when distraction is not distracting: a behavioral and erp study on distraction in adhd

www.elsevier.com/locate/clinph

Clinical Neurophysiology 118 (2007) 1855–1865

When distraction is not distracting: A behavioral and ERP studyon distraction in ADHD

Rosa van Mourik a,*, Jaap Oosterlaan a, Dirk J. Heslenfeld a,Claudia E. Konig b, Joseph A. Sergeant a

a Department of Clinical Neuropsychology, Vrije Universiteit Amsterdam, The Netherlandsb Regional Centre for Child and Adolescent Psychiatry, Gooi en Vechtstreek, The Netherlands

Accepted 7 May 2007

Abstract

Objective: Although an increased distractibility is one of the behavioral criteria of Attention Deficit Hyperactivity Disorder (ADHD),there is little empirical evidence that children with ADHD are in fact more distractible than their normal peers.Methods: We recorded event-related potentials (ERPs) to distracting novel sounds (novels) and standard sounds, (standards) while chil-dren performed a visual two-choice reaction time task. Twenty-five children with ADHD were compared with eighteen normal controls(aged 8–12 years).Results: Children with ADHD showed a larger early P3a (150–250 ms), both in response to the standard and in response to the novel.The late phase of the P3a had a larger amplitude in the ADHD group in the 250–300 ms window compared to the control group, whichwas only present in response to the novel. Interestingly, the novel reduced the errors of omission in the ADHD group to a greater extentthan in the normal control group.Conclusions: Although children with ADHD show an increased orienting response to novels, this distracting information can enhancetheir performance temporarily, possibly by increasing their arousal to an optimal level, as indicated by the reduced omission rate.Significance: These data indicate that distraction is not always distracting in children with ADHD and that distraction can also havebeneficial effects.� 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Keywords: ADHD; Children; ERP; P3a; Distraction; Attention

1. Introduction

One of the behavioral manifestations of children withAttention Deficit Hyperactivity Disorder (ADHD) is theirabnormal apparent distractibility (DSM-IV; AmericanPsychiatric Association, 1994). Especially in the classroom,children with ADHD often pay more attention to eventshappening in and outside the classroom and less attentionto their schoolwork than their normal peers. Surprisingly, anumber of attempts to prove that children with ADHD are

1388-2457/$32.00 � 2007 International Federation of Clinical Neurophysiolo

doi:10.1016/j.clinph.2007.05.007

* Corresponding author. Tel.: +31 20 5982715.E-mail address: [email protected] (R. van Mourik).

abnormally distractible have been unsuccessful (see forreview Douglas and Peters, 1979). Selective or focusedattention tasks, such as visual search paradigms withdistracters, often do not differentiate children with ADHDfrom normal controls (Van der Meere and Sergeant, 1988;Mason et al., 2003; Huang-Pollock et al., 2005). However,it has been frequently reported that children with ADHDhave greater difficulty in inhibiting conflicting stimuli thatare incorporated in a task, i.e. they show poorer perfor-mance on Stroop- and Flanker-tasks (Scheres et al.,2004), although the difference in interference controlbetween normal control groups and ADHD groups onthe Stroop task is small (see for meta-analysis Van Mouriket al., 2005). Apparent distractibility in the classroom may

gy. Published by Elsevier Ireland Ltd. All rights reserved.

mailto:[email protected]

1856 R. van Mourik et al. / Clinical Neurophysiology 118 (2007) 1855–1865

have multiple causes, for example, children with ADHDcould have less intrinsic motivation (Carlson et al., 2002),suffer from a failure to inhibit stimuli extraneous to thetask (Barkley, 1997), or distractibility could be a functionalattempt to modulate underarousal by seeking stimulation(Zentall and Zentall, 1983). Alternatively, children withADHD could have an abnormally low threshold for thebreakthrough of unattended (and usually irrelevant) infor-mation, which is found even in children with subclinicalattentional problems (Kilpelainen et al., 1999).

Most distraction or selective attention tasks are not eco-logically valid measures of distraction in daily life situa-tions, because in daily life, the distraction that has to beinhibited is outside the task and not conflicting with taskdemands, for example, a child is doing schoolwork whileother children are talking. Escera et al. (1998) developeda paradigm to measure this kind of distraction withevent-related potentials (ERPs). This paradigm has beenadapted by Gumenyuk et al. (2001) to measure distractionin children. In this paradigm, the child performs a visualtask while listening to standard tones and occasionally anovel environmental sound, such as a mooing cow, anengine, or a bell. These sounds should be ignored by thechild. Novel and unexpected stimuli are hard to ignoreand cause distraction. Behaviorally, this distraction isobserved as deterioration in performance in the task asshown by increased reaction times and/or decreased perfor-mance accuracy (Escera et al., 1998; Gumenyuk et al.,2004).

At the electrophysiological level, unattended and task-irrelevant novel stimuli elicit a P3a component in adults(Squires et al., 1975; Escera et al., 1998) and children(Cycowicz and Friedman, 1997; Ceponien _e et al., 2004;Gumenyuk et al., 2004). The P3a component is thoughtto reflect an evaluative, conscious aspect of the orientingresponse and an attentional switch to the novel informa-tion (Friedman et al., 2001). The dorsolateral prefrontalcortex has an important role in response to auditorydistraction (Campbell, 2005). Both the noradrenergic(Missonnier et al., 1999) and the dopaminergic (Kahkonenet al., 2002) systems have been found to modulate the P3a.Previous studies indicated that the P3a has two subcompo-nents in adults (see for review Escera et al., 2000) and inchildren (Ceponien _e et al., 2004; Gumenyuk et al., 2004):an early P3a (eP3a) component with its peak latencyaround 200 ms and a late P3a (lP3a), which peaks ataround 300 ms. Ceponien _e et al. (2004) suggested that theeP3a might be a receiver of the sensory information andgoverns the direction of the attentional focus as reflectedby the lP3a. The lP3a is thus more closely related to theactual orienting of attention (Escera et al., 2000). Sincethe latency of the eP3a component is similar to that ofthe auditory P2 peak, it could be argued that the eP3a isactually an enhanced P2 in response to the physicalfeatures of the novel sounds. Although the scalp topogra-phy of these two positivities differs (fronto-central for theeP3a and centro-parietal for the P2, see Ceponien _e et al.,

2002) further studies are needed to clarify whether thegenerators of the auditory P2 differ from those of the eP3a.

An enhanced lP3a may indicate that too much attentionis attributed to the novel stimuli, which may result inincreased distractibility at the behavioral level. Anenhanced lP3a has been found in children with majordepression (Lepisto et al., 2004) and in adults with closedhead injury (Kaipo et al., 1999). Inconsistent findings havebeen reported with regard to children with ADHD. Twostudies found no differences in P3a response to novelsounds (Holcomb et al., 1986; Kenmer et al., 1996). Onestudy found a reduced P3a in the ADHD group in responseto novel visual stimuli (Keage et al., 2006) and anotherstudy found an enhanced lP3a and a reduced eP3a in theADHD group in response to novel sounds (Gumenyuket al., 2005). A possible explanation for these inconsistentresults is that one study (Gumenyuk et al., 2005) used avariety of novels, whereas in other studies the same novelwas repeated, reducing the novelty effect with trials.

The P3a to distracting sounds is sometimes followedby a frontally distributed negativity with a latency of400–700 ms. This negativity was interpreted by Schrogerand Wolff (1998) as reflecting the reorienting of attentionback to the main task after distraction, and it was labeledas the reorienting negativity (RON). A recent fMRI studyindicated that a prefrontal-temporal network including theleft superior and right middle temporal cortex, right frontaleye fields, the left inferior frontal gyrus and the right precu-neus underlies reorienting (Mayer et al., 2006). In children,a similar frontal late negativity (hereafter LN) was found inresponse to distracting novel sounds, which is sometimesreferred to as LN (Gumenyuk et al., 2004), NegativeComponent (Ceponien _e et al., 2004; Maata et al., 2005),or RON (Wetzel et al., 2004). The LN in children withADHD has been found to be reduced in comparison tocontrol children (Gumenyuk et al., 2005). Interestingly,Konrad et al. (2006) demonstrated that children withADHD tend to recruit more fronto-striatal-insular activa-tion than normal controls during reorienting in the absenceof behavioral differences, which is explained in the contextof neural compensation. Using a different paradigm, it hasalso been found that children with ADHD have behavioraldifficulties in reorienting of attention (Pearson et al., 1991)and in disengaging attention when voluntary control isrequired (Wood et al., 1999).

Although behavioral distractibility is a major clinicalfeature of ADHD, little research has been conducted toelucidate the neural mechanisms that underlie this behav-ioral distractibility in the disorder. The only ERP studythat investigated auditory distraction during a visual taskin children with ADHD is by Gumenyuk et al. (2005).Their results indicated that children with ADHD showedan enhanced distractibility, both at the behavioral as wellas at the electrophysiological levels. This very importantfinding needs to be replicated and extended, since theseauthors had only small sample size (10 children in eachgroup) and a small age range (only 8- to 10-year-old

R. van Mourik et al. / Clinical Neurophysiology 118 (2007) 1855–1865 1857

children were included). Therefore, the present study aimedat examining both distraction and reorienting in a largergroup of children with a broader age range (8- to 12-year-old) with and without ADHD, by recording ERPsfrom task-irrelevant standard tones and novel sounds,while children performed a visual demanding task. Thevisual task was different from the task used in the studyof Gumenyuk et al. (2005) in order to make it moresuitable for older children. Behaviorally, the hypothesiswas tested that novel sounds compared to standard toneswould result in a larger deterioration in performance (i.e.increased reaction times and/or decreased performanceaccuracy) in the ADHD group than in the normal controlgroup. At the psychophysiological level, ERPs were mea-sured to elucidate the neural mechanisms of the presumeddistractibility in the ADHD group. Differences in the earlyP3a after the novel compared with the standard (as foundin the study of Gumenyuk et al. (2005)) could be inter-preted as abnormalities in directing attention, a relativelylarger late P3a in the ADHD group would be evidence thatchildren with ADHD attribute too much attention to irrel-evant and distracting information, while a relativelyreduced LN would suggest that children with ADHD areless capable of reorienting attention back to the main taskafter temporary distraction.

2. Method

2.1. Participants

Twenty-five children aged between 8 and 12 years withADHD were compared with eighteen normal control chil-dren. Subject characteristics are summarized in Table 1.

The ADHD group was recruited via an advertisementon a website and via a university affiliated outpatientdepartment for ADHD. They all had a formal clinicaldiagnosis of ADHD by their health care professional.The control children were recruited from primary schools.None of the children had any neurological, sensory ormotor impairment or any other developmental psychiatricdisorder. Written informed consent was obtained from thechildren’s parents prior to the study, and children also hadto agree by writing down their name on a permission form.

Table 1Subject characteristics for the ADHD and normal control groups

Whole groups

NC (n = 18)

Boys/girls 17/1Age Mean (SD) 10.5 (1.1)

Median (Range) 10.4 (3.4)

IQ Mean (SD) 117 (16)Median (Range) 120 (59)

Asterisks indicate significant differences between Normal Control and the ADHgroups matched on IQ (fifth and last column).** p < .01.

The Ethical Committee of the Vrije Universiteit MedicalCentre approved the study.

Parents and teachers completed the Dutch version ofthe Disruptive Behavior Disorder rating scale (DBD; Pel-ham et al., 1992; Oosterlaan et al., 2000), which allowedthe assessment of symptoms of ADHD and comorbidOppositional Defiant Disorder of Conduct Disorder. Par-ent and teacher ratings for the ADHD group had to fallwithin the clinical range (95th–100th percentile) for theInattention and/or the Hyperactivity/Impulsivity subscale.Control children were included if they received scoresbelow the 90th percentile on all subscales. In order toconfirm the DSM-IV diagnosis of ADHD, the DiagnosticInterview Schedule for Children Version IV (DISC-IV,Shaffer et al., 2000) was administered to the parents ofthe children with ADHD. Only those children with aDSM-IV diagnosis of ADHD participated in the study.Within the clinical group twenty-two children met theDISC-IV criteria for the ADHD combined subtype andthree for the ADHD inattentive subtype. Fourteen chil-dren with ADHD were also diagnosed with ODD, andtwo other children with ADHD also received a diagnosisof ODD and CD.

Hearing was screened at 20 dB. All children had normalhearing. IQ was estimated with two performance and twoverbal subtests of the Dutch version of the Wechsler Intel-ligence Scale for Children, third edition (Weschler, 1991;Kort et al., 2002): Picture Arrangement, Block Design,Arithmetic and Vocabulary. All children had an estimatedIQ greater than 70. The mean estimated IQ in the controlgroup (M = 117, SD = 16.20) was higher than in theADHD group (M = 97, SD = 10.28) [t(41) = 5.02,p < .001]. The children with ADHD taking methylpheni-date stopped their medication at least 36 h before testingallowing a complete washout (Pelham et al., 1999). Thechildren were rewarded for their participation with a giftvoucher of €7.50.

2.2. Distraction paradigm

The present study employed a modification of the dis-traction paradigm of Gumenyuk et al. (2001) that is sche-matically depicted in Fig. 1.

IQ-matched subgroups

ADHD (n = 25) NC (n = 14) ADHD (n = 14)

22/3 13/1 11/310.2 (1.2) 10.6 (1.1) 10.0 (1.2)9.9 (4.6) 10.4 (3.4) 9.8 (4.4)

97 (10)** 112 (13) 104 (7)97 (42) 114 (43) 101 (26)

D group, for the whole groups (third and fourth column) and the selected

Fig. 1. Schematic diagram of the distraction paradigm. The sounds consisted of standards and novel sounds and children performed a forced choice visual(left/right) task.


The experiment consisted of a visual two-choice task inwhich an irrelevant sound preceded the visual stimulus.The visual stimulus was a colored picture of a runner,which was either turned to the left or to the right. Childrenwere asked to indicate the direction of the runner with abutton press. The runner was displayed in the middle ofa white screen. The irrelevant sound was presented priorto the visual stimulus through a speaker. The sound waseither a 600 Hz tone (p = .80) or a novel sound (p = .20).All sounds had an intensity of approximately 60 dB. Trialswere completely randomized, with the exception that atleast three standard tones had to occur between any twosuccessive novel sounds. The novel sounds were 99 differentenvironmental sounds, such as a dog barking or a bell ring-ing. Each novel sound was presented once.

Every trial began with the presentation of a small fixa-tion cross in the centre of the screen for a random timeinterval ranging from 0 to 200 ms. The sound was then pre-sented for 200 ms followed by a 100 ms delay. The fixationcross was on during the sound and the delay. Thereafter,the runner was presented for 300 ms. Immediately afterpresentation of the runner, the fixation cross reappearedfor 1200 ms. Children could respond to the runner bypressing a response button with their left or right thumb.The interstimulus interval was on average 1900 ms (range:1800–2000 ms).

The task consisted of three blocks each of 166 trials.After each block (5 min and 32 seconds) a short breakwas given. Performance feedback (mean reaction time, per-centage correct and percentage incorrect) was given aftereach block. After the first block, children were told thatthey did well, but were not in the top three, following thesecond block they were told that they were third and afterthe last block that they had won the first prize.

2.3. Electrophysiological recordings

The electroencephalogram (EEG; 0.05–200 Hz, sam-pling rate 1000 Hz) was recorded at 60 scalp sites usingelectrode caps with tin electrodes referenced to one earlobe. The ground electrode was placed on the cheek. Hor-izontal and vertical eye movements (EOG) were recorded

from the outer canthi of each eye and below and abovethe left eye. Impedances were maintained below 10 kX.Pre-processing of the EEG data was performed with scan4.3 software (Compumedics). After additional filtering(0.25–30 Hz), vertical ocular artifacts were corrected usinga subtraction algorithm (Semlitsch et al., 1986). The EEGwas re-referenced to the mean of both ear lobes. Epochswere extracted from the continuous data file over a1000 ms period starting 100 ms before each sound onset.Epochs containing EEG or horizontal EOG artifacts thatexceed ±100 lV at any electrode were excluded. ERPs wereobtained separately for the tones occurring before a novel(standards) and the novel sounds (novels). Averaged ERPsfor standards and novels consisted of 85 epochs on average(at least 43 epochs because one child missed a block) percondition per child. In contrast to earlier studies (e.g.Gumenyuk et al., 2004; Schroger and Wolff, 1998), it wasdecided to analyze both the standards and the novelsinstead of the difference wave. The reason for this was thatwe were interested if there were any differences at baseline(standards) between the groups. Because the overlap of thevisual stimuli after the sound was the same after the noveland the standard, this could not cause differences betweenthe novel and the standard. After visual inspection of thegrand average ERPs, it was decided to include the elec-trodes Fz and Cz, and to analyze the mean amplitude ofthe standards and the novels in seven separate windowsof 50 ms starting 150 ms until 500 ms after sound onset.These windows gave better insight into the temporaldynamics of the differences than only one window per com-ponent and covered the components of interest: the eP3a,the lP3a and the late negativity.

2.4. Procedure

Following the attachment of the electrode cap and EOGelectrodes, the children sat comfortably in a chair in anacoustically and electrically shielded room, which wasdimly lit. Stimuli were displayed on a 17-inch monitor at2.4 meter distance from the child’s eyes. Children weremonitored by video during the entire experiment and couldcommunicate with the experimenter in the adjacent room


via an intercom. Before the experimental task, the childrenparticipated in one or two short practice blocks, includingtwenty trials with feedback on each response and withoutthe novel sounds, until the child fully understood the taskrequirements. After the practice session, the child wasinstructed to respond as accurately and as fast as possibleto the runner and to ignore the sounds. The experimenterleft the room and initiated the task.

2.5. Data analysis

Performance measures included mean response times(MRT), percentage of commission errors and percentageof omission errors. Strategy effects were tested by correlatingthe percentage of commission errors with MRT. In order totest the hypothesis that the performance of children withADHD was more easily disrupted by the novels comparedto the standards, a repeated measures analysis of variance(ANOVA) was performed with group (ADHD – normalcontrols) as between subjects factor and condition (standard– novel) as within subjects factor for MRT, percentage ofcommission errors and percentage of omission errors sepa-rately. IQ was not entered in the analysis as a covariatebecause a lower IQ is highly associated with ADHD (Kuntsiet al., 2004). ANCOVA is only appropriate when individualshave been randomly assigned to groups but when naturallyoccurring groups are compared, this method can yieldspurious results (Miller and Chapman, 2001). The influenceof IQ on group effects was investigated by computingcorrelation coefficients between performance measures(differences between the novel and the standard with regardto MRT, errors of commission and errors of omission) andIQ for groups separately. Furthermore, IQ-matchedsubgroups were selected from the whole sample and allanalyses were also performed with these subgroups.

In order to assess possible differences in electrophysiolog-ical responding to the novels compared to the standards, theaverage ERP amplitudes from the novels and standards at

Table 2Performance data for the ADHD and the Normal Control Groups

Performance Stimuli Whole groups (Mean (SD))

NC (n = 18) AD

RT (ms) Novel 610 (65) 648Standard 577 (68) 616Difference 33 (26) 31

Errors (%) Novel 11 (7) 16Standard 15 (9) 19Difference �4 (6) �3

Misses (%) Novel 0.6 (1) 4Standard 1 (2) 9Difference �0.6 (0.8) �6

Note. Difference, novel minus standard; NC, Normal Controls asterisks indicatethe whole groups (third and fourth column) and the selected groups matched

* p < .05.** p < .01.

Fz and Cz were analyzed with a repeated measures ANOVAincluding the following factors: Group (ADHD – normalcontrols) as a between subjects factor and Sound (standard– novel) as within subjects factor for each of the sevenwindows separately. The influence of IQ on group effectswas investigated by computing correlation coefficientsbetween the mean ERP amplitudes (differences betweenthe novel and the standard for each of the seven windowsseparately) and IQ for each group separately. Furthermore,partial correlation coefficients were calculated for groupsseparately to explore if there was a relation between theeffect of the novel sound on MRT relative to the standard(novel minus standard) and the mean amplitudes of theERPs (novel minus standard, all seven windows) whilecontrolling for IQ. Because multiple correlations areperformed, only correlations with a p-value < .01 arereported, with an exception of the correlations betweenMRT and commission errors where alpha was set at .05,because in that analysis, only two correlations were tested.

3. Results

3.1. Performance measures

In the ADHD group, there was no significant correlationbetween IQ and the difference scores (novel minus standard)on mean reaction time, percentage of commission and omis-sion errors. Thus, IQ did not influence performance differ-ences between the conditions in the ADHD group. In thecontrol group there was a significant correlation betweenthe difference in the percentage of omission errors and IQ(r = .60, p < .01), but no significant correlations betweenthe other performance measures and IQ. In order to controlfor possible effects of IQ on the performance measures, thegroups were matched for IQ (n = 14 in each group), allanalyses were performed with these subgroups and withthe entire group. Although the results did not differ in termsof significant and nonsignificant effects from the results

IQ-matched subgroups (Mean (SD))

HD (n = 25) NC (n = 14) ADHD (n = 14)

(100) 625 (64) 642 (69)(99) 588 (73) 613 (62)(40) 37 (26) 29 (38)

(10) 10 (6) 17 (9)(10) 15 (9) 19 (9)(7) �5 (7) �2 (7)

(5)** 0.7 (1) 3 (3)*

(12)** 1 (2) 9 (9)*

(8)** �0.7 (0.9) �5 (6)*

significant differences between Normal Control and the ADHD group, foron IQ (fifth and last column).


obtained in the entire group, for the sake of completeness,the results of the subgroups are summarized in Table 1.The results described in the text apply to the entire group.In both groups, there was evidence for a speed accuracytrade-off: the correlation between the mean reaction timeand errors was �.42, p < .05 in the ADHD group and�.59, p < .01 in the control group, respectively. Childrenwho reacted faster made more errors. However, there wasno significant difference between these correlations (correla-tion test; Preacher, 2002), thus possible performance differ-ences between the groups could not be explained by adifference in response strategy between the groups.

Table 2 presents the mean reaction times, and percent-ages of commission and omission errors. Condition effectswere found for mean reaction time [F(1,41) = 45.63,p < .0005], commission errors [F(1, 41) = 11.63, p < .001]and omission errors [F(1, 41) = 11.89, p < .001]. The chil-dren responded slower on the trials, after the novels werepresented compared to the trials after the standards butmade fewer errors of commission and omission on thesetrials. The ADHD group committed more omission errorsoverall than the control group [F(1,41) = 7.34, p < .01],while there was no difference in mean reaction time andcommission errors. A group by condition effect was found

Fig. 2. ERPs in response to the novel and the standard (left) and the differencethe windows that were analyzed.

for omission errors [F(1, 41) = 7.87, p < .01]. The novelsound was associated with a decrease in errors of omissionwith 6% in the ADHD group and only with 0.6% in thecontrol group. Although the difference in omission errorsbetween the groups was smaller after novels than afterstandards, children with ADHD still made more omissionerrors than controls [t(41) = 2.89, p < .01].

3.2. Event-related potentials

Fig. 2 displays the ERPs at Fz and Cz for the standard, thenovel and the difference wave (novel minus standard) for bothgroups. The auditory stimulus was presented at 0 ms and thevisual stimulus at 300 ms. As can be seen in Fig. 2, the P3ahad a biphasic structure with an early phase, the eP3a(150–250 ms) and a late phase, the lP3a (250–400 ms). TheeP3a had its maximum amplitude over the fronto-centralscalp. The lP3a was more widely distributed than the eP3aand had a central maximum. The LN (400–800 ms) had a widedistribution with a frontal maximum.

Table 3 displays the significant effects of condition,group and their interaction for all selected windows at Fzand Cz and the scalp distribution for the difference wavein both groups. Only significant F-values and effect sizes

wave (novel minus standard, right) for both groups. Vertical lines indicate

Table 3Summary of significant effects of the ANOVAs of the ERPs after a novel compared with the standard at Fz and Cz and the scalp distribution of the difference waves for both groups within seven 50-mstime window

Electrode: Fz Time Window

150–200 200–250 250–300 300–350 350–400 400–450 450–500

EffectGroup

g2p

F(1,41)Condition

g2p .72 .56 .54 .66 .28 .27 .75

F(1,41) 104.3** 57.7** 47.6** 79.8** 15.7** 14.8** 123.1**

Group · Conditiong2

p .15F(1,41) 7.2**

Electrode: CzGroup

g2p .09 .16

F(1,41) 4.2* 7.8**

Conditiong2

p .67 .62 .69 .79 .36 .36 .76F(1,41) 84.8** 68.0** 61.8** 151.9** 22.9** 23.0** 129.1**

Group · Conditiong2

p

F(1,41)

Scalp distribution difference wave NC

Scalp distribution difference wave ADHD

Scale (lV)

* p < .05.** p < .01, NC, Normal Controls.

R.

van

Mo

urik

eta

l./

Clin

ical

Neu

rop

hy

siolo

gy

11

8(

20

07

)1

85

5–

18

65

1861


are reported. Condition effects were found for all selectedwindows: The novel elicited a larger positivity in thewindows of the eP3a (150–250 ms) and the lP3a(250–400 ms) and a larger negativity in the windows ofthe LN (400–500 ms). A group effect was found for thewindows of the eP3a (150–250 ms) at Cz: the ADHD grouphad a larger positivity, both in response to the standardand to the novel. An interaction between Group · Condi-tion was found for the 250–300 ms window at Fz: Thenovel elicited a larger positivity in the ADHD group in thiswindow than in the control group. This increased positivityreflects that the first part of lP3a is larger in the ADHDgroup at Fz. The scalp distribution for this window(250–300 ms) shows that this positivity is more widespreadin the ADHD group than in the control group.

In order to test whether there was a relation between IQand the mean ERP amplitudes, correlations were computedin each group between total IQ scores and the difference inamplitude in all selected windows. None of these correla-tions was significant. These difference windows were alsocorrelated with the difference in reaction time (novel minusstandard), while controlling for IQ in order to test if therewas a relation between distraction effects at the behaviorallevel and at the electrophysiological level. Again, none ofthese correlations reached significance, indicating thatthere is no direct relationship between performance mea-sures and ERPs.

4. Discussion

The main findings of the present study were that thenovel sounds reduced the percentage of errors of omissionsin the ADHD group more than in the normal controlgroup and enhanced the mean amplitude of the second partof the P3a (lP3a) to a greater extent in the ADHD groupthan in the normal control group. First, the performanceresults are discussed and then the ERP results.

4.1. Performance

Our study showed that irrelevant novel sounds distractchildren’s performance by increasing reaction time afterthe occurrence of a novel compared with a standard. Thisis in line with previous studies on distraction (Esceraet al., 1998; Gumenyuk et al., 2004). The increase in reac-tion time was similar in children with ADHD and normalcontrol children; thus children with ADHD did not dispro-portionally slow down when distracted. Importantly, it wasfound that, after a novel sound, both groups committedfewer errors. This finding contrasts with earlier studies ondistraction in children, which reported that more errorsare committed after distracting sounds (Gumenyuk et al.,2004, 2005). A possible explanation for these contrastingfindings might be that, because the task here was designedas a runner’s game, the children were especially focused onspeed and, therefore, committed more errors. When a novelsound captured their attention, they automatically slowed

down, resulting in less fast guesses. The speed accuracytrade-off effect in both groups supports the idea that chil-dren tend to make less errors, if they slow down on thistask.

An intriguing finding was that children with ADHDcommitted fewer omission errors after the occurrence ofa novel. Overall, children with ADHD made more omis-sion errors, a finding that has been frequently reported incontinuous performance tasks (see for meta-analysis Losieret al., 1996) and is related to various ADHD symptomsincluding difficulties sustaining attention and being easilydistracted (Epstein et al., 2003). The novel sounds in thistask could serve as stimulation for children with ADHDby making them more alert and focused on the task result-ing in a decreased number of omission errors. That chil-dren with ADHD benefit from extra-task distraction hasbeen established in several studies (Zentall and Meyer,1987; Abikoff et al., 1996; Leung et al., 2000) and can beconsidered as support for the optimal stimulation theoryof ADHD (Zentall and Zentall, 1983) and the cognitiveenergetic model of ADHD (Sergeant et al., 1999; Sergeant,2005). The optimal stimulation theory postulates that theperformance of children with ADHD benefits from extra-task distraction because this increases their arousal to anoptimal level. The cognitive energetic model emphasizesthat children with ADHD might suffer from an energeticaldysfunction and are, therefore, unable to adjust their acti-vation to meet task demands. The reduction of omissionerrors after a novel can be interpreted as the result ofincreased activation to a more adequate level. However,it should be noted that there might have been a ceilingeffect in the normal control group. The normal controlgroup did not commit many omission errors and thus theyhad no room for improvement.

4.2. Event-related potentials

The ERPs consisted of biphasic P3a with an early anda late phase (eP3a and lP3a) followed by a LN compo-nent, which were visible, both in the raw ERPs as wellas in the difference wave. The ADHD group showed alarger positivity at Cz in the 150–250 ms window inresponse to the standard and to the novel. As statedearlier, it is difficult to separate the eP3a from the P2component. Enhanced P2 components in ADHD groupscompared with normal controls in various tasks werereported by Robaey et al. (1992) and Oades et al.(1996) and might be related to altered automatic informa-tion processing in ADHD. Specifically, abnormalities inthe P2 amplitude topography and latency in the ADHDgroup have been interpreted as atypical inhibition ofsensory input from further processing (Johnstone et al.,2001). However, the eP3a has been described as acomponent that is related to govern the direction ofthe attentional move, which, in turn, would be reflectedby the lP3a. Thus, although the latency of the P2 andthe eP3a is the same, the functional interpretation is


somewhat different. Following other studies of distraction(Escera et al., 2000; Gumenyuk et al., 2004, 2005), thepositivity in the 150–250 ms window in this study is inter-preted as reflecting the eP3a, and not the P2. A largereP3a in the ADHD group could indicate that ‘the callfor attention’ is stronger in the ADHD group, both inresponse to the standard as to the novel. Contrary toour findings, Gumenyuk et al. (2005) found a reducedeP3a in the difference wave (novel minus standard) inthe ADHD group compared with the normal controlgroup. Since in that study only the results of the differ-ence wave were presented, it is difficult to compare thatfinding with our results.

The lP3a in response to the novel compared to thestandard was enhanced in the ADHD group at Fz in the250–300 ms window. This positivity was more widespreadthan in the control group. This larger positivity in thelP3a window points to a stronger involuntary switchingof attention to the novel in the ADHD group. This findingis in line with the results of Gumenyuk et al. (2005). TheP3a is known to be modulated by the noradrenergic system(Missonnier et al., 1999) and the dopaminergic system(Kahkonen et al., 2002). A dysregulation in the noradren-ergic and dopaminergic systems has been implicated in thepsychopathology of ADHD (Biederman and Spencer,1999; Solanto, 2002; Pliszka, 2005).

Berridge and Waterhouse (2003) suggested that the nor-adrenergic system might enhance cognitive functioningunder ‘noisy’ conditions in which irrelevant stimuli couldimpair performance, by reducing ‘noise’ and/or facilitatingprocessing of relevant sensory signals. Following this lineof reasoning, the enhanced lP3a in response to the novelsin the ADHD group could be the result of insufficient nor-adrenergic modulation of the fronto-subcortical pathways,which could lead to a greater sensitivity to irrelevantstimuli.

Polich and Criado (2006) developed a theoretical modelof the P3a and the P3b. They stated that the P3a originatesfrom stimulus-driven disruption of frontal attentionengagement during task processing, while the P3b origi-nates when temporal-parietal mechanisms process the rele-vant stimulus information for memory storage. A P3a canthus be elicited by novel and distracting stimuli acrossmodalities, but also by non-novel stimuli in a difficult odd-ball paradigm (target/standard discrimination) in responseto infrequent, irrelevant, but non-novel distracters (Comer-chero and Polich, 1998; Polich and Comerchero, 2003).Both saliency of the distracter as well as task difficulty ofthe primary task rather than novelty per se contribute toeliciting of the P3a (Polich and Criado, 2006). A P3b is typ-ically elicited by infrequent target stimuli. Interestingly,this task relevant P3b has been found to be reduced inADHD (Satterfield et al., 1994) indicating that childrenwith ADHD suffer from deficient preferential processingof to be attended stimuli. Methylphenidate, which increasesnoradrenergic and dopaminergic neurotransmission in theprefrontal cortex (Berridge et al., 2006), has been found

to enhance the amplitude of the P3b and to improve per-formance in ADHD (Jonkman et al., 1997). It is possiblethat a dysfunction in the noradrenergic and dopaminergicsystems in ADHD could lead to an altered balance betweenthe attentional resources in which target stimuli receive lessattention, while irrelevant novel stimuli elicit more atten-tion compared to normal controls. Future research is nec-essary to test this hypothesis and to examine the influenceof methylphenidate on the P3a in response to distractingirrelevant stimuli.

No differences were found between the groups for theLN component, which suggests that children with ADHDdid not have more difficulty than their normal peers inreorienting their attention back to the task after havingbeen distracted. Konrad et al. (2006) reported that childrenwith ADHD tend to recruit more fronto-striatal-insularactivation than normal controls during reorienting in theabsence of behavioral differences in reorienting, which isexplained in the context of neural compensation. Gumen-yuk et al. (2005) did find differences in the LN (a largerLN in an early time window and a smaller LN in a latertime window) in an ADHD group compared to a controlgroup. In their study, the differences in the LN in theADHD group might be related to their increased omissionrate after a novel. Perhaps children with ADHD do nothave a functional deficit in reorienting of attention, buttheir reorienting capability may be more dependent on taskdemands.

Taken together, it may be concluded that children withADHD show an increased orienting to novel auditoryinformation as indicated by the larger positivity in thelP3a window, but they seem to have normal reorientingabilities. Thus, the distractibility observed in the classroomin children with ADHD could be caused by an enhancedorienting reaction to unattended and irrelevant informa-tion, which is probably modulated by the dopaminergicand noradrenergic systems. However, this increased orient-ing to novels does not necessarily lead to larger behavioraldistraction effects in the ADHD group. Instead, the presentresults provide evidence that the performance of childrenwith ADHD can be improved by temporary distraction,as indicated by the reduced omission rate. Possibly, novelscan increase the arousal of children with ADHD to an opti-mal level (Zentall and Zentall, 1983), which results inimproved performance. In this case, the distraction is notdetrimental, but has a stimulating effect on the perfor-mance accuracy (specifically on omission errors) in theADHD group.

Acknowledgments

We thank all the children who participated in this study,Durk Talsma for his advice on the paradigm, Eric vanRossum for editing the novel sounds, Paul Groot for hisassistance in programming the task and Joost Witlox forhis assistance with the electrophysiological recording.


References

Abikoff H, Courtney ME, Szeibel PJ, Koplewicz HS. The effects ofauditory stimulation on the arithmetic performance of children withADHD and nondisabled children. J Learn Disabil 1996;29:238–46.

American Psychiatric Association. Diagnostic and statistical manual ofmental disorders. 4th ed. Washington, DC: American PsychiatricPress; 1994.

Barkley RA. Behavioral inhibition, sustained attention and executivefunctions: constructing a unifying theory of ADHD. Psychol Bull1997;121:65–94.

Berridge CW, Waterhouse BD. The locus-coeruleus noradrenergic system:modulation of behavioral state and state-dependent cognitive pro-cesses. Brain Res Rev 2003;42:33–84.

Berridge CW, Devilbiss DM, Andrzejewski ME, Arnsten AF, KelleyAE, Schmeichel B, et al. Methylphenidate preferentially increasescatecholamine neurotransmission within the prefrontal cortex at lowdoses that enhance cognitive function. Biol Psychiatry2006;60:1111–20.

Biederman J, Spencer T. Attention-deficit/hyperactivity disorder (ADHD)as a noradrenergic disorder. Biol Psychiatry 1999;46:1234–42.

Campbell T. The cognitive neuroscience of auditory distraction. TrendsCogn Sci 2005;9:3–5.

Carlson CL, Booth JE, Shin M, Canu WH. Parent-, teacher -, and self-rated motivational styles in ADHD subtypes. J Learn Disabil2002;35:104–13.

Ceponien_e R, Rinne T, Naatanen R. Maturation of cortical soundprocessing as indexed by event-related potentials. Clin Neurophysiol2002;113:870–82.

Ceponien_e R, Lepisto T, Soininen M, Aronen E, Alku P, Naatanen R.Event-related potentials assiociated with sound discrimination versusnovelty detection in children. Psychophysiology 2004;41:130–41.

Comerchero MD, Polich J. P3a, perceptual disticntiveness, and stimulusmodality. Cogn Brain Res 1998;7:41–8.

Cycowicz YM, Friedman D. A developmental study of the effect oftemporal order on the ERPs elicited by novel environmental sounds.Electroencephalogr Clin Neurophysiol 1997;103:304–18.

Douglas VI, Peters KG. Toward a clearer definition of the attentionaldeficit of hyperactive children. In: Hale GA, Lewis M, editors.Attention and the development of cognitive skills. New York: PlenumPress; 1979. p. 173–246.

Epstein JN, Erkanli A, Conners CK, Klaric J, Costello JE, Angold A.Relation between continuous performance test performance measuresand ADHD behaviors. J Abnorm Child Psychol 2003;31:543–54.

Escera C, Alho K, Winkler I, Naatanen R. Neural mechanisms ofinvoluntary attention to acoustic novelty and change. J Cogn Neurosci1998;10:590–604.

Escera C, Alho K, Schroger E, Winkler I. Involuntary attention anddistractibility as evaluated with event-related brain potentials. AudiolNeurootol 2000;5:151–66.

Friedman D, Cycowicz YM, Gaeta H. The novelty P3: an event-relatedbrain potential (ERP) sign of the brain’s evaluation of novelty.Neurosci Biobehav Rev 2001;25:355–73.

Gumenyuk V, Korzyukov O, Alho K, Escera C, Schroger E, IlmoniemiRJ, et al. Brain activity index of distractibility in normal school-agechildren. Neurosci Lett 2001;314:147–50.

Gumenyuk V, Korzyukov O, Alho K, Escera C, Naatanen R. Effects ofauditory distraction on electrophysiological brain activity and perfor-mance in children. Psychophysiology 2004;41:30–6.

Gumenyuk V, Korzyukov O, Escera C, Hamalainen M, Huotilainen M,Hayrinen T, et al. Electrophysiological evidence of enhanced distract-ibility in ADHD children. Neurosci Lett 2005;374:212–7.

Holcomb PJ, Ackerman PT, Dykman RA. Auditory event-relatedpotentials in attention and reading disabled boys. Int J Psychophysiol1986;3:263–73.

Huang-Pollock CL, Nigg JT, Carr TH. Deficient attention is hard to find:applying the perceptual load model of selective attention to attention

deficit hyperactivity disorder subtypes. J Child Psychol Psychiatry2005;46:1211–8.

Jonkman LM, Kemner C, Verbaten MN, Koelega HS, Camfferman G,van der Gaag RJ, et al. Effects of methylphenidate on event-relatedpotentials and performance of attention-deficit hyperactivity disorderchildren in auditory and visual selective attention tasks. Biol Psychi-atry 1997;41:690–702.

Johnstone SJ, Barry RJ, Anderson JW. Topographic distribution anddevelopmental timecourse of auditory event-related potentials in twosubtypes of attention-deficit hyperactivity disorder. Int J Psychophys-iol 2001;42:73–94.

Kahkonen S, Ahveninen J, Pekkonen E, Kaakkola S, Huttunen J,Ilmoniemi RJ, et al. Dopamine modulates involuntary attentionshifting and reorienting: an electromagnetic study. Clin Neurophysiol2002;113:1894–902.

Kaipo ML, Alho K, Winkler I, Escera C, Surma-aho O, Naatanen R.Event-related brain potentials reveal covert distractibility in closedhead injuries. Neuroreport 1999;10:2125–9.

Keage HD, Clark CR, Hermens DF, Kohn MR, Clarke S, Williams LM,et al. Distractibility in AD/HD predominantly inattentive and com-bined subtypes: the P3a ERP component, heart rate and performance.J Integr Neurosci 2006;5:139–58.

Kenmer C, Verbaten MN, Koelega HS, Buitelaar JK, van der Gaag RJ,Camfferman G, et al. Event-related brain potentials in children withattention-deficit and hyperactivity disorder: effects of stimulus devi-ancy and task relevance in the visual and auditory modality. BiolPsychiatry 1996;40:522–34.

Kilpelainen R, Luoma L, Herrgard E, Sipila P, Ypparila H, Partanen J,et al. Distractible children show abnormal orienting to non-attendedauditory stimuli. Neuroreport 1999;10:1869–74.

Kuntsi J, Eley TC, Taylor A, Hughes C, Asherson P, Caspi A, et al. Co-occurence of ADHD and low IQ has genetic origins. Am J Med Gen BNeuropsychiatr Genet 2004;124:41–7.

Konrad K, Neufang S, Hanisch C, Fink GR, Herpetz-Dahlmann B.Dysfunctional attentional networks in children with attention deficit/hyperactivity disorder: evidence from an event-related functionalmagnetic resonance imaging study. Biol Psychiatry 2006;59:643–51.

Kort W, Compaan EL, Bleichrodt N, Resing WCM, Schittekatte M,Bosman M, et al. WISC-IIINL. [Dutch translation of the WechslerIntelligence Scales for Children, 3rd ed.]. Amsterdam NL: NDC/NIP,2002.

Lepisto T, Soininen M, Ceponien_e R, Almqvist F, Naatanen R, AronenET. Auditory event-related potential indices of increased distractibilityin children with major depression. Clin Neurophysiol 2004;115:620–7.

Leung JP, Leung PWL, Tang CSK. A vigilance study of ADHD andcontrol children: event rate and extra-task stimulation. J Dev PhysicalDisabil 2000;12:187–201.

Losier BJ, McGrath PJ, Klein RM. Error patterns on the continuousperformance test in non-medicated and medicated samples of childrenwith and without ADHD: a meta-analytic review. J Child PsycholPsychiatry 1996;37:971–87.

Maata S, Saavalainen P, Kononen M, Paakkonen A, Muraja-Murro A,Partanen J. Processing of highly novel auditory events in children andadults: an event-related potential study. Neuroreport 2005;16:1443–6.

Mason DJ, Humphreys GW, Kent LS. Exploring selective attention inADHD: visual search through space and time. J Child PsycholPsychiatry 2003;44:1–20.

Mayer AR, Harrington D, Adair JC, Lee R. The neural networksunderlying endogenous auditory covert orienting and reorienting.Neuroimage 2006;30:938–49.

Miller GA, Chapman JP. Misunderstanding analysis of covariance. JAbnorm Psychol 2001;110:40–8.

Missonnier P, Ragot R, Derouesne C, Guez D, Renault B. Automaticattentional shifts induced by a noradrenergic drug in Alzheimer’sdisease: evidence from evoked potentials. Int J Psychophysiol1999;33:243–51.


Oades RD, Dittmann-Balcar A, Schepker R, Eggers C, Zerbin D.Auditory event-related potentials (ERPs) and mismatch negativity(MMN) in healthy children and those with attention-deficit ortourette/tic symptoms. Biol Psychol 1996;43:163–85.

Oosterlaan J, Scheres A, Antrop I, Roeyers H, Sergeant JA. Vragenlijstvoor Gedragsproblemen bij Kinderen (VvGK). Nederlandse bewer-king van de Disruptive Behavior Disorders Rating Scale [Dutchtranslation of the Disruptive Behavior Disorders Rating Scale]. Lis-se: Swets & Zeitlinger; 2000.

Pearson DA, Lane DM, Swanson JM. Auditory attention switching inhyperactive children. J Abnorm Child Psychol 1991;19:479–92.

Pelham WE, Aronoff HR, Midlam JK, Shapiro CJ, Gnagy EM, ChronisAM, et al. A comparison of ritalin and adderal: efficacy and time-course in children with attention-deficit/hyperactivity disorder. Pedi-atrics 1999;103:1–14.

Pelham WE, Gnagy EM, Greenslade KE, Milich R. Teacher ratings ofDSM-III-R symptoms for the disruptive behavior disorders. J AmAcad Child Adolesc Psychiatry 1992;31:210–8.

Pliszka SR. The neuropsychopharmacology of attention-deficit/hyperac-tivity diosorder. Biol Psychiatry 2005;57:1385–90.

Polich J, Comerchero MD. P3a from visual stimuli: typicality, task andtopography. Brain Topogr 2003;15:141–52.

Polich J, Criado JR. Neuropsychology and neuropharmacology of P3aand P3b. Int J Psychophysiol 2006;60:172–85.

Preacher KJ. Calculation for the test of the difference between twoindependent correlation coefficients [Computer software], 2002. Avail-able from <http://www.quantpsy.org>.

Robaey P, Bretton F, Dugas M, Renault B. An event-related potentialstudy of controlled and automatic processes in 6–8 year old boys withattention deficit hyperactivity disorder. Electroencephalogr clin Neu-rophysiol 1992;82:330–40.

Satterfield JH, Schell AM, Nicholas T. Preferential neural processing ofattended stimuli in attention-deficit hyperactivity disorder and normalboys. Psychophysiology 1994;31:1–10.

Sergeant JA. Modeling attention-deficit/hyperactivity disorder: a criticalappraisal of the cognitive-energetic model. Biol Psychiatry2005;57:1248–55.

Sergeant JA, Oosterlaan J, van der Meere JJ. Information processing andenergetic factors in attention-deficit/hyperactivity disorder. In: Quay

HC, Hogan S, editors. Handbook of disruptive behaviour disor-ders. New York: Plenum Pres; 1999. p. 75–104.

Scheres A, Oosterlaan J, Geurts H, Morein-Zamir S, Meiran N, Schut H,et al. Executive functioning in boys with ADHD: primarily aninhibition deficit?. Arch Clin Neuropsychol 2004;19:569–94.

Schroger E, Wolff C. Attentional orienting and reorienting is indicated byhuman event-related potentials. Neuroreport 1998;9:3355–8.

Semlitsch HV, Anderer P, Schuster P, Presslich O. A solution for reliableand valid reduction of ocular artifacts applied to the P300 ERP.Psychophysiology 1986;23:695–703.

Shaffer D, Fisher P, Lucas CP, Dulcan MK, Schwab-Stone ME. NIMHdiagnostic interview schedule for children version IV (NIMH DISC-IV): description, differences from previous versions, and reliability ofsome common diagnoses. J Am Acad Child Adolesc Psychiatry2000;39:28–38.

Solanto MV. Dopamine dysfunction in AD/HD: integrating clinical andbasic neuroscience research. Behav Brain Res 2002;130:65–71.

Squires N, Squires KC, Hillyard A. Two varieties of long-latency positivewaves evoked by unpredictable auditory stimuli. Electroencep ClinNeurophysiol 1975;38:387–401.

Van Mourik R, Oosterlaan J, Sergeant JA. The stroop revisited: a meta-analysis of interference control in AD/HD. J Child Psychol Psychiatry2005;46:150–65.

Van der Meere J, Sergeant JA. Focused attention in pervasivelyhyperactive children. J Abnorm Child Psychol 1988;16:627–39.

Weschler D. Weschler intelligence scales for children. 3rd ed. SanAntonio, TX: The Psychological Corporation; 1991.

Wetzel N, Berti S, Widmann A, Schroger E. Distraction and reorientationin children: a behavioral and ERP study. Neuroreport 2004;15:1355–8.

Wood C, Maruff P, Levy F, Farrow M, Hay D. Covert orienting of visualspatial attention in attention deficit hyperactivity disorder: doescomorbidity make a difference? Arch Clin Neuropsychol1999;14:179–89.

Zentall SS, Meyer MJ. Self-regulation of stimulation for ADD-H childrenduring reading and vigilance performance. J Abnorm Child Psychol1987;15:519–36.

Zentall SS, Zentall TR. Optimal stimulation: a model of disorderedactivity and performance in normal and deviant children. Psychol Bull1983;94:446–71.

http://www.quantpsy.org

when distraction is not distracting: a behavioral and erp study on distraction in adhd

Documents