- Email: [email protected]
The effects of cue control relaxation on adults with severe mental retardation
Research in Developmental
Disabilities,
Vol. 15, No. 6, pp. 425-437.
Copyright
Pergamon
0
1994 Elsevier
Printed in the USA.
1994
Science Ltd
All tights reserved
0891-4222194
$6.00 + 00
0891-4222(94)00017-4
The Effects of Cue Control Relaxation on Adults With Severe Mental Retardation William
R. Lindsay, Mairi Fee, Amanda Michie, and lmelda Heap Strathmartine
Hospital
This study was designed to assess two aspects of behnvioural relaxation training with subjects who have severe mental retardation. Thefirst was whether or not cue words could be linked effectively to training so that they would eventually produce a relaxation effect in the absence of a full relaxation procedure. The second was to determine the effects of relaxation training and subsequent cue control on concentration and attention to an occupational task. Five subjects participated, and individual case designs were used. Cue Control, behavioural relaxation training (BRT), and new therapists were introduced at different times for each subject to ascertain the effects of each variable. Subjects were assessed at baseline and following each training session on a behavioural relaxation scale to judge the direct effects of relaxation training and on the amount of time spent concentrating on an occupational task. BRT produced reductions in rated anxiety and improvements in concentration for all subjects. The cue control words were effective only ajier they had been linked to BRT In most cases, the introduction of a new therapist had no appreciable effect on anxiety or concentration. There was some suggestion that in the final cue only phases concentration was better and more consistent than during the BRTphases of the study.
Psychological treatments designed to alleviate anxiety and agitation may be helpful for people with severe mental retardation. Calamari, Geist, and Shahbazian (1987) found a combined package of treatment, including biofeedback, produced reductions in anxiety in subjects with a range of intellectual deficits. Lindsay, Baty, Michie, and Richardson (1989) found Requests for reprints should be sent to William R. Lindsay, Tayside Area Clinical Department, Stmthmartine Hospital, Dundee, DD3 OPG, Scotland.
425
Psychology
426
W! R. Lindsay et al.
both abbreviated progressive relaxation (APR) and behavioural relaxation training (BRT) to be effective in reducing rated anxiety in people with severe mental retardation. APR is a long-established therapy in which the client is asked to engage in a series of tension and release cycles for each part of the body. The client is taught to understand the difference between tension and relaxation and is subsequently able to adopt a relaxed state spontaneously. BRT was developed by Schilling and Poppen (1983) when APR proved ineffective with learning disabled boys. Originally subjects were asked to imitate unrelaxed and relaxed postures in each part of the body. Therefore, an understanding of internal states of muscular tension and relaxation is not required. Lindsay et al. (1989) found that BRT produced quicker and more substantial improvements than APR. Lindsay and Baty (1989) further simplified BRT by asking clients with severe mental retardation only to imitate unrelaxed postures. They suggested that, in contrast to APR, BRT is not a paradoxical technique requiring tensing of the muscles before relaxation. This leaves BRT less open to misinterpretation as a technique designed to produce excitement rather than relaxation, which they felt was very important for individuals with severe learning disabilities. McPhail and Chamov (1989) compared a group of six subjects receiving 12 sessions of APR (4 sessions per week over 3 weeks) with a control group who were in a story-telling condition. The subjects’ disabilities were reported to range from profound to mild mental retardation, and they were chosen because they displayed disruptive behaviour. As well as being assessed for immediate relaxation effects, they were also rated on 17 categories of disruptive behaviour. A significant and substantial reduction in disruptive behaviour in subjects receiving APR occurred with no changes in the controls. Lindsay, Richardson, and Michie (1989) also studied the generalised effects of relaxation training. Agitated movement, agitated speech, and general anxiety were measured 1 h after the completion of each of 12 treatment sessions and at a 3-week follow-up. Subjects receiving both APR and BRT showed a gradual reduction in behaviours indicating anxiety and agitation, which were maintained at follow-up. Papers by Williams (1990) and Turk and Francis (1990) serve to underscore the effectiveness of such procedures with the client group. The data support the conclusion reached by Calamari et al. (1987) that the results “call into question the sometimes held assumption that developmentally disabled persons are not good candidates for psychological treatments involving techniques other than simple behavioural management procedures” (p. 69). Now that effectiveness has apparently been established, it is important to adapt the methods to prove more useful in the lives of people with mental retardation. Lindsay (199 1) has written that the straightforward implemen-
Cue Control Relaxation
427
tation of powerful methods could be used as a simple (and possibly unethical) management technique to keep certain individuals quiet and undemanding. Serious consideration should be made when embarking on a programme of relaxation training concerning whether it should be used in conjunction with another programme, for example, educational or occupational activities, increased social contact, and so forth. Therefore, relaxation training would be used as an enabling technique to allow individuals to become more accessible to their environment. This article describes an intervention using (a) BRT as an enabling technique to help very agitated individuals to concentrate and (b) cue control to reduce the length of time needed to achieve relaxation. Very little research has been conducted on cue control relaxation for individuals with mental retardation. One very well controlled case study by Wells, Turner, Bellack, and Hersen (1978) did suggest potential value. Using a reversal design, they established a functional relationship between cue controlled relaxation (linked to APR) and psychomotor seizures in a young woman functioning in the borderline range of intelligence (IQ = 71). She was instructed to vocalise the cue word covertly at the onset of each preseizure aura, and this reduced the frequency of seizures. When she stopped using the cue word during the experimental reversal phase, seizure frequency increased to baseline levels. Reinstatement of cue control relaxation procedures reduced seizures once again, and improvements maintained at a 3-month follow-up. This study links cue control to BRT and investigates the effects of this procedure on subjects’ attending to various table-top tasks. METHOD Subjects
Five subjects each diagnosed with severe mental retardation (IQ < 40 as measured by the WATS-R) participated. All had spent most of their lives in a hospital and were currently hospitalised. All subjects scored below the 15th percentile on all scales of the Adaptive Behaviour Scale Part I (American Association on Mental Deficiency, 1975) except physical development. No subject was currently taking antianxiety medication. Subjects were a man, age 42; a woman, age 35; a man, age 48; a woman, age 45; and a man, age 29. Subjects showed a variety of agitation and anxiety symptoms, such as constant pacing; frequent facial, verbal, and bodily mannerisms; perseverative talking and mumbling; frequent shouting and screaming; avoidance of others; tremor; and high-frequency nondirected activity. When approached by others, these behaviours would almost always increase.
428
U! R. Lindsay et al.
Design
A single case design was employed to test the variables in this study. All subjects underwent a baseline (BL) condition in which they were simply asked to sit and relax in a chair and then engage in the task. BL consisted of five assessment sessions. Following this, three subjects went on to a cue only condition (CO) to ascertain the power of the cue words alone. In this condition, subjects were encouraged to sit in the chair for a period of time equivalent to that of subjects in the BRT condition. Two subjects went on to the BRT procedures. The three subjects in the CO condition then went on to the BRT condition. Following BRT, all subjects went on to CO. For all subjects, the cue was introduced (IC) on the sixth session of BRT. A new therapist (NT) replaced the original therapist (OT) at some point for each subject to determine whether or not changes were person-specific or generalised to new individuals. For one subject, the NT was introduced in the first CO condition to determine whether or not a therapist change per se had an effect. For this subject, the OT was reintroduced at the end of training. In this design, generalisation and maintenance are considered. Because the CO conditions do not employ active therapeutic procedures, they can be considered as a naturalistic follow-up and maintenance condition. Procedure
Cues were tied to BRT, which is a technique concentrating on observable states of relaxation in 10 areas of the body. The instructor demonstrates the relaxed states in each area and then helps the person copy the relaxed behaviours. The aim is to reduce muscle activity and help the individual to be able to adopt a relaxed state spontaneously, in any part of the body, culminating in an overall relaxed posture and quiet mood. The person must be seated in a comfortable chair, rather than lying down. He or she is asked first to watch and then to imitate the relaxed behaviour. Simple direct feedback on the accuracy of the copying and manual guidance are given. With extremely agitated subjects, verbal and physical prompting may be necessary. For example, with a subject who moved her hands frequently, a light physical prompt on the forearm would accompany the instructions to keep the arms quiet and still. Training involved demonstrating an overall relaxed and then unrelaxed posture, breathing when unrelaxed and relaxed (quick vs. slow), vocalisation (noisy vs. silent), trunk movements (moving and fidgeting or rocking vs. still), head movements (held tense or moving vs. leaning back and still), eyes (closed tight vs. lightly closed or open and still), lips (pressed tight vs. relaxed), throat (movement/swallowing vs. still), shoulders (hunched vs. sloped and even), arms (moving vs. still), hands (tense fist vs. relaxed and
Cue Control Relaxation
429
open), legs (moving or tense vs. still), feet (moving vs. still). No reinforcement schedules were used at any time throughout the study. All sessions took place at the same time of day for each subject. Sessions occurred daily except for weekends. The verbal cue linked to BRT was “quiet and still.” This was introduced at various stages in each individual design. Following introduction of the cue, the relaxation procedure was gradually shortened and eventually eliminated. Following cue control or BRT, a simple dexterity task was introduced for the subject from a variety of tasks, including colour matching, shape sorting, shape matching, and peg boards. These are generally widely available in establishments for people with mental retardation. A simple attainable task was chosen because the individuals involved had such poor concentration and short attention spans. Subjects were simply asked to put the pieces in the appropriate place, for example, “Put the shapes in their place” or “Put the pegs in the holes.” Therapists were instructed to prompt subjects at no more than 10-s intervals. No reinforcers were given to subjects at any time. If clients became disruptive, they were asked to continue, but if they persisted, the session was ended. Assessment Subjects were video taped after each session. Tapes were later rated on a scale designed to assess degree of anxiety. Raters judged 10 body parts on a 5-point scale from no signs of anxiety or agitation through moderate levels of anxiety and agitation to extreme agitation and anxieo. Each rating point was explained with guidelines on the scale. The items assessed were voice, trunk, head, eyes, lips, throat, shoulders, hands, feet, and a general rating of relaxation/tension. Lindsay et al. (1989) previously reported extremely high reliability between trained observers on the total behavioural anxiety score produced on this test (r = 0.95). Reliability was calculated on 84 sets of scores, each rated by two experienced independent raters, blind to the treatment conditions. This produced a Pearson’s correlation coeffrcient of r = 0.89. Further investigation on the 10 component parts of the scale found interobserver agreement to be high using the formula: number of agreements divided by the number of agreements plus disagreements expressed as a percentage. An agreement was only scored if both observers agreed on the degree of anxiety (O-4) observed in the particular body part being rated. This produced 94% absolute agreement between trained raters. Subjects were also video taped while they were engaging in the 20 min of occupational activity. The amount of time spent “on task” was computed later. On task behaviour was defined as any meaningful manipulation of the occupational materials, for example, lifting a piece, looking at it, and trying to fit it into the board. As a check on reliability of this score, two
430
N! R. Lindsay et al.
independent raters judged 20 tapes, and agreement was found to be 98.4%. The method of calculating reliability in this score was less exact. Only the total time “on task” was used in the calculation, that is, total time for Observer A (scoring the lowest amount of “on task” activity) divided by total time for Observer B, expressed as a percentage. However, reliability is so high (almost 100%) and the clinical changes are so considerable that drawbacks in the method of computation may not be crucial. For BL conditions, subjects were video taped after the instruction to sit and relax; for the BRT condition, they were taped at the end of the relaxation session; in the CO condition, they were taped after the use of the cue words. In the BL and CO conditions, attempts were made to ensure that the period of time was similar to that of the BRT condition. RESULTS Figure 1 shows the anxiety ratings for each subject at each session in the experimental trial. The ratings were made at the end of each session and, therefore, show the subjects rated anxiety after BRT or cue control. Subject 1 showed a fair amount of anxiety on all five BL assessments. There was an immediate response to BRT, with a reduction in rated anxiety following the session. His anxiety tended to improve as the sessions progressed. Decreases in anxiety ratings did not decrease with the introduction of the cue at Session 11. There was an increase in rated anxiety following Session 16, but this did not continue to Session 17 or thereafter. When the cue was used without the support of BRT, there was no rated increase in anxiety, and the introduction of an NT on Session 26 produced little effect. At BL, Subject 2 scored between 31 and 39 on the anxiety rating scale. There was a quick and steady response to the introduction of BRT, and from Session 8, rated anxiety was consistently low until the end of the trial. The improvements in anxiety were unaffected by the introduction of the cue, with the gradual withdrawal of BRT and the removal of BRT leaving only a cue to control her anxiety. She was not affected by the introduction of an NT using the same cue control words. The responses of Subject 3 showed high anxiety ratings at BL. These were unaffected by the introduction of the cue in the absence of BRT. Once BRT was introduced on Session 11, there was a gradual reduction in anxiety, which became low and stable until the end of the trial. The improvements in anxiety were unaffected by the reintroduction of the cue, the introduction of an NT, and gradual withdrawal of BRT, leaving cue control only. In contrast to the first introduction of the cue, the final phase of study suggests that linking the cue words to BRT was of crucial importance because they had no effect in the initial stages of the trial.
Cue Control Relaxation
431
Subject
Anxiety
2
2.
Ratings
Rating;
T
Ssrsions
Sessions
Subject
4
Sessions
Subject 40 -
Anxiety
BL
5 BRT
30G 2.
Ratings
10
0
L 0
FIGURE 1. Behavioural anxiety ratings for Subjects 1 through 5 under each condition. BL, baseline; RT, Relaxation training session; CO, Cue only; IC, Introduction of cue for remaining sessions; NT, New therapist for remaining sessions except SS; OT, Original therapist for remaining sessions (SS).
For Subject 4, the BL scores were less stable, and her anxiety began at moderate levels. It was the impression of the therapist that she was somewhat wary of the new settings and, as a result, remained fairly still.
432
N? R. Lindsay et al.
However, her anxiety was at moderate levels and increased during BL to extremely high levels. The introduction of the cue in the absence of BRT had no effect on this, and her anxiety remained high. When BRT was introduced, ratings of anxiety followed each session, falling gradually to low levels. Her rated anxiety continued to fall through the introduction of the cue on Session 16, the introduction of an NT on Session 22, and the gradual removal of BRT, leaving cue control only. Once again, comparisons of the two cue control phases suggested that the effectiveness of the cue lies in its relationship with BRT rather than the semantic nature of the words. For Subject 5, BL anxiety was rated at fairly high levels, and the introduction of the cue had no effect on this. The introduction of an NT had little effect on these high levels of anxiety, suggesting that the previous introduction of an NT was not a factor in maintaining low levels of anxiety. Once BRT commenced, his anxiety ratings fell to low levels. In the case of Subject 5, there seemed to be an increase in anxiety with the introduction of the OT, who was previously uninvolved in training this particular subject. However, his anxiety levels fell once again at Session 23 and remained stably low until the end of the trial. This situation was true despite the withdrawal of BRT leaving only cue control. Again, the comparison of the two CO phases reinforced the fact that the effectiveness of the cue was in relation to its link with BRT. Figure 2 shows the amount of time in a 20min period in which subjects were engaged in the occupational task In all subjects, there was some increase in the amount of time spent on task. For Subject 1 there was a fairly erratic increase, showing initial improvements with reversal back to BL levels. However, when the CO phase was introduced, the amount of time he spent on the task was almost at a maximum for every session. For Subject 2, there was a more gradual increase, with a refusal to engage in the task on Session 17. This appeared unrelated to her anxiety ratings. Once again, during the CO phase, the amount of time spent on task is fairly high at between 60% and 90%. For Subject 3, there was no increase in the amount of time spent on task during the first CO phase. There was then a gradual increase during the BRT phase until at the end he spent almost all of the 10 min on task. During the second CO phase, he was working at a stably high rate. Although Subject 4 worked for around 13 min during the first BL session, this was not repeated at any of the BL phases or during the CO condition. There was no immediate increase during the BRT condition, and it was not until Session 19 that there was a sudden and substantial improvement in time spent on task. These improvements remain stable until the end of the trial despite the introduction of an NT and CO condition. For Subject 5, there was a similar pattern with low on task rates dur-
433
Cue Control Relaxation
Subjsct 1 BL
20
Minr
BRT
I
f -
I I I
I/
I
Subject 20
Minr
BL
20
Minr
IO
0
2
I
d
10
Subject BRT
/1
4
co
10
0
Subject
5
FIGURE 2. The amount of time spent on task for each subject during each 20-min session across conditions. BL, Baseline; RT, Relaxation training session; CO, Cue only; IC, Introduction of cue for remaining sessions; NT, New therapist for remaining sessions except SS; OT, Original therapist for remaining sessions (53).
ing BL and the first CO condition. There was no immediate increase in work rate on the introduction of BRT, but by Session 16, his amount of on task activity had increased somewhat. This remained high for most of the remaining sessions, although there were some dips toward the end of the
434
W R. Lindsay et al.
BRT phase. It was interesting that for all subjects the highest and most stable on task rates appear to be during the final CO phase. DISCUSSION We assessed whether or not cue controlled relaxation can be employed with subjects with mental retardation. Several studies have suggested that relaxation techniques can be effective with this client group (Calamari et al., 1987; Lindsay et al., 1989; McPhail & Chamov, 1989). First, if it were possible to link a cue with the relaxation effect, then the goals of training could be attained more quickly. Second, we ascertained whether or not cue controlled relaxation would facilitate concentration and attention to an occupational task. Five subjects were employed using individual case designs, and the results suggest that cue control can indeed be linked to relaxation. BRT has previously proven effective with the client group by Lindsay et al. (1989). This finding was linked to the cue in this study. Assessments were taken directly of the amount of anxious behaviour displayed by subjects during BL conditions and following training sessions. For all five subjects, there was a decrease in rated anxiety with BRT. The cue used to become associated with the effects of BRT were the words “quiet and still.” The introduction of cue words did not produce any changes in the relaxation effect of BRT and, because the relaxation instructions were phased out leaving only the cue words, there was no decrease in the effect of training. An essential condition of each individual design is the phase in which cue words only are used. When a CO condition was introduced following BL, there appeared to be no relaxation effect, suggesting that the words themselves were of little value to subjects. The CO condition was effective only when it followed the procedure whereby BRT was linked to the cue words, giving them a residual meaning related to relaxation. In all cases, the final CO condition produced effective relaxation. To ascertain whether or not relaxation effects were confined to one therapist or generalised to a new person, an NT was introduced at various stages of the individual designs. For two subjects, the NT was introduced during the final CO condition with no adverse effect on rated anxiety. For a further two subjects, the NT was introduced toward the end of the BRT condition, and, again, there seemed to be little effect on rated anxiety. With Subject 4, there was a slight increase in rated anxiety, although this was nowhere near BL levels and, indeed, did not reach levels shown at the beginning of the BRT phase. For Subject 5, the NT was introduced earlier in treatment. For the other subjects, the NT was introduced to determine whether or not she would have an adverse effect on relaxation. For this subject, the exper-
Cue Control Relaxation
435
imental design checked whether or not the introduction of an NT had a positive effect on relaxation. However, the introduction of the NT at the beginning of treatment had no effect in reducing anxiety. For this final subject, the OT was reintroduced toward the end of treatment and seemed to have an adverse effect on increasing levels of anxiety following the first session. However, even in this case, anxiety levels did not return to BL levels and, in subsequent sessions, reduced to low levels once again. Therefore, it seems that a cue can be linked to behavioural relaxation training for people with mental retardation. In all cases, the relationship between the cue words and BRT has been established, and relaxation effects have been evident when the cue words only are used. Although cue control was proven effective for a subject with mental retardation in a previous study (Wells, Turner, Bellack, & Hersen, 1978), this is the first time that this effect has been shown with subjects with severe mental retardation. This could be an important finding because Lindsay (199 1) suggested that relaxation techniques have a powerful effect on people with mental retardation, It was possible to induce relaxation effects in agitated individuals for up to 2 h. This obviously could be misused as a technique to make people easier to deal with and easily managed. It seems to us unreasonable to have individuals sitting around in a relaxed manner doing nothing for long time periods, rather than being excitable and agitated. We would urge therapists to consider relaxation training as an enabling technique, allowing people to become more accessible to their environment, able to make use of educational, occupational, and social opportunities available to them. In this study, the words “quiet and still” were used. In line with this caution, the words “please relax” or an equivalent might be a more appropriate cue. The results of the measures of on task occupational behaviour are similar to the results on rated anxiety. In all cases, subjects increased the amount of time spent in occupational tasks during the BRT phase. However, this was not a uniform improvement, and Subjects 1, 2, and 5 showed somewhat erratic progress in their improved attention. However, for those three subjects, there were definite improvements when compared to BL. For Subject 3, there was a more gradual improvement during the BRT phase, and for Subject 4, there was somewhat sudden improvement in concentration after eight sessions of treatment. Subjects’ ability to concentrate during the CO conditions again reflected results seen for rated anxiety. When the cue words were linked to BRT, there was a very good effect on concentration. In the absence of this link, for Subjects 3, 4, and 5, there was no effect. One interesting suggestion in the results is that during the final CO phase, concentration was better and more stable than in the BRT phases. Whereas Subjects 1 and 2 showed erratic
436
U! R. Lindsay et al.
progress during the BRT phase, their concentration was better and more consistent during the CO phase. Subjects 3 and 4 also showed good concentration and consistency during the final CO phase. Subject 5 continued to show some consistency, with concentration varying between just over 6 min and just under 16 min. It may be that by not having to go through the full BRT procedure but quickly establishing a relaxed position and beginning work, subjects were better able to concentrate for longer periods of time. This would certainly be an unexpected but positive outcome of this study. Although BRT can be used effectively to decrease agitation and increase concentration in people with severe mental retardation, there are several methodological shortcomings that threaten the validity of this finding. The effect of habituation on being seated has not been investigated. It may have been that if subjects were encouraged simply to sit for an equivalent number of sessions to the BRT condition, this would have produced a similar effect. All subjects were asked to sit in a relaxed fashion during BL, and no one complied. However, the number of trials was fewer than in the BRT condition. Even for the three subjects who had the CO condition following BL, there were only 10 sessions of being asked to sit. This was in contrast to 15 sessions of BRT and 5 to 10 further sessions in the CO condition. However, for these three subjects, increased relaxation from sitting over these 10 sessions did not occur. All three showed some relaxation effect after 4 to 6 sessions of BRT. Therefore, there is some indication that habituation was not a significant factor. Unless formal relaxation procedures were instigated, there appeared to be no treatment effect from BL and CO conditions via habituation. A second confound to treatment validity was the possibility of extinction of agitated behaviour over the course of the study. The earlier argument on habituation would hold similarly in considering the issue of extinction. The three subjects who received the CO condition following BRT were encouraged to engage in the concentration task following a period of sitting. Therefore, attempts were made to engage them in other tasks, rather then have them engage in distracted and agitated behaviour. None of the three showed any sign of extinction over these first 10 sessions. However, they showed signs of improvement after 6 to 8 sessions of BRT. Again, this would give some indication of the potency of BRT as opposed to the alternative hypothesis that extinction was a significant variable. However, neither habituation nor extinction were investigated formally; therefore, the results should be treated cautiously.
REFERENCES American Association on Mental Deficiency. Adults. Washington, DC: Author.
(1975). Adaptive Behaviour Scale For Children And
Cue Control Relaxation
437
Calamari, J. E., Geist, G. 0.. & Shabhazian, M. J. (1987). Evaluation of multiple component relaxation training with developmentally disabled persons. Resenrch in Developmental Disabilities, 8, 55-70. Lindsay, W. R. (1991). Psychological therapies in mental handicap. In W. Fraser, R. MacGillivray, & A. Green (Eds.), Hallas’ - Caring for people with mental handicaps (pp. 225-243). Oxford: Butterworth Heineman. Lindsay, W. R., & Baty, F. J. (1989). Group relaxation training with adults who are mentally handicapped. Behavioural Psychotherapy, 17.43-5 1. Lindsay, W. R., Baty, F. J., Michie, A. M., & Richardson, I. (1989). A comparison of anxiety treatments with adults who have moderate and severe mental retardation. Research in Developmental Disabilities, 10, 129-140. Lindsay, W. R., Richardson, I., & Michie, A. M. (1989). Short-term generalised effects of relaxation training on adults with moderate and Severe mental handicaps. Mental Handicap Research, 2, 197-206. McPhail, C. H., & Chamov, A. S. (1989). Relaxation reduces disruption in mentally handicapped adults. Journal of Mental Deficiency Research, 33,399-406. Schilling, D., Kc Poppen, R. (1983). Behavioural relaxation training and assessment. Journal of Behavior Therapy and Experimental Psychiatry, 16.99-107. Turk, V., & Francis, E. (1990). An anxiety management group: Strengths and pitfalls. Mental Handicap, 18.78-8 1. Wells, K. C., Turner, S. M., Bellack, A. S., & Hersen, M. (1978). Effects of cue controlled relaxation on psychomotor seizures: An experimental analysis. Behaviour Reseamh and Therapy, 16,51-53. Williams, J. (1990). Helping people to relax in over stimulating environments. Mental Handicap, 18, 160-162.
Disabilities,
Vol. 15, No. 6, pp. 425-437.
Copyright
Pergamon
0
1994 Elsevier
Printed in the USA.
1994
Science Ltd
All tights reserved
0891-4222194
$6.00 + 00
0891-4222(94)00017-4
The Effects of Cue Control Relaxation on Adults With Severe Mental Retardation William
R. Lindsay, Mairi Fee, Amanda Michie, and lmelda Heap Strathmartine
Hospital
This study was designed to assess two aspects of behnvioural relaxation training with subjects who have severe mental retardation. Thefirst was whether or not cue words could be linked effectively to training so that they would eventually produce a relaxation effect in the absence of a full relaxation procedure. The second was to determine the effects of relaxation training and subsequent cue control on concentration and attention to an occupational task. Five subjects participated, and individual case designs were used. Cue Control, behavioural relaxation training (BRT), and new therapists were introduced at different times for each subject to ascertain the effects of each variable. Subjects were assessed at baseline and following each training session on a behavioural relaxation scale to judge the direct effects of relaxation training and on the amount of time spent concentrating on an occupational task. BRT produced reductions in rated anxiety and improvements in concentration for all subjects. The cue control words were effective only ajier they had been linked to BRT In most cases, the introduction of a new therapist had no appreciable effect on anxiety or concentration. There was some suggestion that in the final cue only phases concentration was better and more consistent than during the BRTphases of the study.
Psychological treatments designed to alleviate anxiety and agitation may be helpful for people with severe mental retardation. Calamari, Geist, and Shahbazian (1987) found a combined package of treatment, including biofeedback, produced reductions in anxiety in subjects with a range of intellectual deficits. Lindsay, Baty, Michie, and Richardson (1989) found Requests for reprints should be sent to William R. Lindsay, Tayside Area Clinical Department, Stmthmartine Hospital, Dundee, DD3 OPG, Scotland.
425
Psychology
426
W! R. Lindsay et al.
both abbreviated progressive relaxation (APR) and behavioural relaxation training (BRT) to be effective in reducing rated anxiety in people with severe mental retardation. APR is a long-established therapy in which the client is asked to engage in a series of tension and release cycles for each part of the body. The client is taught to understand the difference between tension and relaxation and is subsequently able to adopt a relaxed state spontaneously. BRT was developed by Schilling and Poppen (1983) when APR proved ineffective with learning disabled boys. Originally subjects were asked to imitate unrelaxed and relaxed postures in each part of the body. Therefore, an understanding of internal states of muscular tension and relaxation is not required. Lindsay et al. (1989) found that BRT produced quicker and more substantial improvements than APR. Lindsay and Baty (1989) further simplified BRT by asking clients with severe mental retardation only to imitate unrelaxed postures. They suggested that, in contrast to APR, BRT is not a paradoxical technique requiring tensing of the muscles before relaxation. This leaves BRT less open to misinterpretation as a technique designed to produce excitement rather than relaxation, which they felt was very important for individuals with severe learning disabilities. McPhail and Chamov (1989) compared a group of six subjects receiving 12 sessions of APR (4 sessions per week over 3 weeks) with a control group who were in a story-telling condition. The subjects’ disabilities were reported to range from profound to mild mental retardation, and they were chosen because they displayed disruptive behaviour. As well as being assessed for immediate relaxation effects, they were also rated on 17 categories of disruptive behaviour. A significant and substantial reduction in disruptive behaviour in subjects receiving APR occurred with no changes in the controls. Lindsay, Richardson, and Michie (1989) also studied the generalised effects of relaxation training. Agitated movement, agitated speech, and general anxiety were measured 1 h after the completion of each of 12 treatment sessions and at a 3-week follow-up. Subjects receiving both APR and BRT showed a gradual reduction in behaviours indicating anxiety and agitation, which were maintained at follow-up. Papers by Williams (1990) and Turk and Francis (1990) serve to underscore the effectiveness of such procedures with the client group. The data support the conclusion reached by Calamari et al. (1987) that the results “call into question the sometimes held assumption that developmentally disabled persons are not good candidates for psychological treatments involving techniques other than simple behavioural management procedures” (p. 69). Now that effectiveness has apparently been established, it is important to adapt the methods to prove more useful in the lives of people with mental retardation. Lindsay (199 1) has written that the straightforward implemen-
Cue Control Relaxation
427
tation of powerful methods could be used as a simple (and possibly unethical) management technique to keep certain individuals quiet and undemanding. Serious consideration should be made when embarking on a programme of relaxation training concerning whether it should be used in conjunction with another programme, for example, educational or occupational activities, increased social contact, and so forth. Therefore, relaxation training would be used as an enabling technique to allow individuals to become more accessible to their environment. This article describes an intervention using (a) BRT as an enabling technique to help very agitated individuals to concentrate and (b) cue control to reduce the length of time needed to achieve relaxation. Very little research has been conducted on cue control relaxation for individuals with mental retardation. One very well controlled case study by Wells, Turner, Bellack, and Hersen (1978) did suggest potential value. Using a reversal design, they established a functional relationship between cue controlled relaxation (linked to APR) and psychomotor seizures in a young woman functioning in the borderline range of intelligence (IQ = 71). She was instructed to vocalise the cue word covertly at the onset of each preseizure aura, and this reduced the frequency of seizures. When she stopped using the cue word during the experimental reversal phase, seizure frequency increased to baseline levels. Reinstatement of cue control relaxation procedures reduced seizures once again, and improvements maintained at a 3-month follow-up. This study links cue control to BRT and investigates the effects of this procedure on subjects’ attending to various table-top tasks. METHOD Subjects
Five subjects each diagnosed with severe mental retardation (IQ < 40 as measured by the WATS-R) participated. All had spent most of their lives in a hospital and were currently hospitalised. All subjects scored below the 15th percentile on all scales of the Adaptive Behaviour Scale Part I (American Association on Mental Deficiency, 1975) except physical development. No subject was currently taking antianxiety medication. Subjects were a man, age 42; a woman, age 35; a man, age 48; a woman, age 45; and a man, age 29. Subjects showed a variety of agitation and anxiety symptoms, such as constant pacing; frequent facial, verbal, and bodily mannerisms; perseverative talking and mumbling; frequent shouting and screaming; avoidance of others; tremor; and high-frequency nondirected activity. When approached by others, these behaviours would almost always increase.
428
U! R. Lindsay et al.
Design
A single case design was employed to test the variables in this study. All subjects underwent a baseline (BL) condition in which they were simply asked to sit and relax in a chair and then engage in the task. BL consisted of five assessment sessions. Following this, three subjects went on to a cue only condition (CO) to ascertain the power of the cue words alone. In this condition, subjects were encouraged to sit in the chair for a period of time equivalent to that of subjects in the BRT condition. Two subjects went on to the BRT procedures. The three subjects in the CO condition then went on to the BRT condition. Following BRT, all subjects went on to CO. For all subjects, the cue was introduced (IC) on the sixth session of BRT. A new therapist (NT) replaced the original therapist (OT) at some point for each subject to determine whether or not changes were person-specific or generalised to new individuals. For one subject, the NT was introduced in the first CO condition to determine whether or not a therapist change per se had an effect. For this subject, the OT was reintroduced at the end of training. In this design, generalisation and maintenance are considered. Because the CO conditions do not employ active therapeutic procedures, they can be considered as a naturalistic follow-up and maintenance condition. Procedure
Cues were tied to BRT, which is a technique concentrating on observable states of relaxation in 10 areas of the body. The instructor demonstrates the relaxed states in each area and then helps the person copy the relaxed behaviours. The aim is to reduce muscle activity and help the individual to be able to adopt a relaxed state spontaneously, in any part of the body, culminating in an overall relaxed posture and quiet mood. The person must be seated in a comfortable chair, rather than lying down. He or she is asked first to watch and then to imitate the relaxed behaviour. Simple direct feedback on the accuracy of the copying and manual guidance are given. With extremely agitated subjects, verbal and physical prompting may be necessary. For example, with a subject who moved her hands frequently, a light physical prompt on the forearm would accompany the instructions to keep the arms quiet and still. Training involved demonstrating an overall relaxed and then unrelaxed posture, breathing when unrelaxed and relaxed (quick vs. slow), vocalisation (noisy vs. silent), trunk movements (moving and fidgeting or rocking vs. still), head movements (held tense or moving vs. leaning back and still), eyes (closed tight vs. lightly closed or open and still), lips (pressed tight vs. relaxed), throat (movement/swallowing vs. still), shoulders (hunched vs. sloped and even), arms (moving vs. still), hands (tense fist vs. relaxed and
Cue Control Relaxation
429
open), legs (moving or tense vs. still), feet (moving vs. still). No reinforcement schedules were used at any time throughout the study. All sessions took place at the same time of day for each subject. Sessions occurred daily except for weekends. The verbal cue linked to BRT was “quiet and still.” This was introduced at various stages in each individual design. Following introduction of the cue, the relaxation procedure was gradually shortened and eventually eliminated. Following cue control or BRT, a simple dexterity task was introduced for the subject from a variety of tasks, including colour matching, shape sorting, shape matching, and peg boards. These are generally widely available in establishments for people with mental retardation. A simple attainable task was chosen because the individuals involved had such poor concentration and short attention spans. Subjects were simply asked to put the pieces in the appropriate place, for example, “Put the shapes in their place” or “Put the pegs in the holes.” Therapists were instructed to prompt subjects at no more than 10-s intervals. No reinforcers were given to subjects at any time. If clients became disruptive, they were asked to continue, but if they persisted, the session was ended. Assessment Subjects were video taped after each session. Tapes were later rated on a scale designed to assess degree of anxiety. Raters judged 10 body parts on a 5-point scale from no signs of anxiety or agitation through moderate levels of anxiety and agitation to extreme agitation and anxieo. Each rating point was explained with guidelines on the scale. The items assessed were voice, trunk, head, eyes, lips, throat, shoulders, hands, feet, and a general rating of relaxation/tension. Lindsay et al. (1989) previously reported extremely high reliability between trained observers on the total behavioural anxiety score produced on this test (r = 0.95). Reliability was calculated on 84 sets of scores, each rated by two experienced independent raters, blind to the treatment conditions. This produced a Pearson’s correlation coeffrcient of r = 0.89. Further investigation on the 10 component parts of the scale found interobserver agreement to be high using the formula: number of agreements divided by the number of agreements plus disagreements expressed as a percentage. An agreement was only scored if both observers agreed on the degree of anxiety (O-4) observed in the particular body part being rated. This produced 94% absolute agreement between trained raters. Subjects were also video taped while they were engaging in the 20 min of occupational activity. The amount of time spent “on task” was computed later. On task behaviour was defined as any meaningful manipulation of the occupational materials, for example, lifting a piece, looking at it, and trying to fit it into the board. As a check on reliability of this score, two
430
N! R. Lindsay et al.
independent raters judged 20 tapes, and agreement was found to be 98.4%. The method of calculating reliability in this score was less exact. Only the total time “on task” was used in the calculation, that is, total time for Observer A (scoring the lowest amount of “on task” activity) divided by total time for Observer B, expressed as a percentage. However, reliability is so high (almost 100%) and the clinical changes are so considerable that drawbacks in the method of computation may not be crucial. For BL conditions, subjects were video taped after the instruction to sit and relax; for the BRT condition, they were taped at the end of the relaxation session; in the CO condition, they were taped after the use of the cue words. In the BL and CO conditions, attempts were made to ensure that the period of time was similar to that of the BRT condition. RESULTS Figure 1 shows the anxiety ratings for each subject at each session in the experimental trial. The ratings were made at the end of each session and, therefore, show the subjects rated anxiety after BRT or cue control. Subject 1 showed a fair amount of anxiety on all five BL assessments. There was an immediate response to BRT, with a reduction in rated anxiety following the session. His anxiety tended to improve as the sessions progressed. Decreases in anxiety ratings did not decrease with the introduction of the cue at Session 11. There was an increase in rated anxiety following Session 16, but this did not continue to Session 17 or thereafter. When the cue was used without the support of BRT, there was no rated increase in anxiety, and the introduction of an NT on Session 26 produced little effect. At BL, Subject 2 scored between 31 and 39 on the anxiety rating scale. There was a quick and steady response to the introduction of BRT, and from Session 8, rated anxiety was consistently low until the end of the trial. The improvements in anxiety were unaffected by the introduction of the cue, with the gradual withdrawal of BRT and the removal of BRT leaving only a cue to control her anxiety. She was not affected by the introduction of an NT using the same cue control words. The responses of Subject 3 showed high anxiety ratings at BL. These were unaffected by the introduction of the cue in the absence of BRT. Once BRT was introduced on Session 11, there was a gradual reduction in anxiety, which became low and stable until the end of the trial. The improvements in anxiety were unaffected by the reintroduction of the cue, the introduction of an NT, and gradual withdrawal of BRT, leaving cue control only. In contrast to the first introduction of the cue, the final phase of study suggests that linking the cue words to BRT was of crucial importance because they had no effect in the initial stages of the trial.
Cue Control Relaxation
431
Subject
Anxiety
2
2.
Ratings
Rating;
T
Ssrsions
Sessions
Subject
4
Sessions
Subject 40 -
Anxiety
BL
5 BRT
30G 2.
Ratings
10
0
L 0
FIGURE 1. Behavioural anxiety ratings for Subjects 1 through 5 under each condition. BL, baseline; RT, Relaxation training session; CO, Cue only; IC, Introduction of cue for remaining sessions; NT, New therapist for remaining sessions except SS; OT, Original therapist for remaining sessions (SS).
For Subject 4, the BL scores were less stable, and her anxiety began at moderate levels. It was the impression of the therapist that she was somewhat wary of the new settings and, as a result, remained fairly still.
432
N? R. Lindsay et al.
However, her anxiety was at moderate levels and increased during BL to extremely high levels. The introduction of the cue in the absence of BRT had no effect on this, and her anxiety remained high. When BRT was introduced, ratings of anxiety followed each session, falling gradually to low levels. Her rated anxiety continued to fall through the introduction of the cue on Session 16, the introduction of an NT on Session 22, and the gradual removal of BRT, leaving cue control only. Once again, comparisons of the two cue control phases suggested that the effectiveness of the cue lies in its relationship with BRT rather than the semantic nature of the words. For Subject 5, BL anxiety was rated at fairly high levels, and the introduction of the cue had no effect on this. The introduction of an NT had little effect on these high levels of anxiety, suggesting that the previous introduction of an NT was not a factor in maintaining low levels of anxiety. Once BRT commenced, his anxiety ratings fell to low levels. In the case of Subject 5, there seemed to be an increase in anxiety with the introduction of the OT, who was previously uninvolved in training this particular subject. However, his anxiety levels fell once again at Session 23 and remained stably low until the end of the trial. This situation was true despite the withdrawal of BRT leaving only cue control. Again, the comparison of the two CO phases reinforced the fact that the effectiveness of the cue was in relation to its link with BRT. Figure 2 shows the amount of time in a 20min period in which subjects were engaged in the occupational task In all subjects, there was some increase in the amount of time spent on task. For Subject 1 there was a fairly erratic increase, showing initial improvements with reversal back to BL levels. However, when the CO phase was introduced, the amount of time he spent on the task was almost at a maximum for every session. For Subject 2, there was a more gradual increase, with a refusal to engage in the task on Session 17. This appeared unrelated to her anxiety ratings. Once again, during the CO phase, the amount of time spent on task is fairly high at between 60% and 90%. For Subject 3, there was no increase in the amount of time spent on task during the first CO phase. There was then a gradual increase during the BRT phase until at the end he spent almost all of the 10 min on task. During the second CO phase, he was working at a stably high rate. Although Subject 4 worked for around 13 min during the first BL session, this was not repeated at any of the BL phases or during the CO condition. There was no immediate increase during the BRT condition, and it was not until Session 19 that there was a sudden and substantial improvement in time spent on task. These improvements remain stable until the end of the trial despite the introduction of an NT and CO condition. For Subject 5, there was a similar pattern with low on task rates dur-
433
Cue Control Relaxation
Subjsct 1 BL
20
Minr
BRT
I
f -
I I I
I/
I
Subject 20
Minr
BL
20
Minr
IO
0
2
I
d
10
Subject BRT
/1
4
co
10
0
Subject
5
FIGURE 2. The amount of time spent on task for each subject during each 20-min session across conditions. BL, Baseline; RT, Relaxation training session; CO, Cue only; IC, Introduction of cue for remaining sessions; NT, New therapist for remaining sessions except SS; OT, Original therapist for remaining sessions (53).
ing BL and the first CO condition. There was no immediate increase in work rate on the introduction of BRT, but by Session 16, his amount of on task activity had increased somewhat. This remained high for most of the remaining sessions, although there were some dips toward the end of the
434
W R. Lindsay et al.
BRT phase. It was interesting that for all subjects the highest and most stable on task rates appear to be during the final CO phase. DISCUSSION We assessed whether or not cue controlled relaxation can be employed with subjects with mental retardation. Several studies have suggested that relaxation techniques can be effective with this client group (Calamari et al., 1987; Lindsay et al., 1989; McPhail & Chamov, 1989). First, if it were possible to link a cue with the relaxation effect, then the goals of training could be attained more quickly. Second, we ascertained whether or not cue controlled relaxation would facilitate concentration and attention to an occupational task. Five subjects were employed using individual case designs, and the results suggest that cue control can indeed be linked to relaxation. BRT has previously proven effective with the client group by Lindsay et al. (1989). This finding was linked to the cue in this study. Assessments were taken directly of the amount of anxious behaviour displayed by subjects during BL conditions and following training sessions. For all five subjects, there was a decrease in rated anxiety with BRT. The cue used to become associated with the effects of BRT were the words “quiet and still.” The introduction of cue words did not produce any changes in the relaxation effect of BRT and, because the relaxation instructions were phased out leaving only the cue words, there was no decrease in the effect of training. An essential condition of each individual design is the phase in which cue words only are used. When a CO condition was introduced following BL, there appeared to be no relaxation effect, suggesting that the words themselves were of little value to subjects. The CO condition was effective only when it followed the procedure whereby BRT was linked to the cue words, giving them a residual meaning related to relaxation. In all cases, the final CO condition produced effective relaxation. To ascertain whether or not relaxation effects were confined to one therapist or generalised to a new person, an NT was introduced at various stages of the individual designs. For two subjects, the NT was introduced during the final CO condition with no adverse effect on rated anxiety. For a further two subjects, the NT was introduced toward the end of the BRT condition, and, again, there seemed to be little effect on rated anxiety. With Subject 4, there was a slight increase in rated anxiety, although this was nowhere near BL levels and, indeed, did not reach levels shown at the beginning of the BRT phase. For Subject 5, the NT was introduced earlier in treatment. For the other subjects, the NT was introduced to determine whether or not she would have an adverse effect on relaxation. For this subject, the exper-
Cue Control Relaxation
435
imental design checked whether or not the introduction of an NT had a positive effect on relaxation. However, the introduction of the NT at the beginning of treatment had no effect in reducing anxiety. For this final subject, the OT was reintroduced toward the end of treatment and seemed to have an adverse effect on increasing levels of anxiety following the first session. However, even in this case, anxiety levels did not return to BL levels and, in subsequent sessions, reduced to low levels once again. Therefore, it seems that a cue can be linked to behavioural relaxation training for people with mental retardation. In all cases, the relationship between the cue words and BRT has been established, and relaxation effects have been evident when the cue words only are used. Although cue control was proven effective for a subject with mental retardation in a previous study (Wells, Turner, Bellack, & Hersen, 1978), this is the first time that this effect has been shown with subjects with severe mental retardation. This could be an important finding because Lindsay (199 1) suggested that relaxation techniques have a powerful effect on people with mental retardation, It was possible to induce relaxation effects in agitated individuals for up to 2 h. This obviously could be misused as a technique to make people easier to deal with and easily managed. It seems to us unreasonable to have individuals sitting around in a relaxed manner doing nothing for long time periods, rather than being excitable and agitated. We would urge therapists to consider relaxation training as an enabling technique, allowing people to become more accessible to their environment, able to make use of educational, occupational, and social opportunities available to them. In this study, the words “quiet and still” were used. In line with this caution, the words “please relax” or an equivalent might be a more appropriate cue. The results of the measures of on task occupational behaviour are similar to the results on rated anxiety. In all cases, subjects increased the amount of time spent in occupational tasks during the BRT phase. However, this was not a uniform improvement, and Subjects 1, 2, and 5 showed somewhat erratic progress in their improved attention. However, for those three subjects, there were definite improvements when compared to BL. For Subject 3, there was a more gradual improvement during the BRT phase, and for Subject 4, there was somewhat sudden improvement in concentration after eight sessions of treatment. Subjects’ ability to concentrate during the CO conditions again reflected results seen for rated anxiety. When the cue words were linked to BRT, there was a very good effect on concentration. In the absence of this link, for Subjects 3, 4, and 5, there was no effect. One interesting suggestion in the results is that during the final CO phase, concentration was better and more stable than in the BRT phases. Whereas Subjects 1 and 2 showed erratic
436
U! R. Lindsay et al.
progress during the BRT phase, their concentration was better and more consistent during the CO phase. Subjects 3 and 4 also showed good concentration and consistency during the final CO phase. Subject 5 continued to show some consistency, with concentration varying between just over 6 min and just under 16 min. It may be that by not having to go through the full BRT procedure but quickly establishing a relaxed position and beginning work, subjects were better able to concentrate for longer periods of time. This would certainly be an unexpected but positive outcome of this study. Although BRT can be used effectively to decrease agitation and increase concentration in people with severe mental retardation, there are several methodological shortcomings that threaten the validity of this finding. The effect of habituation on being seated has not been investigated. It may have been that if subjects were encouraged simply to sit for an equivalent number of sessions to the BRT condition, this would have produced a similar effect. All subjects were asked to sit in a relaxed fashion during BL, and no one complied. However, the number of trials was fewer than in the BRT condition. Even for the three subjects who had the CO condition following BL, there were only 10 sessions of being asked to sit. This was in contrast to 15 sessions of BRT and 5 to 10 further sessions in the CO condition. However, for these three subjects, increased relaxation from sitting over these 10 sessions did not occur. All three showed some relaxation effect after 4 to 6 sessions of BRT. Therefore, there is some indication that habituation was not a significant factor. Unless formal relaxation procedures were instigated, there appeared to be no treatment effect from BL and CO conditions via habituation. A second confound to treatment validity was the possibility of extinction of agitated behaviour over the course of the study. The earlier argument on habituation would hold similarly in considering the issue of extinction. The three subjects who received the CO condition following BRT were encouraged to engage in the concentration task following a period of sitting. Therefore, attempts were made to engage them in other tasks, rather then have them engage in distracted and agitated behaviour. None of the three showed any sign of extinction over these first 10 sessions. However, they showed signs of improvement after 6 to 8 sessions of BRT. Again, this would give some indication of the potency of BRT as opposed to the alternative hypothesis that extinction was a significant variable. However, neither habituation nor extinction were investigated formally; therefore, the results should be treated cautiously.
REFERENCES American Association on Mental Deficiency. Adults. Washington, DC: Author.
(1975). Adaptive Behaviour Scale For Children And
Cue Control Relaxation
437
Calamari, J. E., Geist, G. 0.. & Shabhazian, M. J. (1987). Evaluation of multiple component relaxation training with developmentally disabled persons. Resenrch in Developmental Disabilities, 8, 55-70. Lindsay, W. R. (1991). Psychological therapies in mental handicap. In W. Fraser, R. MacGillivray, & A. Green (Eds.), Hallas’ - Caring for people with mental handicaps (pp. 225-243). Oxford: Butterworth Heineman. Lindsay, W. R., & Baty, F. J. (1989). Group relaxation training with adults who are mentally handicapped. Behavioural Psychotherapy, 17.43-5 1. Lindsay, W. R., Baty, F. J., Michie, A. M., & Richardson, I. (1989). A comparison of anxiety treatments with adults who have moderate and severe mental retardation. Research in Developmental Disabilities, 10, 129-140. Lindsay, W. R., Richardson, I., & Michie, A. M. (1989). Short-term generalised effects of relaxation training on adults with moderate and Severe mental handicaps. Mental Handicap Research, 2, 197-206. McPhail, C. H., & Chamov, A. S. (1989). Relaxation reduces disruption in mentally handicapped adults. Journal of Mental Deficiency Research, 33,399-406. Schilling, D., Kc Poppen, R. (1983). Behavioural relaxation training and assessment. Journal of Behavior Therapy and Experimental Psychiatry, 16.99-107. Turk, V., & Francis, E. (1990). An anxiety management group: Strengths and pitfalls. Mental Handicap, 18.78-8 1. Wells, K. C., Turner, S. M., Bellack, A. S., & Hersen, M. (1978). Effects of cue controlled relaxation on psychomotor seizures: An experimental analysis. Behaviour Reseamh and Therapy, 16,51-53. Williams, J. (1990). Helping people to relax in over stimulating environments. Mental Handicap, 18, 160-162.