Emotional and Behavioural Problems in Young Children with Autism Spectrum Disorder

Susie Chandler1, 2, Patricia Howlin,2, 3 , Emily Simonoff2, 4 , Tony O’Sullivan5, Evelin Tseng6, Juliet Kennedy7; Tony Charman2 and Gillian Baird1

Affiliations:1Paediatric Neurosciences, Newcomen Centre, Guy’s & St Thomas’ NHS Foundation Trust, London, UK. King’s Health Partners2Institute of Psychiatry, Psychology & Neuroscience, London, King’s College London, UK. 3Faculty of Health Sciences, University of Sydney, Australia.4NIHR Biomedical Research Centre for Mental Health, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UK.5Lewisham & Greenwich NHS Trust, London, UK.6Bromley Healthcare CIC, London, UK. 7Oxford University Hospitals NHS Trust, Oxford, UK.

Address correspondence to: Professor Gillian Baird, Children’s Neurosciences Centre, Block D, St Thomas’ Hospital, Westminster Bridge Road, London SE1 7EH Email: gillian.baird@gstt.nhs.uk Tel: 0207 188 4645

Funding source: Clothworkers’ Foundation, brokered by Research AutismGrant number: R011217 Autism M10 2011/12

Word count: 3295 (text, excluding abstract and refs)Refs: 32Tables: 3Supporting information: Figure 1; Appendices 1-4

1

ABSTRACT Aim To assess the frequency, pervasiveness, associated features and persistence of emotional and behavioural problems in a community sample of young children with autism spectrum disorder (ASD). Method: Parents (N= 277) and teachers (N=228) of 4-8 year olds completed the Developmental Behaviour Checklist (DBC). Intellectual ability and autism symptomatology were also assessed. A subsample repeated the DBC.Results Three quarters of the cohort scored above the clinical cut-off on the parent questionnaire (DBC-P); almost two thirds of these scored above cut-off on the teacher questionnaire (DBC-T). In 81% problems persisted above threshold 14 months later. Higher DBC-P scores were associated with greater autism symptomatology, higher deprivation index, parental unemployment and more children in the home but not with parental education or ethnicity, or child’s age or sex. Children with IQ > 70 scored higher for disruptive behaviour, depression and anxiety symptoms; those with IQ < 70 scored higher for self-absorption and hyperactivity. Interpretation The DBC identifies a range of additional behaviour problems that are common in ASD and which could be the focus for specific intervention. The results highlight the potential benefit of systematic screening for co-existing problems.

What this paper adds Three quarters of children aged 4-8 years with autism have parent

reported emotional, behavioural and functional problems Two thirds of these also have teacher reported problems. Problems were persistent over time in 80% and identification was possible

at 4-5 years. The Developmental Behaviour Checklist assesses a wide range of

difficulties and is acceptable to parents and teachers of children with ASD. Systematic use of such a questionnaire is recommended in routine clinical

care of children with ASD

Running title: Emotional and behavioural problems in ASD

2

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterised by impairments in reciprocal social interaction, communication and repetitive, stereotyped interests and behaviours1. Recent estimates suggest a prevalence rate of over 1%2. ASD is associated with high morbidity and significant costs to individuals, families and services3. Emotional and behavioural problems in ASD are frequent 4 and rates of social anxiety, attention deficit hyperactivity disorder and oppositional disorder are elevated compared with both the general population5 and children with intellectual disability (ID) 6, 7, 8. Functional problems in sleeping, eating and toileting are also increased. High rates of behavioural and emotional disturbance are evident from pre-school onwards9 and many children have two or more comorbid problems10, 11. In non-ASD populations with and without ID, prevalence and types of emotional and behavioural difficulties are associated with age, sex, IQ, ethnic group and social disadvantage5, 7. In ASD these factors are less clearly linked to psychiatric comorbidity6, 11 although the type of difficulty does vary with IQ, age and sex6. Psychiatric disorders also have a high prevalence in adults12 and, together with functional problems, can have as great an adverse impact as core ASD symptoms. There is, nevertheless, a range of evidence-based interventions that could reduce the effects of these problems 13. Hence systematic assessment of emotional/behavioural difficulties by clinicians is crucial. This paper describes the use of parent and teacher questionnaires to assess comorbid emotional and behavioural problems in a community sample of young children (age 4-9 years) with ASD.

METHODSThe study was approved by Guy’s Hospital Research Ethics Committee

(08/H0804/37) and Bromley and Lewisham Local Research Ethics Committees (RDLEWBR 428). 

Recruitment procedures The target population comprised all children born between 01/09/2000 and

01/09/2004 (aged 4-8 years at time of recruitment) with an ASD diagnosis made by the local multidisciplinary teams and living in two London boroughs with a broad range of socio-economic circumstances and ethnic backgrounds. In both boroughs children suspected of ASD receive a neurodevelopmental assessment from the community paediatrician. A small minority are diagnosed at this appointment but most are referred onto the multidisciplinary communication clinic. Following a speech and language assessment and report from nursery/school,.children then attend the multidisciplinary clinic consisting of paediatrician, speech and language therapist or psychologist for a formal assessment of features specific to ASD. Measures include the Autism Diagnostic Interview-Revised (ADI-R)14, Developmental, Diagnostic and Dimensional Interview (3di)15, Diagnostic Interview for Social Communication Disorder (DISCO)16 or local proforma and the Autism Diagnostic Observation Schedule (ADOS)17.

3

Eligible families (N=447) were mailed invitations to participate together with consent forms. Consenting families were mailed the questionnaire pack. Non-responders were telephoned and re-mailed. Consenting families were mailed the questionnaire pack. If they still failed to respond, researchers telephoned again and offered to re-mail the questionnaires or complete them with parents over the telephone or at home. To assess pervasiveness of problems teachers were asked to complete the teacher version of the questionnaire. To assess the persistence of difficulties a subset of parents completed the questionnaire on two occasions (Times 1 and 2). For the Time 2 assessment girls were over-sampled in order to provide adequate numbers for analysis (further details in Supporting Information, Appendix 1).

MeasuresEmotional and behaviour problems: A focus group of eight parents with children with ASD aged < 10 years was recruited to determine which of a number of questionnaires, commonly used in clinical practice in the UK, best described their child’s emotional and/or behavioural problems and was most acceptable to them. (Details in Supporting Information, Appendix 2). The Developmental Behaviour Checklist (DBC18) was selected as the preferred instrument for the study. Parents rated it positively in terms of acceptability, ease of use, clarity of questions, scoring and range of questions and there are parallel parent and teacher versions that allow assessment of problems across settings.

The DBC Primary Carer Version (DBC-P) is a 96-item behaviour checklist covering a broad range of behaviours, each rated as 0 (“not true as far as I know”), 1 (“somewhat or sometimes true”) or, 2 (“very or often true”). It has satisfactory internal consistency, inter-rater and test-retest reliability and concurrent validity with other psychopathology measures, particularly in samples of children and adolescents with ID19, 20. The DBC has five empirically-derived subscales (Disruptive/Antisocial; Self-absorbed; Communication disturbance; Anxiety; Social-relating), and four scales that can be used to identify specific psychiatric syndromes (Autism Screening Algorithm (ASA); Depression scale; Hyperactivity scale, and Anxious Behaviour Rating Scale (ABRS). The total behaviour problem score (range 0-192) indicates overall level of disturbance; a cut-off score ≥ 46 is recommended for identifying clinically significant emotional and/or behavioural problems. The teacher version (DBC-T) has 94 items; a total score ≥30 is recommended as the cut-off for identifying caseness.

Autism symptomatology was measured using the parent rated Social Communication Questionnaire Lifetime version (SCQ) 21. It comprises 39 items scored 0 or 1; a cut-off of >15 is recommended for identifying potential ASD cases.

Demographic characteristics were collected by means of a parent questionnaire. Postcode data were used to access a deprivation index (The English Indices of Deprivation 2007)22 which ranks small geographical areas

4

across England according to deprivation level, based on income, education, employment and living environment (higher index score = greater deprivation).

Child assessments: IQ was measured by the research team, trained and experienced in psychometric assessments, using either the Wechsler Intelligence Scale for children (WISC-IV)23, Wechsler Preschool and Primary Scale of Intelligence (WPPSI-III)24 or Mullen Scales of Early Learning (MSEL)25 depending on chronological age and developmental level (see Supporting Information, Appendix 2). Parents were also asked “At what age do you think your child is functioning overall?” The British Picture Vocabulary Scale (BPVS-II) 26 was used to measure receptive language for children who had sufficient language to access the test.

Data analysis (see Supporting Information, Appendix 1 for full details) Statistical analyses were carried out using Stata 1127. For some analyses

IQ was treated as continuous and for others as a dichotomous variable (< 70 / > 70). For sleeping, eating and toileting problems, the “somewhat” / “sometimes” and “very often”/ often” categories of the DBC were combined to create a dichotomous score. Associations between DBC-P/T scores and background factors (number of children at home; parent education, employment and ethnicity) were examined using multiple linear regression. Multinomial logistic regression was used to explore factors associated with discrepancies (i.e. above/below clinical cut-off) between primary carer (hereafter referred to as parent) and teacher questionnaires, and also to explore factors associated with changes in parent questionnaires (i.e. above/below clinical cut off) at Times 1 and 2.

RESULTSParticipant characteristics

Responses were received from 362 of 447 (81%) families (see Supporting Information, Figure 1). Parental questionnaires were completed on 277 children (62%), half by post, the remainder by phone or home visit. Non-participating families had a higher mean deprivation index (t (443) =3.39,p<.001) and contained a higher proportion of boys ( X2(1, N=447) = 5.36, p=.021. (See Supporting Information, Appendix 3 for details of participants and non-participants at initial assessment). Mean age of child participants at recruitment was 6.0 years (SD 1.1) and at first assessment 6.8 years (SD 1.2). The majority (77%, n=214) attended mainstream school; 44 attended special schools; 15 attended special units within mainstream schools; three were home-schooled; one attended mainstream nursery. The clinical records of any child with a low score on the SCQ score (i.e. total < 10; n=28) were checked with local clinicians; diagnosis was confirmed in all cases and none was excluded from the sample.

Cognitive assessments were completed for 258 children. Full-scale IQ scores were calculated for 211 (67 on WISC-IV23; 144 on WPPSI-III24). For children tested on the Mullen Scales of Early Learning (MSEL25, the composite standard score was used when possible (n=13); for those above the MSEL age

5

range (n=34) age-equivalents were used to derive a ratio IQ. Those with an MSEL ratio IQ < 20 (n=11) were assigned an IQ of 19 to reflect their very low ability. Parental estimates of functional age were also used to generate a parent-estimated DQ for 15 children. (Details in Supporting Information, Appendix 4)). The mean IQ estimate for the sample was 72.7; 35% had an IQ < 70; 21.6% had IQ<50. Missing cognitive data were due to lack of parental consent or child compliance issues.

Parental reports of emotional and behavioural problems

--- Table 1 about here ---

DBC-P scores are presented in Table 1. The internal consistency of each of the subscales (Disruptive/antisocial, Self-absorbed, Communication Disturbance, Anxiety, Social Relating), the ASA, Depression and Hyperactivity scales, and ABRS was good to excellent (Chronbach’s alpha range = .66 - .91, mean = .80). Mean total behaviour problem score was 71.2 (SD 29.6); 79% scored above cut-off, (95% at age 4; 83% at age 5; 76% at age 6; 75% at age 7; 80% at age 8),significantly higher than the 41% who scored above cut-off in the Australian normative ID sample aged 4-8 years15 (p<.001)

There was no significant difference in total problems scores for children with IQ above or below 70 (t (271) = 0.56, p = .58) but there were some group differences in subscale scores. Thus, children with IQ > 70 scored higher on the scales for Disruptive/Antisocial behaviour (t (271) = -2.70, p =.007), Depression (t (271) = -2.55, p =.012), and Anxious Behaviour (t (271) = 2.30, p =.004). Children with IQ < 70 scored more highly on the Self-absorbed (t (271) = -2.70, p<.0001) and Hyperactivity scales (t (271) = 2.30, p=.022).

Seventy percent of the sample was reported to be fussy eaters/have food fads; disrupted sleep was reported in 59% and soiling/smearing in 24%. Soiling/smearing was more common in younger children (i.e., <7years vs > 7 years, X2 = 3.84, p = .050); no other sleeping/eating/toileting problems were associated with age (all p > .2). No significant sex differences were found between total or subscale scores (all p>.3).

--- Table 2 about here ---

There was no correlation between total DBC-P score and age, IQ, BPVS-II standard score or sex (all correlations < .10; see Table 2). However, there was a significant correlation with greater autism symptoms on the SCQ (r = .58; p < .001), and with higher deprivation index score (although the correlation was small (r=.15; p = .01). Multiple regression also showed DBC-P score to be associated with parental unemployment (β = -11.59 p = .008) and more children in the home (β = 5.94, p = .008). There was no association between DBC-P score and parental education (p = .48) or ethnicity (p = .14).

6

Almost half (44%) the parents indicated that they had sought help for their child’s behaviour or emotional problems. Most (86%) of those who had sought help had children who scored above the DBC-P clinical cut-off. However, over one third (37%) of those whose children scored above cut-off had not sought any help.

Teacher reports of emotional and behaviour problemsDBC-T data were available for 228 of the children with parent data (See

Table 2). Missing data were largely due to teachers not returning the questionnaire. Children lacking teacher data were slightly older than those with teacher data (M = 7.1 years, SD 1.1 and 6.7 years, SD 1.2 respectively, t (275) = 2.1, p = .04), but they did not differ in terms of SCQ score (p = .76) or sex (p = .12).

Mean total score on the DBC-T was 38.1 (SD 21.1); 62% scored above the cut-off for major emotional and/or behavioural disturbance. Mean total and subscale scores for this ASD sample were significantly higher than those for the normative ID sample 15 (p < .05 in each case). Teachers rated children with IQ < 70 as being more self-absorbed (t(226) = 7.48 p < .0001), and having more problems with communication (t (226) = 4.22, p < .0001) and social-relating (t(226) = 3.2, p =.002) than those with IQ > 70 (see Table 2). There was no significant relationship between IQ and DBC-T Disruptive/Antisocial or Anxiety subscale scores (p = .67 and .08, respectively). No sex differences were found on DBC-T subscale scores (t-tests, all p > .1).

Unlike the parent ratings, total DBC-T score was negatively correlated with IQ (r =-. 34; p < .001) and BPVS-II standard scores (r=-.23; p = .002) although these correlations were weak to moderate (See Table 3). There was no correlation between DBC-T scores and autism severity (SCQ. r = .06; p = .33). Multiple linear regression showed higher DBC-T scores were associated with greater family size/more children (F(4, 200) = 2.75, p = .03, β = -5.38, p = .002) but not with any other family factors or school placement.

Agreement between parent and teacher reports Two thirds (64%) of children above clinical cut-off on the DBC-P were also

above the cut-off on the DBC-T, indicating that significantly more parents than teachers rated children as having emotional/behavioural problems (one-sample proportion test, z (N=228) = -6.2, p < .001).

There was no significant correlation between DBC-P and DBC-T total scores (r = .07, p = .29). The discrepancy between caseness (reaching cut-off) on the DBC-P and the DBC-T was examined using multinomial regression for each of the four possible concordance/discordance categories (see Supporting Information, Appendix 3 for details). Being above cut-off on the DBC-P but not the DBC-T, relative to being above cut-off on both, was associated with living in a

7

home with more children (Wald (1) = 17.4, p < .001), higher IQ (Wald (1) =11.4, p < .005) and female sex (Wald (1) = 9.4, p < .005).

Persistence of emotional and behavioural problems over time

--Table 3 about here—

Ninety-three families repeated the DBC-P after an average interval of 14 months, (SD 4.6; range 2-23 months). They did not differ from the other 184 in terms of age or deprivation index (t-tests, p = .12 and .17, respectively) but they had a slightly lower mean IQ (67.6 versus 75.3, p = .02) and differed, by design, in the percentage of males (57% vs. 77%, X2= 58.9, p < .0001) because females were deliberately oversampled to obtain sufficient numbers for gender analysis. There was a reduction in total and subgroup scores from T1 to T2 (see Table 3) but 81% of those above cut-off initially remained above cut-off at T2. Scores at T1 and T2 were highly correlated (r(93) = .79 p < .001). Three children below cut-off at T1 scored above cut-off at T2; 14 children were above cut-off at T1 but not T2. Multinomial regression analysis on T1/T2 concordance (see Supporting Information, Appendix 3) indicated that scoring above DBC-P cut-off at T1 but not T2 relative to those scoring above cut-off at both time-points was associated with lower SCQ total (Wald(1) = 6.12, p = .013). No family variables (including whether or not parents reported seeking help for the problems at T1) were related to change over time (i.e. DBC-P at T1 – DBC-P at T2).

DISCUSSION As in previous research 6, 11 the present study identified high levels of

emotional and behavioural problems in young children with ASD. Three-quarters scored above clinical cut-off on the DBC-P. High rates of functional problems were also reported compared with children of similar age; for example disrupted sleep in 59%; 70% fussy eaters, and 24% and 34% respectively having soiling and urinating problems (despite being toilet trained).

Two-thirds of children with ASD who were above the DBC-P cut- off were also above cut-off on the teacher report, indicating pervasiveness of problems. Problems were persistent. Most children (81%) initially above cut-off remained above cut-off after an average period of 14 months, although total problem scores reduced somewhat over time, consistent with previous studies28, 29. Nevertheless, less than half the parents/carers of these children had sought help for these difficulties.

The lack of association between overall rate of parent-reported emotional and behaviour problems and child factors (i.e. sex, IQ, language, age) is consistent with the literature in ASD6,11 and different from both general population samples and clinical groups with and without intellectual disability5, 7. However, in contrast to Simonoff et al.11, severity of ASD in this study was associated with

8

parental report of problems (although not teacher report). Possible explanations are differences in age of the participants and the measures used.

Consistent with studies of the general population5 and of ID7 and some studies of ASD 30, though not all 11, psychosocial factors (higher deprivation, parental unemployment and number of children in the home but not parental education or ethnicity) were associated with higher parental ratings of emotional and behaviour problems. Also consistent with other studies, type of symptoms varied with IQ. Children with IQ < 70 were more self-absorbed than others6, children with IQ > 70 scored more highly for depression, anxiety31 and ADHD31. Higher DBC-T scores were associated with greater family size but no other family factors. Discrepancy between parental and teacher reports of overall problem severity (DBC-P>DBC-T) tended to be related to female sex, higher IQ and larger family size. It is possible that children with higher IQ respond positively to the structure and expectations of school; alternatively they may contain or suppress any anxiety or distress they experience in school and thus behave differently at home and in school. Different behaviour across home and school situations is well recognised in child psychopathology and there is anecdotal evidence that females with ASD may be more concerned to conform in school; therefore parent experience is of greater behaviour problems.. Type of school was not associated with either parent or teacher reports.

The strengths of this study include: a large, total population derived cohort (although non-participants included more males and had higher deprivation index, thus possibly underestimating total problems); young age of the children; standardised IQ scores available for almost all; data collected from both parents and teachers; and follow-up measures obtained for a subgroup of children. Limitations include: no confirmation of clinical diagnoses using a research diagnostic assessment (although the SCQ was used); the small number of females available to assess possible sex differences; lack of an adaptive behaviour measure; and reliance on checklist data rather than clinician validation of symptoms. Furthermore, the DBC was developed for use with children with intellectual disability. Our sample included children across the full IQ range, indeed 65% had an IQ<70. While it is possible that the DBC may not have been sensitive enough to detect all problems in children with higher IQs, we found no relationship between IQ and overall problem levels and the variation in patterns of problems according to IQ was consistent with other studies.

Autism is currently being diagnosed at ever younger ages32. Our findings, alongside those of others, show that significant comorbid emotional and behaviour problems can be detected as early as age 4 to 5 years. These difficulties occur both at home and school and persist over time. The identification of co-morbid problems is important for parents and crucial in the management of ASD as the impact of these additional problems can be reduced using a range of evidence-based interventions 13. Consistent with NICE guidance 13 , systematic eliciting of co-occurring problems by clinicians as part of routine assessment and care in ASD is recommended. . Although a number of behaviour

9

measures exists, parents of children across the IQ range found the DBC easy to use and understand and it covers a range of behaviours including sleeping, eating and toileting problems. The availability of both parent and teacher versions increases its value for measuring problems in different settings.

10

AcknowledgementsWe thank Lewisham Autism Support and the Complex Communication

Diagnostic Service in Bromley for their support. We thank all the families and schools who participated. This paper represents independent research part funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London and Research Autism. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.

References

1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author; 2013.

2. Centers for Disease Control and Prevention. Prevalence of autism spectrum disorder among children aged 8 Years. Morbidity and Mortality Weekly Report, Surveillance Summaries 2014, 63(SS02):1-21.

3. Buescher AV, Cidav Z, Knapp M, Mandell DS. Costs of autism spectrum disorders in the United Kingdom and the United States. JAMA Paediatr 2014; 168: 721-8.

4. Matson JL & Cervantes PE. Commonly studied comorbid psychopathologies among persons with autism spectrum disorder. Res Dev Disabil 2014; 35: 952-962.

5. Green H, McGinty A, Meltzer H, Ford T, Goodman R. Mental health of children and young people in Great Britain. Basingstoke, UK: Palgrave Macmillan; 2005.

6. Brereton AV, Tonge JT, Einfeld S. Psychopathology in children and adolescents with autism compared to young people with intellectual disability. J Autism Dev Disord 2006; 36: 863-870.

7. Emerson E & Hatton C The mental Health of Children and Adolescents with Intellectual Disabilities in Britain. Institute for Health Research. Lancaster University, UK; 2007

8. Totsika V, Hasting R, Emerson E, Lancaster G, Berridge D. A population-based investigation of behavioural and emotional problems and maternal mental health: associations with autism spectrum disorder and intellectual disability. J Child Psychol Psychiatry 2011; 52: 91-99.

11

9. Herring S, Gray K, Taffe J, Tonge B, Sweeney D, Einfeld S. (2006). Behavioural and emotional problems in toddlers with pervasive developmental disorders and developmental delay and association with parental mental health. J Intellect Disabil Res 2006; 50: 874-82.

10.Maskey M, Warnell F, Parr JR, LeCouteur, A. McConachie, H. Emotional and behavioural problems in children with autism spectrum disorder. J Autism Dev Disord 2013; 43: 851-859.

11.Simonoff E, Pickles A, Charman T, Chandler S, Loucas T, Baird G. Psychiatric disorders in children with autism spectrum disorders: Prevalence, comorbidity, and associated factors in a population-derived sample. J Am Acad Child Adolesc Psychiatry 2008; 47: 921-929.

12. Buck TR, Viskochil J, Farley M, Coon H, McMahon WM., Morgan J, Bilder, DA. Psychiatric comorbidity and medication use in adults with autism spectrum disorder. J Autism Dev Disord 2014; 44: 3063-71.

13. National Institute for Health and Care Excellence. Autism diagnosis in children and young people: recognition, referral and diagnosis of children and young people on the autism spectrum. CG 128 and the management and support of children and young people on the autism spectrum CG 170 . London: National Institute for Health and Care Excellence; 2011 and 2013.

14. Rutter M, LeCouteur A, Lord C. The Autism Diagnostic Interview-Revised (ADI-R). Los Angeles, CA: Western Psychological Services; 2003.

15. Skuse D, Warrington R, Bishop D, et al. The developmental, dimensional and diagnostic interview (3di): A novel computerized assessment for autism spectrum disorders. J Am Acad Child Adolesc Psychiatry 2004;43:548–58.

16. Wing L, Leekam SR, Libby SJ, et al. The Diagnostic Interview for Social and Communication Disorders (DISCO): background, inter-rater reliability and clinical use. J Child Psychol Psychiatry 2002;43:307–25.

17. Lord C, Risi S, Lambrecht L, et al. The Autism Diagnostic Observation Schedule-Generic (ADOS-G): A standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord 2000; 30, 205–23

18. Einfeld SL, Tonge BJ. Manual for the Developmental Behaviour Checklist: Primary Carer Version (DBC-P) and Teacher Version (DBC-T). 2nd ed. Melbourne: Monash University Centre for Developmental Psychiatry and Psychology; 2002

12

19. Einfeld SL, Tonge BJ. The developmental behaviour checklist: the development and validation of an instrument to access behavioural and emotional disturbance in children and adolescents with mental retardation. J Autism Dev Disord 1995; 25:81-104.

20. Einfeld SL, Tonge BJ. Population prevalence of psychopathology in children and adolescents with intellectual disability: II epidemiological findings. J Intellect Disabil Res 1996; 40: 99-109.

21. Rutter M, Bailey A, Lord C. The Social Communication Questionnaire. 1st ed. Los Angeles: Western Psychological Services; 2003.

22. Department for Communities and Local Government. The English Indices of Deprivation 2007.http://webarchive.nationalarchives.gov.uk/20100410180038/http://communities.gov.uk/communities/neighbourhoodrenewal/deprivation/deprivation07/

23. Wechsler D. Wechsler Intelligence Scale for Children (IV-UK edition). London, UK: The Psychological Corporation; 2003.

24. Wechsler D. Wechsler Pre-school and Primary Scale of Intelligence (III-UK edition). San Antonio, TX: Harcourt Assessment Inc; 2002.

25. Mullen EM. Mullen Scales of Early Learning (AGS ed.). Circle Pines, MN: American Guidance Service Inc; 1995.

26. Dunn LM, Dunn LM, Whetton C, Burley J. British picture vocabulary scale, 2nd ed. London, UK: NFER-Nelson; 1997.

27. StatCorp. Stata Statistical Software. Release 11. College Station, TX: StataCorp LP; 2009.

28. Simonoff E, Jones CR , Baird G, Pickles A, Happe F, Charman T. The persistence and stability of psychiatric problems in adolescents with autism spectrum disorders. J Child Psychol Psychiatry 2013; 54: 186-94.

29. Gray K, Keating C, Taffe J, Brereton A. Trajectory of behavior and emotional problems in autism. Am J Intellect Dev Disabil 2012; 117: 121-133.

30. Midouhas E, Yogaratnam A, Flouri E, Charman T. Trajectories of psychopathology in young children with autism spectrum disorder: the role of

13

family poverty and parenting. J Am Acad Child Adolesc Psychiatry 2013; 52:1057-1065.

31. Mazzone L, Ruta L, Reale L. Psychiatric comorbidities in Asperger syndrome and high functioning autism: diagnostic challenges. Ann Gen Psychiatry 2012; 11:16.

32. Daniels, AM, Mandell, DS. Explaining differences in age at autism spectrum disorder diagnosis: A critical review. Autism 2014; 18: 583-597

14

Table 1 : Mean DBC-P and DBC-T scores by IQ group and gender.

IQ<70N=96

IQ>70a

N=177MalesN=227

FemalesN=50

Total sampleN= 277

mean scores (SD)DBC-P:

Total Behaviour Problem Score 72.3 (24.8) 70.2 (32.1) 71.6 (30.4) 69.1 (25.6) 71.2 (29.6)

Disruptive/Antisocial 19.5 (9.1) * 23.2(11.4) * 22.1 (11.0) 21.5 (10.0) 22.0 (10.8)

Self-absorbed 27.4 (11.5)** 19.9 (11.5)** 22.8 (12.2) 21.8 (11.4) 22.6 (12.0)

Communication disturbance 10.6 (4.7) 11.1 (5.1) 11.2 (5.1) 10 (4.6) 11.0 (5.0)

Anxiety 7.2 (3.5) 8.2 (4.5) 7.8 (4.2) 8.0 (4.0) 7.8 (4.2)

Social Relating 6.7 (2.8) 6.0 (3.3) 6.3 (3.3) 6.1 (2.9) 6.3 (3.2)

Autism Screening Algorithm 31.0 (9.8)** 25.8 (11.4)** 28.1 (11.3) 26.5 (10.4) 27.8 (11.1)

Depression Scale 5.2 (3.1) * 6.4 (3.8) * 6.0 (3.7) 6.1 (3.0) 6.0 (3.6)

Hyperactivity Scale 8.3 (2.8) * 7.4 (3.3) * 7.8 (3.2) 7.3 (2.8) 7.7 (3.1)

Anxious Behaviour Rating Scale 5.1 (3.1) * 6.3 (3.6) * 5.9 (3.5) 5.9 (3.6) 5.9 (3.5)

% sleeps too little/disrupted sleep 59.6 57.1 57.8 62.0 58.6

% has nightmares/night terrors/sleepwalks 27.1* 45.5* 27.1 45.8 39.7

% fussy eater/has food fads 66.7 71.0 69.0 72.0 69.6

% gorges on food 31.3* 18.6* 22.9 24.0 23.1

% urinates outside the toilet 39..6 32.2 36.1 26.0 34.3

15

% soils outside the toilet/smears 39.6** 15.8** 25.1 20.0 24.2

DBC-TTotal Behaviour Problem Score 45.7 (18.4)** 34.0 (21.4)** 38.6 (20.2) 36.2 (24.6) 38.1 (21.1)

Disruptive/Antisocial 11.8 (7.2) 11.3 (8.9) 11.5 (8.4) 11.1 (8.4) 11.4 (8.3)

Self-absorbed 17.3 (8.9)** 9.0 (7.5)** 12.1 (8.4) 11.3 (10.8) 12.0 (8.9)

Communication disturbance 6.7 (3.9)** 4.5 (3.7)** 5.5 (3.9) 4.5 (4.1) 5.3 (3.9)

Anxiety 3.8 (2.5) 3.2 (2.7) 3.3 (2.6) 4.0 (2.8) 3.4 (2.6)

Social Relating 5.9 (3.3)* 4.4 (3.3)* 5.0 (3.4) 4.3 (3.4) 4.9 (3.4)

Autism Screening Algorithm 20.7 (9.0)** 12.6 (8.9)** 15.7 (9.2) 14.6 (11.8) 15.5 (9.8)

Depression Scale 2.9 (2.2) 3.1 (2.5) 2.9 (2.4) 3.6 (2.5) 3.0 (2.4)

Hyperactivity Scale 6.0 (3.1)** 4.5 (3.4)** 5.2 (3.3) 4.2 (3.4) 5.0 (3.4)

Anxious Behaviour Rating Scale 2.7 (2.3) 2.8 (2.5) 2.7 (2.3) 3.3 (2.6) 2.8 (2.4)

* t-test (means comparison) or chi-square (% comparison), p < .05

** t-test (means comparison) or chi-square (% comparison), p < .001a Total N with IQ data = 273 (4 cases missing IQ data)

16

Table 2: DBC-P and DBC-T total scores: Pearson’s correlations with age, IQ, SCQ and deprivation index, and regression analysis for effect of background factors

N r value p valueDBC-P total score with:

IQ 273 -.03 .61

BPVS standard score 209 -.05 .83

Child’s age 277 -.02 .75

SCQ total score 277 .58 <.0001

DBC-T total score 228 .07 .29

Deprivation index 227 .13 .01

DBC-T total score with:IQ 228 -.34 <.0001

BPVS standard score 187 -.23 .002

Child’s age 228 .03 .67

SCQ total score 228 .06 .33

Deprivation index 228 -.02 .08

DBC-P total score (n=250) Coefficient (SE) t p-value

Parent education -3.03 (3.80) -0.80 .42

Parent employ -13.29 (4.30) -3.09 .002

Ethnicity -0.006 (3.79) -0.00 .999

Num children 6.42 (2.25) 2.86 .005

DBC-T total score (n=205)

Parent education -2.47 (2.96) -0.84 .41

Parent employ -1.15 (3.30) -0.35 .73

Ethnicity 1.54 (2.91) 0.53 .60

Num children -5.40 (1.75) -3.08 .002

17

Table 3: DBC-P scores at Times 1 and 2 (n=93)

Mean score (SD)Time 1

Mean score (SD)Time 2

r value

Total Behaviour Problem Score 71.8 (27.5)** 59.4 (2.6)** .79***

Disruptive/Antisocial 21.8 (10.3)** 17.8 (9.6)** .79***

Self-absorbed 23.5 (12.2)** 19.2 (10.8)** .78***

Communication disturbance 10.7 (4.7)** 9.1 (2.8)** .66***

Anxiety 8.1 (3.9)** 6.6 (3.4)** .75***

Social Relating 6.3 (3.0)* 5.6 (2.8)* .51***

Autism Screening Algorithm 28.0 (10.6)** 24.0 (10.5)** .78***

Depression Scale 6.0 (3.4)** 4.7 (2.9)** .78***

Hyperactivity Scale 7.7 (3.1)** 6.8 (3.2)** .70***

Anxious Behaviour Rating Scale 6.0 (3.3)** 5.2 (3.1)** .77***

* Paired t-test, p < .05

**Paired t-test, p < .001

***Pairwise correlation, p < .0001

18