affective psychopathology in the general distress dimension, rather than with depressive psychopathology, as examined in previous studies.

Confirmatory factor analyses should be considered to examine the four dimension model of psychopathology. Future studies involving individuals with intellectual disabilities should examine the relevance of global affective models of psychopathology.

• Item Type: Thesis (MD)
• Qualification Level: Doctoral
• Keywords: intellectual disabilities; psychopathology; dimensions; factor analysis; validity
• Subjects: R Medicine > RZ Other systems of medicine; B Philosophy. Psychology. Religion > BF Psychology
• Colleges/Schools: College of Medical Veterinary and Life Sciences > Institute of Health and Wellbeing
• Supervisor's Name: Cooper, Professor Sally-Ann
• Date of Award: 2010
• Depositing User: Dr Craig Melville
• Unique ID: glathesis:2010-2099
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 12 Oct 2010
• Last Modified: 10 Dec 2012 13:51
• URI: http://theses.gla.ac.uk/id/eprint/2099