Severe mood problems in adolescentswith autism spectrum disorder

Emily Simonoff, 1 Catherine R.G. Jones, 2 Andrew Pickles, 3 Francesca Happe, 4

Gillian Baird, 5 and Tony Charman6

1Department of Child and Adolescent Psychiatry, King’s College London, Institute of Psychiatry and NIHR BiomedicalResearch Centre for Mental Health, De Crespigny Park, London, UK; 2Department of Psychology, University of Essex,Wivenhoe Park, Colchester, Essex, UK; 3Department of Biostatistics, King’s College London, Institute of Psychiatry,London, UK; 4MRC SDGP Research Centre, King’s College London, Institute of Psychiatry, London, UK; 5Guy’s & StThomas’ NHS Foundation Trust, Newcomen Centre, London, UK; 6Centre for Research in Autism and Education,

Institute of Education, London, UK

Introduction: Severe mood dysregulation and problems (SMP) in otherwise typically developing youthare recognized as an important mental health problem with a distinct set of clinical features, familyhistory and neurocognitive characteristics. SMP in people with autism spectrum disorders (ASDs) havenot previously been explored. Method: We studied a longitudinal, population-based cohort ofadolescents with ASD in which we collected parent-reported symptoms of SMP that included rage, lowand labile mood and depressive thoughts. Ninety-one adolescents with ASD provided data at age16 years, of whom 79 had additional data from age 12. We studied whether SMP have similar correlatesto those seen in typically developing youth. Results: Severe mood problems were associated withcurrent (parent-rated) and earlier (parent- and teacher-rated) emotional problems. The number of priorpsychiatric diagnoses increased the risk of subsequent SMP. Intellectual ability and adaptivefunctioning did not predict to SMP. Maternal mental health problems rated at 12 and 16 years wereassociated with SMP. Autism severity as rated by parents was associated with SMP, but the relationshipdid not hold for clinician ratings of autistic symptoms or diagnosis. SMP were associated with difficultyin identifying the facial expression of surprise, but not with performance recognizing other emotions.Relationships between SMP and tests of executive function (card sort and trail making) were notsignificant after controlling for IQ. Conclusions: This is the first study of the behavioural and cognitivecorrelates of severe mood problems in ASD. As in typically developing youth, SMP in adolescentswith ASD are related to other affective symptoms and maternal mental health problems. Previouslyreported links to deficits in emotion recognition and cognitive flexibility were not found in thecurrent sample. Further research is warranted using categorical and validated measures ofSMP. Keywords: Severe mood dysregulation, mood disorders, childhood autism, autism spectrumdisorder, SNAP.

IntroductionSevere mood problems (SMP) in children and ado-lescents include high levels of irritability, oftenmanifested by temper tantrums, as well as low andlabile mood; together, these have been identified asan important cause of psychosocial impairment.Debate has raged about the aetiology of mood dys-regulation symptoms, most specifically the extent towhich these are best conceptualized as part of thespectrum of juvenile bipolar disorder, attention def-icit hyperactivity disorder (ADHD) or as a separatesyndrome (Leibenluft, 2011). Leibenluft Cohen,Gorrindo, Brook, & Pine, (2006) argue persuasivelyfor a new diagnostic category, severe mood dysre-gulation (SMD), currently under consideration forDSM-5 (www.dsm5.org). Under current proposals,SMD would include severe and prominent moodabnormalities, hyperarousal and increased reactivityto negative emotional stimuli, with consequent

functional impairment. In support of this new diag-nosis, Leibenluft and colleagues provide evidence fordistinctive presenting and longitudinal clinical fea-tures, family history and neurocognitive profile (Lei-benluft, 2011). They argue that, while featuresaligned with irritability are included in several diag-nostic categories, the syndrome of severe andimpairing irritability with predominant negativemood includes features not adequately captured byother diagnoses. Unlike classic juvenile bipolar dis-order, manic episodes do not appear to be commonadult outcomes in SMD (Brotman et al., 2006),whereas unipolar depression and anxiety are(Stringaris et al., 2009). One small family studyfailed to find elevated rates of bipolar disorder inparents of children with SMD, in contrast with theparents of children with juvenile bipolar disorder(Brotman et al., 2007). Neurocognitive differencesexist in young people with SMD in relation to:labelling facial emotions (Guyer et al., 2007);response to frustration (Rich et al., 2011); andperformance on response reversal paradigmsConflict of interest statement: No conflicts declared.

Journal of Child Psychology and Psychiatry 53:11 (2012), pp 1157–1166 doi:10.1111/j.1469-7610.2012.02600.x

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA

(Dickstein, Finger, Brotman, et al., 2010), withneural circuitry differences on fMRI from juvenilebipolar disorder patients in the first two tasks(Brotman et al., 2010; Rich et al., 2011).

There has recently been an appreciation that otherpsychiatric problems frequently occur in people withautism spectrum disorders (ASDs), with rates ashigh as 60–70%. These co-occurring disordersinclude high rates of ADHD, anxiety and ODD inchildren (Joshi et al., 2010; Leyfer et al., 2006;Simonoff et al., 2008). The emergence and timing ofdifferent psychiatric problems in ASD has not beenexplicity studied, but comparison of cross-sectionalstudies of different age groups suggests thatdepressive and obsessive-compulsive disorder maybe more common in older adolescents and adults(Bakken et al., 2010; Mazefsky et al., 2010) and onelongitudinal clinical study showed that affectivedisorder was amongst the most common newlyemerging psychiatric disorders in adults with autism(Hutton, Goode, Murphy, Le Couteur, & Rutter,2008). Affective disorders in autism have includedbipolar disorder (Bradley & Bolton, 2006; Munesueet al., 2008), although the ascertainment methodsand sample sizes in these studies do not provide aconclusion on whether bipolar disorder is dispro-portionately increased in ASD.

Most of the research on co-occurring psychiatricsymptoms and disorders in people with ASD hasused standardized instruments to measure recog-nized symptom patterns and diagnoses. The syn-drome of SMD has not, to our knowledge, beenpreviously explored. There are several reasons toconsider this a useful concept to explore in peoplewith ASD. First, several of the symptom domainsthat are increased in people with ASD, includingADHD and affective disorder, have been associatedwith SMD. Second, people with ASD have high levelsof psychosocial impairment that are greater thanwould be expected based on their level of intellectualfunctioning. While this has often been attributed tothe core autistic deficits, it is an empirical questionwhether co-occurring psychiatric problems, such asSMD, contribute to this psychosocial impairment.Third, the relationship between low mood and‘challenging’ behaviour has long been recognized inintellectual disability, where communication is sig-nificantly impaired (Hayes, McGuire, O’Neill, Oliver,& Morrison, 2011). Challenging behaviour occurs in10–20% of people with ASDs, affects the entireintellectual ability spectrum (Emerson et al., 2001),but its causes are less well-understood than in thosewith intellectual disability without ASD. One possi-bility is that unrecognized mood problems partiallyexplain high rates of challenging behaviour in ASD.

In the present study, we use data collected fromthe Special Needs and Autism Project (SNAP) cohort(Baird et al., 2006) at age 16 years to create ameasure of SMP. We test whether the psychiatric,family and neurocognitive correlates to this scale are

similar in our ASD sample to those seen in typicallydeveloping populations.

MethodsParticipants

The sample in the present analyses comprisesninety-one 16-year olds with ASD from the 158participants with ASD in the original SNAP cohort. Inaddition, longitudinal data from 12 years wereavailable on 79 of the 91 individuals. As describedpreviously [see Baird et al., (2006) for details], SNAPwas drawn from a total population cohort of 56,946children. All those with a current clinical diagnosis ofpervasive developmental disorder (PDD, N = 255) orconsidered ‘at risk’ for being an undetected case byvirtue of having a statement of Special EducationalNeeds (SEN; N = 1,515) were surveyed using theSocial Communication Questionnaire [SCQ (Rutter,Bailey, & Lord, 2003)]. A diagnostic assessment of astratified sample at 12 years (also 255 individuals)identified 158 young people with ASD. The follow-upassessment at 16 years focussed on the cognitivephenotype of ASD and therefore only those who hadestimated IQs of ‡50 at 12 years were invited toparticipate (Charman et al., 2011). From the SNAPdatabase, 131 possible participants were identifiedon the basis of IQ; of these, 19 indicated they werenot interested in participating, 11 could not be con-tracted and 1 indicated interest, but was not in-cluded before the end of the study leaving 100adolescent participants, for whom 91 had data toprovide an SMP score for these analyses.

For this cohort, consensus clinical ICD-10 ASDdiagnoses at 12 years were made using the AutismDiagnostic Interview-Revised [ADI-R (Le Couteuret al., 1989)] and Autism Diagnostic ObservationSchedule-Generic [ADOS-G (Lord et al., 2000)] aswellas IQ, language and adaptive behaviour measures.The 91 in the contemporaneous analyses included 83male, 8 female; 48 met consensus criteria for child-hoodautismand43 for anotherASD.For the subset of79 included in the longitudinal analyses, 73 weremale and 41 had a diagnosis of autism. The samplehad amean age of 15 years 6 months (SD =5 months;range 14 years 8 months–16 years, 9 months) with amean time interv al from the 12 year assessment of4 years 0 months (SD =11 months, range 1 year7 months–5 years 8 months).

The study was approved by the South EastResearch Ethics Committee (05/MRE01/67) andinformed consent was obtained from all participants.

Measures

Questionnaires and interviews. A scale comprising‘SMP’ was generated a priori from four items on theparent-reported Profile of Neuropsychiatric Symp-toms (PONS), completed at 16 years. The PONS is a

1158 Emily Simonoff et al. J Child Psychol Psychiatry 2012; 53(11): 1157–66

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

62-item questionnaire that assesses the severity andimpact of 31 symptoms commonly reported in chil-dren and young people with neurodevelopmentaldisorders (Santosh, Baird, Pityaratstian, Tavare, &Gringras, 2006). For each symptom, a brief definitionis given and the respondent is asked to endorse theoverall frequency and impact on everyday life. Eachcomponent (frequency and impact) is each scored 0–5(‘not at all’ to ‘all the time’/‘extremely’), with a com-bined score ranging from 0 to 10. Four items wereincluded in the SMP scale, taking into consideration,the proposed DSM-5 criteria: ‘explosive range’, ‘lowmood’, ‘depressive thoughts’ and ‘labile mood’. Adescription of the PONS, the presentation of the indi-vidual items and the means and ranges for the SNAPASD samples are described in the supplementaryonline appendix and Supplementary Table S1. Thescalehadgood internal consistencywithaCronbach’sa of .92. The raw scale was nonnormally distributedwith amean of 7.8 (SD 8.0, range 0–36) and a square-root transformation was applied to generate a morenormally distributed continuous measure withskewness of 0.16 and kurtosis 2.78. A binary classi-fication divided the top 25% of scores (13–36, N = 24)from the rest of the distribution (0–12, N = 67). Thisthresholdwas chosenpragmatically because (a) itwaslikely to have reasonable power to detect mean dif-ferences and (b) it is conservative compared to therates of ADHD (28%) and anxiety disorders (42%) re-ported in this cohort and is therefore, a plausiblethreshold to select.

The Strengths and Difficulties Questionnaire [SDQ(Goodman, Ford, Simmons, Gatward, & Meltzer,2000)], rated by parents at 12 and 16 years andteachers at 12 years, was also used to measuremental health symptoms. The SDQ is a widely usedscreening instrument for child psychiatric problemsand its psychometric properties have been estab-lished in several samples, including UK (Goodmanet al., 2000) and US studies (Bourdon, Goodman,Rae, Simpson, & Koretz, 2005). The present analysesuse the hyperactivity, conduct and emotional sub-scales.

At 12 years, the parent-reported Child and Ado-lescent Psychiatric Assessment [CAPA (Angold &Costello, 2000)] was completed on 69 of the presentsample. The CAPA is a semistructured psychiatricinterview and the following diagnostic areas wereincluded: all anxiety and phobic disorders (includingobsessive-compulsive disorder); major depressionand dysthymic disorder; ODD and conduct disorder(CD); ADHD; tics/Tourette/trichotillomania; enure-sis and encopresis. The prevalence rates and diag-nostic correlates have been reported previously(Simonoff et al., 2008).

Autism severity was assessed in three ways. Weused the diagnostic dichotomy of childhood autism/other PDD. Clinicians undertaking the review ofautism diagnostic information based on the ADI-Rand ADOS-G described above scored the 12 ICD-10

symptoms that comprise the autism spectrum dis-order diagnoses. The Social Responsiveness Scale[SRS (Constantino et al., 2003)] T scores were usedas a quantitative measure of autism severity, scoredat 12 years in 60 participants and at 16 years in 27,where data were missing at 12 years.

Adaptive functioning at 12 years was measuredusing the Vineland Adaptive Behaviour Scales com-posite score (Sparrow, Balla, & Cichetti, 1984). Aquantitative measure of the shortfall, or adaptive‘under-function’, was generated by subtracting theVineland score from the full-scale IQ, both measuredat 12 years and standardized to mean of 100, SD 15and.

Maternal self-reports on the General HealthQuestionnaire [GHQ-30 (Goldberg & Muller, 1988)]when the participants were 12 and 16 years pro-vided a measure of maternal psychiatric symptomswith particular emphasis on mood, anxiety and so-matic difficulties. The Parenting Stress Index [PSIShort Form; (Abidin, 1995)] measures difficulties inthe parent-child relationship on three subscales:disturbed child, parental distress and parent-childdysfunctional interaction. Parental distress wasused, herein, to index the parental component ofstress, as it attempts to measure parental charac-teristics rather than aspects of the parent-childrelationship, which may be affected by the presenceof an ASD in the child.

Neurocognitive measures. IQ was measured at12 years with the Wechsler Intelligence Scales forChildren-Third Edition [WISC-IIIUK (Wechsler,1992)] and at 16 years with the Wechsler Abbrevi-ated Scales of Intelligence [WASI (Wechsler, 1999)].Details of the neurocognitive tasks are given in theSupplementary online appendix. All were adminis-tered at 16 years. The emotion recognition task hasbeen previously described (Jones et al., 2011). In thepresent analysis, we used the Ekman-Friesen test ofaffect recognition (Ekman & Friesen, 1976), as thiswas most similar to tasks undertaken in typicallydeveloping youths with SMD (Brotman et al., 2008).We measured total number of correct responses.

The Card Sort was included as a measure of cog-nitive flexibility and response reversal (Tregay, Gil-mour, & Charman, 2009). The task requires theparticipant to correctly sort cards to one of threealternative sets across three trials, with the correctset varying in each trial. The key variable was thenumber of sorts required to reach criterion. In thepresent analyses, we included only those partici-pants who demonstrated an understanding of therule in the first trial by reaching criterion before theend. The number of sorts required in the second andthird trials was divided into four levels: top half(scores 12–18, N = 42); third quartile (scores 19–24,N = 22), bottom quartile (scores 25–40, N = 17) andthose who did not reach criterion by the end of bothtrials (N = 8).

doi:10.1111/j.1469-7610.2012.02600.x Mood problems in autism 1159

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

Trail Making was included as a measure of atten-tional switching and response reversal. The task wascomprised of three separate trials (Reitan & Wolfson,1985). The participant was asked to ‘join the dots’ innumerical order, then, in a second trial, in alpha-betic order, followed by a third trial switching be-tween numbers and letters. The difference scorebetween the time taken on the first and the third trialcomprised a measure of switching ability. The meandifference score was 57.8 (SD 40.7, range 10.5–229.1). As the data were highly skewed, a square-root transformed score was used in the presentanalyses.

Statistical analysis

Data reduction and statistical analysis wereundertaken in Stata version 11 (StataCorp, 2009).Linear regression was used to examine the contin-uous outcome of the transformed SMP score andlogistic regression for the binary variable of highversus low SMP scores. Ordinal logistic regressionwas used for the Card Sort, where a 4-level scalewas generated, and for the analyses using the totalnumber of psychiatric diagnoses on the CAPA.Multivariate regression, analogous to multipleanalysis of variance, was employed to analyse theemotion recognition profile to allow for tests of an

overall effect and those specific to an emotion.Ordinal logistic regression was also used for ordinaloutcomes, such as number of diagnoses. For sets ofordinal items, such as SDQ items, specificity ofassociation was tested using similar models esti-mated in gllamm (www.gllamm.org) using a gener-alized estimating equations approach with anIndependent Working Model. The models allowedseparate threshold parameters for each item andestimated a common and an item-specific effect inthe manner of testing for differential item function-ing. Significance of effects was determined fromWald tests using the robust form of the parametercovariance matrix.

ResultsParticipant characteristics, according to high/lowSMP are shown in Table 1.

Emotional and behavioural characteristicsassociated with SMP

Examining the contemporaneous relationships be-tween the three parent-rated mental health problemsdomains of the SDQ (Table 2) revealed that hyper-activity, conduct and emotional problems all wereassociated with SMP in bivariate analyses, but that

Table 1 Sample characteristics according to severe mood dysregulation and problems (SMP) classification [M (SD)]

High SMP (N = 24) Low SMPa (N = 67)

Raw PONS scores on individual itemsExplosive rage 4.9 (2.4) 1.2 (1.4)Low mood 4.8 (2.3) 1.2 (1.4)Labile mood 4.8 (2.8) 0.8 (1.6)Depressive thoughts 4.6 (2.8) 0.6 (1.2)

Other characteristics at 16 yearsFull-scale IQ 80.0 (16.6) 85.8 (17.5)SDQc Hyperactivity 6.6 (2.6) 5.7 (2.4)SDQc Conduct problems 2.6 (1.1) 1.5 (1.6)SDQc Emotional problems 5.3 (2.1) 2.9 (2.3)Maternal GHQ score 8.0 (8.5) 4.1 (6.3)

Other characteristics at 12 yearsAdaptive behaviour 50.6 (12.3) 52.1 (14.4)Diagnosed childhood autism N (%) 13 (54) 35 (52)ICD-10 symptom severity 8.4 (2.4) 8.0 (2.5)SRSb 101.3 (25.9) 90.3 (22.4)SDQc Hyperactivity (parent) 7.4 (2.8) 7.5 (2.5)SDQc Conduct problems (parent) 3.5 (1.9) 2.9 (2.1)SDQ3 Emotional problems (parent) 6.3 (2.5) 3.8 (2.4)CAPA Any emotional problem N (%) 11 (57.9) 13 (26.0)CAPA Oppositional defiant/conduct disorder N (%) 8 (42.1) 9 (18.0)CAPA ADHD N (%) 8 (42.1) 9 (16.0)Maternal GHQ score 7.3 (7.3) 4.9 (6.5)

Neurocognitive measures at 16Ekman faces total score 41.8 (8.2) 42.8 (7.7)Card sort errors to criterion 24.5 (8.3) 21.8 (8.7)Trail making difference score 69.1 (48.8) 61.9 (43.7)

ADHD, attention deficit hyperactivity disorder; CAPA, Child and Adolescent Psychiatric Assessment; GHQ, General HealthQuestionnaire; SDQ, Strengths and Difficulties Questionnaire; SRS, Social Responsiveness Scale.aHigh SMP refers to those scoring in the top quartile, whereas low SMP is the rest of the distribution.bMeasured at 12 years in 60, at 16 years in 27.cParent-reported measures at 12 and 16 years.

1160 Emily Simonoff et al. J Child Psychol Psychiatry 2012; 53(11): 1157–66

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

only conduct and emotional problems remainedsignificant in multivariate regression. The conductproblems scale includes amongst its five items ‘oftenhas temper tantrums or hot tempers’, which is clo-sely related to the SMP concept; we hypothesized thisitem might underpin the association of SMP withconduct problems. Fitting an ordinal response modelto the five SDQ conduct items showed a significantassociation with SMP (common proportional oddsratio (OR) 1.43, 95% CIs 1.23, 1.66; p < .001).However, the inclusion of an additional effect specificto tempers reduced the common partial OR to 1.18(95% CIs 1.00, 1.40; p = .055), whereas the specificeffect item was itself large and significant(OR = 2.09, 95% CIs 1.38, 3.16; p < .001). Thisprovides concurrent validity for the concept of SMPas defined in the current study. Furthermore, al-though the emotional subscale of the SDQ includesitems on anxiety as well as mood, this can also beconsidered an index of concurrent validity. To ad-dress the issue of whether the association with theSDQ emotional scale was simply due to overlappingitem content, we repeated the regression with atruncated version of the SMP scale that includedonly explosive rage and labile mood. The results re-mained highly significant (b = .19, p < .001), indi-cating the relationship is due to co-occurrence ofcommon mood and anxiety symptoms and thosespecifically related to severe mood dysregulation.

In examining the prediction of SMP at 16 yearsfrom mental health symptoms at 12 years, bivari-ate analyses showed the same pattern for parentand teacher ratings at 12 years; only SDQ emo-tional symptoms were significantly associated(Table 2); however, these effects were of moderate

size, with correlations for parent and teacheremotional scores at 12 and SMP of 0.44 and 0.18,respectively. These associations remained signifi-cant using the truncated SMP scale of explosiverage and labile mood (b = .16, p = .002; b = .20,p = .001 for parent and teacher SDQ emotionalscale respectively), indicating that common emo-tional symptoms predict to the less emotionalcomponents of SMD. Finally, in the subsample of69, we used CAPA diagnoses at 12 years to predictthe category of high (top quartile) SMP score.Having any CAPA emotional or behavioural disor-der at 12 years substantially increased the odds ofbeing in the top quartile for SMP (OR = 9.6,95%CIs 2.43, 37.00, p < .001), as did a diagnosisin each of the three domains: ADHD (OR = 3.82,95%CIs 1.17, 12.47, p = .03.,); ODD/CD (OR =3.32, 95%CIs 1.04, 10.59, p = .04); and any emo-tional (anxiety/depressive/phobic) disorder (OR =3.92, 95%CIs 1.29, 11.86, p = .02). There werestrong associations among the CAPA diagnoses anda multivariable analysis including all three CAPAdomains as predictors of the high SMP categoryfailed to distinguish an independent domain pre-dicting SMP 4 years later. However, ordinal logisticregression indicated a significant trend in theassociation between the number of the 12 possibleCAPA disorders and being in the high SMP group(p value from ordinal logistic regression = .006). Inrelation to the high SMP category, rates were asfollows: for no CAPA disorder, 3%, 95% CIs 0–19%;one CAPA disorder, 32%, 95% CIs 9–55%; ‡2 dis-orders 44%, 95% CIs 23–65%. The finding repli-cated when the CAPA diagnoses were collapsed tothe three domains described above (ordinal logistic

Table 2 Personal characteristics under various scales associated with severe mood problems [95 per cent confidence intervals]

Unadjusted Adjusted

B (95% CIs) p b (95% CIs) p

Emotional and behavioural problems at 16 (parent-rated)SDQ Hyperactivity .15 (.03, .28) .02 .03 ().08, .14) .56SDQ Conduct problems .37 (.19, .56) <.001 .27 (.11, .44) <.01SDQ Emotional problems .35 (.22, .45) <.001 .30 (.19, .41) <.001

Emotional and behavioural problems at 12 (parent-rated)SDQ Hyperactivity .000 ().13, .13) .98SDQ Conduct problems .14 ().03, .30) .10SDQ Emotional problems .25 (.13, .36) <.001

Emotional and behavioural problems at 12 (teacher-rated)SDQ Hyperactivity ).02 ().19, .13) .75SDQ Conduct problems .11 ().08, .30) .27SDQ Emotional problems .29 (.16, .42) <.001

Other personal characteristicsFull-scale IQ (current) ).01 ().03, .00) .13Adaptive functioning (12 years) ).02 ().04, .01) .17IQ-adaptive functioning discrepancy (12 years) .00 ().02, .02) .99Social Responsiveness Scale (SRS) Autism severity .02 (.01, .03) <.01ICD-10 total autism symptom score .10 ().04, .23) .15

Maternal distressMaternal GHQ, 16 years .07 (.02, .11) <.01Maternal GHQ, 12 years .08 (.03, .13) .01

GHQ, General Health Questionnaire; SDQ, Strengths and Difficulties Questionnaire.

doi:10.1111/j.1469-7610.2012.02600.x Mood problems in autism 1161

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

regression p < .001); the rate of high SMP categoryfor one disorder domain 44%, 95% CIs 19–70%; ‡2disorder domains 50%, 95% CIs 22–78%.

Other participant characteristics

Full-scale IQ, whether measured at 16 (Table 2) or12 years (b = .01, 95%CIs ).02, .01, p = .42) was notassociated with SMP. Similarly, neither adaptivefunctioning at 12 years nor the difference betweencognitive ability and adaptive functioning was sig-nificantly associated with SMP.

Autism symptoms on the parent-reported SRSwere strongly and positively related to SMP. In con-trast, the relationship was not replicated when usingeither the clinician-rated ICD-10 symptom score orthe diagnostic classification of childhood autism/other PDD. For the latter, the mean transformedSMP scores were 2.38 (childhood autism) versus2.32 (other PDD), t = 0.20, p = .84).

Maternal mental health

Using maternal GHQ scores, we found a significantrelationship between SMP and high scores at bothages. The latter relationship remained significantwhen the other contemporaneously assessedparental variables of educational and socioeconomicstatus and contextual factors of family-based andneighbourhood deprivation were included as cova-riates (b = .08, 95% CIs .02, .14, p = .01). To addressthe possibility that maternal mental health wasindexing distress in relation to having a challengingchild with ASD, rather than a characteristic of themothers, we undertook a further regression in whichwe included as independent variables both thematernal GHQ score and the parental distress sub-scale from the Parenting Stress Index, both mea-sures at 12 years. The GHQ retained a similar levelof prediction (b = .08, p = .002) as seen in thebivariate analysis (Table 2) while the index ofparental distress was nonsignificant (b = .02,p = .43).

Neurocognitive performance associated with SMP

Association with all emotions was tested usingmultivariate regression, which revealed a significantoverall association between poorer emotion recogni-

tion and SMP (F(6,86) = 3.00, p = .010), which posthoc tests identified as being due to a specific asso-ciation with surprise (Bonferroni corrected p = .04)(Table 3). Covarying for full-scale IQ and includingonly those 72 participants with IQ ‡70 showed sim-ilar overall significant findings, but the Bonferroni-adjusted significance level is associated with sur-prise fell to 0.11 and 0.37 respectively. Results onthe Card Sort using ordinal logistic regression, inwhich the three response categories were predictedby SMP, showed that SMP was associated with moreerrors or decreased cognitive flexibility (OR = 1.35,95%CIs 1.04, 1.73, p = .02) (Table 4). However, theCard Sort trials to criterion were strongly associatedwith IQ (ordinal logistic regression OR = 0.91, 95%CIs 0.88, 0.94, p < .001), and when this was addedas a covariate, the relationship between Card Sorttrials and SMP became nonsignificant (OR = 1.24,95% CIs 0.94, 1.63, p = .13). A sensitivity analysislimiting the participants to those with IQ ‡70(N = 72) replicated the nonsignificant finding(p = .13).

In the Trail Making, there was no significant rela-tionship between transformed time difference scoreand SMP (b = .17, 95% CIs ).17, .51, p = .33). IQ wasalso strongly related to Trail Making (b = .07, 95%CIs ).10, ).04, p < .001), but its addition, as a co-variate did not alter the pattern of results (p = .63),nor did the exclusion of participants with IQ <70(p = .64).

Discussion and conclusionsThis is the first examination of the mental healthand neurocognitive correlates of SMP in adolescents

Table 3 Multivariate regression of association between severe mood problems and emotion recognition

Emotion Overalltesta

Happiness Sadness Rear Anger Surprise Disgust

F(6) pB (95% CIs) p B (95% CIs) p B (95% CIs) p B (95% CIs) p B (95% CIs) p B (95% CIs) p

).09 ().21, 020) .11 ).01 ().31, .29) .96 ).15 ().54, .24) .45 .05 ().22, .32) .75 ).45 ().77, ).13) <.01 18 ().22, .58) .37 3.00 .01

aBonferroni-corrected test of overall significance.

Table 4 Severe mood problems score according to Card sorttrials to criterion

SMPa

Mean score 95% CIs

Card-sort trials to criterionTop 50% 2.03 1.58, 2.4751–75% 2.57 1.68, 3.2576%+ 2.84 1.68, 3.25Did not meet criterion 3.09 2.23, 3.96

SMP, severe mood problems.aSquare root transformed score.

1162 Emily Simonoff et al. J Child Psychol Psychiatry 2012; 53(11): 1157–66

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

with ASDs. The strongest mental health correlatesare with emotional symptoms both in contempora-neous and predictive analyses. We considered thepossibility that the SMP scale could be dominatedby mood symptoms and therefore be not more thana proxy for emotional symptoms. However, all foursymptoms contribute similar to the high SMP groupand the scale has high internal consistency. Fur-thermore, the specific link to the ‘tempers’ item ofthe conduct scale adds additional support for theSMP construct in our sample. In addition, theassociation with the teacher emotional score at12 years excludes the possibility of a parental raterbias. This finding is consistent with the literature intypically developing children showing that the lon-gitudinal course of SMP involves affective problemsprimarily (Brotman et al., 2006; Stringaris et al.,2009).

The relationship between SMP and ADHD hasbeen previously identified, with a suggestion thatemotional lability may be prominent in people withADHD (Sobanski et al., 2010). Our findings in ASDdo not provide a clear answer. SDQ ADHD symp-toms were neither predictive nor independentlycorrelated with SMP. On the other hand, the pres-ence of any psychiatric disorder at age 12 washighly predictive of being in the top quartile for SMPat 16 (OR>9) and the prediction was equally strongfor disorders in each of the three domains of emo-tional, oppositional/conduct and ADHD. Previously,we showed that, in this sample, more than 80% ofthose with ADHD also had an emotional and/oroppositional/CD (Simonoff et al., 2008), and thismay be consistent with the SMP concept. However,it is interesting that a high SMP score at 16 ispredicted not only by the presence of any disorder,but also by the number of individual diagnoses,further highlighting that there may be a number ofdifferent psychiatric correlates to SMP in peoplewith ASD.

We found that autism severity, as measured onthe parent-reported SRS was strongly associatedwith SMP, whereas the clinician-rated symptomscore generated at 12 years was not. The SRS con-tains many more items with a wider range of scoresthan the other measures. The possibility of corre-lated measurement error in parents’ responses tothe SRS and SMP scale cannot be excluded withoutanother data source. Whatever the explanation forthe discrepancy between parent- and clinician- re-ported autism severity scores and SMP, the findingthat SMP are not associated with childhood autismversus another PDD is important for service provi-sion. As autism is generally considered more severethan Asperger syndrome or atypical autism/ ASDs,there is a temptation for clinicians to assume thatall aspects of psychosocial functioning will be lessimpaired in the latter group. Our previous findingshighlight that this diagnostic dichotomy doesnot predict the rate of co-occurring psychiatric

disorders (Simonoff et al., 2008) or the stability ofpsychiatric symptoms over time (Simonoff et al.,2012).

Neither full-scale IQ nor adaptive functioning wasassociated with SMP. This is consistent with previ-ous findings regarding co-occurring psychiatric dis-orders in SNAP (Simonoff et al., 2008) but it shouldbe noted that other studies report a relationshipbetween lower IQ and psychiatric problems in ASD(Totsika, Hastings, Emerson, Lancaster, & Berridge,2011). Our prediction of an association betweenadaptive under-function and SMP was also notsubstantiated, excluding this as one reason for theIQ-adaptive functioning discrepancy.

We replicate the finding in otherwise typicallydeveloping children with SMD that parents are atincreased risk of affective disorder (Brotman et al.,2007) and show that this is independent of thepossible confounders of family background, neigh-bourhood deprivation and parental distress associ-ated with having a challenging child with ASD.Families of individuals with ASDs have higher ratesof affective disorder, both major depression andbipolar disorder, and the nature of this association isnot fully understood (Bolton, Pickles, Murphy, &Rutter, 1998; Ingersoll, Meyer, & Becker, 2011).However, this relationship does not appear to besolely a consequence of stress induced by raising achild with ASD as some cases of affective disorder inparents commenced prior to the child’s birth andother instances of affective disorder occurred insecond degree, and other relatives not directly in-volved in the care of the autistic child. The currentanalyses suggest a specific relationship betweenmaternal affective symptoms and SMP in offspringwithin the ASD group that warrants additionalexploration with more detailed psychiatric mea-sures, as the GHQ is a nonspecific measure of psy-chiatric symptoms.

We find very little support for the same neuro-cognitive correlates of SMD in our ASD sample asare reported in typically developing youth, i.e.poorer performance on emotion recognition (Guyeret al., 2007) and response reversal tasks (Dicksteinet al., 2007; Dickstein, Finger, Skup, et al., 2010).Previous analyses of the emotion recognition tasksin SNAP demonstrated that participants with ASDwere not generally deficient in emotion recognitioncompared to controls of the same intellectual abil-ity, but rather had specific difficulties correctlyidentifying the emotion of surprise (Jones et al.,2011). We report an association of SMP to the sameemotion, but not for a wider range of emotions, asdescribed in typically developing populations withSMD.

The Card Sort task used in this study is a measureof cognitive flexibility; in fact, the ID/ED task usedby (Dickstein et al., 2007) in youth with SMD is de-scribed as being based on the Wisconsin Card Sort.We found a significant relationship between SMP

doi:10.1111/j.1469-7610.2012.02600.x Mood problems in autism 1163

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

and cognitive rigidity, as measured by increasedtrials to criterion after having learned one rule.However, this relationship disappeared when IQ wasaccounted for. The IQ range in SNAP is wide(50–119), making it particularly important to con-sider its role in the performance of all neurocognitivetasks. Trail Making, also tapping cognitive flexibility,showed no association with SMPs in our ASD sam-ple. Thus, we failed to replicate the neurocognitivefindings from typically developing samples. Ourfindings on neurocognitive tasks raise the possibilitythat different brain mechanisms are involved in SMPin people with ASD.

This studyhasanumberof strengths. The sample ispopulation–based and carefully characterized withrespect to both their autism and general cognitivefeatures. The longitudinal nature of the sample allowsus to examine predictors of SMP as well as correlates.Most importantly, the sample remains unusual inhaving been assessed with respect to a wide range ofboth behavioural characteristics and neurocognitivetasks. It is this feature that allowed us to explore therelationship of SMPs to neurocognitive performancein ASD. We have been conservative in our analyticapproach, limiting these to the tasks (emotion recog-nition, cognitive flexibility) associated with SMD intypically developing children.

The limitations to this study are also important tonote. The study was not originally designed to assessSMD and the bespoke measure used herein, whilecarefully constructed, is not presently standardizedor validated. Therefore, our findings must be inter-preted with caution. Furthermore, although theconcept of SMD is an evolving one, this measure mayimperfectly characterize the current definition, as itdoes not explicitly include an item on irritability(because this was not included in the PONS). Due tothe moderate sample size of SNAP and the lack of anidentified cut-off for the SMP score, most of thepresent analyses use the continuous measure incontrast with the findings reported in non-ASDsamples, which study ‘clinical’ groups.

Despite these limitations, the results are interest-ing and important in showing that SMP in adoles-cents with ASD represent a coherent and primarily

emotional construct. This is the first exploration ofSMD in autism and the findings indicate that SMD isa psychometrically coherent concept in which psy-chiatric and family correlates are similar with thoseseen in typically developing children, but where theneurocognitive basis may be different. It will beimportant to test more definitively in larger samplesthat allow clinically meaningful subgroups whetherthese problems have similar or distinct origins. AsSMD could be a target for intervention in ASD, futurestudies should also clarify its prevalence in ASD, itsrelationship to other measures of ‘challengingbehaviour’ and the additional impairment that itcauses, as well as investigating both the biological/cognitive and environmental risk factors and corre-lates for SMD.

Supporting informationAdditional Supporting Information may be found in theonline version of this article:

Appendix S1 The Profile of Neuropsychiatric Symp-toms (PONS)

Appendix S2 Details of neurocognitive tasksTable S1 Characteristics of individual PONS items for

full ASD samplePlease note: Wiley-Blackwell are not responsible for

the content or functionality of any supporting materialssupplied by the authors. Any queries (other thanmissing material) should be directed to the corre-sponding author for the article.

AcknowledgementsWe thank Ellen Leibenluft for her advice on the itemselection for the SMP scale and Paramala Santosh forpermission to reprint the relevant items from the PONS.

CorrespondenceEmily Simonoff, Department of Child and AdolescentPsychiatry, King’s College London, Institute of Psychi-atry and NIHR Biomedical Research Centre for MentalHealth, De Crespigny Park, London SE58AF, UK;Email: emily.simonoff@kcl.ac.uk

Key points

• Severe mood problems in young people are now considered a separate entity that predicts to subsequentdepressive disorder and has distinct neurocognitve and brain imaging correlates.

• A new diagnostic category of severe mood dysregulation is proposed for DSM-5.• Young people with autism spectrum disorder (ASD) have high rates of psychiatric comorbidities, but the role

of severe mood problems has not previously been studied.• In adolescents with ASD, we found links to emotional and behavioural problems and family history similar to

those reported in otherwise typically developing youth, but did not see the same associations with theneurocognitive deficits in emotion recognition and response reversal.

• Severe mood problems may have different underlying causes in ASD and this requires further investigation.

1164 Emily Simonoff et al. J Child Psychol Psychiatry 2012; 53(11): 1157–66

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

ReferencesAbidin, R.R. (1995). Parenting Stress Index (3rd edn). London:

NFER: Nelson.Angold, A., & Costello, E.J. (2000). The Child and Adolescent

Psychiatric Assessment (CAPA). Journal of the AmericanAcademy of Child & Adolescent Psychiatry, 39, 39–48.

Baird, G., Simonoff, E., Pickles, A., Chandler, S., Loucas, T.,Meldrum, D., & Charman, T. (2006). Prevalence of disordersof the autism spectrum in a population cohort of children inSouth East Thames – The Special Needs and Autism Project.The Lancet, 368, 210–215.

Bakken, T.L., Helverschou, S.B., Eilertsen, D.E., Heggelund,T., Myrbakk, E., & Martinsen, H. (2010). Psychiatric disor-ders in adolescents and adults with autism and intellectualdisability: A representative study in one county in Norway.Research in Developmental Disabilities, 31, 1669–1677.

Bolton, P.F., Pickles, A., Murphy, M., & Rutter, M. (1998).Autism, affective and other psychiatric disorders: Patterns offamilial aggregation. Psychological Medicine, 28, 385–395.

Bourdon, K.H., Goodman, R., Rae, D.S., Simpson, G., &Koretz, D.S. (2005). The Strengths and Difficulties Ques-tionnaire: U.S. normative data and psychometric properties.Journal of the American Academy of Child & AdolescentPsychiatry, 44, 557–564.

Bradley, E., & Bolton, P. (2006). Episodic psychiatric disordersin teenagers with learning disabilities with and withoutautism. British Journal of Psychiatry, 189, 361–366.

Brotman, M.A., Guyer, A.E., Lawson, E.S., Horsey, S.E., Rich,B.A., Dickstein, D.P., ... & Leibenluft E. (2008). Facialemotion labeling deficits in children and adolescents at riskfor bipolar disorder. American Journal of Psychiatry, 165,385–389.

Brotman, M.A., Kassem, L., Reising, M.M., Guyer, A.E.,Dickstein, D.P., Rich, B.A., ... & Leibenluft E. (2007).Parental diagnoses in youth with narrow phenotype bipolardisorder or severe mood dysregulation. American Journal ofPsychiatry, 164, 1238–1241.

Brotman, M.A., Rich, B.A., Guyer, A.E., Lunsford, J.R.,Horsey, S.E., Reising, M.M., ... & Leibenluft E. (2010).Amygdala activation during emotion processing of neutralfaces in children with severe mood dysregulation versusADHD or bipolar disorder. American Journal of Psychiatry,167, 61–69.

Brotman, M.A., Schmajuk, M., Rich, B.A., Dickstein, D.P.,Guyer, A.E., Costello, E.J., ... & Leibenluft E. (2006).Prevalence, clinical correlates, and longitudinal course ofsevere mood dysregulation in children. Biological Psychiatry,60, 991–997.

Charman, T., Jones, C.R.G., Pickles, A., Simonoff, E., Baird,G., & Happe, F. (2011). Defining the cognitive phenotype ofautism. Brain Research, 1380, 10–21.

Constantino, J.N., Davis, S.A., Todd, R.D., Schindler, M.K.,Gross, M.M., Brophy, S.L., .. & ReichW. (2003). Validation ofa brief quantitative measure of autistic traits: Comparison ofthe social responsiveness scale with the autism diagnosticinterview-revised. Journal of Autism & Developmental Disor-ders, 33, 427–433.

Dickstein, D.P., Finger, E.C., Brotman, M.A., Rich, B.A., Pine,D.S., Blair, J.R. & Leibenluft E. (2010). Impaired probabi-listic reversal learning in youths with mood and anxietydisorders. Psychological Medicine, 40, 1089–1100.

Dickstein, D.P., Finger, E.C., Skup, M., Pine, D.S., Blair, J.R.,& Leibenluft, E. (2010). Altered neural function in pediatricbipolar disorder during reversal learning. Bipolar Disorders,12, 707–719.

Dickstein, D.P., Nelson, E.E., McClure, E.B., Grimley, M.E.,Knopf, L., Brotman, M.A., ... & Leibenluft E. (2007). Cognitiveflexibility in phenotypes of pediatric bipolar disorder. Journal ofthe American Academy of Child & Adolescent Psychiatry, 46,341–355.

Ekman, P., & Friesen, W.V. (1976). Pictures of facial affect. PaloAlto, CA: Consulting Psychologists Press.

Emerson, E., Kiernan, C., Alborz, A., Reeves, D., Mason, H.,Swarbrick, R., ... & Hatton C. (2001). The prevalence ofchallenging behaviors: A total population study. Research inDevelopmental Disabilities, 22, 77–93.

Goldberg, D., & Muller, P. (1988). A user’s guide to the GeneralHealthQuestionnaire GHQ.Windsor, England:NFER-Nelson.

Goodman, R., Ford, T., Simmons, H., Gatward, R., & Meltzer,H. (2000). Using the Strengths and Difficulties Question-naire (SDQ) to screen for child psychiatric disorders in acommunity sample. British Journal of Psychiatry, 177, 534–539.

Guyer, A.E., McClure, E.B., Adler, A.D., Brotman, M.A., Rich,B.A., Kimes, A.S., ... & Leibenluft E. (2007). Specificity offacial expression labeling deficits in childhood psychopa-thology. [Research Support, N.I.H., Intramural]. Journal ofChild Psychology & Psychiatry & Allied Disciplines, 48, 863–871.

Hayes, S., McGuire, B., O’Neill, M., Oliver, C., & Morrison, T.(2011). Low mood and challenging behaviour in people withsevere and profound intellectual disabilities. Journal ofIntellectual Disability Research, 55, 182–189.

Hutton, J., Goode, S., Murphy, M., Le Couteur, A., & Rutter, M.(2008). New-onset psychiatric disorder in individuals withautism. Autism, 12, 373–390.

Ingersoll, B., Meyer, K., & Becker, M.W. (2011). Increased ratesof depressed mood in mothers of children with ASD associ-ated with the presence of the broader autism phenotype.Autism Research, 4, 143–148.

Jones, C.R., Pickles, A., Falcaro, M., Marsden, A.J., Happe, F.,Scott, S.K., ... & Charman T. (2011). A multimodal approachto emotion recognition ability in autism spectrum disorders.Journal of Child Psychology & Psychiatry & Allied Disci-plines, 52, 275–285.

Joshi, G., Petty, C., Wozniak, J., Henin, A., Fried, R., Galdo,M., ... & Biederman J. (2010). The heavy burden ofpsychiatric comorbidity in youth with autism spectrumdisorders: A large comparative study of a psychiatricallyreferred population. Journal of Autism & DevelopmentalDisorders, 40, 1361–1370.

Le Couteur, A., Rutter, M., Lord, C., Rios, P., Robertson, S.,Holdgrafer, M., & McLennan J. (1989). Autism diagnosticinterview: A standardized investigator-based instrument.Journal of Autism and Developmental Disorders, 19, 363–387.

Leibenluft, E. (2011). Severe mood dysregulation, irritability,and the diagnostic boundaries of bipolar disorder in youths.American Journal of Psychiatry, 168, 129–142.

Leibenluft, E., Cohen, P., Gorrindo, T., Brook, J.S., & Pine,D.S. (2006). Chronic versus episodic irritability in youth: Acommunity-based, longitudinal study of clinical and diag-nostic associations. Journal of Child & Adolescent Psycho-pharmacology, 16, 456–466.

Leyfer, O.T., Folstein, S.E., Bacalman, S., Davis, N.O., Dinh,E., Morgan, J., ... & Lainhart JE. (2006). Comorbid psychi-atric disorders in children with autism: Interview develop-ment and rates of disorders. Journal of Autism andDevelopmental Disorders, 36, 849–861.

Lord, C., Risi, S., Lambrecht, L., Cook, E.H., Leventhal, B.L.,DiLavore, P.C., ... & Rutter M. (2000). The Autism Diagnos-tic Observation Schedule-Generic: A standard measure ofsocial and communication deficits associated with thespectrum of autism. Journal of Autism and DevelopmentalDisorders, 30, 205–223.

Mazefsky, C.A., Conner, C.M., Oswald, D.P., Mazefsky, C.A.,Conner, C.M., & Oswald, D.P. (2010). Association betweendepression and anxiety in high-functioning children withautism spectrum disorders and maternal mood symptoms.[Research Support, Non-U.S. Gov’t]. Autism Research, 3,120–127.

doi:10.1111/j.1469-7610.2012.02600.x Mood problems in autism 1165

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.

Munesue, T., Ono, Y., Mutoh, K., Shimoda, K., Nakatani, H., &Kikuchi, M. (2008). High prevalence of bipolar disordercomorbidity in adolescents and young adults with high-functioning autism spectrum disorder: A preliminary studyof 44 outpatients. [Comparative Study]. Journal of AffectiveDisorders, 111, 170–175.

Reitan, R.M., & Wolfson, D. (1985). The Halstead-Reitanneuropsychological test battery: Theory and clinical interpre-tation. Tucson: Neuropsychology Press.

Rich, B.A., Carver, F.W., Holroyd, T., Rosen, H.R., Mendoza,J.K., Cornwell, B.R., ... & Lainhart JE. (2011). Differentneural pathways to negative affect in youth with pediatricbipolar disorder and severe mood dysregulation. Journal ofPsychiatric Research, 45, 1283–1294.

Rutter, M., Bailey, A., & Lord, C. (2003). The Social Commu-nication Questionnaire (1st edn). Los Angeles: WesternPsychological Services.

Santosh, P.J., Baird, G., Pityaratstian, N., Tavare, E., &Gringras, P. (2006). Impact of comorbid autism spectrumdisorders on stimulant response in children with attentiondeficit hyperactivity disorder: A retrospective and prospec-tive effectiveness study. Child: Care, Health & Development,32, 575–583.

Simonoff, E., Jones, C., Baird, G., Pickles, A., Happe, F., &Charman, T. (2012). The persistence and stability of psy-chiatric problems in adolescents with autism spectrumdisorders. Journal of Child Psychology and Psychiatry.Advance online publication; doi: [JCPP-2012-00232.R2]

Simonoff, E., Pickles, A., Charman, T., Chandler, S., Loucas,T., & Baird, G. (2008). Psychiatric disorders in children withautism spectrum disorders: Prevalence, comorbidity andassociated factors. Journal of the American Academy of Childand Adolescent Psychiatry, 47, 921–929.

Sobanski, E., Banaschewski, T., Asherson, P., Buitelaar, J.,Chen, W., Franke, B., ... & Faraone S.V. (2010). Emotionallability in children and adolescents with attention deficit/hyperactivity disorder (ADHD): Clinical correlates and famil-ial prevalence. Journal of Child Psychology & Psychiatry &Allied Disciplines, 51, 915–923.

Sparrow, S., Balla, D., & Cichetti, D. (1984). Vineland AdaptiveBehavior Scales. Circle Pines, Minnesota: American Guid-ance Services.

StataCorp (2009). Stata statistical software release 11.1.College Station, TX: Stata Corporation.

Stringaris, A., Cohen, P., Pine, D.S., & Leibenluft, E. (2009).Adult outcomes of youth irritability: A 20-year prospectivecommunity-based study. [Comparative Study]. AmericanJournal of Psychiatry, 166, 1048–1054.

Totsika, V., Hastings, R.P., Emerson, E., Lancaster, G.A., &Berridge, D.M. (2011). A population-based investigation ofbehavioural and emotional problems and maternal mentalhealth: Associations with autism spectrum disorder andintellectual disability. Journal of Child Psychology & Psychi-atry, 52, 91–99.

Tregay, J., Gilmour, J., & Charman, T. (2009). Childhoodrituals and executive functions. British Journal of Develop-mental Psychology, 27, 283–296.

Wechsler, D. (1992). The Wechsler Intelligence Scale forchildren (3rd edn). London: Psychological Corporation.

Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence(WASI). London: The Psychological Corporation.

Accepted for publication: 5 July 2012Published online: 22 August 2012

1166 Emily Simonoff et al. J Child Psychol Psychiatry 2012; 53(11): 1157–66

� 2012 The Authors. Journal of Child Psychology and Psychiatry � 2012 Association for Child and Adolescent Mental Health.