- Email: [email protected]
Patterns of emotional and behavioural disturbance associated with autistic traits in young people with severe intellectual disabilities and challenging behaviours
Research in Developmental Disabilities 27 (2006) 517–528
Patterns of emotional and behavioural disturbance associated with autistic traits in young people with severe intellectual disabilities and challenging behaviours§ Jennie Hill a,*, Frederick Furniss b,c a
Nottinghamshire Healthcare NHS Trust, Community Learning Disability Team, 53, Cheapside, Worksop, Nottinghamshire S80 2JD, UK b The Hesley Group, Doncaster, UK c School of Psychology, University of Leicester, UK Received 22 February 2005; received in revised form 16 June 2005; accepted 5 July 2005
Abstract Emotional and behavioural disturbance was assessed in 82 individuals with severe intellectual disabilities and challenging behaviour using the Diagnostic Assessment for the Severely HandicappedII (DASH-II). Levels of disturbance were compared firstly in individuals with and without features of autism as assessed by the DASH-II, and secondly in individuals with varying severities of autism. In both cases levels of ability and overall severity of behaviour disorder were comparable across groups. Individuals with autistic features were found to have significantly higher scores than nonautistic individuals on the DASH-II organic disorder, anxiety, mania, PDD/autism and stereotypies subscales. When participants with autistic features were separated into groups of severe and moderate autism and compared with nonautistic participants, significant effects of group were found for scores on the anxiety, mood, mania, PDD/autism, schizophrenia and stereotypies subscales. Scheffe´ tests were conducted to further evaluate between-group differences. Item analysis showed seven DASH-II items to have a 30% or more difference between levels of endorsement in autistic and nonautistic individuals, with six further items showing a 20% or greater difference in levels of endorsement. Findings are compared to those from previous research and implications for the conceptualisation of emotional and behavioural disorders in individuals with autism are discussed. # 2005 Elsevier Ltd. All rights reserved. Keywords: Autism; Intellectual disability; Emotional and behavioural disturbance; DASH-II
§ A preliminary version of this paper was presented to the 2004 World Congress of the International Association for the Scientific Study of Intellectual Disability. * Corresponding author. Tel.: +44 1909 530 205; fax: +44 1909 530 464. E-mail address: [email protected] (J. Hill).
0891-4222/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.ridd.2005.07.001
518
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
1. Introduction Individuals with intellectual disabilities have consistently been found to have an increased prevalence of emotional and behavioural disturbance compared to the general population. However, there have only been a limited number of studies conducted to assess whether there is an increased prevalence rate of emotional and behavioural disturbances in individuals diagnosed with autism (Ghaziuddin, Alessi, & Greden, 1995; Kim, Szatmari, Bryson, Streiner, & Wilson, 2000; Tsai, 1996). Further, these studies have tended to focus on high functioning and verbal autistic individuals who have been most able to give a verbal account of their symptoms (Bradley, Summers, Wood, & Bryson, 2004). One reason for this is that diagnosing emotional and behavioural disturbances in individuals with intellectual disabilities can be problematic (King, DeAntonio, McCracken, Forness, & Ackerland, 1994). These problems arise because existing diagnostic classificatory systems (DSM-IV, American Psychiatric Association, 1994; ICD-10-CDDG, World Health Organisation, 1992) depend substantially on the self-reported subjective experiences of the individuals being diagnosed (Bradley et al., 2004), a methodology often not feasible for individuals with severe cognitive and communication impairments. One assessment scale that can be utilised to explore emotional and behavioural disturbances in individuals with severe intellectual disabilities is the Diagnostic Assessment for the Severely Handicapped-II (DASH-II) (Matson, 1995; Matson, Gardner, Coe, & Sovner, 1991). The DASHII can be used to assess individuals with severe cognitive and communication impairments because a rater who has known the individual for a significant length of time provides information regarding behaviours associated with emotional and behavioural disturbance. The items contained in the DASH-II were formulated from both the DSM-III-R and from research relating to behavioural disturbances in individuals with severe intellectual disabilities. The DASH-II has been found to be reliable (Matson, Coe, Gardner, & Sovner, 1991; Sevin, Matson, Williams, & Kirkpatrick-Sanchez, 1995) and valid (Paclawskyj, Matson, Bamburg, & Baglio, 1997). Several of the DASH-II subscales have been found to be valid (Matson & Smiroldo, 1997; Matson, Smiroldo, & Hastings, 1998; Matson et al., 1999), although evidence for the validity of the anxiety subscale is still limited (Matson, Smiroldo, Hamilton, & Baglio, 1997). Factor analysis of the DASH-II by Matson et al. (1991b) yielded a six-factor structure with factors relating to emotional lability, aggression/conduct disorder, language disorder/verbal aggression, social withdrawal/stereotypy, eating disorder and sleep disorder. Several previous studies have utilised the DASH-II to assess whether there is an increased prevalence rate of emotional and behavioural disturbances in individuals with a diagnosis of autism or autistic traits, using subscale scores equivalent to a score of more than one standard deviation above the mean for the standardization sample (Matson, 1995) as a cutoff point indicating a level of disturbance meriting further investigation. Matson et al. (1996) compared individuals who scored over the cutoff point on the DASH-II PDD/autism subscale with members of a control group with severe and profound intellectual disabilities but no elevation in scores on the DASH-II PDD/autism subscale, and found differences of over 25% in the percentages who scored above the cutoff on the DASH-II stereotypies, mania, impulse control disorders, and organic disorders subscales. Matson et al. (1999) also found that the most common comorbid disorders as assessed by the DASH-II in a group of adults previously diagnosed with autism were impulse control disorders and stereotypies, with at least 40% of those scoring above the cutoff on the PDD/autism subscale also scoring above the cutoff on the impulse control and stereotypies subscales.
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
519
More recently, Bradley et al. (2004) conducted a similar study on 12 lower functioning autistic individuals and matched controls. They found that the participants with autism scored significantly higher on four of the five DASH-II psychiatric disorder subscales (anxiety, PDD/ autism, mania and depression) and three of the eight DASH-II aberrant behaviour subscales (stereotypies, sleep disorders and organic syndromes). They also found that 50% or more of the autistic participants scored above the cutoff point on seven of the 13 subscales (PDD/autism, mania, depression, stereotypies, self-injury, eating disorders and sleep disorders), whereas only for one subscale (self-injury) did the non-autistic participants present a comparable extent of disorder. The autistic participants each had an average of 5.25 clinically significant disorders (not including PDD/autism), whilst the non-autistic participants each had an average of 1.25 clinically significant disorders. The above research has begun to identify clinical features characteristic of autism in people with severe and profound learning disabilities. Research to date however has compared persons with and without marked autistic features but without controlling for overall severity of behaviour disorder. Given the general association between severity of autistic traits and extent of behavioural disturbance (Bhaumik, Branford, McGrother, & Thorp, 1997), it remains uncertain whether the differences between autistic and non-autistic groups reported above reflect specific patterns of comorbidity with autism or simply reflect greater generalized behavioural disturbance in persons with autism. The aims of the current study were therefore (i) to compare patterns of behavioural disturbance as assessed by the DASH-II in groups of persons with intellectual disabilities scoring above and below the cutoff point on the DASH-II PDD/autism subscale but of comparable intellectual ability and overall level of behavioural disturbance and (ii) to examine whether there is a unique set of psychiatric and behavioural traits in autistic individuals using DASH-II subscale item scores. 2. Method 2.1. Participants Participants were 82 children and young adults with severe intellectual disabilities and challenging behaviour living in four residential establishments in England. The mean age of the participants was 18 years and 2 months (minimum 8 years and 4 months; maximum 29 years and 9 months). Seventeen were female and 65 male. Each participant was assessed using either the AAMR Adaptive Behavior Scale – School (Lambert, Nihira, & Leland, 1993) or the AAMR Adaptive Behavior Scale – Residential (Nihira, Leland, & Lambert, 1993), dependent upon whether they were below or above 18 years of age respectively, to assess their adaptive abilities. The mean independent functioning standard score of the participants was 9 (minimum 1; maximum 15) and the mean language development standard score of the participants was 7 (minimum 2; maximum 16). Of the participants, 25 were taking anti-convulsant medication (32%) and 30 were taking some other psychoactive medication (39%). Medication was unknown for five of the participants. 2.2. Instruments 2.2.1. The Diagnostic Assessment for the Severely Handicapped-II (DASH-II) The DASH-II (Matson, 1995; Matson et al., 1991b) is an 84-item broad-band (Rojahn, Aman, Matson, & Mayville, 2003) behaviour rating scale. The DASH-II is a general screening
520
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
instrument aimed at assessing psychopathological symptoms relating to heterogeneous conditions such as schizophrenia, mania and autism in individuals with severe and profound intellectual and developmental disabilities. The 13 subscales of the DASH-II are (i) impulse control (17 items), (ii) organic problems (9 items), (iii) anxiety (8 items), (iv) mood disorders (15 items), (v) mania (7 items), (vi) pervasive developmental disorders/autism (6 items), (vii) schizophrenia (7 items), (viii) stereotypies (7 items), (ix) self-injurious behaviour (5 items), (x) elimination disorders (2 items), (xi) eating disorders (6 items), (xii) sleep disorders (5 items) and (xiii) sexual disorders (3 items). Each subscale item is scored on three scales: (a) a three-point frequency scale based on occurrence of the behaviour in the last 2 weeks (zero, 1–10, or over 10 occurrences), (b) a three-point duration scale concerning how long the behaviour has been a problem (less than 1 month, 1–12 months, or over 12 months), and a three-point severity scale concerning how serious the behaviour has been in the last 2 weeks (caused no disruptions or damages, caused no damages but at least one disruption to others, or caused injury or property damage at least once). 2.2.2. The Aberrant Behavior Checklist – community The ABC was used to examine overall psychological and behavioural disturbance in the participants. The ABC (Aman & Singh, 1994) is a factor analytically derived 58-item broad-band (Rojahn et al., 2003) behaviour rating scale. The ABC was originally designed to aid in the assessment of psychotropic drug treatment in individuals with intellectual and developmental disabilities, although it can also be used as a general assessment tool for problem behaviours. The five subscales of the ABC are (i) irritability, agitation, crying (15 items), (ii) lethargy, social withdrawal (16 items), (iii) stereotypic behaviour (7 items), (iv) hyperactivity, non-compliance (16 items), and (v) inappropriate speech (4 items). Each subscale item is scored on a four-point overall severity scale ranging from not at all a problem through to the problem is severe in degree. The behaviour must have appeared at least once over the past 4 weeks in order for it to be rated (i.e. given a severity score greater than 0). 2.3. Interviewers and respondents Psychology graduates administered both the ABC and the DASH-II. The respondents were direct support staff members who were either participants’ keyworkers or had worked with them for an extended period of time. 2.4. Data analysis Data analysis was undertaken in two stages. Firstly, to replicate the approach of previous studies, participants were divided into two groups, those scoring above and below the conventional (1 S.D. above the standardization sample mean) cutoff score on the DASH-II PDD/ Autism subscale. Secondly, participants were divided into three groups, those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and finally those scoring more than two standard deviations above the standardization sample mean. As in previous research (Bradley et al., 2004; Matson et al., 1996), analysis of DASH-II subscale scores was performed on frequency scores. Additionally, analysis was limited to those subscales (impulse control, organic problems, anxiety, mood disorders, mania, schizophrenia, and stereotypies), where the DASH-II subscale score is based on a sum of frequency scores, thus excluding those
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
521
subscales (self-injurious behaviour, elimination disorders, eating disorders, sleep disorders and sexual disorders) where the DASH-II is scored categorically on the basis of any single subscale item receiving a severity score greater than zero. 3. Results 3.1. Ages, abilities and levels of behavioural disturbance For the two-group comparison, the group of 13 participants scoring below the cutoff on the DASH-II PDD/autism subscale (hereafter the ‘‘nonautistic’’ group) had a mean score on that subscale of 3.38 (S.D. 0.77), while the group of 69 participants scoring above the cutoff (hereafter the ‘‘autistic’’ group) had a mean score of 7.65 (S.D. 1.91). Table 1 shows means and standard deviations for the two groups for age, Adaptive Behavior Scale Part 1 ‘‘Independent Functioning’’ and ‘‘Language Development’’ domain standard scores, and raw scores for each Aberrant Behavior Checklist subscale, together with the numbers of females and males in each group. Independent samples t-tests confirmed that the groups did not differ significantly in age (t = 0.94, d.f. = 80, p = 0.35), Independent Functioning score (t = 1.28, d.f. = 80, p = 0.21), Language Development score (t = 1.41, d.f. = 80, p = 0.16), ABC Irritability subscale score (t = 1.57, d.f. = 25.78, p = 0.13), ABC Lethargy subscale score (t = 0.68, d.f. = 80, p = 0.5), ABC Stereotypy subscale score (t = 1.31, d.f. = 80, p = 0.19), or ABC Hyperactivity subscale score (t = 1.65, d.f. = 80, p = 0.1). The autistic group however had significantly higher scores on the ABC Inappropriate Speech subscale (t = 5.35, d.f. = 55.12, p < 0.001). Chi-square analysis showed no significant association between participant gender and group membership (x2 = 1.81, d.f. = 1, n.s.). For the three-group comparison, the group of 34 participants with scores on the PDD/autism subscale greater than one standard deviation above the normative sample mean, but not more than two standard deviations above that mean (hereafter the ‘‘moderately autistic’’ group), had a mean score of 6.06 (S.D. 0.78), whereas the 35 members of the ‘‘severely autistic’’ group scoring more than two standard deviations above the normative sample mean had a mean score of 9.2 (S.D. 1.3). Table 1 again shows mean ages, relevant ABS and ABC scores, and the numbers of females and males in each group. One-way analyses of variance showed that the groups did not differ significantly in age (F (2,79) = 1.00, p = 0.37), ABS Independent Functioning score (F (2,79) = 0.95, p = 0.39), ABS Language Development score (F (2,79) = 2.3, p = 0.11), ABC Irritability subscale score (F (2,79) = 0.76, p = 0.47), ABC Lethargy subscale score (F (2,79) = 0.33, p = 0.72), ABC Stereotypy subscale score (F (2,79) = 1.01, p = 0.37), or ABC Hyperactivity subscale score (F (2,79) = 2.31, p = 0.11). The groups however differed significantly on the ABC Inappropriate Speech subscale (F (2,79) = 10.03, p < 0.001). Post hoc Scheffe´ tests showed that the severely autistic group had significantly higher scores on this subscale than the other two groups, which did not differ significantly on this measure. Chi-square analysis again showed no significant association between participant gender and group membership (x2 = 3.25, d.f. = 2, n.s.). 3.2. DASH-II subscale scores When the sample was divided into those scoring more or less than one standard deviation above the normative mean on the DASH-II PDD/autism subscale, significant differences were found between the means of the two (nonautistic and autistic) groups on the DASH-II organic
522
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Table 1 Ages and adaptive and aberrant behaviour scores of participant groups Measure
Age (months) Number (%) female/male ABS Independent Functioning (standard score) ABS Language Development (standard score) ABC Irritability (raw score) ABC Lethargy (raw score) ABC Stereotypy (raw score) ABC Hyperactivity (raw score) ABC Inappropriate Speech (raw score)
Group Nonautistic (n = 13)
Autistic (n = 69)
Moderately autistic (n = 34)
Severely autistic (n = 35)
229.92 (63.26) 5 (38%) female, 8 (62%) male 8.15 (2.44)
215.54 (47.80) 12 (17%) female, 57 (83%) male 9.0 (2.15)
209.06 (49.62) 5 (15%) female, 29 (85%) male 8.85 (2.36)
221.83 (45.81) 7 (20%) female, 28 (80%) male 9.14 (1.94)
6.82 (3.64)
8.06 (2.9)
16.88 (9.60) 11.00 (7.13) 5.82 (5.21) 17.71 (11.18) 2.41 (3.02)
17.97 (10.05) 11.83 (8.50) 5.20 (3.96) 14.51 (8.21) 4.89 (3.75)
6.08 (2.6) 14.23 (6.02) 9.85 (6.49) 3.69 (4.46) 11.23 (9.2) 0.69 (1.25)
7.45 (3.32) 17.43 11.42 5.51 16.09 3.67
(9.77) (7.81) (4.6) (9.85) (3.60)
subscale (t = 2.05, d.f. = 80, p < .05), the DASH-II anxiety subscale (t = 2.06, d.f. = 80, p < .05), the DASH-II mania subscale (t = 3.77, d.f. = 80, p < .001), the DASH-II PDD/autism subscale (t = 13.62, d.f. = 45.34, p < .001), and the DASH-II stereotypies subscale (t = 4.64, d.f. = 80, p < .001) The means of the two sub samples did not differ significantly on the DASH-II Impulse, Mood, or Schizophrenia subscales. When the sample was divided into those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and finally those scoring more than two standard deviations above the standardization sample mean, significant effects of group were seen for scores on the DASH-II anxiety subscale (F (2,79 = 3.18, p < 0.05), the DASH-II mood subscale (F (2,79) = 5.52, p < 0.01), the DASH-II mania subscale (F (2,79) = 8.69, p < 0.001), the DASH-II PDD/autism subscale (F (2,79) = 172.55, p < 0.001), the DASH-II schizophrenia subscale (F (2,79) = 3.96, p < 0.05), and the DASH-II stereotypies subscale (F (2,79) = 20.47, p < 0.001). Scheffe´ tests showed that for Anxiety subscale scores, the nonautistic and severely autistic groups differed significantly, but neither differed significantly from the moderately autistic group. For Mood subscale scores, the severely autistic group scored higher than the nonautistic and moderately autistic groups, which did not differ significantly. For Mania subscale scores, the moderately and severely autistic groups did not differ significantly, but both scored significantly higher than the nonautistic group. For the PDD/autism and Stereotypies subscales, the severely autistic group scored higher than the moderately autistic group, who in turn scored higher than the nonautistic group. Finally, for the Schizophrenia subscale, the severely autistic group scored significantly higher than the moderately autistic group but not the nonautistic group, and the nonautistic and moderately autistic groups did not differ significantly. There was no significant effect of group on scores on the DASH-II Impulse or Organic subscales (Figs. 1 and 2). Table 2 shows the percentages of participants in each group (and, for the autistic participants, the moderately autistic and severely autistic subgroups) who scored above the conventional (1 S.D. above the standardization sample mean) cutoff score on each of the DASH-II subscales. If a difference between percentages of groups exceeding the cutoff of 30% or more is taken to indicate a critical difference (Matson et al., 1996), then critical or near-critical differences were
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Fig. 1. Mean DASH-II subscale scores for autistic and nonautistic participants.
Fig. 2. Mean DASH-II subscale scores for nonautistic, moderately autistic, and severely autistic participants.
523
524
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Table 2 Percentages of participants scoring above cutoff points on DASH-II subscales Subscale
Percent nonautistic (n = 13)
Percent autistic (n = 69)
Percent moderately autistic (n = 34)
Percent severely autistic (n = 35)
Impulse Organic Anxiety Mood Mania PDD/autism Schizophrenia Stereotypies
46 31 39 46 39 0 8 39
75 46 77 64 77 100 30 86
77 35 79 47 71 100 21 79
74 57 74 80 83 100 40 91
observed between both moderately and severely autistic groups and the nonautistic group on the impulse, anxiety, mania, PDD/autism, and stereotypies subscales. On the Mood subscale, only the severely autistic group showed a critical difference from the nonautistic group; on the Schizophrenia subscale, only the severely autistic group showed a critical difference from the nonautistic group; and on the Organic subscale there were no critical differences between groups. 3.3. Means and item endorsement within the DASH-II subscales The final analyses conducted aimed to assess whether the elevated DASH-II subscale scores for individuals with high DASH-II PDD/autism subscale scores are suggestive of the presence of disorders as characterised in the DSM system (in which case one might expect all the items in a particular subscale to contribute roughly equally to the elevation in scores), or whether the elevations in scores are related to specific autism-related difficulties (in which case elevated scores might be produced by consistently high scores on a relatively small number of items). In order to do this, the number of people for whom each DASH-II item had been endorsed (i.e. score more than zero) was calculated. The percentage of individuals who endorsed each DASH-II item was compared across the ‘‘nonautistic’’ and ‘‘autistic’’ groups, i.e. the groups were defined by using the cutoff of 1 standard deviation above the PDD/autism mean for the normative sample. Items that had a 30% or more difference in endorsement between them were item 21, ‘‘amuses self with a limited set of objects or highly repetitive activities’’ (stereotypies subscale; endorsed for 87% of the autistic group and 46% of the nonautistic group), item 31, ‘‘engages in repetitive body movements’’ (stereotypies subscale; endorsed for 75% of the autistic group and 31% of the nonautistic group), item 41 ‘‘repeats the same words or sounds’’ (stereotypies subscale; endorsed for 81% of the autistic group and 46% of nonautistic group), item 49, ‘‘talks about the same subject or concern over and over’’ (stereotypies subscale; endorsed for 62% of the autistic group and 23% of the nonautistic group), item 62, ‘‘talks loudly’’ (mania subscale; endorsed for 46% of the autistic group and 8% of the nonautistic group), item 72, ‘‘complains about the absence of particular individuals’’ (mood disorders subscale; endorsed for 35% of the autistic group and 0% of the nonautistic group) and item 76, ‘‘displays rapid change in mood’’ (organic disorders subscale; endorsed for 77% of the autistic group and 46% of the nonautistic group). Items that had a 20% or more difference between them were item 5, ‘‘exhibits excessive need for attention or approval from others’’ (impulse subscale; endorsed for 59% of the autistic group and 31% of the nonautistic group), item 28, ‘‘collects or hoards objects’’ (stereotypies subscale; endorsed for 38% of the autistic group and 15% of the nonautistic group), item 35, ‘‘becomes agitated or cries
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
525
when separated from familiar people’’ (anxiety subscale; endorsed for 41% of the autistic group and 15% of the nonautistic’ group), item 36, ‘‘exhibits a period of sudden motor or vocal activity’’ (stereotypies subscale; endorsed for 62% of the autistic group and 38% of the nonautistic group), item 47, ‘‘is extremely happy or cheerful for no obvious reason’’ (mania subscale; endorsed for 72% of the autistic group and 46% of the nonautistic group) and item 50, ‘‘engages in unprovoked screaming or yelling’’ (impulse subscale; endorsed for 58% of the autistic group and 38% of the nonautistic group). 4. Discussion 4.1. Emotional and behavioural problems in autism Previous research using the DASH-II to investigate the issue of emotional and behavioural problems associated with autism in persons with severe or profound intellectual difficulties has consistently shown higher stereotypies subscale scores in persons with autism when compared with nonautistic controls (Bradley et al., 2004; Matson et al., 1996, 1999). Elevated mania, organic, anxiety and impulse subscale scores have been reported in two of the above three studies, and Bradley et al. (2004) further reported elevated mood subscale scores in their participants with autism. Using the cutoff score of one standard deviation above the normative sample mean to define autistic and nonautistic groups, the current study found the autistic group to score significantly higher than the nonautistic group on the stereotypies, mania, anxiety and organic subscales. When the sample was divided into those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and those scoring at or more than two standard deviations above the standardization sample mean, effects of group were found for the stereotypies, mania, mood, anxiety, and schizophrenia subscales. Further, the pattern of between group differences suggested that for the anxiety subscale, only the nonautistic and severely autistic groups showed significant differences in scores; for the mania subscale, both the moderately and severely autistic groups scored significantly higher than the nonautistic group; for the stereotypies subscale, all between-group differences were significant, while for the mood and schizophrenia subscales, the nonautistic and moderately autistic groups did not differ significantly in mean score, but the mean score of the severely autistic group was significantly higher than that of the nonautistic group for mood and for the moderately autistic group for schizophrenia. The fact that there were significant differences between the autistic and nonautistic group scores, and between nonautistic, moderately autistic, and severely autistic group scores on the DASH-II PDD/autism subscale is to be expected, as participants were divided into these groups on the basis of scores on this subscale. Indeed, significant differences between the groups on this subscale simply indicate that the analyses were correctly implemented. The current findings thus are consistent with those of earlier studies in suggesting that persons with severe intellectual disabilities and autism show higher stereotypies, mania, organic, mood, and anxiety DASH-II subscale scores than people with comparable levels of intellectual disability without autism. Furthermore, these results extend those of previous studies by demonstrating that these differences remain when groups are matched not only on ability (Bradley et al., 2004), but even when they are matched on overall level of behavioural disturbance as assessed by the Aberrant Behavior Checklist. In contrast to the findings of the studies by Matson et al. (1996, 1999), but consistent with those of Bradley et al. (2004), we found
526
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
however no differences between groups on the impulse subscale. The earlier pioneering studies of Matson et al. (1996) and Matson et al. (1999) compared groups of autistic and nonautistic participants with severe or profound intellectual disabilities but did not explicitly match groups either using a specific measure of level of functioning or on overall severity of behavioural disturbance. The failure of this study and that of Bradley et al. (2004) to find differences between autistic and nonautistic groups on impulse subscale scores may therefore suggest that the differences reported in earlier studies may relate to severity of intellectual disability rather than severity of autistic tendencies. Alternatively this difference in findings may relate to differences in methodology, since Matson et al. (1996, 1999) compared percentages of autistic and nonautistic participants scoring above the cutoff points on DASH-II subscales, whereas we compared subscale severity scores. Probably both factors contributed to the difference in findings. Thus, we found a 29% difference between percentages of autistic and nonautistic participants scoring above the impulse cutoff, despite failing to find a significant difference between the groups when comparing frequency scores on this subscale. Bradley et al. (2004) however used both approaches in analysing their data and found no difference between their autistic and nonautistic participants in levels of impulse subscale scores using either approach. The results of the three-group comparison may shed further light on the relationship between autistic features and emotional and behavioural disturbance. Of the five subscales showing significant effects of group, one (stereotypies) showed significant differences between each group. This result may however be largely explained by the fact that four DASH-II items contribute to both PDD/autism and stereotypies subscales. Two further items from this subscale however also showed substantial differences in endorsement between autistic and nonautistic groups (see further below), suggesting between-group differences beyond those related to the shared items. The anxiety subscale yielded a significant difference between the nonautistic and highly autistic groups, with the moderately autistic group obtaining an intermediate score, suggesting that level of anxiety increases with severity of autism. For the mania subscale, the moderately and severely autistic groups did not differ, but both scored higher than the nonautistic group, suggesting that features of mania are associated with all severities of autism. For the mood disorder and schizophrenia subscales, the severely autistic group scored significantly higher than one of the other two groups, which did not differ, suggesting that features of these disorders are associated only with more severe cases of autism. Matson et al. (1996) identified eight DASH-II items where the difference in rates of endorsement in autistic and nonautistic groups equalled or exceeded 20%, while Matson et al. (1999) identified 14 items, excluding those from the PDD/autism subscale, with endorsement rates of 40% or over for an autistic group; seven items were identified as ‘‘associated symptoms’’ in both studies. Only three of the items showing differences in rates of endorsement of 20% or more in the current study (item 47, ‘‘Is extremely happy or cheerful for no obvious reason’’, item 50 ‘‘Engages in unprovoked screaming or yelling’’, and item 76 ‘‘Displays rapid changes in mood’’) had been identified as ‘‘associated symptoms’’ earlier studies. Furthermore, this study identified four items showing high differential endorsement rates not identified as ‘‘associated symptoms’’ in earlier studies. These were item 5 (‘‘Exhibits excessive need for attention or approval from others’’), item 35 (‘‘Becomes agitated or cries when separated from familiar people’’), item 62 (‘‘Talks loudly’’) and item 72 (‘‘Complains about absence of particular individuals’’). It may be that the smaller number of ‘‘associated symptoms’’ identified in the current study by comparison with those of Matson et al. (1996, 1999) is again related to the fact that our ‘‘nonautistic’’ group was matched for overall level of behaviour disorder with the autistic group. Many of the ‘‘associated symptoms’’ identified in the earlier studies might then be
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
527
associated with presence of behavioural disturbance rather than autism per se, while the critical symptoms identified in the present study, which mainly relate to lability of mood and difficulties in interpersonal relationships, might be more specifically associated with autism. The differences between the autistic and nonautistic subgroups in patterns of item endorsement also suggest an alternative interpretation of DASH-II subscale score differences between the two groups. Although the critical ‘‘associated symptoms’’ load on to a variety of DASH-II subscales including impulse, organic, anxiety, mood, and mania, between-group differences on these subscales may relate mainly to differential endorsement of a relatively small number of items concerning lability in mood and disorders in social relationships, particularly difficulties in social attachment to specific individuals. Gillott, Furniss and Walter (2001), who compared anxiety levels in ‘‘high-functioning’’ autistic children with those in typically developing children and those with specific language disorders using the Spence Children’s Anxiety Scale (Spence, 1997), noted that the highest subscale scores for the autistic group were Separation Anxiety and Obsessive–compulsive disorder. It may be therefore that core features of autism including repetitive behaviours, difficulties in social relationships, and mood lability contribute substantially to differences between people with autism and those without on standardized measures of psychopathology across varying ability levels. With respect to stereotyped behaviours, it is of note that in both the two- and three-group comparisons between-group differences emerged on the DASH-II stereotypies subscale but not the ABC Stereotypy subscale. The ABC subscale items relate almost entirely to simple motor stereotypies, whereas the DASH-II subscale items are also concerned with restricted interests and repetitive conversational patterns which may be more specifically related to autism. Future research could usefully compare autistic and nonautistic groups on more detailed assessments of stereotyped, repetitive and obsessional behaviours. 4.2. Strengths and weaknesses of the current study The current study aimed to overcome some of the limitations highlighted in previous work in this area. Indeed, some of the main strengths of the study were that it assessed a large number of individuals with severe intellectual disabilities using a scale that has been found to be both valid and reliable for exploring emotional and behavioural disturbances in this population. However, the participants in the current study all lived in residential homes for individuals with severe intellectual disabilities and challenging behaviour and thus it is not possible to generalise results to all individuals with a severe intellectual disability. Another weakness of the current study is that no formal PDD/autism diagnostic assessment was conducted on the participants. Classification of participants as ‘autistic’ or ‘non-autistic’ was achieved solely by utilising the participants’ scores on the DASH-II PDD/autism subscale. Although this method provided a means of creating ‘autistic’ and ‘non autistic’ sub samples for the purpose of the current study, one would not expect to diagnose an individual with autism in a clinical setting based solely upon this subscale score. Future studies should utilise more formal assessments to diagnose an individual with or without autism. References Aman, M. G., & Singh, N. N. (1994). Aberrant Behavior Checklist – community: Supplementary manual. East Aurora (NY): Slosson Educational Publications.
528
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, DC: Author. Bhaumik, S., Branford, D., McGrother, C. W., & Thorp, C. F. (1997). Autistic traits in adults with learning disabilities. British Journal of Psychiatry, 170, 502–506. Bradley, E. A., Summers, J. A., Wood, H. L., & Bryson, S. E. (2004). Comparing rates of psychiatric and behavior disorders in adolescents and young adults with severe intellectual disability with and without autism. Journal of Autism and Developmental Disorders, 34, 151–161. Ghaziuddin, M., Alessi, N., & Greden, J. F. (1995). Life events and depression in children with pervasive developmental disorders. Journal of Autism and Developmental Disorders, 25, 495–502. Gillott, A., Furniss, F., & Walter, A. (2001). Anxiety in high-functioning children with autism. Autism, 5, 277–286. Kim, J. A., Szatmari, P., Bryson, S. E., Streiner, D. L., & Wilson, F. J. (2000). The prevalence of anxiety and mood problems among children with autism and Asperger syndrome. Autism, 4, 117–132. King, B. H., DeAntonio, C., McCracken, J. T., Forness, S. R., & Ackerland, V. (1994). Psychiatric consultation in severe and profound mental retardation. American Journal of Psychiatry, 151, 1802–1808. Lambert, N., Nihira, K., & Leland, H. (1993). Adaptive behavior scale – school (2nd ed.). Texas: Austin. Pro-Ed. Matson, J. L. (1995). The Diagnostic Assessment for the Severely Handicapped-Revised (DASH-II). Baton Rouge, LA: Scientific Publishers. Matson, J. L., Baglio, C. S., Smiroldo, B. B., Hamilton, M., Paclawskyj, T., Williams, D., et al. (1996). Characteristics of autism as assessed by the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 17, 135–143. Matson, J. L., Coe, D. A., Gardner, W. I., & Sovner, R. (1991a). A factor analytic study of the Diagnostic Assessment for the Severely Handicapped. Journal of Nervous and Mental Disease, 179, 553–557. Matson, J. L., Gardner, W. I., Coe, D. A., & Sovner, R. (1991b). A scale for evaluating emotional disorders in severely and profoundly mentally retarded persons. Development of the Diagnostic Assessment for the Severely Handicapped (DASH) scale. British Journal of Psychiatry, 159, 404–409. Matson, J. L., Rush, K. S., Hamilton, M., Anderson, S. J., Bamburg, J. W., Baglio, C. S., et al. (1999). Characteristics of depression as assessed by the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 20, 305–313. Matson, J. L., & Smiroldo, B. B. (1997). Validity of the mania subscale of the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 18, 221–225. Matson, J. L., Smiroldo, B. B., Hamilton, M., & Baglio, C. S. (1997). Do anxiety disorders exist in individuals with severe and profound mental retardation? Research in Developmental Disabilities, 18, 39–44. Matson, J. L., Smiroldo, B. B., & Hastings, T. L. (1998). Validity of the autism/pervasive developmental disorders subscale of the Diagnostic Assessment for the Severely Handicapped-II. Journal of Autism and Developmental Disorders, 28, 77–81. Nihira, K., Leland, H., & Lambert, N. (1993). AAMR adaptive behavior scale—residential and community (2nd ed.). Texas: Austin. [Pro-Ed]. Paclawskyj, T. R., Matson, J. L., Bamburg, J. W., & Baglio, C. S. (1997). A comparison of the Diagnostic Assessment for the Severely Handicapped-II (DASH-II) and the Aberrant Behavior Checklist (ABC). Research in Developmental Disabilities, 18, 289–298. Rojahn, J., Aman, M. G., Matson, J. L., & Mayville, E. (2003). The Aberrant Behavior Checklist and the Behavior Problems Inventory: Convergent and divergent validity. Research in Developmental Disabilities, 24, 391–404. Sevin, J. A., Matson, J. L., Williams, D., & Kirkpatrick-Sanchez, S. (1995). Reliability of emotional problems with the Diagnostic Assessment for the Severely Handicapped (DASH). British Journal of Clinical Psychology, 34, 93–94. Spence, S. H. (1997). The spence children’s anxiety Scale. In I. Sclare (Ed.), Child psychology portfolio. Windsor: NFERNelson. Tsai, L. Y. (1996). Brief report: Comorbid psychiatric disorders of autistic disorder. Journal of Autism and Developmental Disorders, 26, 159–163. World Health Organization. (1992). The ICD-10 classification of mental and behavioural disorders – clinical descriptions and diagnostic guidelines. Geneva: Author.
Patterns of emotional and behavioural disturbance associated with autistic traits in young people with severe intellectual disabilities and challenging behaviours§ Jennie Hill a,*, Frederick Furniss b,c a
Nottinghamshire Healthcare NHS Trust, Community Learning Disability Team, 53, Cheapside, Worksop, Nottinghamshire S80 2JD, UK b The Hesley Group, Doncaster, UK c School of Psychology, University of Leicester, UK Received 22 February 2005; received in revised form 16 June 2005; accepted 5 July 2005
Abstract Emotional and behavioural disturbance was assessed in 82 individuals with severe intellectual disabilities and challenging behaviour using the Diagnostic Assessment for the Severely HandicappedII (DASH-II). Levels of disturbance were compared firstly in individuals with and without features of autism as assessed by the DASH-II, and secondly in individuals with varying severities of autism. In both cases levels of ability and overall severity of behaviour disorder were comparable across groups. Individuals with autistic features were found to have significantly higher scores than nonautistic individuals on the DASH-II organic disorder, anxiety, mania, PDD/autism and stereotypies subscales. When participants with autistic features were separated into groups of severe and moderate autism and compared with nonautistic participants, significant effects of group were found for scores on the anxiety, mood, mania, PDD/autism, schizophrenia and stereotypies subscales. Scheffe´ tests were conducted to further evaluate between-group differences. Item analysis showed seven DASH-II items to have a 30% or more difference between levels of endorsement in autistic and nonautistic individuals, with six further items showing a 20% or greater difference in levels of endorsement. Findings are compared to those from previous research and implications for the conceptualisation of emotional and behavioural disorders in individuals with autism are discussed. # 2005 Elsevier Ltd. All rights reserved. Keywords: Autism; Intellectual disability; Emotional and behavioural disturbance; DASH-II
§ A preliminary version of this paper was presented to the 2004 World Congress of the International Association for the Scientific Study of Intellectual Disability. * Corresponding author. Tel.: +44 1909 530 205; fax: +44 1909 530 464. E-mail address: [email protected] (J. Hill).
0891-4222/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.ridd.2005.07.001
518
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
1. Introduction Individuals with intellectual disabilities have consistently been found to have an increased prevalence of emotional and behavioural disturbance compared to the general population. However, there have only been a limited number of studies conducted to assess whether there is an increased prevalence rate of emotional and behavioural disturbances in individuals diagnosed with autism (Ghaziuddin, Alessi, & Greden, 1995; Kim, Szatmari, Bryson, Streiner, & Wilson, 2000; Tsai, 1996). Further, these studies have tended to focus on high functioning and verbal autistic individuals who have been most able to give a verbal account of their symptoms (Bradley, Summers, Wood, & Bryson, 2004). One reason for this is that diagnosing emotional and behavioural disturbances in individuals with intellectual disabilities can be problematic (King, DeAntonio, McCracken, Forness, & Ackerland, 1994). These problems arise because existing diagnostic classificatory systems (DSM-IV, American Psychiatric Association, 1994; ICD-10-CDDG, World Health Organisation, 1992) depend substantially on the self-reported subjective experiences of the individuals being diagnosed (Bradley et al., 2004), a methodology often not feasible for individuals with severe cognitive and communication impairments. One assessment scale that can be utilised to explore emotional and behavioural disturbances in individuals with severe intellectual disabilities is the Diagnostic Assessment for the Severely Handicapped-II (DASH-II) (Matson, 1995; Matson, Gardner, Coe, & Sovner, 1991). The DASHII can be used to assess individuals with severe cognitive and communication impairments because a rater who has known the individual for a significant length of time provides information regarding behaviours associated with emotional and behavioural disturbance. The items contained in the DASH-II were formulated from both the DSM-III-R and from research relating to behavioural disturbances in individuals with severe intellectual disabilities. The DASH-II has been found to be reliable (Matson, Coe, Gardner, & Sovner, 1991; Sevin, Matson, Williams, & Kirkpatrick-Sanchez, 1995) and valid (Paclawskyj, Matson, Bamburg, & Baglio, 1997). Several of the DASH-II subscales have been found to be valid (Matson & Smiroldo, 1997; Matson, Smiroldo, & Hastings, 1998; Matson et al., 1999), although evidence for the validity of the anxiety subscale is still limited (Matson, Smiroldo, Hamilton, & Baglio, 1997). Factor analysis of the DASH-II by Matson et al. (1991b) yielded a six-factor structure with factors relating to emotional lability, aggression/conduct disorder, language disorder/verbal aggression, social withdrawal/stereotypy, eating disorder and sleep disorder. Several previous studies have utilised the DASH-II to assess whether there is an increased prevalence rate of emotional and behavioural disturbances in individuals with a diagnosis of autism or autistic traits, using subscale scores equivalent to a score of more than one standard deviation above the mean for the standardization sample (Matson, 1995) as a cutoff point indicating a level of disturbance meriting further investigation. Matson et al. (1996) compared individuals who scored over the cutoff point on the DASH-II PDD/autism subscale with members of a control group with severe and profound intellectual disabilities but no elevation in scores on the DASH-II PDD/autism subscale, and found differences of over 25% in the percentages who scored above the cutoff on the DASH-II stereotypies, mania, impulse control disorders, and organic disorders subscales. Matson et al. (1999) also found that the most common comorbid disorders as assessed by the DASH-II in a group of adults previously diagnosed with autism were impulse control disorders and stereotypies, with at least 40% of those scoring above the cutoff on the PDD/autism subscale also scoring above the cutoff on the impulse control and stereotypies subscales.
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
519
More recently, Bradley et al. (2004) conducted a similar study on 12 lower functioning autistic individuals and matched controls. They found that the participants with autism scored significantly higher on four of the five DASH-II psychiatric disorder subscales (anxiety, PDD/ autism, mania and depression) and three of the eight DASH-II aberrant behaviour subscales (stereotypies, sleep disorders and organic syndromes). They also found that 50% or more of the autistic participants scored above the cutoff point on seven of the 13 subscales (PDD/autism, mania, depression, stereotypies, self-injury, eating disorders and sleep disorders), whereas only for one subscale (self-injury) did the non-autistic participants present a comparable extent of disorder. The autistic participants each had an average of 5.25 clinically significant disorders (not including PDD/autism), whilst the non-autistic participants each had an average of 1.25 clinically significant disorders. The above research has begun to identify clinical features characteristic of autism in people with severe and profound learning disabilities. Research to date however has compared persons with and without marked autistic features but without controlling for overall severity of behaviour disorder. Given the general association between severity of autistic traits and extent of behavioural disturbance (Bhaumik, Branford, McGrother, & Thorp, 1997), it remains uncertain whether the differences between autistic and non-autistic groups reported above reflect specific patterns of comorbidity with autism or simply reflect greater generalized behavioural disturbance in persons with autism. The aims of the current study were therefore (i) to compare patterns of behavioural disturbance as assessed by the DASH-II in groups of persons with intellectual disabilities scoring above and below the cutoff point on the DASH-II PDD/autism subscale but of comparable intellectual ability and overall level of behavioural disturbance and (ii) to examine whether there is a unique set of psychiatric and behavioural traits in autistic individuals using DASH-II subscale item scores. 2. Method 2.1. Participants Participants were 82 children and young adults with severe intellectual disabilities and challenging behaviour living in four residential establishments in England. The mean age of the participants was 18 years and 2 months (minimum 8 years and 4 months; maximum 29 years and 9 months). Seventeen were female and 65 male. Each participant was assessed using either the AAMR Adaptive Behavior Scale – School (Lambert, Nihira, & Leland, 1993) or the AAMR Adaptive Behavior Scale – Residential (Nihira, Leland, & Lambert, 1993), dependent upon whether they were below or above 18 years of age respectively, to assess their adaptive abilities. The mean independent functioning standard score of the participants was 9 (minimum 1; maximum 15) and the mean language development standard score of the participants was 7 (minimum 2; maximum 16). Of the participants, 25 were taking anti-convulsant medication (32%) and 30 were taking some other psychoactive medication (39%). Medication was unknown for five of the participants. 2.2. Instruments 2.2.1. The Diagnostic Assessment for the Severely Handicapped-II (DASH-II) The DASH-II (Matson, 1995; Matson et al., 1991b) is an 84-item broad-band (Rojahn, Aman, Matson, & Mayville, 2003) behaviour rating scale. The DASH-II is a general screening
520
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
instrument aimed at assessing psychopathological symptoms relating to heterogeneous conditions such as schizophrenia, mania and autism in individuals with severe and profound intellectual and developmental disabilities. The 13 subscales of the DASH-II are (i) impulse control (17 items), (ii) organic problems (9 items), (iii) anxiety (8 items), (iv) mood disorders (15 items), (v) mania (7 items), (vi) pervasive developmental disorders/autism (6 items), (vii) schizophrenia (7 items), (viii) stereotypies (7 items), (ix) self-injurious behaviour (5 items), (x) elimination disorders (2 items), (xi) eating disorders (6 items), (xii) sleep disorders (5 items) and (xiii) sexual disorders (3 items). Each subscale item is scored on three scales: (a) a three-point frequency scale based on occurrence of the behaviour in the last 2 weeks (zero, 1–10, or over 10 occurrences), (b) a three-point duration scale concerning how long the behaviour has been a problem (less than 1 month, 1–12 months, or over 12 months), and a three-point severity scale concerning how serious the behaviour has been in the last 2 weeks (caused no disruptions or damages, caused no damages but at least one disruption to others, or caused injury or property damage at least once). 2.2.2. The Aberrant Behavior Checklist – community The ABC was used to examine overall psychological and behavioural disturbance in the participants. The ABC (Aman & Singh, 1994) is a factor analytically derived 58-item broad-band (Rojahn et al., 2003) behaviour rating scale. The ABC was originally designed to aid in the assessment of psychotropic drug treatment in individuals with intellectual and developmental disabilities, although it can also be used as a general assessment tool for problem behaviours. The five subscales of the ABC are (i) irritability, agitation, crying (15 items), (ii) lethargy, social withdrawal (16 items), (iii) stereotypic behaviour (7 items), (iv) hyperactivity, non-compliance (16 items), and (v) inappropriate speech (4 items). Each subscale item is scored on a four-point overall severity scale ranging from not at all a problem through to the problem is severe in degree. The behaviour must have appeared at least once over the past 4 weeks in order for it to be rated (i.e. given a severity score greater than 0). 2.3. Interviewers and respondents Psychology graduates administered both the ABC and the DASH-II. The respondents were direct support staff members who were either participants’ keyworkers or had worked with them for an extended period of time. 2.4. Data analysis Data analysis was undertaken in two stages. Firstly, to replicate the approach of previous studies, participants were divided into two groups, those scoring above and below the conventional (1 S.D. above the standardization sample mean) cutoff score on the DASH-II PDD/ Autism subscale. Secondly, participants were divided into three groups, those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and finally those scoring more than two standard deviations above the standardization sample mean. As in previous research (Bradley et al., 2004; Matson et al., 1996), analysis of DASH-II subscale scores was performed on frequency scores. Additionally, analysis was limited to those subscales (impulse control, organic problems, anxiety, mood disorders, mania, schizophrenia, and stereotypies), where the DASH-II subscale score is based on a sum of frequency scores, thus excluding those
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
521
subscales (self-injurious behaviour, elimination disorders, eating disorders, sleep disorders and sexual disorders) where the DASH-II is scored categorically on the basis of any single subscale item receiving a severity score greater than zero. 3. Results 3.1. Ages, abilities and levels of behavioural disturbance For the two-group comparison, the group of 13 participants scoring below the cutoff on the DASH-II PDD/autism subscale (hereafter the ‘‘nonautistic’’ group) had a mean score on that subscale of 3.38 (S.D. 0.77), while the group of 69 participants scoring above the cutoff (hereafter the ‘‘autistic’’ group) had a mean score of 7.65 (S.D. 1.91). Table 1 shows means and standard deviations for the two groups for age, Adaptive Behavior Scale Part 1 ‘‘Independent Functioning’’ and ‘‘Language Development’’ domain standard scores, and raw scores for each Aberrant Behavior Checklist subscale, together with the numbers of females and males in each group. Independent samples t-tests confirmed that the groups did not differ significantly in age (t = 0.94, d.f. = 80, p = 0.35), Independent Functioning score (t = 1.28, d.f. = 80, p = 0.21), Language Development score (t = 1.41, d.f. = 80, p = 0.16), ABC Irritability subscale score (t = 1.57, d.f. = 25.78, p = 0.13), ABC Lethargy subscale score (t = 0.68, d.f. = 80, p = 0.5), ABC Stereotypy subscale score (t = 1.31, d.f. = 80, p = 0.19), or ABC Hyperactivity subscale score (t = 1.65, d.f. = 80, p = 0.1). The autistic group however had significantly higher scores on the ABC Inappropriate Speech subscale (t = 5.35, d.f. = 55.12, p < 0.001). Chi-square analysis showed no significant association between participant gender and group membership (x2 = 1.81, d.f. = 1, n.s.). For the three-group comparison, the group of 34 participants with scores on the PDD/autism subscale greater than one standard deviation above the normative sample mean, but not more than two standard deviations above that mean (hereafter the ‘‘moderately autistic’’ group), had a mean score of 6.06 (S.D. 0.78), whereas the 35 members of the ‘‘severely autistic’’ group scoring more than two standard deviations above the normative sample mean had a mean score of 9.2 (S.D. 1.3). Table 1 again shows mean ages, relevant ABS and ABC scores, and the numbers of females and males in each group. One-way analyses of variance showed that the groups did not differ significantly in age (F (2,79) = 1.00, p = 0.37), ABS Independent Functioning score (F (2,79) = 0.95, p = 0.39), ABS Language Development score (F (2,79) = 2.3, p = 0.11), ABC Irritability subscale score (F (2,79) = 0.76, p = 0.47), ABC Lethargy subscale score (F (2,79) = 0.33, p = 0.72), ABC Stereotypy subscale score (F (2,79) = 1.01, p = 0.37), or ABC Hyperactivity subscale score (F (2,79) = 2.31, p = 0.11). The groups however differed significantly on the ABC Inappropriate Speech subscale (F (2,79) = 10.03, p < 0.001). Post hoc Scheffe´ tests showed that the severely autistic group had significantly higher scores on this subscale than the other two groups, which did not differ significantly on this measure. Chi-square analysis again showed no significant association between participant gender and group membership (x2 = 3.25, d.f. = 2, n.s.). 3.2. DASH-II subscale scores When the sample was divided into those scoring more or less than one standard deviation above the normative mean on the DASH-II PDD/autism subscale, significant differences were found between the means of the two (nonautistic and autistic) groups on the DASH-II organic
522
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Table 1 Ages and adaptive and aberrant behaviour scores of participant groups Measure
Age (months) Number (%) female/male ABS Independent Functioning (standard score) ABS Language Development (standard score) ABC Irritability (raw score) ABC Lethargy (raw score) ABC Stereotypy (raw score) ABC Hyperactivity (raw score) ABC Inappropriate Speech (raw score)
Group Nonautistic (n = 13)
Autistic (n = 69)
Moderately autistic (n = 34)
Severely autistic (n = 35)
229.92 (63.26) 5 (38%) female, 8 (62%) male 8.15 (2.44)
215.54 (47.80) 12 (17%) female, 57 (83%) male 9.0 (2.15)
209.06 (49.62) 5 (15%) female, 29 (85%) male 8.85 (2.36)
221.83 (45.81) 7 (20%) female, 28 (80%) male 9.14 (1.94)
6.82 (3.64)
8.06 (2.9)
16.88 (9.60) 11.00 (7.13) 5.82 (5.21) 17.71 (11.18) 2.41 (3.02)
17.97 (10.05) 11.83 (8.50) 5.20 (3.96) 14.51 (8.21) 4.89 (3.75)
6.08 (2.6) 14.23 (6.02) 9.85 (6.49) 3.69 (4.46) 11.23 (9.2) 0.69 (1.25)
7.45 (3.32) 17.43 11.42 5.51 16.09 3.67
(9.77) (7.81) (4.6) (9.85) (3.60)
subscale (t = 2.05, d.f. = 80, p < .05), the DASH-II anxiety subscale (t = 2.06, d.f. = 80, p < .05), the DASH-II mania subscale (t = 3.77, d.f. = 80, p < .001), the DASH-II PDD/autism subscale (t = 13.62, d.f. = 45.34, p < .001), and the DASH-II stereotypies subscale (t = 4.64, d.f. = 80, p < .001) The means of the two sub samples did not differ significantly on the DASH-II Impulse, Mood, or Schizophrenia subscales. When the sample was divided into those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and finally those scoring more than two standard deviations above the standardization sample mean, significant effects of group were seen for scores on the DASH-II anxiety subscale (F (2,79 = 3.18, p < 0.05), the DASH-II mood subscale (F (2,79) = 5.52, p < 0.01), the DASH-II mania subscale (F (2,79) = 8.69, p < 0.001), the DASH-II PDD/autism subscale (F (2,79) = 172.55, p < 0.001), the DASH-II schizophrenia subscale (F (2,79) = 3.96, p < 0.05), and the DASH-II stereotypies subscale (F (2,79) = 20.47, p < 0.001). Scheffe´ tests showed that for Anxiety subscale scores, the nonautistic and severely autistic groups differed significantly, but neither differed significantly from the moderately autistic group. For Mood subscale scores, the severely autistic group scored higher than the nonautistic and moderately autistic groups, which did not differ significantly. For Mania subscale scores, the moderately and severely autistic groups did not differ significantly, but both scored significantly higher than the nonautistic group. For the PDD/autism and Stereotypies subscales, the severely autistic group scored higher than the moderately autistic group, who in turn scored higher than the nonautistic group. Finally, for the Schizophrenia subscale, the severely autistic group scored significantly higher than the moderately autistic group but not the nonautistic group, and the nonautistic and moderately autistic groups did not differ significantly. There was no significant effect of group on scores on the DASH-II Impulse or Organic subscales (Figs. 1 and 2). Table 2 shows the percentages of participants in each group (and, for the autistic participants, the moderately autistic and severely autistic subgroups) who scored above the conventional (1 S.D. above the standardization sample mean) cutoff score on each of the DASH-II subscales. If a difference between percentages of groups exceeding the cutoff of 30% or more is taken to indicate a critical difference (Matson et al., 1996), then critical or near-critical differences were
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Fig. 1. Mean DASH-II subscale scores for autistic and nonautistic participants.
Fig. 2. Mean DASH-II subscale scores for nonautistic, moderately autistic, and severely autistic participants.
523
524
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
Table 2 Percentages of participants scoring above cutoff points on DASH-II subscales Subscale
Percent nonautistic (n = 13)
Percent autistic (n = 69)
Percent moderately autistic (n = 34)
Percent severely autistic (n = 35)
Impulse Organic Anxiety Mood Mania PDD/autism Schizophrenia Stereotypies
46 31 39 46 39 0 8 39
75 46 77 64 77 100 30 86
77 35 79 47 71 100 21 79
74 57 74 80 83 100 40 91
observed between both moderately and severely autistic groups and the nonautistic group on the impulse, anxiety, mania, PDD/autism, and stereotypies subscales. On the Mood subscale, only the severely autistic group showed a critical difference from the nonautistic group; on the Schizophrenia subscale, only the severely autistic group showed a critical difference from the nonautistic group; and on the Organic subscale there were no critical differences between groups. 3.3. Means and item endorsement within the DASH-II subscales The final analyses conducted aimed to assess whether the elevated DASH-II subscale scores for individuals with high DASH-II PDD/autism subscale scores are suggestive of the presence of disorders as characterised in the DSM system (in which case one might expect all the items in a particular subscale to contribute roughly equally to the elevation in scores), or whether the elevations in scores are related to specific autism-related difficulties (in which case elevated scores might be produced by consistently high scores on a relatively small number of items). In order to do this, the number of people for whom each DASH-II item had been endorsed (i.e. score more than zero) was calculated. The percentage of individuals who endorsed each DASH-II item was compared across the ‘‘nonautistic’’ and ‘‘autistic’’ groups, i.e. the groups were defined by using the cutoff of 1 standard deviation above the PDD/autism mean for the normative sample. Items that had a 30% or more difference in endorsement between them were item 21, ‘‘amuses self with a limited set of objects or highly repetitive activities’’ (stereotypies subscale; endorsed for 87% of the autistic group and 46% of the nonautistic group), item 31, ‘‘engages in repetitive body movements’’ (stereotypies subscale; endorsed for 75% of the autistic group and 31% of the nonautistic group), item 41 ‘‘repeats the same words or sounds’’ (stereotypies subscale; endorsed for 81% of the autistic group and 46% of nonautistic group), item 49, ‘‘talks about the same subject or concern over and over’’ (stereotypies subscale; endorsed for 62% of the autistic group and 23% of the nonautistic group), item 62, ‘‘talks loudly’’ (mania subscale; endorsed for 46% of the autistic group and 8% of the nonautistic group), item 72, ‘‘complains about the absence of particular individuals’’ (mood disorders subscale; endorsed for 35% of the autistic group and 0% of the nonautistic group) and item 76, ‘‘displays rapid change in mood’’ (organic disorders subscale; endorsed for 77% of the autistic group and 46% of the nonautistic group). Items that had a 20% or more difference between them were item 5, ‘‘exhibits excessive need for attention or approval from others’’ (impulse subscale; endorsed for 59% of the autistic group and 31% of the nonautistic group), item 28, ‘‘collects or hoards objects’’ (stereotypies subscale; endorsed for 38% of the autistic group and 15% of the nonautistic group), item 35, ‘‘becomes agitated or cries
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
525
when separated from familiar people’’ (anxiety subscale; endorsed for 41% of the autistic group and 15% of the nonautistic’ group), item 36, ‘‘exhibits a period of sudden motor or vocal activity’’ (stereotypies subscale; endorsed for 62% of the autistic group and 38% of the nonautistic group), item 47, ‘‘is extremely happy or cheerful for no obvious reason’’ (mania subscale; endorsed for 72% of the autistic group and 46% of the nonautistic group) and item 50, ‘‘engages in unprovoked screaming or yelling’’ (impulse subscale; endorsed for 58% of the autistic group and 38% of the nonautistic group). 4. Discussion 4.1. Emotional and behavioural problems in autism Previous research using the DASH-II to investigate the issue of emotional and behavioural problems associated with autism in persons with severe or profound intellectual difficulties has consistently shown higher stereotypies subscale scores in persons with autism when compared with nonautistic controls (Bradley et al., 2004; Matson et al., 1996, 1999). Elevated mania, organic, anxiety and impulse subscale scores have been reported in two of the above three studies, and Bradley et al. (2004) further reported elevated mood subscale scores in their participants with autism. Using the cutoff score of one standard deviation above the normative sample mean to define autistic and nonautistic groups, the current study found the autistic group to score significantly higher than the nonautistic group on the stereotypies, mania, anxiety and organic subscales. When the sample was divided into those scoring below the conventional cutoff score on the PDD/autism subscale, those scoring above that cutoff but less than two standard deviations above the standardization sample mean, and those scoring at or more than two standard deviations above the standardization sample mean, effects of group were found for the stereotypies, mania, mood, anxiety, and schizophrenia subscales. Further, the pattern of between group differences suggested that for the anxiety subscale, only the nonautistic and severely autistic groups showed significant differences in scores; for the mania subscale, both the moderately and severely autistic groups scored significantly higher than the nonautistic group; for the stereotypies subscale, all between-group differences were significant, while for the mood and schizophrenia subscales, the nonautistic and moderately autistic groups did not differ significantly in mean score, but the mean score of the severely autistic group was significantly higher than that of the nonautistic group for mood and for the moderately autistic group for schizophrenia. The fact that there were significant differences between the autistic and nonautistic group scores, and between nonautistic, moderately autistic, and severely autistic group scores on the DASH-II PDD/autism subscale is to be expected, as participants were divided into these groups on the basis of scores on this subscale. Indeed, significant differences between the groups on this subscale simply indicate that the analyses were correctly implemented. The current findings thus are consistent with those of earlier studies in suggesting that persons with severe intellectual disabilities and autism show higher stereotypies, mania, organic, mood, and anxiety DASH-II subscale scores than people with comparable levels of intellectual disability without autism. Furthermore, these results extend those of previous studies by demonstrating that these differences remain when groups are matched not only on ability (Bradley et al., 2004), but even when they are matched on overall level of behavioural disturbance as assessed by the Aberrant Behavior Checklist. In contrast to the findings of the studies by Matson et al. (1996, 1999), but consistent with those of Bradley et al. (2004), we found
526
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
however no differences between groups on the impulse subscale. The earlier pioneering studies of Matson et al. (1996) and Matson et al. (1999) compared groups of autistic and nonautistic participants with severe or profound intellectual disabilities but did not explicitly match groups either using a specific measure of level of functioning or on overall severity of behavioural disturbance. The failure of this study and that of Bradley et al. (2004) to find differences between autistic and nonautistic groups on impulse subscale scores may therefore suggest that the differences reported in earlier studies may relate to severity of intellectual disability rather than severity of autistic tendencies. Alternatively this difference in findings may relate to differences in methodology, since Matson et al. (1996, 1999) compared percentages of autistic and nonautistic participants scoring above the cutoff points on DASH-II subscales, whereas we compared subscale severity scores. Probably both factors contributed to the difference in findings. Thus, we found a 29% difference between percentages of autistic and nonautistic participants scoring above the impulse cutoff, despite failing to find a significant difference between the groups when comparing frequency scores on this subscale. Bradley et al. (2004) however used both approaches in analysing their data and found no difference between their autistic and nonautistic participants in levels of impulse subscale scores using either approach. The results of the three-group comparison may shed further light on the relationship between autistic features and emotional and behavioural disturbance. Of the five subscales showing significant effects of group, one (stereotypies) showed significant differences between each group. This result may however be largely explained by the fact that four DASH-II items contribute to both PDD/autism and stereotypies subscales. Two further items from this subscale however also showed substantial differences in endorsement between autistic and nonautistic groups (see further below), suggesting between-group differences beyond those related to the shared items. The anxiety subscale yielded a significant difference between the nonautistic and highly autistic groups, with the moderately autistic group obtaining an intermediate score, suggesting that level of anxiety increases with severity of autism. For the mania subscale, the moderately and severely autistic groups did not differ, but both scored higher than the nonautistic group, suggesting that features of mania are associated with all severities of autism. For the mood disorder and schizophrenia subscales, the severely autistic group scored significantly higher than one of the other two groups, which did not differ, suggesting that features of these disorders are associated only with more severe cases of autism. Matson et al. (1996) identified eight DASH-II items where the difference in rates of endorsement in autistic and nonautistic groups equalled or exceeded 20%, while Matson et al. (1999) identified 14 items, excluding those from the PDD/autism subscale, with endorsement rates of 40% or over for an autistic group; seven items were identified as ‘‘associated symptoms’’ in both studies. Only three of the items showing differences in rates of endorsement of 20% or more in the current study (item 47, ‘‘Is extremely happy or cheerful for no obvious reason’’, item 50 ‘‘Engages in unprovoked screaming or yelling’’, and item 76 ‘‘Displays rapid changes in mood’’) had been identified as ‘‘associated symptoms’’ earlier studies. Furthermore, this study identified four items showing high differential endorsement rates not identified as ‘‘associated symptoms’’ in earlier studies. These were item 5 (‘‘Exhibits excessive need for attention or approval from others’’), item 35 (‘‘Becomes agitated or cries when separated from familiar people’’), item 62 (‘‘Talks loudly’’) and item 72 (‘‘Complains about absence of particular individuals’’). It may be that the smaller number of ‘‘associated symptoms’’ identified in the current study by comparison with those of Matson et al. (1996, 1999) is again related to the fact that our ‘‘nonautistic’’ group was matched for overall level of behaviour disorder with the autistic group. Many of the ‘‘associated symptoms’’ identified in the earlier studies might then be
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
527
associated with presence of behavioural disturbance rather than autism per se, while the critical symptoms identified in the present study, which mainly relate to lability of mood and difficulties in interpersonal relationships, might be more specifically associated with autism. The differences between the autistic and nonautistic subgroups in patterns of item endorsement also suggest an alternative interpretation of DASH-II subscale score differences between the two groups. Although the critical ‘‘associated symptoms’’ load on to a variety of DASH-II subscales including impulse, organic, anxiety, mood, and mania, between-group differences on these subscales may relate mainly to differential endorsement of a relatively small number of items concerning lability in mood and disorders in social relationships, particularly difficulties in social attachment to specific individuals. Gillott, Furniss and Walter (2001), who compared anxiety levels in ‘‘high-functioning’’ autistic children with those in typically developing children and those with specific language disorders using the Spence Children’s Anxiety Scale (Spence, 1997), noted that the highest subscale scores for the autistic group were Separation Anxiety and Obsessive–compulsive disorder. It may be therefore that core features of autism including repetitive behaviours, difficulties in social relationships, and mood lability contribute substantially to differences between people with autism and those without on standardized measures of psychopathology across varying ability levels. With respect to stereotyped behaviours, it is of note that in both the two- and three-group comparisons between-group differences emerged on the DASH-II stereotypies subscale but not the ABC Stereotypy subscale. The ABC subscale items relate almost entirely to simple motor stereotypies, whereas the DASH-II subscale items are also concerned with restricted interests and repetitive conversational patterns which may be more specifically related to autism. Future research could usefully compare autistic and nonautistic groups on more detailed assessments of stereotyped, repetitive and obsessional behaviours. 4.2. Strengths and weaknesses of the current study The current study aimed to overcome some of the limitations highlighted in previous work in this area. Indeed, some of the main strengths of the study were that it assessed a large number of individuals with severe intellectual disabilities using a scale that has been found to be both valid and reliable for exploring emotional and behavioural disturbances in this population. However, the participants in the current study all lived in residential homes for individuals with severe intellectual disabilities and challenging behaviour and thus it is not possible to generalise results to all individuals with a severe intellectual disability. Another weakness of the current study is that no formal PDD/autism diagnostic assessment was conducted on the participants. Classification of participants as ‘autistic’ or ‘non-autistic’ was achieved solely by utilising the participants’ scores on the DASH-II PDD/autism subscale. Although this method provided a means of creating ‘autistic’ and ‘non autistic’ sub samples for the purpose of the current study, one would not expect to diagnose an individual with autism in a clinical setting based solely upon this subscale score. Future studies should utilise more formal assessments to diagnose an individual with or without autism. References Aman, M. G., & Singh, N. N. (1994). Aberrant Behavior Checklist – community: Supplementary manual. East Aurora (NY): Slosson Educational Publications.
528
J. Hill, F. Furniss / Research in Developmental Disabilities 27 (2006) 517–528
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, DC: Author. Bhaumik, S., Branford, D., McGrother, C. W., & Thorp, C. F. (1997). Autistic traits in adults with learning disabilities. British Journal of Psychiatry, 170, 502–506. Bradley, E. A., Summers, J. A., Wood, H. L., & Bryson, S. E. (2004). Comparing rates of psychiatric and behavior disorders in adolescents and young adults with severe intellectual disability with and without autism. Journal of Autism and Developmental Disorders, 34, 151–161. Ghaziuddin, M., Alessi, N., & Greden, J. F. (1995). Life events and depression in children with pervasive developmental disorders. Journal of Autism and Developmental Disorders, 25, 495–502. Gillott, A., Furniss, F., & Walter, A. (2001). Anxiety in high-functioning children with autism. Autism, 5, 277–286. Kim, J. A., Szatmari, P., Bryson, S. E., Streiner, D. L., & Wilson, F. J. (2000). The prevalence of anxiety and mood problems among children with autism and Asperger syndrome. Autism, 4, 117–132. King, B. H., DeAntonio, C., McCracken, J. T., Forness, S. R., & Ackerland, V. (1994). Psychiatric consultation in severe and profound mental retardation. American Journal of Psychiatry, 151, 1802–1808. Lambert, N., Nihira, K., & Leland, H. (1993). Adaptive behavior scale – school (2nd ed.). Texas: Austin. Pro-Ed. Matson, J. L. (1995). The Diagnostic Assessment for the Severely Handicapped-Revised (DASH-II). Baton Rouge, LA: Scientific Publishers. Matson, J. L., Baglio, C. S., Smiroldo, B. B., Hamilton, M., Paclawskyj, T., Williams, D., et al. (1996). Characteristics of autism as assessed by the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 17, 135–143. Matson, J. L., Coe, D. A., Gardner, W. I., & Sovner, R. (1991a). A factor analytic study of the Diagnostic Assessment for the Severely Handicapped. Journal of Nervous and Mental Disease, 179, 553–557. Matson, J. L., Gardner, W. I., Coe, D. A., & Sovner, R. (1991b). A scale for evaluating emotional disorders in severely and profoundly mentally retarded persons. Development of the Diagnostic Assessment for the Severely Handicapped (DASH) scale. British Journal of Psychiatry, 159, 404–409. Matson, J. L., Rush, K. S., Hamilton, M., Anderson, S. J., Bamburg, J. W., Baglio, C. S., et al. (1999). Characteristics of depression as assessed by the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 20, 305–313. Matson, J. L., & Smiroldo, B. B. (1997). Validity of the mania subscale of the Diagnostic Assessment for the Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 18, 221–225. Matson, J. L., Smiroldo, B. B., Hamilton, M., & Baglio, C. S. (1997). Do anxiety disorders exist in individuals with severe and profound mental retardation? Research in Developmental Disabilities, 18, 39–44. Matson, J. L., Smiroldo, B. B., & Hastings, T. L. (1998). Validity of the autism/pervasive developmental disorders subscale of the Diagnostic Assessment for the Severely Handicapped-II. Journal of Autism and Developmental Disorders, 28, 77–81. Nihira, K., Leland, H., & Lambert, N. (1993). AAMR adaptive behavior scale—residential and community (2nd ed.). Texas: Austin. [Pro-Ed]. Paclawskyj, T. R., Matson, J. L., Bamburg, J. W., & Baglio, C. S. (1997). A comparison of the Diagnostic Assessment for the Severely Handicapped-II (DASH-II) and the Aberrant Behavior Checklist (ABC). Research in Developmental Disabilities, 18, 289–298. Rojahn, J., Aman, M. G., Matson, J. L., & Mayville, E. (2003). The Aberrant Behavior Checklist and the Behavior Problems Inventory: Convergent and divergent validity. Research in Developmental Disabilities, 24, 391–404. Sevin, J. A., Matson, J. L., Williams, D., & Kirkpatrick-Sanchez, S. (1995). Reliability of emotional problems with the Diagnostic Assessment for the Severely Handicapped (DASH). British Journal of Clinical Psychology, 34, 93–94. Spence, S. H. (1997). The spence children’s anxiety Scale. In I. Sclare (Ed.), Child psychology portfolio. Windsor: NFERNelson. Tsai, L. Y. (1996). Brief report: Comorbid psychiatric disorders of autistic disorder. Journal of Autism and Developmental Disorders, 26, 159–163. World Health Organization. (1992). The ICD-10 classification of mental and behavioural disorders – clinical descriptions and diagnostic guidelines. Geneva: Author.