University of Oxford Section of Child and Adolescent Psychiatry, Park Hospital for Children, Headington, Oxford, UK

Transcription

1 J. Sleep Res. (1998) 7, Behavioural treatment for sleep problems in children with severe learning disabilities and challenging daytime behaviour: Effect on sleep patterns of mother and child LUCI WIGGS andgregory STORES University of Oxford Section of Child and Adolescent Psychiatry, Park Hospital for Children, Headington, Oxford, UK Accepted in revised form 30 March 1998; received 17 July 1997 SUMMARY Children with sleep problems present serious management problems to their parents. Such children are also more likely to have additional problems, behavioural disturbance being particularly common. This randomized controlled trial of behavioural interventions for the children s sleep problems was conducted to explore the efficacy and mechanisms of treatment in children with the most extreme forms of problems: severe learning disabilities, severe sleep problems and severe daytime challenging behaviour. Fifteen index families received behavioural advice for the child s sleep problem and were compared with 15 matched controls who received no such advice. Repeat assessments of the children s and mothers sleep were made by parental report as well as actometry. Objective changes in the children s sleep quality and quantity were not seen after treatment. However, mothers in the treatment group reported improvements in the children s sleep problems and had an increased sleeping time themselves following treatment. The results indicate that sleep problems can be successfully treated in this group of children, although the mechanisms of treatment may not be as direct as supposed. This has implications for understanding of sleep problems in children with learning disabilities and also for clinical practice, when considering ways of offering help to these highly challenged families. KEYWORDS actometry, behavioural treatment, challenging behaviour, sleep INTRODUCTION Behavioural interventions have been used successfully to treat the sleep problems of children with learning disabilities Consistently high rates of sleep problems are reported in (e.g. Quine 1992). However, there have been no investigations children with learning disabilities (Bartlett et al. 1985; Quine of children s sleep which have focused solely on children with 1991; Wiggs and Stores 1996). Work has indicated that children challenging behaviour. Previous behavioural intervention with severe learning disabilities and sleep problems are more studies, both in the general population and in children with likely to have more severe and more coexistent challenging learning disabilities, have selected subjects with sleep problems, behaviour than children without sleep problems (Wiggs and of which only a minority have had behaviour disorders at the Stores 1996). Challenging behaviour refers to behaviour which onset (e.g. Seymour et al. 1983; Sanders et al. 1984; Richman includes injury to the child or others, damage or destruction et al. 1985; Quine 1992; Minde et al. 1994). Therefore, it is not to the environment or social disruption affecting the lives of known whether it is possible to successfully treat sleep problems the child or others (Emerson et al. 1987; Mansell 1994). by behavioural methods in children who also have challenging behaviour and what the effects of successful treatment might be upon the child and parent. Correspondence: Dr Luci Wiggs, University of Oxford Section of Child and Adolescent Psychiatry, Park Hospital for Children, Old Road, Given the effects of sleep disturbance on children s daytime Headington, Oxford OX3 7LQ, UK. Tel: ; Fax: functioning, the potential effects of disturbed sleep are ; lucinda.wiggs@psychiatry.oxford.ac.uk important for both clinical practice and research. Interventions 1998 European Sleep Research Society 119

2 120 L. Wiggs and G. Stores to reduce a child s sleep problem may be beneficial since the intervention began for no given reason) and 15 in the matched family stresses and child behaviour problems associated with control group. In total, all 51 families were asked to participate childhood sleep problems, (or at least a component of the and 30 accepted. Of the 20 who did not, 10 considered them), may arise as a direct result of impaired sleep. themselves too busy, seven felt that their child s sleep had The methodology of previous work has not allowed this improved since completing the original questionnaire (and aspect of treatment to be investigated because earlier studies therefore they no longer fulfilled the inclusion criteria) and relied almost exclusively on parental reports of the children s three declined with no reason given. The remaining children sleep. Only one study has investigated the objective effects were matched for sex and for the duration of their sleep of intervention on behavioural sleep (Sadeh 1994), but the problem to within 6 months. Since the children s ages were differences between the subjects in Sadeh s study to those in not necessarily a reflection of their developmental level, the the present investigation (e.g. in terms of age, duration of sleep duration of their sleep problems, rather than their age, was problem, developmental level, family factors etc), makes it considered a more important variable for which to match. difficult to generalize from Sadeh s work. However, it does highlight an important distinction which needs to be made when evaluating any treatment trials for sleep problems, namely Definition of sleep problems between those interventions which lead to objective Sleep problems of settling, night waking and early waking were improvements in the child s sleep and those treatments which defined as follows according to standardized criteria (Richman do not objectively affect the child s sleep but which teach the and Graham 1971): child to no longer disturb parents during any wakings. This 1 Settling problems were defined as severe if they occurred distinction is important as both could contribute to the three or more times per week, if the child took more than improvements in family and child functioning associated with one hour to settle and fall asleep, and if the parents were successful behavioural interventions; the former by directly disturbed during this time. reversing the effects of the sleep disruption and leading to 2 Night waking was defined as severe if it occurred three or improvements in daytime functioning, the latter by reversing more times per week, and if the child woke for more than only the effects of the parents sleep disruption which may a few minutes and disturbed the parents or went into the make them better able to cope with problems, should they parents room or bed. arise. Objective recording of both the child s and the parents 3 Severe early waking was defined as waking before sleep would help to clarify this. hours three or more times per week. It is important to try to explore these mechanisms of intervention as they have practical and theoretical implications. Clinically, it may be possible to develop or refine treatments Definition of challenging behaviour to focus on the most salient mechanisms; theoretically, such The Aberrant Behavior Checklist (Aman and Singh 1986) was work may contribute further to understanding of the function used to assess the following types of challenging behaviour: self of sleep and the effects of sleep disruption. injury, aggression, screaming, temper tantrums, noncompliance This study aimed to answer the following basic questions by and impulsivity. Scoring details for the assessment of means of a randomized controlled trial using behavioural challenging behaviour are given elsewhere (Wiggs and Stores interventions to treat the children s sleep problems. 1996). 1 Can behavioural treatment for sleep problems be used successfully in children with severe learning disabilities and Procedure daytime challenging behaviour? To avoid the nature of the intervention being discussed between 2 If so, what are the effects of treatment on the sleep patterns the parents in the control and intervention groups, it was of the child and the mother? decided that the children s schools, rather than the families would be randomly allocated to either the treatment or the METHOD control group. Families were approached by a letter explaining Subjects the study and the randomization procedure to either immediate treatment or a waiting list for treatment when the therapist Children were recruited from respondents to a questionnaire was available. A follow-up telephone call was made a few days survey of special schools conducted by the authors (Wiggs and later. Sleep diaries were completed by parents and were used Stores 1996). Inclusion criteria for this study (as defined below) as a clinical tool to aid design and monitoring of treatment. were that the child had both a severe sleep problem and at Graphical presentation of each subject s diary data for visual least one form of daytime challenging behaviour. Of the 209 inspection is not reported in this paper for reasons of brevity, families who completed a questionnaire (43% response rate) especially since the diary data was positively correlated with 51 children fulfilled these criteria. questionnaire data about the children s sleep (the composite Thirty-one children were recruited into the study: 16 in the sleep index described later). intervention group (one family dropped out of the study before Six visits were made to the family home for each subject.

3 Behavioural treatment and sleep patterns 121 Apart from the behavioural intervention and progress The behavioural programmes telephone calls, the same procedure was followed and The following is a summary of the intended therapeutic process. assessments performed for both the treatment and control Details are available from the authors on request. groups. Progress telephone calls, the prolonged social support element of the intervention, was not given to the control group 1 Functional analysis of problem as this alone has been suggested as a major component of the 2 Identification of parents aims of treatment intervention itself (e.g. Richman et al. 1985; Scott and Richards 3 Discussion of possible mechanisms maintaining settling and 1990). waking problems 4 Discussion of therapeutic techniques i.e. extinction, graded Visit one extinction procedures such as checking and gradual withdrawal, stimulus control procedures and positive This was an introductory visit for the family to ask any reinforcement. The positive and negative aspects of each questions, to meet the researcher and for the researcher to technique, and the practical application to their situation, meet the child. The questionnaires for the mother (and/or were discussed so that parents could make an informed father if appropriate) were explained to the family and left for choice about whether or not they would be able to carry completion. An activity monitor for both the child and the out the technique mother was given to the family along with instructions for use 5 Identification and anticipation of particular problems with and left with a monitor diary. intervention Visit two 6 Identification of target (or targets) for the first stage. Two or three days later a semi-structured interview took place so that a detailed sleep history was taken. This interview lasted between an hour and a half and two and a half hours. For the intervention group, a tailored behavioural programme was discussed (see below for a description) and agreed with the parents. The questionnaires and monitors were collected. It was requested that if possible both parents should be present for this visit. This was possible for nine families in the treatment group and seven in the control group. For the remainder, only mother was present. After visit two, the parents in the treatment group were sent a written outline of the agreed behavioural programme. Parents in the treatment group were telephoned at least weekly to monitor progress, encourage, discuss any problems and amend the programme as necessary. They were also told that they could telephone the researcher at any time, although this rarely happened. Visit three One month later the activity monitors and questionnaires were delivered to the family again. Visit four Within three days, the questionnaires and monitors were collected. Telephone contact continued with the treatment group as necessary. Visit five Two months later, the activity monitors and questionnaires were delivered to the family again. Visit six Two or three days later, the questionnaires and monitors were collected. Assessments 1 Composite sleep index This index of the children s sleep problems was calculated from parental questionnaire information obtained by means of a modification of the Simonds and Parraga Sleep Questionnaire (Simonds and Parraga 1982). Settling and night waking were scored in terms of frequency and duration and early waking and sleeping in the parents bed for frequency only. Frequency: problems occurring once or twice a week were given a score of one and problems occurring more than several times a week were given a score of two. Duration: settling problems lasting up to one hour were given a score of one; over one hour was scored as two. Night wakings lasting a few minutes were scored as one; two if they lasted longer than that. The possible scores, therefore, ranged from zero to 12. For consistency with previous studies, and to allow comparisons with other reports, this was taken to be the measure of treatment efficacy. 2 Body movements Activity monitors (Gaehwiler Electronics, Hombrechtikon, Switzerland) were worn by the child and the mother usually for three nights (five children and one mother only wore the monitors for two nights). The activity monitors used were wrist-watch size movement sensors, worn on the non-dominant wrist between getting into bed and getting out of bed the next morning. An amount of movement was calculated for every 30 s during the recording period. Parents completed monitor diaries to note the times at which the monitors were put on the wrist and taken off, times got into bed and got out of bed, approximate sleep onset and wake-up times and any events of importance. The mean scores for each recording period were examined

4 122 L. Wiggs and G. Stores and the following variables were calculated: sleep period (time varied in duration) rather than the beginning of treatment. from sleep onset to waking), activity score (mean value of Such an approach biases the results in the direction of a movement during sleep period), movement index (percentage favourable outcome and, especially when comparing results of sleep period spent moving) and fragmentation index with a control group, a more standardized approach to the (percentage of immobile phases during sleep period which were timing of the post-intervention investigation is necessary. 30 s duration or less) (Aubert Tulkens et al. 1987). They were Two post-intervention time periods were chosen for use in calculated as follows: this study to allow for the individual variability one might expect to see in improvement in this population. The decision Sleep period: time from sleep onset to time woke-up. Sleep of when these measurements should take place was guided onset and time woke-up were defined by parental monitor by the rate of change seen in the study by Quine (1992) but diaries. The activity monitor data was visually inspected to increased slightly to take account of the more severe learning identify any nights where there was a gross disparity (i.e. disabilities and daytime behaviour problems experienced by greater than approximately 30 min) between diary records and the present group of children. activity data, so that these nights could be excluded from the Because there has been no previous work looking at an current analyses. Following this procedure, one night (of three) untreated control group of children with sleep problems, it was excluded for one subject (treatment) at baseline and one was of interest to examine not only any differences between night (of three) was excluded for another subject (control) at the groups but also within the groups. This is particularly baseline and visit six. In each recording, the activity monitor relevant since Szydler and Bell (19920 report improvements data suggested that the child may have woken much earlier in maternal well-being, even when intervention is than indicated by the parents diary. unsuccessful. It could be that merely focusing more on the Activity score: the sum of the epoch scores divided by the child s sleep (which the control group are also forced to do) total number of epochs in the sleep period. would be therapeutic. Movement index: the number of 30 s epochs with a value 2 Quine (1991) has noted that there is likely to be a complicated greater than zero (i.e. with movement), divided by the total interaction between several factors which determines any number of 30 s epochs in the sleep period multiplied by 100. change in the family dynamics. This is perhaps especially true Fragmentation index: the number of discrete 30 s epochs with when examining a group of children who also have various a value of zero (i.e. with no movement) divided by the total forms of disability with associated problems and features of number of immobile phases of any duration multiplied by 100. possible relevance. This being so, the mechanisms involved in any change within the child or mother are likely to be less 3 Epilepsy straightforward and conventional hypothesis testing may be Seizure type was assessed by means of a structured interview too simplistic. (Reutens et al. 1992) comprising of 26 main questions, including The statistical procedures used are described in each section. nested questions when a symptom was present. An agreement of Parametric tests were used when exploration of the data 0.76 (Cohen s K statistic) between physician and questionnaire suggested they were suitable (Lilliefors and Levene tests for diagnoses has been shown by the authors (Reutens et al. normality of scores and homogeneity of variance respectively). 1992). The diagnoses were made by a neuropsychiatrist with All analyses of variance tests (ANOVA) were of a two-way considerable clinical and research experience in the epilepsy 2x3 mixed design. Independent variables were the subjects field (GS). Seizure frequency was assessed by pages in the sleep group (treatment or control) (between) and the time of study diary on which parents could note down the time and a (baseline, visit 4 and visit 6) (within). Averaged tests of description of any seizures. significance are used for all ANOVA tests unless otherwise specified. Where Mauchly Sphericity tests were significant, the Statistical analysis Hotellings multivariate test of significance was used instead. This is indicated in the results by the symbol. Post hoc For the following reasons, it was decided to make no explicit contrasts were made using Scheffe s test to compare, for both hypotheses but to analyse the data in a more exploratory between and within groups, the change between baseline and manner. Therefore all significance levels given are two-tailed. visit four, baseline and visit six, and between visit four and six. Confidence limits (95%) for each contrast are indicated by the 1 All possible comparisons (i.e. changes over each study period letters CL. for both groups and within groups) are of interest for reasons described below. The rate and timing of any changes in the child or their RESULTS parent are important. Previous studies (e.g. Quine 1992; Richman et al. 1985) took post-intervention measurements The subjects at different times across subjects, when personal targets had Both groups contained 9 males and 6 females. The mean age been reached. Also, Minde et al. (1994) anchored their postintervention of the treatment group was 8.21 years (s.d. 2.7), and the mean measurements to the end of treatment (which duration of their sleep problems was estimated to be 6.21 years

5 Behavioural treatment and sleep patterns 123 Sleep problem Control group (n=15) Treatment group (n=15) Settling 5 5 Settling and night waking 4 2 Settling, night waking and sleeping 2 3 in parental bed Night waking 1 0 Settling and sleeping in parental 1 1 bed Night waking and early waking 1 1 Night waking and sleeping in 0 2 parental bed Settling, night waking and early 0 1 waking Table 1 Number of children in each group showing the various sleep problems and combinations of sleep problems (s.d. 2.44). The children s basic diagnoses were stated as follows: child s sleep to have been significantly improved by treatment, unknown (6); unknown (+autism) (4); Down syndrome (1); within one month of its commencement. These levels of meningitis (1); microcephaly (1); cerebral palsy (1); CHARGE improvement were maintained two months later. The control association (1). Five children had current uncontrolled epilepsy group s sleep remained the same throughout the study period. characterized by the following types of seizure: absence (1), The mean composite sleep index scores of the two groups at tonic clonic (2), absence and tonic clonic (1), atonic (1). A the different time points can be seen in Table 2. further two had been seizure free for between 2 and 4 years. The mean age of the control group was years (s.d. 3.81) and the mean duration of their sleep problems was Body movements estimated to be 6.91 years (s.d. 4.07). The diagnoses of the The mean scores for each activity variable for both groups of children s disabilities were as follows: unknown (4); unknown children can be seen in Table 3. (+autism) (3); cerebral palsy (2); Down syndrome (3); agenesis A similar pattern was seen for each of the variables: there of the corpus callosum (1); Sanfillipo syndrome (1); Ring 15 were no overall differences between the groups, nor differences chromosome disorder (1). Six children had current uncontrolled over time between the groups (statistical data available from epilepsy with seizures as follows: absence (1), tonic clonic (2), the authors on request) but there was an effect of time on each tonic (1), tonic clonic and complex partial (1), tonic, complex variable (sleep period F(2,56) = 5.30, P = 0.008; activity score partial and absence (1). One had been seizure free for 3 years. F(2,27) = 3.64, P = 0.040; movement index F(2,27) = 6.72, For both groups combined, the children s age and the P = 0.004; fragmentation index F(2,56) = 12.84, P<0.000). duration of their sleep problems were positively correlated Both groups showed an increased sleep period between (Pearson s R= , P<0.001). baseline and visit 4 (t = 2.45, P = 0.020, 95% CL = to The types of sleep problems suffered by children in the two 0.462) and baseline and visit 6 (t = 2.89, P = 0.007, 95% CL = groups can be seen in Table 1. There was no significant to 0.440) but no change between visits 4 and 6 (t = 0.07, difference between the composite sleep scores of the two groups P = 0.941, 95% CL = to 0.1). at baseline (t = 0.60, P = 0.554, df = 28). The activity scores of both groups showed a decrease in activity between baseline and visit 4 (t = 2.74, P = 0.010, The children s sleep 95% CL = to 0.182) and baseline and visit 6 (t = 2.66, P = 0.012, 95% CL = to 0.167) but no Composite sleep index change between visits 4 and 6 (t = 0.99, P = 0.994, 95% CL = ANOVA was used to investigate the composite sleep scores to 0.135). There was a significant overall difference between the groups Movement indexes also reduced for both groups between (F(1,23) = 14.2, P = 0.001), an effect of time of study (F(2, baseline and visit 4 (t = 3.54, P = 0.001, 95% CL = )=13.12, P<0.001) and a significant interaction between to 0.975) and baseline and visit 6 (t = 2.73, P = 0.010, the subjects grouping and the time of study (F(2,46)=5.03, 95% CL = 3.45 to 0.494) but did not change between P<0.011). Contrasts confirmed that there was a difference visits 4 and 6 (t = 1.12, P = 0.271, 95% CL = to between baseline and visit 4 for the treatment group and not 0.954). for the control group (t = 2.35, P = 0.027, 95% CL = Fragmentation indexes showed a reduction between baseline to 3.07) and between baseline and visit 6 (t = 3.50, P = 0.001, and visit 4 (t = 5.19, P = 0.002, 95% CL = 5.17 to 2.24) 95% CL = to 2.94), but no further changes for either and this reduction from baseline was sustained at visit 6 (t = group between visits 4 and 6 (t = 3.14, P = 0.756, 95% CL = 2.16, P = 0.039, 95% CL = 3.05 to 0.082) but there 1.18 to 1.61). This suggested that parents reported their was a significant increase in fragmentation, for both groups,

6 124 L. Wiggs and G. Stores Table 2 Mean composite sleep index scores (and standard deviations) of the control and treatment groups at the three study periods Baseline (B) Visit Four (V4) Visit Six (V6) Control (n=15) 7.23 (2.26) 6.62 (1.89) 6.29 (2.70) Treatment (n=15) 6.73 (2.31) 3.79 (1.89) 2.96 (2.24) Table 3 Children: treatment and control groups mean scores for each sleep score (and standard deviations), before and after treatment Treatment Control B V4 V6 B V4 V6 Sleep period (h) (0.3) (0.6) (0.6) (0.9) (0.9) (0.7) Activity score (1.6) (0.5) (0.9) (2.9) (0.8) (0.4) Movement index (%) (5.1) (2.7) (4.6) (6.4) (3.0) (3.2) Fragmentation index (%) (3.7) (4.1) (4.8) (5.0) (5.7) (3.1) Table 4 Mothers mean scores in the treatment and control groups for each sleep score (and standard deviations), before and after treatment Treatment Control B V4 V6 B V4 V6 Sleep period (h) (0.9) (0.7) (0.7) (0.8) (1.4) (1.1) Activity score (1.2) (0.5) (1.3) (0.5) (1.7) (0.6) Movement index (%) (3.7) (4.2) (4.6) (6.0) (2.4) (2.7) Fragmentation index (%) (4.5) (3.9) (5.0) (3.9) (3.2) (6.6) between visit 4 and 6 (t = 2.80, P = 0.009, 95% CL = Correlation between mothers and children s sleep to 3.70). Pearson s Correlational Tests showed significant associations between the sleep periods of the mothers and children (R = The mothers sleep 0.416, P = 0.022) but none of the other variables were Body movements correlated. The association between the sleep periods of mothers and their children continued for the control group at visit 4 The mean scores for each activity variable for both groups of (R = 0.815, P<0.000) and visit 6 (R = 0.507, P = 0.054) but mothers can be seen in Table 4. for the treatment group this association was not present at There were no differences between the groups, but the visit 4 (R = 0.261, P = 0.346) or visit 6 (R = 0.117, P = movement index showed a significant reduction over time for 0.677). both groups (F(2,27) = 5.32, P = 0.011) between baseline and visit 4 (t = 2.87, P = 0.007, 95% CL = 3.07 to 0.512). The mothers sleep period showed significant differences over DISCUSSION time between the groups (F(2,56) = 3.75, P = 0.030). Mothers The results of study indicate that it is possible to successfully in the treatment group showed an increased sleep period treat the sleep problems of children with severe learning between baseline and visit 4 (t = 2.56, P = 0.016, 95% CL = disabilities and current challenging daytime behaviour, using to 1.10) which was sustained at visit 6 (t = 2.16, P = behavioural intervention techniques. 0.03, 95% CL = 0.02 to 0.78). There were no further increases Within one month of starting treatment (and still at three between visits 4 and 6 (t = 0.84, P = 0.40, 95% CL = 0.72 months), the composite sleep index scores of the treatment to 0.31). The non-significant statistical data is available from group were lower than for the control group. Objective the authors on request. recording of the children s sleep however, failed to demonstrate

7 Behavioural treatment and sleep patterns 125 changes in their sleep which were related to the treatment. This seizure types and, given the somewhat subjective nature of suggests that treatment may affect the children s signalling of seizure categorization, such an approach would also have been their awake state to their parents rather than affecting the spuriously detailed. Therefore, information about the children s children s sleep quality or quantity. epilepsy is merely provided for the reader to note that the However, the objective variables under study were mainly presence of epilepsy does not appear to preclude successful concerned with restlessness. There is high inter-subject treatment. variability in levels of restlessness, which is thought to reflect It might have been preferable to have used a sample of individual differences, rather than sleep disturbances (Sadeh et children with specific forms of disability (rather than this mixed al. 1995; American Sleep Disorders Association 1995). If this group) and to have divided the children according to those is the case, changes following treatment might be minimal. with settling, those with night waking and those with early There were, interestingly, improvements in the children s waking sleep problems. Whether this is practical, when trying sleep which were seen in both groups over time; an increased to investigate a further area of interest (i.e. challenging sleep period, decreases in magnitude, frequency and pattern of behaviour), is doubtful. The present sample of 30 children with movement during sleep. Why these variables should be affected sleep problems and challenging behaviour were selected from in both groups, and whether the mechanisms underlying the an original group of 486 children. Unwillingness to take part increase is the same in both groups, is unclear. Non-specific and ineligibility due to not fulfilling all the necessary criteria effects of participation in the study (e.g. altered expectations reduced the number available for study. Such figures illustrate and perhaps increased awareness of the children s sleep pattern) the difficulties inherent in investigating children with learning may have led the parents in the control group to alter their disabilities and in trying to select children who are matched management of their child s sleep, in a way similar to the on a number of key features. These figures also indicate that treatment group, resulting in a longer sleep period. However, the study sample is a highly self-selected group which may not mothers in the control group still reported high composite sleep be representative of this particular population as a whole, index scores. If changes in management were an explanation, although the proportion of children with the various sleep detailed behaviour analysis of the parents would be required problems and with challenging behaviours are similar to those to see why effects were not generalized to the children s sleep found in other studies (see Wiggs and Stores 1996 for problems per se. The approximate nature of the measurement discussion), indicating that the sample was not unduly of sleep period must also be considered and this finding needs representative in these respects. There was an age difference to be replicated using more precise assessment tools. between the groups because the groups were matched for the The changes in movement during sleep seem less easy to duration of their sleep problems rather than for age and this explain in terms of parental handling. Developmental changes is a factor which needs to be considered as possibly affecting are unlikely given that all significant changes were confined to outcome. one time (between baseline and visit 4) although the children The design of the intervention trial also needs to be were all at various developmental levels and ages. The test- considered. The use of a control group has highlighted some retest reliability of the activity monitors, as an explanation of interesting and important issues which might be pursued by changes, does not seem to be an issue (Sadeh et al. 1991). further research. However, it is also important to recognize Treatment for the child s sleep problem appears to have had that a control group in a learning disabled population is likely more beneficial effects on the mothers sleep pattern than that to be different from the treatment group in many respects of the children. The mothers in the treatment group showed other than simply whether or not they received treatment. an increased sleep period within one month of starting Despite these limitations, the findings in this study have a treatment and this was sustained two months later. Such a number of implications for the clinical care of children with result suggests that the relationship between successful severe learning disabilities. intervention and improved child sleep (and maternal wellbeing) is not as direct as might be imagined. The mechanisms 1 It is possible to successfully and quickly resolve their sleep underlying treatment appear to be many-faceted and it is only problems, even if these problems are long standing. Contrary by further experimental manipulation of each component in to widely held belief, the high prevalence of sleep problems in isolation and objective monitoring of the families sleep that this population should not lead parents (or health professionals) the absolute contribution of each can be assessed. to unquestioningly conclude that sleep problems are an It would have been interesting to examine the effects of inevitable and untreatable part of disability. epilepsy upon treatment outcome. The parents were asked to 2 The behavioural approach to treatment used for children record and describe their child s seizures in the diary record without learning disabilities is acceptable for use in this but only one family completed the seizure pages of the diary population. This suggests that the sleep patterns of the learning and even then the records were only made for the baseline disabled group were substantially learned behaviours rather period. It is acknowledged to be difficult to obtain an accurate than due to unalterable organic abnormalities. measure of seizure frequency (Cramer 1993). The numbers of 3 The presence of severe daytime challenging behaviour and children with the various different types of seizure were not epilepsy are not barriers to successful treatment. Clinicians large enough to allow investigations of the effects of different should feel able to tackle one of the children s problems (i.e.

8 126 L. Wiggs and G. Stores sleep) even whilst there are other recurrent problems of a effects of treatment on their daytime behavior. J. Am. Acad. Child Adol. Psychiatry, 1994, 33: different nature. Quine, L. Sleep problems in children with severe mental handicap. J. 4 Caution must be applied to these results. The families who Ment. Defic. Res., 1991, 35: took part in this study were all willing to undertake intervention Quine, L. Helping parents to manage children s sleep disturbance. An for their child s sleep problem, an act which in itself can be intervention trial using health professionals. In: J. Gibbons (Ed) The demanding. There are likely to be families who do not feel Children Act 1989 and Family Support: Principles into Practice. HMSO, London, 1992: able to use behavioural treatments and alternative approaches Reutens, D.C., Howell, A.R., Gebert, K.E. and Berkovic, S.F. must also be pursued. Validation of a questionnaire for clinical seizure diagnosis. Epilepsia, 1992, 33: Richman, N., Douglas, J., Hunt, H., Lansdown, R. and Levere, R. ACKNOWLEDGEMENTS Behavioural methods in the treatment of sleep disorders a pilot study. J. Child Psychol. Psychiat., 1985, 26: The authors would like to thank the families who took part Richman, N. and Graham, P.J. A behaviour screening questionnaire in this study and also The Mental Health Foundation who for use with three year old children. J. Child Psychol. Psychiat., provided financial support to LW for the study. 1971, 12: Sadeh, A. Assessment of intervention for infant night waking: parental reports and activity-based home monitoring. J. Consult. Clin. REFERENCES Psychology, 1994, 62: Sadeh, A., Hauri, P., Kripke, D.F. and Lavie, P. The role of actigraphy Aman, M.G. and Singh, N.N. Aberrant Behavior Checklist Manual. in the evaluation of sleep disorders. Sleep, 1995, 18: Slosson Educational Publications Inc., New York, Sadeh, A., Lavie, P., Scher, A., Tirosh, E. and Epstein, R. Actigraphic American Sleep Disorders Association. Practice parameters for the use home-monitoring sleep disturbed and control infants and young of actigraphy in the clinical assessment of sleep disorders. Sleep, children: a new method for pediatric assessment of sleep wake 1995, 18: patterns. Pediatrics, 1991, 87: Aubert Tulkens, G., Culee, C., Harmant-Van Rijckevorsel, K. and Sanders, M.R., Bor, B. and Dadds, M. Modifying bedtime disruptions Rodenstein, D.O. Ambulatory evaluation of sleep disturbance and in children using stimulus control and contingency management therapeutic effects in sleep apnea syndrome by wrist activity techniques. Behav. Psychother., 1984, 12: Scott, G. and Richards, M.P.M. Night waking in infants: effects of monitoring. Am. Rev. Respir. Dis., 1987, 136: providing advice and support for parents. J. Child Psychol. Psychiat., Bartlett, L.B., Rooney, V. and Spedding, S. Nocturnal difficulties in a 1990, 31: population of mentally handicapped children. Br. J. Ment. Subnorm., Seymour, F.W., Bayfield, G., Brock, P. and During, M. Management 1985, 31: of night waking in young children. Aus. J. Fam. Ther., 1983, 4: Cramer, J.A. A clinimetric approach to assessing quality of life in epilepsy. Epilepsia, 1993, 34: (Suppl. 4) S8 S13. Simonds, J. and Parraga, H. Prevalence of sleep disorders and sleep Emerson, E., Barrett, S., Bell, C., Cummings, R., McCool, C., Toogood, behaviors in children and adolescents. J. Am. Acad. Child Psychiatry, A. and Mansell, J. Developing Services for People with Severe 1982, 21: Learning Difficulties and Challenging Behaviours. Institute of Social Szyndler, J. and Bell, G. Are groups for parents of children with sleep and Applied Psychology, Canterbury, problems effective? Health Visit., 1992, 65: Mansell, J. Challenging behaviour: the prospect for change. Br. J. Wiggs, L. and Stores, G. Severe sleep disturbance and daytime Learn. Dis., 1994, 22: 2 5. challenging behaviour in children with severe learning disabilities. Minde, K., Faucon, A. and Falkner, S. Sleep problems in toddlers, J. Intellect. Disabil. Res., 1996, 40: