Cambridge Catalogue  
  • Help
Home > Catalogue > Psychiatric and Behavioural Disorders in Intellectual and Developmental Disabilities
Psychiatric and Behavioural Disorders in Intellectual and Developmental Disabilities
Google Book Search

Search this book


  • 9 b/w illus. 12 tables
  • Page extent: 440 pages
  • Size: 247 x 174 mm
  • Weight: 0.87 kg

Library of Congress

  • Dewey number: n/a
  • Dewey version: n/a
  • LC Classification: RC451.4.M47 P77 2007
  • LC Subject headings:
    • People with mental disabilities--Mental health
    • People with mental disabilities--Mental health services

Library of Congress Record


 (ISBN-13: 9780521608251)

Part I

Assessment and diagnosis

Diagnosis of mental disorders in people
with intellectual disabilities

Peter Sturmey


Classification of mental disorders in people with intellectual disabilities (ID) continues to be a very active field since the last edition of this book (Sturmey, 1999). This chapter reviews the importance and functions of classification as well as psychometric properties that any adequate classification system must have. It then goes on to review classification in relation to intellectual disabilities, mental health and classification of mental health issues in people with ID. The final section highlights the changes that have occurred and areas for future development.

The importance and functions of classification

Classification is often seen as a hallmark of science: atomic theory and the Linnaean system advanced their respective folk technologies into sciences. Yet, nineteenth-century Alienists were reluctant to classify madness beyond insanity and mental deficiency. As psychiatry strived towards scientific respectability, classification progressed vigorously, especially in the twentieth century. Kraepelin (Kihlstrom, 2002) developed taxonomy of insanity into 15 aetiological classes. In the end he had to admit that the effort was in vain, and that classification by course and prognosis was more promising. The precursors of the Diagnostic and Statistical Manual (DSM) implied aetiology, but were based on presenting symptoms, as putative causes could not be observed. Presenting symptoms continue to be cardinal features of DSM and the International Classification of Disease (ICD). Critics of the early DSM-Ⅰ and Ⅱ noted that they were both explicitly psychoanalytical in orientation and did not guide the diagnostician as to the specific symptoms and were hence unreliable. Later editions of DSM made three important changes. First, it was not explicitly theoretical (although implicitly it adopted a medical model of mental disorders). Second, it addressed the issue of reliability by operationalizing symptoms, and the number and combinations of symptoms necessary for diagnosis. Third, instead of adopting a model of diseases where all similar diseases shared common features, it adopted a stochastic model wherein a certain number of symptoms were required, but no one symptom was required. Hence, in DSM-Ⅱ all psychoses were characterized by problems in reality testing, whereas in DSM-Ⅳ there is no such commonality among psychotic disorders (Kihlstrom, 2002).

   Classification of mental disorders and ID has become longer and progressively more detailed, whereas this trend is not observed in classification of physical illnesses. Houts (2002) noted that from DSM-Ⅰ to DSM-Ⅳ the number of diagnoses increased from 106 to 365 and the number of pages increased from 128 to 886. (From these unhappy data we can observe that DSM grows at the rate of approximately six new diagnoses and 16 pages per year. If the growth is mercifully linear then we might predict a DSM-Ⅴ in say 2010 containing 461 diagnoses, including 96 new ones, with 1142 pages!)

   Classification serves several, diverse purposes (see Table 1.1). Problems arise in a single classification system that accommodates multiple functions. For ex-ample, a system of classification of ID based on aetiology might be highly useful for prevention, but might have no use for guiding habilitation.

Necessary properties of classification systems

Any adequate classification scheme must have certain properties. These include being publicly verifiable and open to inspection by more than one person, measuring what it purports to measure, practicality, completeness, acceptability, and usefulness. Reliability refers to a number of related but distinguished properties, including stability across observers, stability over time and internal consistency. Validity refers to the extent to which a measure accurately and coherently measures what it purports to measure. Table 1.2 gives definitions, examples and associated statistics of these properties. Reliability and validity criteria can be applied to a variety of aspects of measurement including diagnostic decisions, such as whether or not someone has a mental disorder, or what type of disorder it is. They can also be applied to psychometric instruments (e.g. Aman, 1990; Deb, et al., 2001, Finlay, 2005; Sturmey, et al., 1991), structured diagnostic interviews (Moss, et al., 1993; Patel et al., 2001) as well as to clinical decision-making, such as a decision to place someone on a psychotropic medication.

   Most research emphasizes reliability and validity, but there are also other important properties of a classification system. Classification systems must be both complete and efficient to use. Practitioners face the paradoxical needs of a classification system that efficiently addresses the relatively small number of common

Table 1.1 Some of the functions of psychiatric classification.

Purpose Example

Service record keeping Diagnoses or levels of disability are entered into a record to indicate reasons for treatment or change in disability in response to intervention.
Statistical A school district surveys the numbers of students with ID and self-injury in order to determine staff training and other policy needs.
Communicating with third parties A label such as ‘Mood Disorder NOS’ is used to summarize a cluster of presenting symptoms and to explain aggressive behaviour.
Reimbursement for services Coverage for in-patient services is denied because the client’s symptoms do not meet covered diagnosis.
Research A research protocol specifies that all participants will meet DSM-Ⅳ criteria for mild or moderate ID and post-traumatic stress disorder.
Basis for screening instrument Structured interview such as the PASADD (Moss et al., 1993; Patel et al., 2001) and psychometric instruments are based on standardized diagnostic criteria in order to screen for those people who truly meet diagnostic criteria.
Summarizing symptoms Instead of listing a large number of specific observable behaviours, a client is said to be manic.
Aetiology The cause of ID is classified as ‘Non-Organic/cultural familial’ or ‘Organic’. If ‘Organic’ is coded then one of several classes is checked such as infections, acquired brain damage, or genetic; if ‘genetic’ is coded, one of 800 or so specific causes is coded.
Prognosis Down syndrome has an increased risk of Alzheimer’s disease after age 40; Lesch–Nyhan syndrome does not.
Legal purposes A client is found incompetent to stand trial because of mild ID and personality disorder.
Service planning, monitoring and evaluation Statistical returns to the US Department of Education indicate that the number of students with ID has dropped by nearly 50% in ten years, but the number of children with autism spectrum disorders has greatly increased. Additional resources are provided for a variety of services related to autism spectrum disorders.
Determining blameworthiness A child with ID is removed from school because it is determined that his aggressive behaviour was not caused by ID or any mental health diagnosis.

Table 1.2 The various types of reliability and validity of classification systems.

Property Definition Examples Associated statistics

Inter-rater reliability Independent observers agree on the presence or absence of a symptom or diagnosis. Two interviewers agree that the client has depressed mood. Two researchers scoring a videotaped standardized interview agree that a client meets research diagnostic criteria for a psychotic disorder. Two team members, both present in an interview, agree that a client no longer meets diagnostic criteria after treatment was successful. Cohen’s kappa between two diagnosticians recording the presence or absence of a diagnosis exceeds 0.7. Pearson correlations between ratings of number of depressive symptoms is 0.9 or greater.
Test–retest reliability When a condition truly does not change over time the observer gives the same result. A client who has Antisocial Personality Disorder or Alzheimer’s disease continues to meet diagnostic criteria at each monthly review. A client who has a mood disorder is correctly diagnosed by a general physician and the diagnosis is confirmed one week later by a consultant psychiatrist. Cohen’s kappa between diagnoses at first and second evaluations by one diagnostician exceed 0.7. Pearson correlations between scores generated by a psychometrician ratings of mood one week apart correlate 0.9 or greater.
Internal consistency Individual criteria that go to make a diagnosis are all closely related to the presence or absence of a diagnosis. All people who present with an individual symptom of depression, score higher on a scale to measure depression and vice versa. All item-total (minus item) point biserial correlations exceed 0.3 and Cronbach’s alpha exceeds 0.6 for a newly developed scale to measure depression. Behavioural items that do not correlate with presence of depression are dropped from a new scale.
Face validity Content appears to measure what the test purports to measure. Members of a committee all agree that the proposed items on a test appear to measure psychosis. Not applicable.
Content validity The instrument samples all the relevant domains of the construct it purports to measure. (Usually relates to tests of achievement.) A measure of depression samples cognitive, affective, and vegetative symptoms of depression. Not applicable.
Criterion validity Scores on test predict some other relevant aspect of behaviour. Scores on a measure of depression are correlated with a measure of social skills. Pearson correlations between the predictor and the criterion exceed 0.7.
Construct validity The degree to which a test score measures the construct it purports to measure. People with an anxiety disorder diagnosis respond better to anxiolytic than an antidepressant medication. People with a mood disorder diagnosis respond better to an antidepressant than an anxiolytic medication. Inferred from a wide range of information about a measure.

disorders and the need for a comprehensive assessment. A classification system must be acceptable to a wide range of audiences with different needs and interests and must be quick, efficient and useful. For example, a classification system should rapidly guide clinical decision-making and lead to better client outcome than would have occurred without the classification system.

Criteria for ID

ID is defined in a variety of legal, professional and research contexts. Most criteria for ID usually refer to (a) significantly below average intelligence, (b) deficits in adaptive behaviours that (c) occur during the developmental period. These criteria have been operationalized in a number of ways. The intelligence quotient (IQ) cut-off was previously operationalized as 85, and more recently as up to 75, 70 and 69. Further, the application of these cut-offs varies dramatically depending upon the test used, the validity of the sample used to standardize the test, the years since the test was standardized (Flynn, 1984, 1985, 1987; Lynn et al., 1987) and attention to cultural and linguistic variables in the assessment. The developmental period has been defined as 18, 21 and 22 years. Measures of adaptive behaviour may also suffer from psychometric and conceptual inadequacies similar to measures of intelligence.

   There is also evidence accumulating that practitioners and services are highly susceptible to social trends in diagnosis. In the United States the number of children classified as having ID has reduced dramatically and the use of other diagnoses such as specific learning disabilities, developmental disabilities, autism and speech and language delays have greatly increased (US Department of Education, 2002). Children who do in fact meet the psychometric criteria for having ID are knowingly and erroneously classified as have specific learning disabilities unless they have maladaptive behaviour or fail to learn to read or write (Keogh et al., 1998; MacMillan et al., 1996, 1998). MacMillan has argued that special education practitioners are not concerned with the accuracy of labels. Rather, as the client fails to obtain the needed services with the current diagnosis, a new diagnosis is obtained to gain access to additional resources. Thus, practitioners, unconcerned about the niceties of diagnostic criteria, but very concerned about providing services to clients, use diagnostic labels as heuristic devices to access progressively more services for clients in need. In contrast to the changes in the USA, in England the numbers of full-time children in segregated special education fell from 90 000 to only 86 777 and the number of segregated schools fell from 1171 to 1088 from 1997 to 2003 (Department of Education and Skills, 2003: Table 15). We might ask how did the USA lose so many children with ID and the United Kingdom lose so few?

   These findings suggest that diagnosis of ID is continually changing. This reflects developing social practices and service ideologies, changing test characteristics, population changes and changes in incidence and case finding.

Criteria for mental disorders

The ICD-10 defines mental disorders as

. . . clinically significant conditions characterised by alterations in thinking, mood (emotions) or behaviour associated with personal distress and/or impaired functioning. Mental and behavioural disorders are not just variations within the range of “normal”, but are clearly abnormal or pathological phenomena. One incidence of abnormal behaviour or a short period of abnormal mood does not, of itself, signify the presence of a mental or behavioural disorder . . . such abnormalities must be sustained or recurring and they must result in some personal distress or impaired functioning. . . . [and] are also characterized by specific symptoms and signs, and usually follow a more or less predictable natural course, unless interventions are made.

The DSM-Ⅳ contains a broadly similar definition. Both definitions invoke a number of similar concepts, such as changes from pre-morbid functioning, personal distress, patterns of symptoms and recognizable courses of illnesses and exclusion of culturally acceptable deviations in behaviour.

   The application of the concept of a mental disorder to people with ID presents several challenges. First, the diagnosis must be made in a person whose behaviour is already restricted and unusual compared to the general population because they have ID. Hence, a decision must be made that a further change in behaviour has occurred beyond that which was already present prior to the onset of the alleged mental disorder. The difficulty of this decision is reflected in a number of potential problems such as intellectual distortion, psychosocial masking, cognitive disintegration, baseline exaggeration (Sovner, 1986), diagnostic overshadowing, and other confusions regarding the difficulty in accurately recognising psychoses (Hurley, 1996), (see Table 1.3).

   A second problem is that making a diagnosis of a mental disorder involves many subjective judgments including judgments concerning the clinical significance of the change in behaviour, the certainty that other causes for the change, such as provocative practices by carers or undetected physical illnesses, have accurately been ruled out, the accuracy of third-party reports, the tolerance of third parties for deviant behaviour, and judgments over the cultural appropriateness of behaviour.

   A third problem is that mental health referrals typically are initiated by the client themselves. However, in the case of referrals from people with ID, most referrals are initiated by distressed carers, rather than distressed clients, usually in response to externalizing maladaptive behaviours, such as aggression, self-injury and tantrums. Indeed, these challenging behaviours, rather than typical mental disorders, are often the reason for the prescription of psychotropic medication in this population (Holden & Gitelson, 2004; Singh et al., 1997). Hence, internalizing disorders, such as anxiety and depression, may be missed as they do not cause distress for carers, and clinicians may receive pressure to make an inappropriate diagnosis in order to justify the use of psychotropic medication to treat a challenging behaviour.

   A fourth problem relates to the presentation of clients with chronic and severe challenging behaviours who take multiple psychotropic medications and who continue to present significant problems. Often such clients do not meet standardized diagnostic criteria, take multiple psychotropic medications, have long but incomplete and contradictory records, and perhaps do not receive adequate behavioural services. Management of such cases is further complicated by the distress and frustration of carers. Disentangling such diagnostic problems is extremely difficult.

Tabl 1.3 Some common problems in making psychiatric diagnoses in people with intellectual disabilities.

Phenomenon Definition Example

Intellectual distortion (Sovner, 1986) Concrete thinking and impaired communication result in poor communication about their own experience. Client describes self as ‘scared’ instead of ‘mad’ because of poor verbal skills.
Psychosocial masking (Sovner, 1986) Impoverished social skills and life experiences result in unsophisticated presentation of a disorder or misdiagnosis of unusual behaviour as a psychiatric disorder. Giggling and silliness is misdiagnosed as psychosis.
Cognitive disintegration (Sovner, 1986) Bizarre behaviour is presented in response to minor stressors that could be misdiagnosed as a psychiatric disorder. A client is highly disruptive and complains a lot after a preferred staff member leaves, but is diagnosed with schizophrenia.
Baseline exaggeration (Sovner, 1986) Prior to the onset of a disorder there are high levels of unusual behaviours, making it difficult to recognize the onset of a new disorder. A person who already had poor social skills and was withdrawn becomes more so and begins to experience other signs and symptoms of depression. This is missed because staff reports are inaccurate and staff turn-over means that no-one is aware of the overall change in the person’s functioning.
Misdiagnosis of developmentally appropriate phenomenon (Hurley, 1996) Developmentally appropriate behaviours that are unusual for the client’s chronological age are misdiagnosed as a psychiatric disorder. Solitary play, talking to oneself and imaginary friends are taken as evidence of psychosis.
Passing (Edgerton, 1967) People with ID learn to cover up disability and pass for normal. Unusual personal experiences are not reported or are ascribed to physical problems.
Diagnostic overshadowing Unusual behaviour is erroneously ascribed to ID, rather than a true mental disorder. Poor social skills and withdrawal are ascribed to ID rather than a psychosis.

printer iconPrinter friendly version AddThis