While the survey did incorporate the British Picture Vocabulary Scales, these data were not contained in the file deposited in the ESRC Data Archive and were not available upon request from the ONS.
Prevalence of psychiatric disorders in children and adolescents with and without intellectual disability†
Article first published online: 10 JAN 2003
Journal of Intellectual Disability Research
Volume 47, Issue 1, pages 51–58, January 2003
How to Cite
Emerson, E. (2003), Prevalence of psychiatric disorders in children and adolescents with and without intellectual disability. Journal of Intellectual Disability Research, 47: 51–58. doi: 10.1046/j.1365-2788.2003.00464.x
- Issue published online: 10 JAN 2003
- Article first published online: 10 JAN 2003
- psychiatric disorders
Background There have been very few population-based studies of the prevalence of psychiatric disorders among children and adolescents with and without intellectual disability (ID).
Methods Secondary analysis of the 1999 Office for National Statistics survey of the Mental Health of Children and Adolescents in Great Britain was performed. This survey collected information on a multistage, stratified, random sample of 10 438 children between 5 and 15 years of age across 475 postcode sectors in England, Scotland and Wales.
Results The prevalence of any diagnosed ICD-10 disorder, conduct disorder, anxiety disorder, hyperkinesis and pervasive developmental disorders were significantly greater among children with ID than among their non-ID peers. There were no statistically significant differences between children with and without ID with regard to the prevalence of depressive disorders, eating disorders or psychosis. Factors associated with an increased risk of psychopathology among children and adolescents with ID included age, gender, social deprivation, family composition, number of potentially stressful life events, the mental health of the child's primary carer, family functioning and child management practices.
Conclusions Children and adolescents with ID are at significantly increased risk of certain forms of psychiatric disorder. Careful consideration of the social and economic adversity facing such families will be necessary to ensure that support services are responsive to both the needs of child as well as the needs of the family in which they are living.
Over the past few decades increased attention has been paid to identifying and responding to the mental health needs of children and adolescents with intellectual disability (ID) (Lindsey 1997; RCP 1998; Tonge 1999; Dykens 2000). The few prevalence studies which have been undertaken have suggested that children and adolescents with ID are at significantly increased risk of psychiatric disorder when compared with their non-ID peers (Dykens 2000). For example, recent studies have identified significant levels of psychopathology among: 40% of a population-based sample of 454 Australian children with ID (Einfeld & Tonge 1996a, b); 38% of a sample of 143 children with severe ID in Lothian, Scotland (Hoare et al. 1998); 50% of a sample of 123 children attending schools for children with severe ID in Southampton and the New Forest area of England (Cormack et al. 2000); and 31% of a sample of children attending special schools in Cape Town, South Africa (Molteno et al. 2001). These rates are comparable to those reported by Rutter et al. (1976) in the classic Isle of Wight study, and recent Norwegian (Stromme & Diseth 2000) and Finnish (Linna et al. 1999) population-based studies.
However, all but the Finnish study (Linna et al. 1999) may be criticized on the grounds of unrepresentative (e.g. Hoare et al. 1998; Cormack et al. 2000) or geographically restricted (e.g. Rutter et al. 1976; Stromme & Diseth 2000) sampling, and all but the Norwegian study (Stromme & Diseth 2000) may be faulted on the grounds of employing screening tools for ‘caseness’ which fail to give diagnoses within an accepted classificatory system.
The aim of the present paper is to present data on the prevalence of ICD-10 psychiatric disorders in children and adolescents with ID. The Social Survey Division of the UK Office for National Statistics (ONS) collected these data in England, Scotland and Wales in 1999 (Meltzer et al. 2000). Data from this survey on problem behaviours reported by the children themselves and the mental health of mothers of children with ID will be presented in associated papers (Emerson, in press).
Materials and methods
Secondary analysis was undertaken on the 1999 ONS survey of the Mental Health of Children and Adolescents in Great Britain (Meltzer et al. 2000). Relevant data files (ONSSSD 1999) were obtained from the ESRC (Economical Social Research Council) Data Archive.
Full details of the design, instrumentation and sampling procedure are presented in Meltzer et al. (2000). Briefly, the ONS survey collected information on a multistage, stratified, random sample of 10 438 children between 5 and 15 years of age (83% of the target sample of 12 529 eligible children). The sample was stratified by age and sex within 475 postcode sectors in England, Scotland and Wales. Information was collected by interview with the child's primary carer (in 94% of cases the child's mother) and, wherever possible for children aged 11 years or over, with the child themselves. Information was also collected by postal questionnaire from the child's teacher.
The presence of psychiatric disorders among children and adolescents was identified through the use of the Development and Well Being Assessment (DAWBA; Goodman et al. 2000). The DAWBA consists of: two structured interviews (one undertaken with the child's primary carer, and the other undertaken, for children aged 11 years or more, with the children themselves); a questionnaire used with the child's teacher; and a computer-assisted diagnostic rating system that provides diagnoses against DSM-IV and ICD-10 criteria (Meltzer et al. 2000). Goodman et al. (2000) has reported that the DAWBA: (1) discriminated well between a sample of 491 children drawn from population-based child benefit registers and 39 children drawn from those attending child psychiatric clinics; (2) shows good convergent validity with the Strengths and Difficulties Questionnaire (SDQ; Goodman 1997); (3) predicts contact with health services and prognosis over a 4–6-month period; and (4) shows acceptable levels of agreement with diagnoses derived from case note review.
A number of measures of family functioning were also included within the survey. These included: the General Health Questionnaire (GHQ; Goldberg & Williams 1988), to estimate psychiatric morbidity among parental informants; the General Functioning Scale of the MacMaster Family Activity Device (FAD; Miller et al. 1985; Byles et al. 1988), to identify ‘unhealthy’ patterns of family functioning; a nine-item scale exploring parental reward strategies and punishment regimes (Meltzer et al. 2000); and a 10-item measure of stressful life events (Meltzer et al. 2000).
Definition of intellectual disability
The ONS survey did not measure whether children had an ID.1 As a result, it was necessary to combine survey items to operationally identify children who were likely to have ID. A child was defined as having ID if either: (1) their parents reported that their child had ‘learning difficulties’ and that they had been seriously concerned about their child's language development during the first 3 years of their life; or (2) the child attended a school for children with ‘learning difficulties’. Children were subsequently excluded from this group if information collected from teachers in the ONS survey indicated that any of the following three conditions were met: (1) the child was reported by teachers in mainstream schools (or special schools for children who did not have ‘learning difficulties’) to be of average ability in either reading, mathematics or spelling; (2) their developmental age (as estimated by their class teacher) was greater than 80% of their chronological age; (3) they were at least 7 years of age and had not progressed beyond Stage 2 in the process of the identification of Special Educational Needs (DoE 1994).
Prevalence of intellectual disability
Use of the operational definition described above identified a total of 264 children as having ID. This is equivalent to an overall prevalence of ID in the total sample of 2.6%. Out of the 264 children, 98 (37%) were in mainstream educational provision, and 166 (63%) attended a school or unit for children with ID. The percentage of children attending special schools or units increased with age from 48% of 5–7-year-olds to 78% of 13–15-year-olds.
The validity of this classification was evaluated by examining the extent to which the prevalence of ID varied as a result of age, gender, social class (the Registrar General's classification of the occupation of the identified head of household) and ethnicity. The results of these analyses are presented in Table 1.
|Social class of head of household:|
|(I & II) professional and managerial occupations||74||1.9%|
|(III) skilled occupations||93||2.4%|
|(IV & V) partly skilled and unskilled occupations||78||3.8%|
Prevalence of psychiatric disorders
The DAWBA (Goodman et al. 2000) produces diagnoses against both ICD-10 and DSM-IV criteria. Preliminary analysis indicated that virtually identical results were obtained by the two classification systems. Therefore, only results using ICD-10 criteria will be reported. The percentage of children for whom diagnosable conditions were identified is given in Table 2.
|Psychiatric disorder||Prevalence||Odds ratio*||95% CI|
|Any diagnosed disorder||39.0%||8.1%||7.3||5.6–9.4|
|Any emotional disorder||9.5%||4.1%||2.4||1.6–3.7|
|Any anxiety disorder:||8.7%||3.6%||2.5||1.6–4.0|
|post-traumatic stress disorder||0.8%||0.2%||4.8||1.1–20.9|
|generalized anxiety disorder||1.5%||0.6%||2.6||1.0–7.4|
|other anxiety disorder||1.9%||1.2%||1.7||0.7–4.1|
|Any conduct disorder:||25.0%||4.2%||7.6||5.7–10.2|
|oppositional defiant disorder||13.3%||2.3%||6.5||4.5–9.5|
|unsocialized conduct disorder||3.4%||0.3%||10.4||4.9–21.9|
|socialized conduct disorder||3.8%||0.8%||4.8||2.5–9.3|
|other conduct disorder||4.9%||1.6%||8.1||4.4–14.9|
|Pervasive developmental disorder||7.6%||0.1%||74.7||35.4–157.7|
|Number of diagnosed disorders|
|Three or more||3%||1%||9.4|
A series of bivariate analyses was undertaken to identify associations between child characteristics (i.e. age, gender and diagnosis of cerebral palsy), family characteristics (i.e. social class, household income, family composition and family functioning), stressful life events and the presence of psychiatric disorders (i.e. any diagnosed disorder, any emotional disorder, any anxiety related disorder, depressive disorder, any conduct disorder, pervasive developmental disorder and hyperkinesis) among children with ID.
Older children (aged 11 years and over) were more likely to be identified as having a depressive disorder [odds ratio (OR) = 10.5, χ2 = 4.6, d.f. = 1, P < 0.05]. Younger children were more likely to be identified as having a pervasive developmental disorder (OR = 2.9, χ2 = 4.2 d.f. = 1, P < 0.05). Boys were more likely to be identified as having a diagnosed disorder (OR = 1.8, χ2 = 4.0, d.f. = 1, P < 0.05) and a conduct disorder (OR = 2.2, χ2 = 5.0, d.f. = 1, P < 0.05). There was no significant association between diagnosis of cerebral palsy and ascertained prevalence of any psychiatric disorder.
Children living in a family where the head of household was classified as being in social classes III–V were more likely to be identified as having a conduct disorder (OR = 2.7, χ2 = 7.2, d.f. = 1, P < 0.01), hyperkinesis (OR = 9.5, χ2 = 7.0, d.f. = 1, P < 0.01) and less likely to be identified as having a pervasive developmental disorder (OR = 5.6, χ2 = 14.7, d.f. = 1, P < 0.001). Children living in lower-income households were more likely to have a diagnosed disorder (Mann–Whitney z = 2.25, P < 0.05), an emotional disorder (Mann–Whitney z = 1.99, P < 0.05), depression (Mann–Whitney z = 2.10, P < 0.05) and a conduct disorder (Mann–Whitney z = 3.16, P < 0.01), but were less likely to have a pervasive developmental disorder (Mann–Whitney z = 2.84, P < 0.05). Children of lone parents were more likely to have a diagnosed disorder (OR = 2.3, χ2 = 9.3, d.f. = 1, P < 0.01), an emotional disorder (OR = 3.0, χ2 = 7.0, d.f. = 1, P < 0.01), an anxiety disorder (OR = 2.4, χ2 = 4.2, d.f. = 1, P < 0.05), depression (OR = 7.5, χ2 = 4.1, d.f. = 1, P < 0.05) and a conduct disorder (OR = 2.9, χ2 = 13.5, d.f. = 1, P < 0.001).
Children whose primary carer had a higher GHQ score were more likely to have a diagnosed disorder (Mann–Whitney z = 3.10, P < 0.01), an emotional disorder (Mann–Whitney z = 2.46, P < 0.05), an anxiety disorder (Mann–Whitney z = 2.18, P < 0.05), a conduct disorder (Mann–Whitney z = 2.58, P < 0.01) and hyperkinesis (Mann–Whitney z = 2.38, P < 0.05). Children whose family was characterized by more ‘unhealthy’ functioning on the FAD were more likely to have a diagnosed disorder (Mann–Whitney z = 2.47, P < 0.05), an emotional disorder (Mann–Whitney z = 2.63, P < 0.01) and an anxiety disorder (Mann–Whitney z = 2.10, P < 0.05). Children whose primary carer reported more frequent use of punitive strategies were more likely to have a diagnosed disorder (Mann–Whitney z = 3.90, P < 0.001) and a conduct disorder (Mann–Whitney z = 4.97, P < 0.001).
Children who had experienced a greater number of potentially stressful life events were more likely to have a diagnosed disorder (Mann–Whitney z = 2.20, P < 0.05), an emotional disorder (Mann–Whitney z = 3.55, P < 0.001), an anxiety disorder (Mann–Whitney z = 3.14, P < 0.01) and a conduct disorder (Mann–Whitney z = 3.16, P < 0.01).
There were no differences in the ascertained prevalence of psychiatric disorders between children attending special schools for children with ID and children identified as having ID who were in mainstream schools.
Finally, a series of multivariate analyses (forward stepwise logistic regression: P-variable entry < 0.05, P-variable removal < 0.1) was undertaken to identify the unique contribution of the variables identified through the bivariate analyses to children having any diagnosis and having a diagnosis of conduct disorder. The small percentages of children diagnosed precluded the use of such methods with other disorders (Tabachnick & Fidell 1996). The results of these analyses are presented in Table 3.
|Variable||P-value||Corrected odds ratio|
|Any ICD-10 diagnosis*|
|(Higher) parental GHQ score||0.003||2.3|
|(Two or fewer) adults living in the household||0.037||2.3|
|(Greater) use of punishment||0.008||2.1|
|ICD-10 diagnosis of conduct disorder (any form)†|
|Parent not a home owner||0.004||2.7|
|(Greater) use of punishment||0.003||2.7|
|(Higher) parental GHQ score||0.020||2.1|
These results identify moderate associations between the assessed mental health status of the child and: (1) the mental health status of parental informant; (2) the use of punitive child management strategies; and (3) the social circumstances of the family. Corrected ORs assess the contribution of one variable while controlling for the effects of other variables in the equation. These may be combined multiplicatively to assess the impact of combinations of variables. Thus, for example, the odds of a child having an ICD-10 diagnosis are increased by 4.8 (i.e. 2.3 × 2.1) if the parent informant screens positive for mental health problems her- or himself, and reports greater use of punitive child management strategies (when compared with the children of parents who show none of these characteristics).
Secondary analysis of the 1999 ONS survey of the Mental Health of Children and Adolescents in Great Britain (Meltzer et al. 2000) suggests that: (1) rates for conduct disorders, anxiety disorders, attention deficit hyperactivity disorder/hyperkinesis and pervasive developmental disorders are higher among children with ID than among their non-ID peers; (2) there were no statistically significant differences between children with and without ID with regard to rates of depressive disorders, eating disorders or psychosis; and (3) factors associated with increased risk of psychopathology include age, gender, social deprivation, family composition, child management practices, family functioning and the mental health of the child's primary carer.
This discussion will address three issues: (1) the validity of the operational definition of ID employed in the study and the DAWBA for assessing psychiatric disorders in children and adolescents with ID; (2) the reported prevalence of psychiatric disorders among children and adolescents with ID; and (3) the association between psychiatric disorders and child and family characteristics, and the implication of these results for the provision of supports to children and families.
The confidence that may be placed in the results of this study is dependent on the validity of the operational definition of ID employed, and the validity of the DAWBA for assessing psychiatric disorders in children and adolescents with ID. The operational definition of ID used in the present study was based on information provided by the child's primary carer, type of school attended and teacher reports. Three strands of evidence support the validity of this operational definition.
First, the majority of children (63% overall and 78% of 13–15-year-olds) were reported to be attending special schools or units for children with ID. This is broadly consistent with UK educational policy and practice. Secondly, the overall prevalence rate reported in the present study (2.6%) is consistent with prevalence rates reported by studies which have included both ‘severe’ and ‘mild’ ID (McLaren & Bryson 1987; Roeleveld et al. 1997). Finally, the observed increase in prevalence of ID among boys, older children and children brought up in conditions of social disadvantage (see Table 1) is also consistent with the results of previous research (McLaren & Bryson 1987; Roeleveld et al. 1997). However, the observed association between ID and ethnicity is somewhat unexpected. Previous research has suggested an increased prevalence of ID among Asian children (Emerson et al. 1997; Kerr 2001). The discrepancy between these findings and the results of the present study may reflect the interviewing procedures adopted in the ONS survey (face-to-face interviews with parents were only undertaken in English). For parents whose first language was not English, translated text versions of the SDQ (Goodman 1997) were used. Unfortunately, this questionnaire did not contain items which could be used to determine whether the child had ID.
The validity of the DAWBA for assessing psychiatric disorders in children and adolescents with ID is more difficult to assess. As noted above, the DAWBA has been shown to discriminate well between community-based and clinic-based samples of children, and in the clinic-based sample, to produce acceptable levels of agreement between diagnoses derived from the DAWBA and those derived from analysis of case notes (Goodman et al. 2000). However, the validity of the DAWBA has not been specifically established for children and adolescents with ID. While the close correspondence between the results of the present study and those of previous research (see below) lends some support for the use of the DAWBA with this group of children and adolescents, the well-established difficulty of assessing psychopathology in this group (cf. Dykens 2000) suggests that the specific prevalence rates reported above should be treated with some degree of caution.
The overall prevalence of psychiatric disorder reported in the present study (39%) is consistent with the results of previous population-based studies (e.g. Rutter et al. 1976; Einfeld & Tonge 1996a, b; Hoare et al. 1998; Linna et al. 1999; Stromme & Diseth 2000; Molteno et al. 2001). Data presented in the present paper also provide very strong evidence to suggest that presence of ID must be considered a highly significant risk factor for the development of some specific forms of psychiatric disorder (see also Rutter et al. 1976; Linna et al. 1999). Overall, children with ID were over seven times more likely to have a diagnosed psychiatric disorder than their non-ID peers.
Unfortunately, it is not possible to compare the prevalence rates for specific disorders obtained within the present study with the results of previous research. The one previous study to have reported ICD-10 diagnoses only did so for the child's primary presenting problem (Stromme & Diseth 2000). Failure to assess the nature of comorbid disorders (present in 50% of children in the present study) may account for the discrepancies between the results of the two studies.
Finally, the results suggested a number of links between child characteristics, family characteristics, life events and the presence of psychiatric disorders. The association between male gender, age and psychiatric disorders is consistent with previous research (Einfeld & Tonge 1996b; Cormack et al. 2000; Stromme & Diseth 2000). The present study also reported a number of associations between indicators of social disadvantage (i.e. lower social class and reduced household income), being cared for by a lone parent, ‘unhealthy’ patterns of family functioning, punitive child management practices, carer mental distress and an increased prevalence of psychiatric disorders. These associations are consistent with the existing literature on the mental health of children and adolescents in general (e.g. Meltzer et al. 2000; McMunn et al. 2001).
Given the cross-sectional nature of the ONS study, it cannot be assumed that such factors have a causal relationship to either the onset or persistence of psychiatric disorders. Thus, for example, psychiatric disorders shown by the child may lead to the break up of existing relationships (Emerson, in press). Alternatively, such effects may be bi-directional (e.g. a damaging and unhelpful reliance on punitive child management practices may represent the desperate attempt of parents to exercise some control over a child with conduct disorder). Finally, any observed effects may reflect the operation of unknown third variables (e.g. coping strategies which may vary as a function of social class or a shared genetic propensity to develop psychiatric disorder).
Nevertheless, the results are of importance in characterizing the social situation of children with ID and psychiatric disorders. Families supporting children with ID are generally more likely to be subject to social and material disadvantage when compared to families supporting ‘typically developing’ children (Beresford 1995; Fujiura 1998; Emerson, in press). The present data indicate that families supporting children with ID and psychiatric disorders are even more disadvantaged. Careful consideration of the social and economic adversity facing such families will be necessary to ensure that support services are responsive not only to the needs of child, but also to the needs of the family in which they are living (e.g. Albin et al. 1996).
- 1996) Contextual fit for behavior support plans: a model of ‘goodness of fit’. In: Positive Behavioral Support: Including People with Difficult Behavior in the Community (eds L. K.Koegel, R. L.Koegel & G.Dunlap), pp. 81–98. Paul H. Brookes, Baltimore, MD. , , & (
- 1995) Expert Opinions: A National Survey of Parents Caring for a Severely Disabled Child. The Policy Press, Bristol. (
- 1988) Ontario Child Health Study: reliability and validity of the General Functioning Scale of the MacMaster Family Assessment Device. Family Process 30, 116–23. , , & (
- 2000) Behavioural and emotional difficulties in students attending schools for children and adolescents with severe intellectual disability. Journal of Intellectual Disability Research 44, 124–9. , & (
- Department for Education (DoE) (1994) Code of Practice on the Identification and Assessment of Special Educational Needs. Department for Education, London.
- 2000) Psychopathology in children with intellectual disability. Journal of Child Psychology and Psychiatry 41, 407–17. (
- 1996a) Population prevalence of psychopathology in children and adolescents with intellectual disability. I. Rationale and methods. Journal of Intellectual Disability Research 40, 91–8. & (
- 1996b) Population prevalence of psychopathology in children and adolescents with intellectual disability. I. Epidemiological findings. Journal of Intellectual Disability Research 40, 99–109. & (
- Mothers of children with intellectual disabilities: social and economic situation, mental health status and self-assessed social and psychological impact of child's difficulties. Journal of Intellectual Disability Research. (In press)
- 1997) Is there an increased prevalence of severe learning disabilities amongst British Asians? Ethnicity and Health 2, 317–21. , , , , & (
- 1998) Demography of family households. American Journal on Mental Retardation 103, 225–35. (
- 1988) A Users Guide to the General Health Questionnaire. NFER-Nelson, Windsor. & (
- 1997) The Strengths and Difficulties Questionnaire: a research note. Journal of Child Psychology and Psychiatry 38, 581–6. (
- 2000) The Development and Well Being Assessment: description and initial validation of an integrated assessment of child and adolescent psychopathology. Journal of Child Psychology and Psychiatry and Allied Disciplines 41, 645–56. , , , & (
- 1998) A community survey of children with severe intellectual disability and their families: psychological adjustment, carer distress and the effect of respite care. Journal of Intellectual Disability Research 42, 218–27. , , & (
- 2001) Assessing the needs of learning disabled young people with additional disabilities: implications for planning services. Journal of Learning Disabilities 5, 157–74. (
- 1997) Emotional, behavioural and psychiatric disorders in children. In: Seminars in the Psychiatry of Learning Disabilities (ed. O.Russell), pp. 81–104 . Gaskell Press, London. (
- 1999) Psychiatric symptoms in children with intellectual disability. European Child and Adolescent Psychiatry 8, 77–82. , , , & (
- 1987) Review of recent epidemiological studies of mental retardation: prevalence, associated disorders, and etiology. American Journal of Mental Retardation 92, 243–54. & (
- 2001) Children's emotional and behavioural well-being and the family environment: findings from the Health Survey for England. Social Science and Medicine 53, 423–40. , , , & (
- 2000) Mental Health of Children and Adolescents in Great Britain. The Stationery Office, London. , , & (
- 1985) The MacMaster Family Assessment Device: reliability and validity. Journal of Marital and Family Therapy 11, 345–56. , , & (
- 2001) Behavioural and emotional problems in children with intellectual disability attending special schools in Cape Town, South Africa. Journal of Intellectual Disability Research 45, 515–20. , , & (
- Office for National Statistics Social Survey Division (ONSSSD) (1999) Mental Health of Children and Adolescents in Great Britain, 1999[computer file]. SN: 4227, 13 December 2000. The Data Archive, Colchester.
- 1997) The prevalence of mental retardation: a critical review of recent literature. Developmental Medicine and Child Neurology 39, 125–32. , & (
- Royal College of Psychiatrists (RCP) (1998) Psychiatric Services for Children and Adolescents with a Learning Disability. Council Report (CR 70). Royal College of Psychiatrists, London.
- 1976) Research report: Isle of Wight Studies 1964–1974. Psychological Medicine 6, 313–32. , , , & (
- 2000) Prevalence of psychiatric diagnoses in children with mental retardation: data from a population-based study. Developmental Medicine and Child Neurology 42, 266–70. & (
- 1996) Using Multivariate Statistics, 3rd edn. Harper Collins, New York, NY. & (
- 1999) Psychopathology of children with developmental disabilities. In: Psychiatric and Behavioural Disorders in Developmental Disabilities and Mental Retardation (ed. N.Bouras), pp. 157–74 . Cambridge University Press, Cambridge. (
Accepted 17 July 2002