Intended for healthcare professionals


Psychiatric problems in children with hemiplegia: cross sectional epidemiological survey

BMJ 1996; 312 doi: (Published 27 April 1996) Cite this as: BMJ 1996;312:1065
  1. Robert Goodman, reader in brain and behavioural medicinea,
  2. Philip Graham, emeritus professor of child psychiatryb
  1. a Department of Child and Adolescent Psychiatry, Institute of Psychiatry, London SE5 8AF
  2. b Behavioural Sciences Unit, Institute of Child Health, London WC1N 1EH
  1. Correspondence to: Dr Goodman.
  • Accepted 24 January 1996


Objective: To examine the prevalence and predictors of psychiatric problems in children with hemiplegia.

Design: Cross sectional questionnaire survey of an epidemiological sample with individual assessments of a representative subgroup. The questionnaire survey was repeated on school age subjects four years later.

Subjects: 428 hemiplegic children aged 2 1/2-16 years, of whom 149 (aged 6-10 years) were individually assessed.

Main outcome measures: Psychiatric symptom scores and the occurrence of psychiatric disorder.

Results: Psychiatric disorders affected 61% (95% confidence interval 53% to 69%) of subjects as judged by individual assessments and 54% (49% to 59%) and 42% (37% to 47%) as judged from parent and teacher questionnaires, respectively. Few affected children had been in contact with child mental health services. The strongest consistent predictor of psychiatric problems was intelligence quotient (IQ), which was highly correlated with an index of neurological severity; age, sex, and laterality of lesion had little or no predictive power.

Conclusion: Though most hemiplegic children have considerable emotional or behavioural difficulties, these psychological complications commonly go unrecognised or untreated. Comprehensive health provision for children with chronic neurodevelopmental disorders such as hemiplegia should be psychologically as well as physically oriented.

Key messages

  • They are almost equally common accompani- ments of left and right hemiplegias

  • They are best predicted by IQ, possibly as a marker for underlying neurobiological abnormalities

  • These difficulties often go untreated, com- pounding the child's other difficulties


Previous clinical and epidemiological studies have shown that children with chronic cerebral disorders such as cerebral palsy have a substantially increased rate of emotional and behavioural difficulties—an increase far greater than that seen in chronic non-cerebral disorders that result in comparable disability and social impact.1 2 3 About one in every 200 children in the general population has a psychiatric disorder in association with unequivocal brain disorders (primarily cerebral palsy, epilepsy, and severe mental retardation).1 In many instances, the psychiatric problems result in more handicap and distress for the child and family than the physical or cognitive disabilities. A better understanding of brain-behaviour links may lead to improved treatment or prevention strategies.

Childhood hemiplegia may provide a particularly useful model for studying brain-behaviour links in childhood. Thus hemiplegia affects up to one child in 1000 and provides the opportunity to examine whether psychiatric consequences vary with the laterality of lesion or the age at onset (which ranges from the prenatal period to later childhood). As most affected children are of normal intelligence and attend mainstream schools, it is possible to examine psychiatric problems that are not secondary to intellectual impairment or segregated schooling. Finally, the relatively minor motor disability does not bar the use of ordinary psychiatric assessment techniques. When assessing hyperactivity, for example, it is no harder to ask about overactivity and fidgetiness in a hemiplegic child than in any ordinary child, whereas it would make little sense to ask the same questions about a child with athetoid cerebral palsy who was restricted to a wheelchair.

Subjects and methods SUBJECTS

The London Hemiplegia Register used multiple sources to ascertain a representative sample of 458 London children with hemiplegia (plus three hemiplegic children who lived just outside the Greater London boundary).4 The present study involved the 428 children who were aged 2 1/2 to 16 years at the time of first assessment.


Questionnaire measures of psychiatric caseness were used for the entire age range,5 6 7 8 with detailed individual psychiatric assessments being carried out on a representative subsample of the 6 to 10 year olds. Questionnaire measures of psychopathology corresponded well with comparable measures derived from individual assessments.8 In the initial cross sectional survey, questionnaires completed by parents were obtained for 90% (386/428) of the sample,5 8 and questionnaires completed by teachers (or other preschool professionals) were obtained for 89% (381/428),6 7 with at least one questionnaire being obtained in all 428 cases. Because cross sectional data cannot be used to distinguish between age and cohort effects, additional longitudinal data are presented from a follow up of the same sample an average of four years later using the same parent and teacher questionnaires. A total of 328 children were eligible for follow up, being aged between 2 1/2 and 12 years at the time of the initial survey and therefore still of compulsory school age four years later; parent questionnaires were obtained for 84% (276/328) of the sample, and teacher questionnaires were obtained for 85% (278/328), with at least one of these being obtained for 90% (296/328).

A subgroup of 149 children aged between 6 and 10 were individually assessed an average of six months after the original questionnaire survey; they were representative of hemiplegic children of their age in this study as judged from demographic, medical, cognitive, and behavioural measures.4 The individual psychiatric evaluation drew on three standardised measures: a semistructured interview with parents,9 a teacher questionnaire,7 and an interview with the child.10 Scores for conduct problems, hyperactivity, and emotional symptoms were derived from interviews with the parents and summed to generate a total parent based symptom score. All items on the teacher questionnaire were summed to generate a total teacher based symptom score. Information from all sources was combined to categorise each child as having a psychiatric disorder or not—a distinction that could be made reliably and validly on our population with identified cases having a level of symptoms comparable with those of attenders of child psychiatric clinics.11 For ease of comparison our symptom scores were scaled such that the mean was 1.0 for children who were free from psychiatric disorder.

The intelligence quotient (IQ) of the individually assessed children was judged from a full version of the revised Wechsler intelligence scale for children,12 except in the case of 19 children with severe cognitive impairments whose IQs were calculated from their mental ages.13 All 149 children were neurologically examined. Five neurological variables—relating to degree of hemiparesis, presence and type of seizure disorder, presence of any signs of bilateral involvement, head circumference, and time of onset—were combined to generate a neurological severity index.13 A family adversity index was generated by summing four z transformed items: a questionnaire rating of maternal psychiatric morbidity (malaise inventory)14 and three standardised interview based ratings covering parental criticism of the child, lack of parental warmth for the child, and poor parental child management skills,9 prorating the total score when data were missing for one item.


For the individually assessed children the significance of the group differences presented in table 2 was established with χ2 tests, t tests, and analysis of variance. Significant univariate predictors of psychiatric problems were entered into stepwise multiple regression analyses to assess independent predictive power, with the standardised regression coefficients for significant effects presented to facilitate comparisons of effect sizes.


Among the children aged less than 5 the rate of psychiatric caseness was 51% (40/78) as judged by parent questionnaire5 and 35% (23/66) as judged from a questionnaire completed by preschool professionals.6 Among the 5 to 16 year olds, the comparable rates were 55% (168/308) as judged by parent questionnaire8 and 43% (137/315) as judged by teacher questionnaire.7 When the two age bands were combined the rates were 54% (95% confidence interval 49% to 59%) and 42% (37% to 47%) as judged from parents and teachers (or other preschool professionals), respectively. As can be seen from figures 1 and 2 there were no consistent trends with age. There was some evidence, however, that different cohorts within our sample maintained fairly constant rates of parent based caseness over time—for example, the children who were 5 or 6 at the time of the initial survey had a particularly high rate of disorder initially and also when reassessed aged 9 or 10 at the time of the follow up.

Fig 1
Fig 1

Degree of emotional or behavioural problems in children with hemiplegia as seen by parents according to child's age

Fig 2
Fig 2

Degree of emotional or behavioural problems in children with hemiplegia as seen by teachers according to child's age

Psychiatric disorders were present in 91 (61%; 95% confidence interval 53% to 69%) of the 149 individually assessed children with hemiplegia. In 75 instances the psychiatric disorder resulted in substantial social impairment for the children; though not socially impaired, the 16 remaining children either had emotional symptoms that resulted in substantial distress or conduct problems that resulted in substantial disruption for others. The proportion of children who had ever been in contact with child mental health services was 18% (16/91) for the children with current psychiatric disorders and 2% (1/58) for children currently free from psychiatric disorders (continuity adjusted χ2=6.6, df=1, P<0.01).

Conduct, emotional, and hyperactivity disorders predominated (table 1). Conduct disorders were typically dominated by irritability and oppositionality rather than by deliberately antisocial behaviours such as stealing or bullying. Emotional disorders usually involved anxieties and fears; depression was uncommon. Pervasive and situational hyperactivity are reported separately because, although pervasive hyperactivity is recognised as a disorder by both the major contemporary systems of classification,15 16 situational hyperactivity is currently recognised as a disorder only by the American system.16

Table 1

Types of psychiatric disorder in 149 intensively studied children

View this table:

Psychiatric problems were associated with greater neurological severity, lower IQ, special schooling, and family adversity (table 2). No measure of psychopathology was significantly influenced by whether the birth occurred prematurely or at term.

Table 2

Psychiatric disorders and symptoms in children aged 6-10 years with hemiplegia

View this table:

Table 3 shows variables that independently predicted psychiatric problems in multivariate analyses. The most consistent predictor of psychiatric problems was IQ. Once this had been taken into account the neurological severity was not a significant predictor of psychiatric problems. As the correlation between IQ and the index of neurological severity was -0.73 it is not surprising that when IQ was omitted from the regression analyses, neurological severity took the place of IQ as the most consistent predictor of psychiatric problems.

Table 3

Significant predictors in multivariate analyses (6-10 year olds), values are standardised regression coefficients (95% confidence intervals) for significant effects

View this table:

Though some subgroups had particularly high rates of psychiatric problems, it is noteworthy that even children with a “good prognosis” were still at substantial psychiatric risk. For example, the rate of psychiatric disorder was still 39% among the 28 individually assessed children who attended a normal school, had an IQ over 90 (mean 108), had never had a seizure, and had a mild hemiparesis with no hint of bilateral involvement.


Psychiatric problems were extremely common in our large and representative sample of children with hemiplegia. Over half of the children studied individually had psychiatric disorders. The measure of psychiatric disorder that we used selected children with a level of psychiatric problems equivalent to that of children attending a psychiatric clinic11; the rate of comparable psychiatric disorders in the general child population assessed by using similar methods is unlikely to exceed 15%.17 18 It is particularly striking that the rate of psychiatric disorder was 39% even among the most mildly affected children who were of normal intelligence and attended mainstream schools. This increased psychiatric risk is far greater than that seen in chronic non-cerebral disorders that are at least as disabling and stigmatising,1 2 3 strongly suggesting direct and powerful brain-behaviour links.


In this sample, greater neurological severity was a powerful predictor of lower IQ. Once this effect had been allowed for, however, neurological severity was not an independent predictor of psychiatric problems. An association between lower IQ and more psychiatric problems has been a consistent finding of studies of neurologically impaired and normal children.1 3 19 20 21 Consequently, the simplest explanation for the brain-behaviour link in hemiplegia is that neurological damage leads to lower IQ, and this in turn is the primary reason for the children's greater psychiatric vulnerability (fig 3). Against this explanation, however, is the fact that the rate of psychiatric disorder in our sample was substantially higher than would be found in controls matched for IQ. For example, a 29% rate of psychiatric disorder among hemiplegic children with above average intelligence is probably at least three times higher than would be expected among similarly intelligent children in general; and the same applies to the rate of 57% among hemiplegic children with normal but below average intelligence. This leads us to conclude that IQ is primarily a marker for underlying neurobiological factors that influence psychopathology rather than being the main risk factor in itself (fig 4).

Fig 3
Fig 3

One causal model of psychiatric problems in children with hemiplegia

Fig 4
Fig 4

Alternative causal model

There was no compelling evidence that age, sex, or laterality influenced liability to psychopathology. The rate of psychopathology was high from the age of 2 1/2 to 16. The apparent influence of age in the intensively studied sample seemed, as judged from the longitudinal data, more likely to be due to random variation between successive cohorts than to a true age effect. The lack of an effect of sex is in line with the results of previous studies of children with congenital and acquired brain lesions,1 22 but in contrast with the greater psychiatric vulnerability of boys in the general population.14 17 Though the rate of psychiatric disorder in the intensively studied sample was higher in children with right hemiplegia, the trends were in the opposite direction for the two other measures of psychopathology. The lack of a consistent effect of laterality is in line with previous findings from large and representative samples of children.23

Psychiatric problems in hemiplegic children were associated with adverse family factors such as parental depression or a high level of parental criticism of the child. In the absence of longitudinal and intervention studies there is no justification for concluding that adverse family factors caused the child's problems. It is equally plausible that the child's difficulties have resulted in parental criticism or ill health; or that the problems in parents and children have a shared origin.


The psychiatric disorders that accompany childhood hemiplegia have rarely been identified and treated. This is regrettable as in our experience these disorders are often helped by specific treatments ranging from cognitive therapy to medication and from family work to individual counselling. In addition, parents are often relieved to hear that their child's problems are similar to those of many other hemiplegic children; the consequent reduction in the parents' sense of guilt and isolation can be very therapeutic for the entire family. Support and information is also available from Hemi-Help, the parents' organisation that grew out of the present study.

We thank Carole Yude, Bob Adak, Suzanne Pemberton, and Judith Elliott for their parts in the project and the children, parents, and teachers who gave so willingly of their time. The telephone number of Hemi-Help for further literature and help is 0181 672 3179.


  • Funding The Wellcome Trust, Scope, and the Medical Research Council.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
View Abstract