ORIGINAL ARTICLE

Emotion recognition in children and adolescents withattention-deficit/hyperactivity disorder (ADHD)

Christina Schwenck • Thekla Schneider • Jutta Schreckenbach •

Yvonne Zenglein • Angelika Gensthaler • Regina Taurines •

Christine M. Freitag • Wolfgang Schneider • Marcel Romanos

Received: 31 January 2013 / Accepted: 23 February 2013 / Published online: 9 March 2013

� Springer-Verlag Wien 2013

Abstract Children with attention-deficit/hyperactivity

disorder (ADHD) are impaired in social adaptation and

display deficits in social competence. Deficient emotion

recognition has been discussed to underlie these social

problems. However, comorbid conduct problems have not

been considered in the majority of studies conducted so far,

and the influence of medication on emotion recognition has

rarely been studied. Here, emotion recognition perfor-

mance was assessed in children with ADHD without

medication compared with children with ADHD under

stimulant medication and a matched control group. In order

to rule out confounding by externalizing symptoms,

children with comorbid conduct problems were excluded.

Video clips with neutral faces developing a basic emotion

(happiness, sadness, disgust, fear and anger) were pre-

sented in order to assess emotion recognition. Results

indicated between-group differences neither concerning the

number of correctly identified emotions nor concerning

reaction times and their standard deviations. Thus, we

suggest that ADHD per se is not associated with deficits in

emotion recognition.

Keywords ADHD � Emotion recognition �Conduct problems � Stimulant medication �Psychopharmacotherapy

Introduction

Attention-deficit/hyperactivity disorder (ADHD) comprises

symptom clusters of inattention, hyperactivity and impul-

sivity. The disorder is rather common in childhood and

adolescents and affects about 5–12 percent of this age

group (Biederman and Faraone 2005; Schachar and Tan-

nock 1995). Psychopharmacotherapy with stimulant med-

ication such as methylphenidate is highly effective in

reducing the core symptoms and is recommended to be

applied in severe cases combined with behavioral thera-

peutic measures according to most existing guidelines

(Kendall et al. 2008).

Children with ADHD show a high rate of social prob-

lems, and reduced social competence was found to be

highly associated with the disorder (Lee et al. 2012). This

result was observed by self-ratings as well as in parent and

teacher ratings. More precisely, children with ADHD were

found to be less accepted by peers and they are perceived

as less socially competent than comparison groups (DuPaul

C. Schwenck � Y. Zenglein � A. Gensthaler � C. M. Freitag

Department of Child and Adolescent Psychiatry, Psychosomatics

and Psychotherapy, Goethe-University, Frankfurt/M., Germany

C. Schwenck � T. Schneider � R. Taurines � M. Romanos

Department of Child and Adolescent Psychiatry,

Psychosomatics and Psychotherapy, University of Wurzburg,

Wurzburg, Germany

C. Schwenck (&)

Klinik fur Psychiatrie, Psychosomatik und Psychotherapie des

Kindes- und Jugendalters Johann, Wolfgang Goethe-Universitat

Frankfurt/M., Deutschordenstraße 50 (Haus 92),

60528 Frankfurt a. M., Germany

e-mail: christina.schwenck@kgu.de

T. Schneider

Department of Psychology, University of Konstanz,

Constance, Germany

J. Schreckenbach � W. Schneider

Department of Psychology, University of Wurzburg,

Wurzburg, Germany

J. Schreckenbach

Department of Child and Adolescent Psychiatry, Psychosomatics

and Psychotherapy, University of Rostock, Rostock, Germany

123

ADHD Atten Def Hyp Disord (2013) 5:295–302

DOI 10.1007/s12402-013-0104-z

et al. 2004; Hoza et al. 2005). Furthermore, these problems

are known to be a significant predictor of serious negative

outcomes in later adolescence and adulthood (Greene et al.

1997; Mrug et al. 2012). Social problems were found at

higher rates in those individuals with ADHD who exhibit

comorbid externalizing disorders, such as conduct disorder

and oppositional defiant disorder (Booster et al. 2012). This

association is of particular relevance considering the large

percentage of children with ADHD showing symptoms of

conduct disorder and vice versa (Hinshaw 1987; King et al.

2005; Taurines et al. 2010). In regard to therapeutic options

for the treatment of social problems, pharmacological

treatment with methylphenidate was found to reduce these

problems in children and adolescents with ADHD to a

significant extent (Pelham et al. 2001), although surpris-

ingly, psychotherapeutic treatments such as social skills

trainings did not have an additional effect on social diffi-

culties of the children (Abikoff et al. 2004; Storebø et al.

2011). The poor effectiveness of psychotherapy on social

difficulties could be attributed to the fact that the core

deficiencies underlying social problems in ADHD have not

been identified so far and, therefore, were not targeted by

psychotherapeutic treatments.

Different factors linked to emotional processing have

been discussed to underlie the social malfunctioning of

children with ADHD. Among others, deficient emotion

recognition has been suggested to play a central role for

social malfunctioning (Dadds et al. 2012). According to the

model of the ‘‘violence inhibition mechanism’’ (VIM),

intact emotion recognition is required for the inhibition of

aggressive behavior, and deficiencies might lead to

aggressive reactions toward others (Blair et al. 2004). A

large number of studies have been conducted on emotion

recognition in ADHD so far, the majority showing defi-

ciencies in ADHD compared to controls (Taurines et al.

2012). This applies to a larger number of incorrect

responses in identifying an emotion (Singh et al. 1998),

higher reaction times (RT) (Kats-Gold et al. 2007), simple

emotion recognition in static pictures (Corbett and Glidden

2000; Yuill and Lyon 2007) as well as complex emotion

recognition in a social context (Fonseca et al. 2008). Some

of the studies found a deficient performance for specific

emotions, respectively, joy and surprise (Sinzig et al. 2008)

as well as anger (Singh et al. 1998), but other studies did

not differentiate for specific emotion dimensions (Cadesky

et al. 2000; Dyck et al. 2001).

Despite the large number of studies in this field of

research, important factors that may have a significant

influence on emotion recognition in children and adoles-

cents with ADHD have been neglected so far: first of all,

only one of the studies controlled for symptoms of conduct

problems (Cadesky et al. 2000). As children with ADHD

and comorbid conduct problems were found to show higher

rates of social problems (Booster et al. 2012), the

assumption that comorbid conduct problems might have an

influence on emotion recognition deficits seems to be

likely. The authors of the only study controlling for con-

duct problems (Cadesky et al. 2000) differentiated a group

with pure ADHD, a group with pure conduct problems, a

group with combined symptomatology and a control group

(CG). Surprisingly, results indicated that the groups with

ADHD only and conduct problems only were outperformed

by controls, although children with combined symptom-

atology did not differ from the CG. However, the authors

presented static pictures of emotional faces to their par-

ticipants, which allowed for evaluating the number of

correctly identified emotions/number of errors without any

reaction time information. Another problem of the pre-

sentation of static pictures to assess emotion recognition is

the low ecological validity: in social interaction of daily

life, facial expressions change rapidly and usually do not

maintain the same. Interestingly, the errors children with

ADHD made were random and comparable to controls, but

children with conduct problems specifically displayed high

error rates for the emotion ‘‘anger.’’ Further studies either

did not report comorbidity (Corbett and Glidden 2000;

Kats-Gold et al. 2007; Shapiro et al. 1993) or did not

consider comorbid disorders as independent factors in their

analyses (Fonseca et al. 2008; Sinzig et al. 2008; Yuill and

Lyon 2007). To our knowledge, in none of the studies,

conduct problems were an exclusion criterion. Therefore,

though externalizing comorbidity was found to have a large

influence on the degree of social problems in children with

ADHD (Booster et al. 2012), this factor has not been

considered by the majority of studies on emotion recog-

nition, and the only study (Cadesky et al. 2000) conducted

on this research question has not been replicated yet.

Second, only one previous study has investigated the

influence of stimulant medication on emotion recognition

performance (Hall et al. 1999). The authors of the study

compared a group with ADHD, a group with ADHD and

comorbid learning disorders, and a CG. Children with

ADHD were tested twice, once with medication and once

without. Results indicated no differences in the recognition

of emotions in human faces, neither between groups nor as

a function of medication. Again, a set of static pictures

showing different emotions was used in the study to

measure emotion recognition performance. The results of

the study seem to be surprising taking the positive influ-

ence of pharmacological treatment on social problems into

account (Pelham et al. 2001), and results have not been

replicated so far. All other studies either consistently tested

children without medication (Fonseca et al. 2008; Kats-

Gold et al. 2007; Sinzig et al. 2008; Yuill and Lyon 2007)

or do not report about the procedure in regard to medical

status (Shapiro et al. 1993; Singh et al. 1998).

296 C. Schwenck et al.

123

In summary, deficient emotion recognition may play an

important role in increased rates of social problems of

children and adolescents with ADHD. Although various

studies have been conducted reporting deficient emotion

recognition in patients with ADHD, important factors that

might influence emotion recognition have yet insufficiently

been accounted for. Therefore, the aim of the current study

is to assess emotion recognition performance in children

and adolescents with ADHD excluding comorbid conduct

problems to avoid confounding. Furthermore, we assess the

influence of psychopharmacological treatment on emotion

recognition performance. Based on previous research, we

expected to find deficient emotion recognition performance

in children with ADHD without medication compared to

controls. Due to the lack of studies on the influence of

stimulant medication on emotion recognition in children

with ADHD, no specific hypothesis was formulated, and

effects were assessed in an exploratory way. For our study,

we adopted a research paradigm with moving instead of

static emotional expressions, which, on the one hand,

allowed us to assess RT additionally to the number of

correctly identified emotions and, on the other hand, yiel-

ded a higher ecological validity. Past studies on looking

behavior in children with autism have shown substantial

differences in regard to the material—static pictures or film

clips—that had been applied, with differences primarily

found in studies with moving stimuli (Rice et al. 2012; Van

der Geest et al. 2002).

Methods

Participants

Participants were part of a large sample of children and

adolescent with different psychiatric disorders and a non-

clinical CG. The current study focuses on the subsample

(n = 84) of children, adolescents with ADHD (n = 56) and

participants without any psychiatric diagnosis (n = 28). A

part of the non-clinical CG served as CG in another publi-

cation (Schwenck et al. 2012). Children with ADHD were

recruited from the Department of Child and Adolescent

Psychiatry, University of Wurzburg, Germany, and through

local Child and Adolescent Psychiatrists. Children had been

assessed by mental health professionals. The standard

clinical assessment consisted of a clinical examination and

psychological tests. Furthermore, observational information

of the children’s behavior, objective tests as well as self-

reports, school and parent reports were considered in the

diagnostic process. Diagnoses of ADHD were made

according to DSM-IV TR criteria (314.00 and 314.01)

(American Psychiatric Association 2000), and 10 children

with combined type, 2 children with predominantly

hyperactive-impulsive type (ADHD-C) and 44 children

with predominantly inattentive type (ADHD-I) took part in

the study. Children of the CG had scores below the cutoff

for both clusters at the FBB-ADHD.

Most of the children (n = 47) in the clinical groups

were taking medication targeting ADHD symptomatol-

ogy, and only six participants were drug naıve. Children

of the clinical groups who were medically treated were

randomly asked either to take their medication or to taper

medication at least 24 h (48 h if medicated with extended

release methylphenidate) prior to testing. We decided to

use an independent study plan to avoid learning effects as

a consequence of repeated examinations and in order to

avoid drop outs due to motivational deficits which might

have resulted from an additional measurement point after

the clinical characterization and intelligence test and the

first trial of the emotion recognition task. All children

who were tested under medication were taking methyl-

phenidate in the recommended dose range and with sat-

isfying clinical effects, and one child was additionally

taking atomoxetine. One child of the group that was

tested without medication was usually taking dexamfeta-

mine, all other children, who were medically treated, were

taking methylphenidate. The diagnosis of conduct disor-

der/oppositional defiant disorder or autism served as an

exclusion criterion, because these disorders are known to

be associated with deficits in emotion recognition. The

CG was recruited by an ad in the local newspaper. All

Table 1 Number of participants, gender, medication status (stimu-

lants or atomoxetine) and means (standard deviations) of age and IQ

in regard to group

Group

ADHD-

MED-

ADHD-

MED?

CG

n (girls/boys) 28 (9/19) 28 (9/19) 28 (9/19)

Currently treated with

ADHD-relevant

medication

19 28 0

Medication naive 6 0 28

Age (months) 148.43

(28.60)

147.75

(32.46)

149.93

(30.63)

IQ 105.89

(12.68)

100.46

(11.66)

103.93

(11.72)

FBB-ADHD inattention raw

sum (max. = 27)

18.41

(4.22)

19.49

(3.58)

3.88

(3.62)

FBB-ADHD hyperactivity/

impulsivity raw sum

(max. = 33)

9.92

(8.07)

13.55

(8.28)

3.66

(4.32)

ADHD-MED- attention-deficit/hyperactivity disorder no medication,

ADHD-MED? attention-deficit/hyperactivity disorder with medica-

tion, CG control group, FBB-ADHD observer rating scale for ADHD

Emotion recognition in ADHD 297

123

participants and their parents agreed to participate in the

study, and written informed consent was obtained from

parents. The study was approved by the local Ethics

Committee of the University of Wurzburg.

In total, the sample consisted of 84 children aged

8.2–17.3 years (mean age 12.39 years, SD = 2.52 years).

There were no group differences for age (F(2,81) = .04,

p = .964) or IQ (F(2,81) = 1.46, p = .238), measured with

the Wechsler Intelligence Scale for Children (Petermann

and Petermann 2007), and groups were matched for gender

and subtype. Descriptive statistics in regard to the number

of participants, medication status, age, IQ and ADHD

symptomatology are demonstrated in Table 1.

In regard to the FBB-ADHD, groups differed on the

scale inattention (F(2,80) = 142.37, p \ .001), and post hoc

tests indicated that the two clinical groups showed more

symptoms of inattention than the CG, although they did not

differ from each other. Furthermore, between-group com-

parisons showed differences in regard to the scale hyper-

activity/impulsivity (F(2,80) = 13.42, p \ .001). Again,

children of the two clinical groups showed more symptoms

of hyperactivity/impulsivity compared to the CG, although

they did not differ from each other.

Materials

Clinical assessment

Twenty-eight children with ADHD without medical treat-

ment (ADHD-MED-), 28 children with ADHD and

medical treatment (ADHD-MED?), and 28 typically

developing children (CG) participated in the current study.

All participants of the clinical group had been diagnosed

previously by experienced clinicians according to DSM-IV

TR criteria. The children were additionally assessed by the

following diagnostic measures:

Observer rating scale for ADHD (Dopfner et al. 2008)

The FBB-ADHD (translation: ‘‘observer rating scale for

ADHD’’) consists of 20 items containing symptoms of

inattention, hyperactivity and impulsivity based on the

diagnostic criteria of DSM-IV (American Psychiatric

Association 2000) and ICD-10 (World Health Organization

1993). Observers (in this study parents) have to rate each

item on a 0–3 point scale. Answers with scores of 2 or 3 are

evaluated as clinically relevant. A cutoff of 6 critically

answered items for each cluster of symptoms (inattention, 9

items; hyperactivity and impulsivity, 11 items) is recom-

mended for categorical diagnosis. Parents were instructed

to rate the behavior of their children as if they had not

taken their medication.

Child behavior checklist (Achenbach 1991)

The child behavior checklist (CBCL) is a parent-reported

measure for screening of child psychiatric symptomatology

and contains 113 items that have to be rated on a three-

point Likert scale. For this study, the broad-band categories

internalizing and externalizing problems as well as the total

score were used to exclude participants with psychiatric

symptoms of the non-clinical CG. Children from the CG

were excluded from the study if they had T-scores larger

than 63 in at least one of the broad-band categories or the

total score.

Assessment of emotion recognition: Morphing Task

(MT)

The MT is a self-developed task in which children were

shown 60 film clips each of 9-s length with a neutral facial

expression changing continuously to an emotional expres-

sion. Participants were supposed to press a key as soon as

they had recognized which emotion was presented and to

name the correct emotion. Stimuli were presented with the

computer program presentation (version 0.71), and RT and

the identified emotions were recorded.

Pictures from the Karolinska Directed Emotional Faces

Set (Lundqvist et al. 1998) served as stimulus material.

These pictures were found to be valid and reliable in past

Table 2 Means (SD) of the dependent measures of the Morphing

Task

ADHD-MED- ADHD-MED? CG

Reaction times in ms

Happiness 3,681 (594) 3,537 (781) 3,773 (685)

Sadness 6,822 (1,245) 6,806 (1,360) 6,684 (1,575)

Disgust 5,634 (1,054) 5,407 (1,065) 5,401 (939)

Fear 7,380 (1,794) 7,069 (1,749) 7,194 (1,633)

Anger 7,216 (1,220) 7,112 (1,662) 7,277 (2,170)

Standard deviations of reaction times

Happiness 933 (313) 862 (369) 978 (485)

Sadness 2,100 (657) 2,258 (1,232) 2,274 (1,500)

Disgust 1,487 (522) 1,566 (727) 1,619 (1,074)

Fear 2,353 (846) 2,875 (2,239) 2,750 (1,714)

Anger 2,094 (730) 2,448 (1,704) 2,679 (2,593)

Number of correctly identified emotions (max. = 12)

Happiness 11.86 (.45) 11.89 (.32) 11.78 (.64)

Sadness 8.36 (2.60) 8.89 (2.22) 8.67 (1.75)

Disgust 6.50 (2.93) 6.71 (2.85) 6.79 (2.69)

Fear 7.54 (2.95) 7.71 (3.17) 7.07 (3.11)

Anger 9.61 (1.81) 9.29 (2.05) 9.64 (1.79)

ADHD-MED- ADHD without medication, ADHD-MED? ADHD

with medication, CG control group

298 C. Schwenck et al.

123

research (Goeleven et al. 2008). In the current study, we

selected a neutral and an emotional picture of 30 different

individuals and morphed the neutral expression to the

emotional one using the computer program Winmorph

3.01. The target emotions were anger, happiness, sadness,

fear and disgust, and we produced six different films for

each emotion. All film clips were presented twice, and

within each episode, we used a random order of film clips

that was the same for all participants. RT, standard devi-

ations of reaction times (SDRT) and the number of correctly

identified emotions (NC) within each emotional category

served as dependent measures.

Prior to the presentation of the target stimuli, we pre-

sented ten stimuli without any facial expression but geo-

metric figures (e.g. circle, triangle) changing into animals

(e.g. elephant, duck). Employing this procedure, we were

able to ensure that between-group differences on RT were

specific to emotion recognition in human faces of the target

films.

Results

Means and standard deviations of the dependent measures

are presented in Table 2, and correlation analyses of the

dependent variables are provided in Table 3. Descriptively,

clear ceiling effects can be observed for the number of

correctly identified happy facial expressions, which were

identified with the smallest RT and SDRT by all groups.

Whereas children reacted most slowly to angry and fearful

facial expressions, they conducted the highest number of

errors at the identification of disgusted expressions. Small

to medium intercorrelations were found for reaction time

measures, whereas intercorrelations between the numbers

of correctly identified emotions revealed not existent to

small.

Initial examination of the dependent measures indicated

that all RT measures were normally distributed (KS tests: all

p [ .05). Therefore, groups were compared by planned

contrasts according to the hypothesis (ADHD-MED-

[ ADHD-MED? = CG) in regard to mean RT within each

emotional category. Prior analysis of variance (ANOVA) of

the control items did not result in any RT differences

between groups (F(2,81) = 1.57, p = .214). Therefore, all

potential between-group differences of the test items could

be attributed to the emotional content of the stimuli.

Between-group comparisons indicated comparable emotion

recognition performances for the groups ADHD-MED?,

ADHD-MED- and the CG for the emotion happiness

(t(81) = -1.19, p = .237), sadness (t(81) = .16, p =

.871), disgust (t(81) = -.47, p = .641), fear (t(81) = -.55,

p = .586) and anger (t(81) = -.34, p = .738).

Examination of the distribution of SDRT and the NC

indicated that empirical data significantly differed from

normal distribution for almost all emotional categories.

Therefore, Kruskal–Wallis tests were conducted to com-

pare groups. According to the results of these between-

group comparisons, groups did not differ in SDRT for the

emotions happiness (v2(2) = 1.26, p = .533), sadness

(v2(2) = .62, p = .732), disgust (v2(2) = .01, p = .997),

fear (v2(2) = .36, p = .837) and anger (v2(2) = .04,

p = .978). The same applies for the NC, where there were

no between-group differences for happiness (v2(2) = .32,

p = .854), sadness (v2(2) = .94, p = .627), disgust

(v2(2) = .15, p = .926), fear (v2(2) = .60, p = .742) and

anger (v2(2) = .57, p = .752).

Finally, the relative number of confusions of happiness,

sadness, fear, disgust and anger with the respective other

emotions was compared between groups. Percentages of

actual and identified emotions by group are presented in

Table 4.

Results indicate that groups equally confused happy,

disgusted, fearful and angry facial expressions with all

respective other emotions (all p [ .05). However, there

was a small between-group difference in actual sad facial

expressions identified as angry facial expressions

(F(2,79) = 3.14, p = .049, g2 = .07). Descriptively, chil-

dren of the group ADHD-MED? mistook sad faces less

commonly as angry faces than the group ADHD-MED-.

However, an alpha-corrected post hoc test was only mar-

ginally significant (p = .054).

Discussion

The aim of the current study was to assess emotion rec-

ognition performance in children and adolescents with

Table 3 Correlation analyses of dependent measures reaction times

and number of correctly identified emotions

Happiness Sadness Disgust Fear Anger

Reaction times (ms)

Happiness – .51** .44** .50** .45**

Sadness – .67** .58** .70**

Disgust – .66** .76**

Fear – .60**

Anger –

Number correct

Happiness – .18 .19 .24* .20

Sadness – -.09 .23* -.07

Disgust – .32** .24*

Fear – .12

Anger –

* p \ .05, ** p \ .01

Emotion recognition in ADHD 299

123

ADHD without comorbid conduct disorder or oppositional

defiant disorder and to evaluate the influence of medication

on emotion recognition performance, as emotion recogni-

tion had been shown to be associated with social func-

tioning (Blair et al. 2004; Dadds et al. 2012). A task with

moving instead of static facial expressions was adopted in

order to assess the speed and its variation of the identifi-

cation additionally to the number of correctly identified

emotions and to adopt a measure of high ecological

validity.

No differences were found between children with

ADHD without medication and a non-clinical CG matched

for age, gender and IQ, neither in regard to RT nor in

regard to the variation of the RT or the number of correctly

identified emotions. This result applies for all basic emo-

tions assessed in this study: happiness, sadness, disgust,

fear and anger. These findings are in contrast to most

previous studies (Fonseca et al. 2008; Kats-Gold et al.

2007; Singh et al. 1998; Taurines et al. 2012). A possible

reason for distinct findings may be the use of short film

clips with moving faces rather than static pictures that the

majority of studies had used before. Because emotions

have to be recognized within short time spans in every day

life, we assumed a higher ecological validity in material

with moving faces. Emotion recognition in moving faces

and under time pressure may thus be more difficult than

under conditions with static pictures and without time

constrictions, and past studies on children with autism and

controls have shown differences only when film clips had

been applied, whereas with static pictures, no differences

have been found (Rice et al. 2012; Van der Geest et al.

2002). However, this assumption may have been incorrect,

and the use of distinct material may account for the non-

finding. On the other hand, we did not detect any floor

effects in the number of correctly identified emotions for

neither emotion and, on the contrary, showed ceiling

effects in the identification of happiness. Therefore, this

aspect seems rather unlikely to be responsible for the dif-

fering results. Future studies may directly compare emotion

recognition in materials with static facial expressions and

film clips to bring light into this methodological issue.

Furthermore, the application of eye tracking technology

would shed light on possible differences in the exact

looking and fixation behavior of children with ADHD

compared to other groups. Though this technology is pri-

marily used under laboratory conditions, as future pros-

pects, it would be a tremendous progress to develop the

possibility of eye tracking in the field and therefore every

day life of the children.

Another difference between our study and others is the

exclusion of children with a comorbid conduct disorder or

oppositional defiant disorder. According to recent research

(Booster et al. 2012), social problems of children with

ADHD are particularly high if externalizing disorders are a

comorbid condition. Only one study on emotion recogni-

tion accounted for conduct problems (Cadesky et al. 2000),

and the results were inconclusive, because the groups with

either ADHD or conduct disorder underperformed in the

emotion recognition task, but children with combined

disorders did not differ from healthy controls. In our study,

we treated conduct problems as an exclusion criterion. Our

non-finding in light of the previous investigations may

indicate that conduct problems more than ADHD-associ-

ated functions could be responsible for emotion recognition

deficiencies. However, further studies with specific

research designs are needed to prove this assumption and

replicate our results.

Furthermore, we did not find any influence of medica-

tion on emotion recognition performance. We had decided

to apply an independent study plan with comparable, but

different groups tested with or without medication. We did

that in order to avoid learning effects by repeated mea-

surements of the same tests. However, this procedure might

have led to a larger between-group error variance, which

we could have avoided by dependent groups and by a

counterbalanced measurement procedure. On the other

hand, a comparable study used a dependent and counter-

balanced study plan to assess the influence of medication

Table 4 Actual and identified emotions (in %) by group

Actual emotion Identified as

Happiness Sadness Disgust Fear Anger

ADHD-MED-

Happiness 98.81 .60 .30 .30 .00

Sadness 1.19 69.64 9.82 14.58 4.76

Disgust 2.08 14.58 54.17 2.38 26.79

Fear 3.27 5.36 26.49 62.80 2.08

Anger .60 12.20 3.57 3.57 80.06

ADHD-MED?

Happiness 99.11 .60 .00 .30 .00

Sadness 2.38 74.11 9.23 13.10 1.19

Disgust 5.06 10.12 55.95 1.79 27.08

Fear 4.17 4.76 24.70 64.29 2.08

Anger .30 12.80 5.95 3.57 77.38

CG

Happiness 98.15 .93 .00 .00 .93

Sadness 1.85 72.22 10.19 11.73 4.01

Disgust 3.27 13.10 56.55 2.08 25.00

Fear 4.17 6.55 27.98 58.93 2.38

Anger .30 10.42 6.55 2.38 80.36

Bold values indicate percentage of correctly identified faces by

emotion and group

ADHD-MED- ADHD without medication, ADHD-MED? ADHD

with medication, CG control group

300 C. Schwenck et al.

123

on emotion recognition and did not find any differences

either (Hall et al. 1999). So, it might be concluded that

medical treatment affects the core deficits of ADHD but

does not influence peripheral domains that are rather intact.

The positive effects of methylphenidate on social problems

found in previous studies seem to positively affect basic

processes other than emotion recognition. Of course, this

interpretation is speculative, and more research is neces-

sary to confirm this assumption and to replicate the findings

of our study.

There are general limitations of our study that we would

like to mention: first, we treated comorbid conduct prob-

lems as an exclusion criterion but did not include a group

of children with pure conduct disorder and without any

ADHD symptomatology or a group with combined disor-

ders. Our interpretation that conduct problems might be

more responsible for emotion recognition deficiencies than

ADHD symptomatology is rather indirect, and a study that

compares all three groups with a CG would be helpful to

confirm our interpretation. Furthermore, as already men-

tioned, a dependent study plan might have reduced error

variance in the assessment of the influence of medication.

Finally, the administration of participants to the two clin-

ical groups was not perfectly random. Nine participants

were not pharmacologically treated when the study was

conducted. These children were administered to the group

ADHD-MED-. All other children (n = 47) were ran-

domly assigned. This procedure might have influenced

results. Furthermore, the majority of our participants were

diagnosed with ADHD predominantly inattentive subtype.

This has to be kept in mind because results can be gen-

eralized only in regard to this proportion that is not

representative.

Keeping these limitations in mind, the present study

provides continuative insights into research on emotion

recognition ability in children and adolescents with

ADHD and a CG. To our knowledge, this study for the

first time assessed emotion recognition in ADHD and the

influence of medical treatment adopting a research para-

digm of high ecological validity and controlling for

comorbid conduct problems. In our study, we did not find

any between-group differences neither for the factor

group nor for the factor medication. Groups did not differ

in regard to the number of correctly identified emotions,

RT or the variability of RT. Unlike other studies on this

field of research, we had defined conduct problems as an

exclusion criterion, which might have led to these dif-

fering results. As social problems are known to strongly

predict negative outcomes in later adolescence and

adulthood (Greene et al. 1997; Mrug et al. 2012), we

believe that it is important to study possible underlying

deficient processes such as emotion recognition but also

others such as emotion regulation in future research and

to assess the differential influence ADHD symptomatol-

ogy and comorbid disorders have on these processes, as

deficits in emotion recognition might not be associated

with ADHD symptomatology per se.

Acknowledgments We thank all participants and their families for

their participation in the study. Our sincere thanks go to all colleagues

from collaborating institutes for their cooperation and support.

References

Abikoff H, Hechtman L, Klein RG et al (2004) Social functioning in

children with ADHD treated with long-term methylphenidate

and multimodal psychosocial treatment. J Am Acad Child

Adolesc Psychiatry 43:820–829. doi:10.1097/01.chi.0000128

797.91601.1a

Achenbach TM (1991) Manual for the child behavior checklist/4-18

and 1991 profile. University of Vermont, Burlington

American Psychiatric Association (2000) Diagnostic and statistical

manual of mental disorders (DSM-IV), 4(null) ed. American

Psychiatric Press, Washington

Biederman J, Faraone SV (2005) Attention-deficit hyperactivity

disorder. Lancet 366:237–248

Blair RJR, Mitchell DGV, Peschardt KS et al (2004) Reduced

sensitivity to others’ fearful expressions in psychopathic indi-

viduals. Pers Individ Differ 37:1111–1122. doi:10.1016/

j.paid.2003.10.008

Booster GD, DuPaul GJ, Eiraldi R, Power TJ (2012) Functional

impairments in children with ADHD: unique effects of age and

comorbid status. J Atten Disord 16:179–189. doi:10.1177/10870

54710383239

Cadesky EB, Mota VL, Schachar RJ (2000) Beyond words: how do

children with ADHD and/or conduct problems process nonverbal

information about affect? J Am Acad Child Adolesc Psychiatry

39:1160–1167

Corbett B, Glidden H (2000) Processing affective stimuli in children

with attention-deficit hyperactivity disorder. Child Neuropsychol

(Neuropsychology, Development and Cognition: Section C)

6:144–155. doi:10.1076/chin.6.2.144.7056

Dadds MR, Cauchi AJ, Wimalaweera S et al (2012) Outcomes,

moderators, and mediators of empathic-emotion recognition

training for complex conduct problems in childhood. Psychiatry

Res 199:201–207. doi:10.1016/j.psychres.2012.04.033

Dopfner M, Gortz-Dorten A, Lehmkuhl G et al (2008) Diagnostik-

System fur psychische Storungen nach ICD-10 und DSM-IV fur

Kinder und Jugendliche-II (DISYPS-II). Huber, Bern

DuPaul GJ, Volpe RJ, Jitendra AK et al (2004) Elementary school

students with AD/HD: predictors of academic achievement.

J Sch Psychol 42:285–301. doi:10.1016/j.jsp.2004.05.001

Dyck MJ, Ferguson K, Shochet IM (2001) Do autism spectrum

disorders differ from each other and from non-spectrum

disorders on emotion recognition tests? Eur Child Adolesc

Psychiatry 10:105–116

Fonseca D, Seguier V, Santos A et al (2008) Emotion understanding

in children with ADHD. Child Psychiatry Hum Dev 40:111–121.

doi:10.1007/s10578-008-0114-9

Goeleven E, De Raedt R, Leyman L, Verschuere B (2008) The

Karolinska directed emotional faces: a validation study. Cogn

Emot 22:1094–1118. doi:10.1080/02699930701626582

Greene RW, Biederman J, Faraone SV et al (1997) Adolescent

outcome of boys with attention-deficit/hyperactivity disorder and

social disability: results from a 4-year longitudinal follow-up

study. J Consult Clin Psychol 65:758

Emotion recognition in ADHD 301

123

Hall CW, Peterson AD, Webster RE et al (1999) Perception of

nonverbal social cues by regular education, ADHD, and ADHD/

LD students. Psychol Sch 36:505–514

Hinshaw SP (1987) On the distinction between attentional deficits/

hyperactivity and conduct problems/aggression in child psycho-

pathology. Psychol Bull 101:443

Hoza B, Mrug S, Gerdes AC et al (2005) What aspects of peer

relationships are impaired in children with attention-deficit/

hyperactivity disorder? J Consult Clin Psychol 73:411–423. doi:

10.1037/0022-006X.73.3.411

Kats-Gold I, Besser A, Priel B (2007) The role of simple emotion

recognition skills among school aged boys at risk of ADHD.

J Abnorm Child Psychol 35:363–378. doi:10.1007/s10802-

006-9096-x

Kendall T, Taylor E, Perez A et al (2008) Diagnosis and management of

attention-deficit/hyperactivity disorder in children, young people,

and adults: summary of NICE guidance. BMJ 337:a1239. doi:

10.1136/bmj.a1239

King S, Waschbusch DA, Frankland BW et al (2005) Taxonomic

examination of ADHD and conduct problem comorbidity in

elementary school children using cluster analyses. J Psychopathol

Behav Assess 27:77–88. doi:10.1007/s-10862-005-5382-0

Lee SS, Falk AE, Aguirre VP (2012) Association of comorbid anxiety

with social functioning in school-age children with and without

attention-deficit/hyperactivity disorder (ADHD). Psychiatry Res

197:90–96. doi:10.1016/j.psychres.2012.01.018

Lundqvist D, Flykt A, Ohman A (1998) The Karolinska directed

emotional faces. Karolinska Institute, Department of Clinical

Neuroscience, Stockholm

Mrug S, Molina BSG, Hoza B et al (2012) Peer rejection and

friendships in children with attention-deficit/hyperactivity disor-

der: contributions to long-term outcomes. J Abnorm Child

Psychol 40:1013–1026. doi:10.1007/s10802-012-9610-2

Pelham WE Jr, Waschbusch DA, Hoza B et al (2001) Effects of

methylphenidate and expectancy on performance, self-evalua-

tions, persistence, and attributions on a social task in boys with

ADHD. Exp Clin Psychopharmacol 9:425

Petermann F, Petermann U (2007) Hamburg-Wechsler-Intelligenztest

fur Kinder (HAWIK-IV). Hogrefe, Gottingen

Rice K, Moriuchi JM, Jones W, Klin A (2012) Parsing heterogeneity

in autism spectrum disorders: visual scanning of dynamic social

scenes in school-aged children. J Am Acad Child Adolesc

Psychiatry 51:238–248. doi:10.1016/j.jaac.2011.12.017

Schachar R, Tannock R (1995) Test of four hypotheses for the

comorbidity of attention-deficit hyperactivity disorder and

conduct disorder. J Am Acad Child Adolesc Psychiatry 34:

639–648

Schwenck C, Mergenthaler J, Keller K et al (2012) Empathy in

children with autism and conduct disorder: group-specific

profiles and developmental aspects. J Child Psychol Psychiatry

53:651–659

Shapiro EG, Hughes SJ, August GJ, Bloomquist ML (1993)

Processing of emotional information in children with attention-

deficit hyperactivity disorder. Dev Neuropsychol 9:207–224.

doi:10.1080/87565649309540553

Singh SD, Ellis CR, Winton ASW et al (1998) Recognition of facial

expressions of emotion by children with attention-deficit hyper-

activity disorder. Behav Modif 22:128–142. doi:10.1177/0145

4455980222002

Sinzig J, Morsch D, Lehmkuhl G (2008) Do hyperactivity, impulsiv-

ity and inattention have an impact on the ability of facial affect

recognition in children with autism and ADHD? Eur Child

Adolesc Psychiatry 17:63–72. doi:10.1007/s00787-007-0637-9

Storebø OJ, Skoog M, Damm D et al (2011) Social skills training for

attention deficit hyperactivity disorder (ADHD) in children aged

5 to 18 years. Cochrane Database Syst Rev 7(12):8223

Taurines R, Schmitt J, Renner T et al (2010) Developmental

comorbidity in attention-deficit/hyperactivity disorder. Atten

Def Hyperact Disord 2:267–289. doi:10.1007/s12402-010-

0040-0

Taurines R, Schwenck C, Westerwald E et al (2012) ADHD and

autism: differential diagnosis or overlapping traits? A selective

review. Atten Defic Hyperact Disord 4:115–139. doi:10.1007/

s12402-012-0086-2

Van der Geest JN, Kemner C, Verbaten MN, Van Engeland H (2002)

Gaze behavior of children with pervasive developmental disor-

der toward human faces: a fixation time study. J Child Psychol

Psychiatry 43:669–678

World Health Organization (1993) The ICD-10 classification of

mental and behavioural disorders: diagnostic criteria for

research. World Health Organization, Geneva

Yuill N, Lyon J (2007) Selective difficulty in recognising facial

expressions of emotion in boys with ADHD. Eur Child Adolesc

Psychiatry 16:398–404. doi:10.1007/s00787-007-0612-5

302 C. Schwenck et al.

123