522352 research-article2014

AUT0010.1177/1362361314522352AutismSizoo et al.

Original Article

Temperament and character as endophenotype in adults with autism spectrum disorders or attention deficit/hyperactivity disorder

Autism 2015, Vol. 19(4) 400­–408 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1362361314522352 aut.sagepub.com

Bram B Sizoo1, Rutger Jan van der Gaag2 and Wim van den Brink3

Abstract Autism spectrum disorder and attention deficit/hyperactivity disorder overlap in several ways, raising questions about the nature of this comorbidity. Rommelse et al. published an innovative review of candidate endophenotypes for autism spectrum disorder and attention deficit/hyperactivity disorder in cognitive and brain domains. They found that all the endophenotypic impairments that were reviewed in attention deficit/hyperactivity disorder were also present in autism spectrum disorder, suggesting a continuity model with attention deficit/hyperactivity disorder as “a light form of autism spectrum disorder.” Using existing data, 75 adults with autism spectrum disorder and 53 with attention deficit/hyperactivity disorder were directly compared on autistic symptoms with the autism spectrum quotient, and on the endophenotypic measure of temperament and character, using the Abbreviated (Dutch: Verkorte) Temperament and Character Inventory. Based on the hypothesis that attention deficit/hyperactivity disorder and autism spectrum disorder are disorders on a continuous spectrum, autism spectrum quotient scores and abbreviated Temperament and Character Inventory scores were expected to be different from normal controls in both disorders in a similar direction. In addition, the autism spectrum quotient and abbreviated Temperament and Character Inventory scores were expected to be closely correlated. These conditions applied to only two of the seven Abbreviated Temperament and Character Inventory scales (harm avoidance and self-directedness), suggesting that temperament and character as an endophenotype of autism spectrum disorder and attention deficit/hyperactivity disorder provides only partial support for the continuity hypothesis of autism spectrum disorder and attention deficit/hyperactivity disorder. Keywords attention deficit/hyperactivity disorder, autism spectrum quotient, autism, character, endophenotype, personality, temperament

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition with a large multigenetic component and neurobiological characteristics, which are presumed to have an onset early in gestation (Courchesne et al., 2007; Minshew and Keller, 2010; Schumann et al., 2009, 2010; Wegiel et al., 2010). The prevalence of ASD in the adult population is estimated to be 9.8 per 1000 (Brugha et al., 2011). The diagnosis of ASD is based on persistent deficits in social communication and social interaction (SI) across contexts, and restricted, repetitive patterns of behavior, interests, or activities. Another developmental disorder that usually becomes manifest early in life is attention deficit/hyperactivity disorder (ADHD), characterized by a combination of impairments in attention with impulsivity and hyperactivity. Like ASD, ADHD is also believed to have a neurobiological origin

(Banaschewski et al., 2005; Doyle et al., 2005; Solanto, 2002). Developmental disorders have a major impact on the way people interact with their environment. Eventually this may lead to stable, but not necessarily functional, behavioral patterns. Although ASD and ADHD are usually diagnosed in childhood, novel diagnoses of ASD and ADHD in adults have risen considerably in recent years. However, the assessment of ASD and ADHD in adults is 1Dimence,

The Netherlands Medical Center St Radboud, The Netherlands 3Academic Medical Center University of Amsterdam, The Netherlands 2University

Corresponding author: Bram B Sizoo, Center for Developmental Disorders, Dimence, PO Box 5003, 7400 GC Deventer, The Netherlands. Email: [email protected]

Sizoo et al. hampered by the lack of reliable information on early development, and the heterogenic phenotype of the disorders. In addition, differentiating ASD from ADHD can be challenging in clinical practice, both in children and in adults, because the disorders present with overlapping symptoms. This overlap has contributed to alternating diagnoses in children (Jensen et al., 1997; Keene et al., 1999). The overlap is also reflected in high levels of comorbidity. For example, Goldstein and Schwebach (2004) reported that 33% of the children with ASD also fulfilled the criteria for an inattentive-type ADHD, and 26% for a combined-type ADHD. Conversely, Clark et al. (1999) showed that 65%–80% of the parents of children with ADHD also reported ASD symptoms. Few studies have examined the prevalence of the comorbidity between ASD and ADHD in adults. In a Danish birth cohort 17.9% of the reported ASD cases were comorbid with ADHD (Abdallah et al., 2011). Despite the frequent occurrence of ASD and ADHD symptoms, clinicians have been dissuaded from diagnosing ASD and ADHD as comorbid disorders because of the phrasing in the diagnostic guidelines of Diagnostic and Statistical Manual of Mental Disorders– Fourth Edition (DSM-IV; American Psychiatric Association (APA), 2000). Instead, ADHD symptoms are often considered to be part of the ASD diagnosis (Reiersen and Todd, 2008). The issue of comorbidity in psychiatry has been the subject of discussion. In medicine, comorbidity refers to the co-occurrence of distinct and well-validated diseases. In psychiatry, however, groups of symptoms constitute disorders, which have been systematically classified, but in reality represent stochastic probabilities of symptoms within a continuum. In the distinction between homotypic comorbidity (within a diagnostic group) and heterotypic comorbidity (between diagnostic groups), the issue is raised whether comorbid conditions have to be seen as etiologically independent or rather as a variation within a diagnostic phenotype (Angold et al., 1999). Traditionally, comorbidity is studied by comparing symptoms in the phenotypical domain. Another way to investigate the comorbidity between ASD and ADHD is from the endophenotypical perspective, because heterogeneous developmental disorders present with phenotypes that consist of behavioral adaptations to underlying causal agents, like dispositional temperament and character traits, which are considered to be “closer” to the genetic make-up than the phenotypic symptoms (Gottesman and Gould, 2003). However, the closeness to genes of endophenotypes relative to phenotypes has been challenged (Flint and Munafo, 2007). Generally accepted criteria for endophenotypes are (a) association with illness in the general population, (b) heritability, (c) primarily state-independent, (d) co-segregation of endophenotype and illness in families, and (e) endophenotypes also found in other family members at a higher rate than in the general population (Gottesman and Gould, 2003). With respect to the endophenotypes of ASD and ADHD, Rommelse et al. (2011) reviewed candidate

401 endophenotypes, and their possible role in the way ASD and ADHD relate to each other. The review covered several different endophenotype domains ranging from brain volume, cortical thickness, structural, and functional connectivity to functional brain processes. The authors did not find any functional or structural domain that was impaired in ADHD but not in ASD, except for preliminary data on delay aversion. Consequently, Rommelse et al. discussed the possibility that ASD and ADHD are different manifestations of the same overarching disorder: “… ADHD can thus be seen as a mild less impaired type within the more severe ASD spectrum with less genes and less endophenotypic impairments involved” (p. 1386). Among the candidate endophenotypes in the Rommelse review were temperament and character traits. The authors are among the first to explicitly posit temperament and character as possible endophenotypes, but did not present an elaborate hypothesis on how these temperament and character concepts might relate to the ASD–ADHD overlap. Although temperament and character are not always recognized as typical endophenotypes, they may be considered as such because they are normally distributed in the general population, are more homogenous than the clinical phenotype of psychiatric disorders, are relatively stable, appear early in life, and show high heritability (Caspi et al., 2005; McCrae et al., 2000). Both temperament and character describe “endogeneous basic tendencies of thoughts, emotions, and behavior” from at least preschool onwards (Caspi et al., 2005; De Pauw et al., 2009). Temperament and character should be seen as quantitative measures on a spectrum (Shiner and Caspi, 2003). In family studies, the temperament and character profiles of non-ASD relatives of ASD probands were different compared to the general population, with increased expression of the traits aloof, anxious, impulsive, shy, oversensitive, irritable, and eccentric (Murphy et al., 2000). Although temperament and character measures overlap to a certain extent with autistic symptoms on item level, this does not imply that the constructs temperament/character and autism are one and the same. One measure to assess ASD symptoms is the autism spectrum quotient (AQ). This instrument was developed to quantify autistic symptoms which are believed to lie on a continuum of socialcommunication skills (Baron-Cohen et al., 2001). Whereas people with ASD score at the extreme end of the scale, non-autistic parents of autistic children score lower on the AQ than their autistic children, but they score significantly higher than the general population (Wheelwright et al., 2010). These observations led several authors to speculate on the relationship between the broader autistic phenotype and personality traits. Hans Asperger was among the first to refer to the condition he described (which later became known as the Asperger syndrome) in terms of personality pathology. The relationship between personality traits and the autistic phenotype has also been studied in the general population, and in individuals with ASD. With respect to

402 the relationship between temperament and character of ASD and ADHD, temperament can serve as a “vulnerability” indicator as it can be assessed very early—before an ASD or ADHD diagnosis is made. At the same time, having ASD or ADHD will also influence the interactions with the environment and to some extent “scar” temperament and character. Yet specific personality traits conversely influence how individuals cope with the disorder (pathoplasty effect). In addition, the broader ADHD and ASD phenotypes (and the current absence of major differences) also support to some extent a spectrum conceptualization, suggesting that there might be an underlying genotype influencing both ADHD and ASD expressions. However, only longitudinal studies will be able to clarify the complex interdependence of traits and disorders. In summary, there are indications that the vulnerability, scar, pathoplasty, and spectrum model all apply to some extent for the ASD– ADHD versus temperament–character linkages. Below, an overview is presented of population studies and clinical studies among adults and children on temperament and character in relation to ASD and ADHD.

Population studies Kunihira et al. (2006) evaluated in 1513 adult students the relationship between autistic traits measured with the AQ and four temperament categories, using 60 questions from the 240-item Temperament and Character Inventory (TCI). The AQ was negatively correlated with novelty seeking and reward dependence and was positively correlated with harm avoidance, but not with the temperament category persistence. The authors concluded that high AQ scores were associated with what Cloninger (1987) defined as an obsessional personality trait.

Clinical studies—adults However, none of the above studies were conducted with individuals who had a confirmed ASD diagnosis. The first study among patients with ASD was by Soderstrom et al. (2002), indicating that 31 adults with Asperger syndrome scored significantly higher than normal controls on the temperament category harm avoidance and lower on the character categories self-directedness and cooperativeness of the TCI. Scores on reward dependence tended to be low (Soderstrom et al., 2002). In the same year, Nigg et al. (2002) published data on the relationship between the NEO Five-Factor Inventory and self-reported ADHD symptoms in adults with and without ADHD. The largest positive correlation was with neuroticism, and the largest negative correlation was with agreeableness and conscientiousness (Nigg et al., 2002). Anckarsäter et al. (2006) presented TCI data on 93 adults with ADHD, 66 with ASD, and 47 with ASD and ADHD. The ASD group showed low reward dependence, novelty seeking and self-directedness, and high harm avoidance compared to normal controls,

Autism 19(4) which was consistent with the Soderstrom data of 2002. The ADHD group presented with high novelty seeking, harm avoidance and self-transcendence, and low selfdirectedness and cooperativeness, compared to normal controls. The comorbid group showed high novelty seeking and harm avoidance, and low self-directedness and cooperativeness. Patients in this comorbid group showed more ADHD-like temperament and lower character scores than subjects with only ASD. The authors did not compare the ASD with the ADHD group, nor did they use the AQ as a measure of autistic symptoms. The conclusion was that ASD and ADHD present with specific personality profiles and that neurocognitive development during childhood may impair healthy character development producing personality pathology in adulthood (Anckarsäter et al., 2006). Sizoo et al. (2009a) studied personality in the same group of subjects discussed in this article, using the Abbreviated (Dutch: Verkorte) Temperament and Character Inventory (VTCI). The ASD group was characterized by high harm avoidance and self-transcendence, and low reward dependence, self-directedness, and cooperativeness, compared to the norm groups. The ADHD group was characterized by high novelty seeking, harm avoidance and self-transcendence, and low self-directedness and cooperativeness. This study did not include a comorbid ASD with ADHD group (Sizoo et al., 2009a).

Clinical studies—children Tillman et al. compared 101 early bipolar, 68 ADHD, and 94 normal controls with a junior-TCI version. All participants were between 7 and 16 years old. In both the bipolar and ADHD groups, novelty seeking was significantly higher than in controls. Both groups showed significantly lower reward dependence than controls and higher harm avoidance (Tillman et al., 2003). De Pauw et al. (2011) studied 175 children with ASD (mean age = 10.3 years) reporting higher negative affect, and lower effortful control and surgency, as well as low extraversion and imagination. The same authors studied 84 children with ADHD (mean age = 10.1 years) showing low scores for effortful control and high scores for conscientiousness and benevolence, compared to children in a community sample (De Pauw and Mervielde, 2011). In a direct comparison of 27 children with ASD and 27 children with ADHD (between 10 and 14 years old), Samyn et al. (2011) confirmed that both developmental disorders present with a low measure of effortful control. Effortful control can be conceptualized as the relatively deliberate “top-down” modulation in the service of more distant goals (Eisenberg et al., 2000). It is an interesting endophenotype, and is related to other endophenotypes in the Rommelse review like executive functioning and sustained attention. The broad clinical impression is that although effortful control may be impaired in both ASD and ADHD patients, ADHD patients generally seem to know what the distant goal is but cannot

403

Sizoo et al. pursue it, whereas ASD patients can pursue, but are unable to imagine a distant goal. This needs further investigation. Finally, Poustka et al. (2011) used a child version of the TCI and the Social Responsiveness Scale (SRS) as a measure of autistic symptoms, in a group of 36 children with ASD and 32 children with ADHD between 7 and 11 years old with normal intelligence (IQ between 100 and 103). Children with ASD showed very low reward dependence (T-score = 20.5) and differed significantly from the ADHD children in harm avoidance, reward dependence, selfdirectedness, and cooperativeness. Interestingly, ADHD children scored only moderately high on novelty seeking (T-score = 53.9) and were not statistically different from the ASD group. In these studies, instruments based on the Five-Factor Model (FFM) (e.g. the NEO Personality Inventory– Revised (NEO-PI-R), the Personality Mini Markers) and on Cloninger’s TCI have been used to assess temperament and character traits. Although there is considered to be a substantial relationship between the TCI and the FFM, the two models are not equivalent (De Fruyt et al., 2000). In summary, the above studies point to patterns in temperament and character traits in ASD or ADHD. In all the above studies among adults, scores on harm avoidance and self-directedness consistently differ from the normal population scores in the same direction, for both ASD and ADHD. If the continuity hypothesis for ADHD as a light form of ASD, as proposed by Rommelse et al., is true for temperament and character as endophenotype, we would expect to find (a) elevated AQ scores in ASD and ADHD relative to normal controls, (b) VTCI scores for ASD and ADHD that differ in the same direction from normal controls, and (c) a clear correlation between AQ scores and VTCI scores. To explore this hypothesis, data that were collected in the context of a broader study were used.

Methods Subjects As part of a broader study described in detail elsewhere (Sizoo et al., 2009b), 75 adults with ASD and 53 adults with ADHD completed the Dutch versions of the AQ and the VTCI. All participants had an IQ above 80, as measured with the Wechsler Adult Intelligence Scale–III (WAIS-III). Subjects were recruited as a consecutive sample from two specialized outpatient centers for developmental disorders in adults between November 2005 and June 2007. All diagnoses were first established in adulthood, but prior to inclusion. In both centers, ASD diagnoses were based on semi-structured clinical interviews such as the Autism Diagnostic Interview–Revised (ADI-R; Lord et al., 1994) or the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000) as well as on DSM-IV checklists, and all available information from schools and child psychiatric services concerning the development during childhood. ADHD diagnoses were made according

to a national protocol, including a semi-structured developmental history and a DSM-IV criteria checklist for adult and childhood ADHD symptoms. The group consisted of individuals with and without a (former) comorbid substance-use disorder. This comorbidity is described in detail elsewhere (Sizoo et al., 2009b). For further details on the diagnostic procedure, the reader is referred to Sizoo et al. (2009a). Approval for the study was obtained from the regional medical ethical committee. After complete description of the study to the subjects, written informed consent was obtained.

Instruments Several publications by different research groups have established the psychometric properties and validity of the AQ (Baron-Cohen et al., 2001; Kurita et al., 2005). The AQ is a self-report instrument consisting of 50 statements reflecting personal beliefs, views, and preferences. Subjects respond on a 4-point Likert scale ranging from 1 to 4 (definitely agree, slightly agree, slightly disagree, definitely disagree), resulting in scores ranging from 50 to 200. In some studies, the answers are dichotomized (0 or 1), resulting in scores ranging from 0 to 50. In our study, we used the Likert scale scores in order to relate our results to those obtained from the general population in the validation study by Hoekstra et al. (2008). The 50 items were divided by the original authors into five theoretical subscales of 10 items each: social skills, communication, imagination, attention to details, and attention switching. Hoekstra et al. (2008) conducted a factor analysis resulting in two factors. The factor SI consists of the subscales social skills, communication, imagination, and attention switching, while the second factor attention to details (AD) is equivalent to the fifth subscale (Hoekstra et al., 2008). The original TCI consists of four temperament factors (novelty seeking, harm avoidance, reward dependence, and persistence) and three character factors (self-directedness, cooperativeness, and self-transcendence; Cloninger et al., 1993). The abbreviated version (VTCI, Dutch) has 105 instead of 238 items, but consists of the same primary scales with similar psychometric properties as the TCI (Duijsens et al., 2000). The VTCI has excellent psychometric qualities, based on data of 657 Dutch healthy controls from the general population (43% men, age = 43.7 years, standard deviation (SD) = 15.7 years, range = 16–90 years; Duijsens et al., 2000).

Statistics The ASD and ADHD groups were compared with respect to age, total IQ, and gender. To allow for comparison of the VTCI scores with data from previous studies, raw scores were converted into gender-corrected T-values, using the Dutch norm group (Duijsens et al., 2000). The AQ scores

404

Autism 19(4)

Table 1.  Comparison of demographic characteristics of adults with ASD or ADHD. ASD (N = 75)  

n (%)a

Age Total IQ Gender (female) Living alone With work/study Years of education Comorbid SUDb

ADHD (N = 53) Mean

SD

34.3 102.9

11.87 13.64

14 (19) 48 (67) 46 (66)

n (%)

Test statistic

Mean

SD

32.1 104.0

11.37 10.53

14.0

3.08

18 (34) 42 (79) 41 (77) 15.3

2.97

21 (26)

32 (60)

F126 = 1.114, p = .293 F85 = 0.133, p = .717 χ2(1) = 3.875, p = .049 χ2(1) = 1.751, p = .186 χ2(1) = 1.975, p = .160 F122 = 5.547, p = .02 χ2(1) = 16.053, p = .000

SD: standard deviation; ASD: autism spectrum disorder; ADHD: attention deficit/hyperactivity disorder. aNumbers in the N column represent frequency (percentage). bSUD: current or former substance-use disorder.

Table 2.  Mean scores for adult patients with ASD or ADHD on two factors of the autism spectrum quotient (AQ; Hoekstra et al., 2008) and the VTCI.

  VTCI             AQ    

Novelty seeking Harm avoidance Reward dependence Persistence Self-directedness Cooperativeness Self-transcendence Total AQ score Social Interaction Attention to details

Population

ASD (N = 75)

ADHD (N = 53)

Mean (SD)

Mean (SD)

Mean (SD)

50 (10)a 50 (10)a 50 (10)a 50 (10)a 50 (10)a 50 (10)a 50 (10)a 104.2 (11.29) 79.9 (10.68) 24.3 (4.97)

48.4 (9.08) 61.9 (10.34)b 47.2 (11.62) 51.7 (9.54) 35.3 (13.16)b 46.2 (12.53) 54.9 (11.79)b 133.8 (19.86)b 108.0 (18.58)b 25.8 (6.45)

60.9 (9.68)b 55.7 (11.44)b 50.0 (10.67) 50.6 (10.73) 34.4 (14.65)b 45.9 (13.24) 53.8 (10.80) 113.5 (15.55)b 89.1 (13.96)b 24.5 (6.34)

p

0.000 0.002 0.160 0.557 0.708 0.889 0.576 0.000 0.000 0.239

ASD: autism spectrum disorder; ADHD: attention deficit/hyperactivity disorder; AQ: autism spectrum quotient with Likert scale scores (range = 50–200); VTCI: Abbreviated (Dutch: Verkorte) Temperament and Character Inventory with T-scores (Duijsens et al., 2000); SD: standard deviation. Bold print indicates Bonferroni-corrected significant difference (p < .005). aFor the VTCI scales all T-scores are based on a population mean of 50 and SD of 10. bIndicates that score differs significantly from the mean in the general population.

were compared to the norm group data from Hoekstra et al. (2008) using T-tests. In the between-group analyses, the scores on the two AQ factors and the seven VTCI scales were compared between the ASD and ADHD groups. The significance level for this comparison was set at α = 0.005, accounting for multiple testing of the nine scales/factors (α = 0.05/9). In the within-group analyses, Pearson correlation coefficients were computed between the VTCI scales and the AQ factors separate for the groups of ASD and ADHD subjects. The results were normally distributed in the ASD and ADHD samples. SPSS-15 software was used for all computations.

3.875, p = .049) and this group also had more educational experience than the ADHD group (F122 = 5.547, p = .02). Substance-use disorders were more prevalent in the ADHD than the ASD group (χ2(1) = 16.053, p = .000). Because correcting for these group differences yielded similar results as the uncorrected computations, the results that follow have not been corrected for gender, years of education, and substance-use disorder. In this study, internal consistencies of the AQ factors and VTCI scales were good in both groups (Table 3) and similar to those found in Dutch norm groups for the AQ (Hoekstra et al., 2008) and the VTCI (Duijsens et al., 2000).

Results

Between-group differences (ASD vs ADHD) in VTCI subscale and AQ factor scores

The groups were comparable with respect to age, IQ, partner status, and current work or study status (Table 1). Males were overrepresented in the ASD group (χ2(1) =

The mean scores for the AQ factors and the VTCI scales are shown in Table 2. The total AQ score and the AQ factor score SI were significantly higher in the ASD compared to

405

Sizoo et al.

Table 3.  Pearson’s product–moment correlations between the AQ scores and the VTCI scores with scale reliability presented as internal consistency (α = Cronbach’s alpha) for adults with ASD and for adults with ADHD. ASD (N = 75)   NS HA RD P SD CO ST

α .674 .859 .765 .669 .790 .819 .809

ADHD (N = 53)

TAQ

SI

AD

.795 −.218 .348** −.507** −.062 −.368** −.469** −.089

.790 −.236 .430** −.531** −.156 −.387** −.437** −.187

.689 .010 −.167 −.029 .257 −.019 −.185 .265

α .753 .866 .763 .740 .849 .848 .786

TAQ

SI

AD

.661 −.263 .556** −.222 .085 −.383** −.278 −.121

.631 −.269 .630** −.283 −.015 −.356** −.306 −.128

.601 −.053 −.023 .079 .240 −.157 −.008 −.015

AQ: autism spectrum quotient; VTCI: Abbreviated (Dutch: Verkorte) Temperament and Character Inventory; ASD: autism spectrum disorder; ADHD: attention deficit/hyperactivity disorder; TAQ: total autism spectrum quotient; SI: social interaction; AD: attention to details; NS: novelty seeking; HA: harm avoidance; RD: reward dependence; P: persistence; SD: self-directedness; CO: cooperation; ST: self-transcendence. **p < .001.

the ADHD group, whereas the factor score AD did not differ between the ASD and the ADHD group. Using data from the AQ validation study by Hoekstra et al. (2008), T-tests indicated that in the ASD and the ADHD group, the total AQ score and the factor score SI differed significantly from the general population mean. Compared to the ASD group, the ADHD group scored significantly higher on novelty seeking, pointing to the presence of more curious, impulsive, quick tempered, easily bored, and careless traits in the ADHD group. Novelty seeking in the ADHD group was also significantly higher than in the norm group, but this was not so in the ASD group. In contrast, the ASD group scored higher on harm avoidance than the ADHD group, indicating more pessimism and anxiousness. However, compared to the norm group, harm avoidance was significantly higher in both ASD and ADHD. Reward dependence, persistence, and cooperativeness did not differ between ASD, ADHD, or the norm group. The scores for self-directedness were similar in the ASD and the ADHD group, and significantly lower than in the norm group. Finally, the score on selftranscendence in the ASD group was significantly higher than in the norm group, but this was not the case for the ADHD group.

Within-group associations between AQ factors and VTCI scales The Pearson’s correlations coefficients between the AQ total, and factor scores and the VTCI scores were calculated separately for the ASD and ADHD group (Table 3). Within the ASD group, the total AQ score is positively correlated with harm avoidance, and negatively correlated with reward dependence, self-directedness, and cooperation. A similar pattern of correlations is seen for the AQ factor SI. In the ADHD group, the AQ total and SI scores are positively correlated with harm avoidance, and

negatively with self-directedness. The AQ factor AD does not correlate with any of the VTCI scales. In summary, only harm avoidance and self-directedness correlated with the AQ scores in both ASD and ADHD groups.

Discussion The aim of the study was to investigate the continuity hypothesis posited by Rommelse et al., suggesting that ADHD can be seen as “a light form of ASD,” based on a review of endophenotypes. There is sufficient support in the literature to regard temperament and character traits as endophenotypes of ASD and ADHD. We hypothesized that if ADHD is indeed continuous with ASD, the autistic symptoms in both groups should be higher, and the temperament and character traits in both the ASD and ADHD groups should be either higher or lower than the norm groups, showing clear correlations between the autistic symptoms and the temperament and character traits in both groups. The high score for total AQ and the AQ factor score SI in both diagnostic groups corresponds with earlier results. It appears, however, that two of the seven VTCI traits show similar scores in ASD and ADHD, judging by the high scores for harm avoidance and the low scores for self-directedness in both diagnostic groups, relative to the norm groups. This result is in line with what could be expected from earlier studies described in literature that show a consistent pattern of high harm avoidance (pessimistic and anxious versus optimistic and risk-taking) and low self-directedness (blaming and inept vs responsible and resourceful) in both ASD and ADHD. Two temperament traits (reward dependence and persistence) and the character trait cooperativeness present with inconclusive scores that do not differ from norm groups in either diagnostic group. Novelty seeking is a temperament trait that is prominent in ADHD, but not in ASD, indicating that the tendency to be curious, careless, easily bored, and

406 impulsive could be overrepresented in individuals with ADHD compared to individuals with ASD. Although the findings of high novelty seeking in only ADHD and high self-transcendence in only ASD are not in line with the continuity hypothesis, they do not eliminate the possibility of continuity either. Whereas previous articles on this study sample explored similarities and differences between ASD and ADHD patients in separate domains, this study attempts to make a cross-dimensional comparison between the temperament and character domain and the autism spectrum domain from an endophenotype perspective. The results in this and other studies are cross-sectional. The only sound way to study interrelationships between different domains (personality, neuropsychology, genetics), however, is through multidisciplinary longitudinal designs. In terms of temperament and character as endophenotypes, it would be interesting to explore the genetic underpinnings of temperament and character traits that differ in ASD and ADHD, such as novelty seeking and self-transcendence. In this respect, the reported overrepresentation in ADHD of the T-allele (marker rs1843809) in the brainexpressed gene tryptophan hydroxylase (TPH2, a rate limiting enzyme in the synthesis of serotonin) is of interest (Sheehan et al., 2005). Sizoo et al. (2010), using the same study sample as presented here, found that the TPH2 gene, among two other candidate genes, was expressed differently in ASD and ADHD with nominal significance. Future research would need to examine the relationship between temperament and character endophenotypes in ASD and ADHD and genetic markers. This study has a number of limitations. First, the study is restricted to patients with either ASD or ADHD. This is in line with the DSM-IV algorithm, but it is well possible that some subjects would have been diagnosed with both ASD and ADHD if exclusion rules would have been ignored. This may have attenuated the differences between both diagnostic groups, but it does not explain the observed differences in novelty seeking between the diagnostic groups. However, we found that only 6 of the 53 individuals with ADHD had a clinical score on the AQ of 26 or more (range = 27–35). Second, the subjects that participated were all diagnosed in adulthood in two closely related diagnostic centers prior to inclusion in the study. Any ADHD symptoms in subjects with ASD were at that time regarded as attributable to the ASD diagnosis (Reiersen and Todd, 2008). Third, the sample consisted of individuals with and without a (former) substance-use disorder and this comorbidity was twice as high in the ADHD group (60%) compared to the ASD group (26%). This imbalance may have influenced our findings. However, post hoc analysis without the comorbid cases yielded similar results. Fourth, we used the VTCI which does not include subscales, and although the shorter version has advantages when used for patients with a limited capacity to focus for a long period of

Autism 19(4) time, it would have been very useful to explore the data at the facet level, using the TCI. Finally, all subjects were recruited in clinical settings and therefore the results cannot be simply translated to ASD and ADHD in the general population. More research is needed to verify the findings and explore the endophenotypes of ASD and ADHD, preferably in multi-domain longitudinal studies. In conclusion, taking the approach of temperament and character measures as endophenotype offers new opportunities for studying the overlap between neurodevelopmental disorders like ASD and ADHD. The results in this study suggest that there is some support for the hypothesis that ADHD is continuous with ASD, although ASD and ADHD also present with disorder-specific endophenotypical patterns in temperament and character. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References Abdallah MW, Greaves-Lord K, Grove J, et al. (2011) Psychiatric comorbidities in autism spectrum disorders: findings from a Danish Historic Birth Cohort. European Child & Adolescent Psychiatry 20: 599–601. American Psychiatric Association (APA) (2000) Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (4th edn). Arlington, VA: American Psychiatric Association. Anckarsäter H, Stahlberg O, Larson T, et al. (2006) The impact of ADHD and autism spectrum disorders on temperament, character, and personality development. The American Journal of Psychiatry 163: 1239–1244. Angold A, Costello EJ and Erkanli A (1999) Comorbidity. Journal of Child Psychology and Psychiatry 40: 57–87. Banaschewski T, Hollis C, Oosterlaan J, et al. (2005) Towards an understanding of unique and shared pathways in the psychopathophysiology of ADHD. Developmental Science 8: 132–140. Baron-Cohen S, Wheelwright S, Skinner R, et al. (2001) The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders 31: 5–17. Brugha TS, McManus S, Bankart J, et al. (2011) Epidemiology of autism spectrum disorders in adults in the community in England. Archives of General Psychiatry 68: 459–465. Caspi A, Roberts BW and Shiner RL (2005) Personality development: stability and change. Annual Review of Psychology 56: 453–484. Clark T, Feehan C, Tinline C, et al. (1999) Autistic symptoms in children with attention deficit-hyperactivity disorder. European Child & Adolescent Psychiatry 8: 50–55. Cloninger CR (1987) Neurogenetic adaptive mechanisms in alcoholism. Science 236: 410–416. Cloninger CR, Svrakic DM and Przybeck TR (1993) A psychobiological model of temperament and character. Archives of General Psychiatry 50: 975–990.

Sizoo et al. Courchesne E, Pierce K, Schumann CM, et al. (2007) Mapping early brain development in autism. Neuron 56: 399–413. De Fruyt F, Van De Wiele L and Van Heeringen C (2000) Cloninger’s psychobiological model of temperament and character and the five-factor model of personality. Personality and Individual Differences 29: 441–452. De Pauw SS and Mervielde I (2011) The role of temperament and personality in problem behaviors of children with ADHD. Journal of Abnormal Child Psychology 39: 277–291. De Pauw SS, Mervielde I and Van Leeuwen KG (2009) How are traits related to problem behavior in preschoolers? Similarities and contrasts between temperament and personality. Journal of Abnormal Child Psychology 37: 309–325. De Pauw SS, Mervielde I, Van Leeuwen KG, et al. (2011) How temperament and personality contribute to the maladjustment of children with autism. Journal of Autism and Developmental Disorders 41: 196–212. Doyle AE, Willcutt EG, Seidman LJ, et al. (2005) Attentiondeficit/hyperactivity disorder endophenotypes. Biological Psychiatry 57: 1324–1335. Duijsens IJ, Spinhoven P, Goekoop JG, et al. (2000) The Dutch Temperament and Character Inventory (TCI): dimensional structure, reliability and validity in a normal and psychiatric outpatient sample. Personality and Individual Differences 28: 487–499. Eisenberg N, Fabes RA, Guthrie IK, et al. (2000) Dispositional emotionality and regulation: their role in predicting quality of social functioning. Journal of Personality and Social Psychology 78: 136–157. Flint J and Munafo MR (2007) The endophenotype concept in psychiatric genetics. Psychological Medicine 37: 163–180. Goldstein S and Schwebach AJ (2004) The comorbidity of Pervasive Developmental Disorder and Attention Deficit Hyperactivity Disorder: results of a retrospective chart review. Journal of Autism and Developmental Disorders 34: 329–339. Gottesman II and Gould TD (2003) The endophenotype concept in psychiatry: etymology and strategic intentions. The American Journal of Psychiatry 160: 636–645. Hoekstra RA, Bartels M, Cath DC, et al. (2008) Factor structure, reliability and criterion validity of the autism-spectrum quotient (AQ): a study in Dutch population and patient groups. Journal of Autism and Developmental Disorders 38: 1555–1566. Jensen VK, Larrieu JA and Mack KK (1997) Differential diagnosis between attention-deficit/hyperactivity disorder and pervasive developmental disorder–not otherwise specified. Clinical Pediatrics (Phila) 36: 555–561. Keene J, Willner P and Love AC (1999) The relevance of problems and models to treatment outcome: a comparative study of two agencies. Substance Use & Misuse 34: 1347–1369. Kunihira Y, Senju A, Dairoku H, et al. (2006) “Autistic” traits in non-autistic Japanese populations: relationships with personality traits and cognitive ability. Journal of Autism and Developmental Disorders 36: 553–566. Kurita H, Koyama T and Osada H (2005) Autism-spectrum quotientJapanese version and its short forms for screening normally intelligent persons with pervasive developmental disorders. Psychiatry and Clinical Neurosciences 59: 490–496. Lord C, Risi S, Lambrecht L, et al. (2000) The autism diagnostic observation schedule-generic: a standard measure of social

407 and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders 30: 205–223. Lord C, Rutter M and Le Couteur A (1994) Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders 24: 659–685. McCrae RR, Costa PT Jr, Ostendorf F, et al. (2000) Nature over nurture: temperament, personality, and life span development. Journal of Personality and Social Psychology 78: 173–186. Minshew NJ and Keller TA (2010) The nature of brain dysfunction in autism: functional brain imaging studies. Current Opinion in Neurology 23: 124–130. Murphy M, Bolton PF, Pickles A, et al. (2000) Personality traits of the relatives of autistic probands. Psychological Medicine 30: 1411–1424. Nigg JT, John OP, Blaskey LG, et al. (2002) Big five dimensions and ADHD symptoms: links between personality traits and clinical symptoms. Journal of Personality and Social Psychology 83: 451–469. Poustka L, Bender F, Bock M, et al. (2011) Personality and social responsiveness in autism spectrum disorders and attention deficit/hyperactivity disorder. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie 39: 133–141. Reiersen AM and Todd RD (2008) Co-occurrence of ADHD and autism spectrum disorders: phenomenology and treatment. Expert Review of Neurotherapeutics 8: 657–669. Rommelse NN, Geurts HM, Franke B, et al. (2011) A review on cognitive and brain endophenotypes that may be common in autism spectrum disorder and attention-deficit/hyperactivity disorder and facilitate the search for pleiotropic genes. Neuroscience and Biobehavioral Reviews 35: 1363–1396. Samyn V, Roeyers H and Bijttebier P (2011) Effortful control in typically developing boys and in boys with ADHD or autism spectrum disorder. Research in Developmental Disabilities 32: 483–490. Schumann CM, Barnes CC, Lord C, et al. (2009) Amygdala enlargement in toddlers with autism related to severity of social and communication impairments. Biological Psychiatry 66: 942–949. Schumann CM, Bloss CS, Barnes CC, et al. (2010) Longitudinal magnetic resonance imaging study of cortical development through early childhood in autism. The Journal of Neuroscience 30: 4419–4427. Sheehan K, Lowe N, Kirley A, et al. (2005) Tryptophan hydroxylase 2 (TPH2) gene variants associated with ADHD. Molecular Psychiatry 10: 944–949. Shiner R and Caspi A (2003) Personality differences in childhood and adolescence: measurement, development, and consequences. Journal of Child Psychology and Psychiatry, and Allied Disciplines 44: 2–32. Sizoo B, van den Brink W, Franke B, et al. (2010) Do candidate genes discriminate patients with an autism spectrum disorder from those with attention deficit/hyperactivity disorder and is there an effect of lifetime substance use disorders? The World Journal of Biological Psychiatry 11: 699–708. Sizoo B, van den Brink W, Gorissen van Eenige M, et al. (2009a) Personality characteristics of adults with autism spectrum disorders or attention deficit hyperactivity disorder with and

408 without substance use disorders. The Journal of Nervous and Mental Disease 197: 450–454. Sizoo B, van den Brink W, Koeter M, et al. (2009b) Treatment seeking adults with autism or ADHD and co-morbid substance use disorder: prevalence, risk factors and functional disability. Drug and Alcohol Dependence. Epub ahead of print 27 September. DOI: 10.1016/j.drugalcdep.2009.09.003. Soderstrom H, Rastam M and Gillberg C (2002) Temperament and character in adults with Asperger syndrome. Autism 6: 287–297. Solanto MV (2002) Dopamine dysfunction in AD/HD: integrating clinical and basic neuroscience research. Behavioral Brain Research 130: 65–71.

Autism 19(4) Tillman R, Geller B, Craney JL, et al. (2003) Temperament and character factors in a prepubertal and early adolescent bipolar disorder phenotype compared to attention deficit hyperactive and normal controls. Journal of Child and Adolescent Psychopharmacology 13: 531–543. Wegiel J, Kuchna I, Nowicki K, et al. (2010) The neuropathology of autism: defects of neurogenesis and neuronal migration, and dysplastic changes. Acta Neuropathologica 119: 755–770. Wheelwright S, Auyeung B, Allison C, et al. (2010) Defining the broader, medium and narrow autism phenotype among parents using the autism-spectrum quotient (AQ). Molecular Autism 1: 10.