Auditory Event-related Potentials in Children With Attention Deficit Hyperactivity Disorder

Background:Recording of event-related potentials (ERPs) from the scalp is a noninvasive technique reflecting the sensory and cognitive processes associated with attention tasks. Attentiondeficit hyperactivity disorder (ADHD) is a disorder involving deficits in attention and behavioralcontrol. The aim of this study was to investigate the difference in ERPs between normalchildren and those with ADHD.

ADHD met the full criteria for ADHD according to Diagnostic and Statistical Manual of MentalDisorders, fourth edition (DSM-IV). The auditory oddball paradigm was applied, and eventrelated long-latency components (N1, P2, N2 and P3) from Fz, Cz and Pz were measured ineach test subject.

Results:Children with ADHD showed a significantly longer latency and a lower amplitude of P3compared to normal control children (p<0.01). Delayed N2 latency at the Pz electrode was shownin children with ADHD compared to normal controls (p<0.01). No differences in other ERP indiceswere found between children with ADHD and controls. When divided into four age groups, thelatency of P3 was significantly increased in all age groups and a significantly smaller amplitudein P3 over the central region was found in children with ADHD>10 years of age (p<0.05).

Conclusion:We found that the endogenous ERPs (P3 and N2) were significantly affected in childrenwith ADHD, compared to exogenous ERPs (N1 and P2). Increased latency of P3 suggests a slowerprocessing speed, and decreased P3 amplitude is interpreted as disruption of inhibitory controlin children with ADHD. These results indicate a neurocognitive abnormality in ADHD, as presentedby a reduction in ERP response.

Copyright©2012, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rightsreserved.

1. Introduction

The recording of event-related potentials (ERPs) from thescalp is a noninvasive technique providing informationregarding neural activity associated with sensory, cognitive, attention and decision-making processes.1ERPs havebeen used as an electrophysiological tool for studyingneural bases of cognitive activities and in clinical applications for patients with psychopathological and neurologicaldiseases, disorders of learning and attention, dementia,and other cognitive deficits.

Attention deficit hyperactivity disorder (ADHD) is characterized by developmentally inappropriate attention,behavioral and cognitive impulsivity, and restlessness.3Evidence has shown that ADHD is associated with a deficitin response selection, motor adjustment, and inhibition ofprepotent responses.4A variety of studies of children withADHD have reported abnormalities in visual, auditory, andalternating visual and auditory ERPs.

Abnormalities in auditory ERP waves are a reflection oftasks associated with selective attention and the categorization of stimuli involved in cognitive functions.6,7Studies onERP have identified N1 and P2 components in response tofrequent tones and P3 components in response to rare tones.8P3, namely the P300 wave, is a late positive waveformoccurring with a latency of approximately 300 milliseconds ormore after an infrequently presented target stimulus (socalled “oddball paradigm”). P3 is thought to be an endogenous potential generated in the medial cortical or subcorticalregion,2,9sensitive to the delivery of task-relevant information requiring a decision or response from the participant.Research has shown that P3 is also a presentation of anupdating of working memory.10Understanding alterations ofthe ERPs in ADHD children helps to take into account theircognitive control processes and pathophysiology.

Many investigations into attention and cognition haveconducted analysis of ERPs, and P3 responses are the mostwidely investigated waveform. As mentioned by Barryet al,5the most common ERP-related discovery associatedwith ADHD is a significant reduction in the amplitude of P3during the performance of oddball tasks. The results concerning latency in P3 with regard to ADHD have beenmixed.11Changes in other ERP waveforms such as N2, P2and N1 have not been thoroughly studied in children withADHD. The results of behavioral performance have beeninconsistent when reported by measurements of reactiontime, total hits, and false alarms.

Behavioral studies of children of various ages haveshown the importance of age with respect to the regulationand direction of attention.13,14Age-dependent changes inthe ERPs of normal children have been reported in severalstudies.15,16Some reports have found that the age speci-ficity of lower P3 amplitude in ADHD patients exists mostlyin children,6,13but not in adolescent subjects.6,17Ageeffect should be considered in studies of the ERP in childrenwith ADHD.

In Taiwan, changes in ERP waveforms in children withADHD have rarely been reported.18The aim of this studywas to investigate the differences in ERP responses,focusing on the differences in the latencies and amplitudesof P3, N2, P2 and N1 among children with ADHD and normal children between 6 and 13 years of age. Furthermore, thesubjects in this study were divided into different age groupsto evaluate the developmental effects of age on ERPs inchildren with ADHD.

2. Materials and Methods

2.1. Subjects

The ADHD group consisted of 50 children, aged 6e13 yearsold (42 boys and 8 girls) and recruited from our outpatientclinic. Children were divided into four age groups fortesting: (1) 6-7 years; (2) 8-9 years; (3) 10-11 years; and(4) 12-13 years.All children with ADHD completed both theparents and teacher versions of the Child Activity Checklistin Chinese19with their scores over P85 (85 percentile). Theteacher version of the Child Activity Checklist includedchildren’s behavioral control in general, during classroom,in groups, and responses to teachers such as “staying in seataccording to classroom rules”, “complying with usualrequest or direction of teachers”, and school performance.The parents’ version of Child Activity Checklist includedbehavior control and attention at home or other dailyactivity such as “interrupting another person’s conversationor activity”, or “unable to stick with one game or toy”.

All children with ADHD met the full criteria of ADHDaccording to Statistical Manual of Mental Disorders, fourthedition (DSM-IV; American Psychiatric Association, 1994).20Of 50 children who were diagnosed with ADHD, 48 (96%) hadcombined-typed ADHD and two had predominantly inattentive type of ADHD. Full-scale Wechsler Intelligence Scalefor Children (WISC) IV in children with ADHD was 80 orhigher. A detailed history and physical and psychiatricexamination were conducted, and the diagnosis wasconfirmed by a child psychiatrist, and none of the childrenwere taking any medication. The patients with Tourette’ssyndrome, seizure disorders, learning disabilities, autism,Asperger diseases, mental retardation, and other psychoticdisorders were excluded.

A comparable age- and sex-matched control group of 51children aged 6e13 years old (40 boys, 11 girls) wasassembled from patients who had previously visited ouroutpatient clinic suffering from various acute illnesses orfor health examinations unconnected with neurological orpsychiatric illness. Subjects with hearing problems werealso excluded. Informed consent was obtained fromparticipants’ parents or guardians in accordance with therequirement of the ethics boards of the Cathay GeneralHospital (CGH-CT9762) and Cheng-Hsin General Hospital(CHGH-IRB-165-98-49).

2.2. Methods

The subjects were tested in a relaxed sitting position withtheir eyes closed in a silent room after cleaning the skin andscalp. Bioelectrical signals were measured by placinga surface electrode (plate-shape electrode, 11 mm indiameter; Dantec Electronics A/S, Skovlunde, Denmark)along the midline frontal (Fz), central (Cz) and parietal (Pz)region, according to the 10e20 international system of EEG electrode placement and grounding, using a surface electrode located midway between the Fz and the midlinefrontopolar (FPz) points. An electrode was placed infraorbitally to monitor eye movement. A reference electrodewas placed on the mastoid, and the impedance wasmeasured at less than 5 kU. The filter band pass was set at0.1e50 Hz and the analysis time was 1 second. Waveformswere averaged, and any electroencephalograms or electrooculograms over 100mV were automatically rejected.

We applied the “oddball paradigm” of auditory stimulation (Medtronic Keypoint V3.22; Medtronic FunctionalDiagnostic A/S, Skovlunde, Denmark). ERPs were elicitedbinaurally through headphones with a typical intensity of 60dB above the hearing level, depending upon the subject. Intotal, 200 tones were delivered. According to the paradigm,20% of the tones were “target” (rare), while the remainingwere “nontarget” (frequent), and the delivery sequences offrequent and rare tones were randomized. The target toneswere 3000 Hz, while the nontarget tones were 2000 Hz,delivered at a rate of 0.7 Hz. Instructions were given by thetechnician before the test, with the subject tasked to pressthe button when they heard a rare tone or count thenumber of rare tones presented. The test was repeatedtwice for each subject.

All artifact-free ERP responses related to the rare toneswere analyzed visually. Latency windows for potentialswere designated in the following ranges: 75-150 milliseconds for N1, 120-250 milliseconds for P2, 150-350 milliseconds for N2 and 250-700 milliseconds for P3. In theanalysis of potentials, the amplitude of P3 was measuredfrom the peak of N2 to peak of P3 (N2eP3); that of N2 wasmeasured from the peak of P2 to the peak of N2 (P2eN2);that of P2 was measured from the peak of N1 to the peak ofP2 (N1eP2); and that of N1 was measured from the firstdeflection to the peak of N1.

3. Results

The age range of the ADHD group (42 male, 8 female) andcontrol group (40 male, 11 female) was between 6 and 13years. The mean age of the ADHD group was 8.9±1.9years, and the mean age of the control group was 9.0±2.0years. No significant difference in the ages or sexes of thechildren was noted between the ADHD and control groups.

Table 1shows the mean latency and amplitude of eachERP component with standard deviation in the ADHD andcontrol groups at each of the electrode sites. The P3latency at Pz in patients with ADHD was 384.6±51.1milliseconds, which was significantly longer than that of329.0±32.3 milliseconds in normal, age-matched controls(p<0.01,Table 1andFigure 1). P3 latencies at Fz and Czelectrode sites were also significantly longer in childrenwith ADHD compared to the controls (p<0.01,Table 1).The N2 at the Pz electrode was found to be longer in children with ADHD (254.7±22.6 milliseconds) compared tonormal controls (238.4±28.9 milliseconds,p<0.01,Table 1). However, no statistically significant differencewas found in the latencies of N2 at Fz or Cz. The latenciesof P2 and N1 did not differ at any of the electrode sitesbetween the children with ADHD and the controls.

The amplitude of P3 was 12.5±4.5μV at Pz in childrenwith ADHD and 14.7±4.6μV at Pz in the control children.Children with ADHD showed significantly lower amplitudein P3 for all electrodes compared to normal controls(p<0.05,Table 1). However, the amplitudes of other ERPs,including N2, P2 and N1, did not show a statisticallysignificant difference in any of the electrodes for childrenwith ADHD compared to the matched control groups.Figure 1illustrates longer P3 latencies and smaller amplitude of P3 in children with ADHD compared to normalchildren.

Linear regression analysis demonstrated a significantnegative linear correlation of P3 latency at Pz with age inthe normal control group (correlation coefficient r=-0.33,p=0.02;Figure 2); however, children with ADHDdid not show significant difference in linear correlationbetween P3 latency and age (coefficient r=-0.17,p=0.24;Figure 2). Similar findings were observed at otherelectrode locations (Fz and Cz).

When divided into four age groups, the P3 latencyappeared to be significantly increased in each age group ofthe ADHD children compared to the controls (Table 2). TheN2 latencies of Pz in the age groups 8-9 years, 10-11 years and 12-13 years appeared significantly longer in the ADHDgroup compared with the normal controls, but not in the 6-7 years group (Table 2). The P3 amplitude at Cz wassignificantly lower in the age groups 10e11 years and 12e13years (p<0.05), and with a trend of significantly loweramplitude at Pz in the age groups 10-11 years and 12-13years for children with ADHD compared to the corresponding age groups of normal controls (p=0.07 and 0.09,respectively,Table 3). The P3 amplitude in the age groups6e7 years and 8-9 years did not differ between childrenwith ADHD and controls. A lower P2 amplitude was alsofound in 10e11-year-old and 12e13-year-old children with ADHD compared to controls (p<0.05,Table 3). Other ERPindices in each age group did not differ significantlycompared to those of the normal controls.

4. Discussion

In recent decades, auditory ERP and P300 (P3) studies havebeen widely used as a noninvasive method to evaluate thefunction of cognition and attention in children. For clinicalpurposes, the most convenient protocol to elicit ERPs is the 

so-called oddball paradigm,1,8which is now available witheach clinical evoked-potential diagnostic machine. The P3response has a large amplitude in its response to rare tones,and is easily identified in single trials. Our study demonstrated a significantly increased P3 latency and lower P3amplitude in 6-13-year-old children with ADHD comparedto normal controls. When divided into four age groups, thelatency of P3 was significantly increased in each age groupof the children with ADHD, and a significantly smaller amplitude in P3 over the central area was found in childrenwith ADHD>10 years of age.

P3 latency is thought to reflect the timing involved in thecategorization of stimuli.6,7Increased P3 latency in a set ofERP responses reflects a defect in the cerebral processingof attention and a reduction in the speed of processing inchildren with ADHD.21Several ERP studies have suggestedthe existence of abnormal sensory or cognitive informationprocessing in patients with ADHD.21,22Our results haveshown an increase in the latency of P3, which is consistentwith the results of previous studies.22e24However,a number of studies have not reported a difference inlatency between ADHD and normal controls.25,26Sucha discrepancy might have been due to different caseinclusion criteria and methodologies. Our case selectionincluded mostly combined-typed ADHD (96%), whereasother studies have recruited patients with only inattentivetype or other ADHD-comorbid groups.25,27Further studiesare needed to investigate the effect of comorbid disordersassociated with ADHD on ERPs.

P3 amplitude is believed to provide a psychophysiological signature of deficits related to inhibitory control.Reduction in P3 amplitude elicited from an auditory oddballtask is specific for children with ADHD,5,21in contrast tohealthy children and children with autism or dyslexia.23,28A number of children with ADHD have never producedP3 waveforms.29A small P3 amplitude is explained asa reflection of behavioral disinhibition,30a failure ofbehavioral control,31and CNS hyperexcitability.32We founda decrease of P3 amplitude in children with ADHD, whichcorresponded to previous results10,21,24,25; however, somestudies have found no difference in P3 amplitude betweencontrol and ADHD subjects.

A negative correlation has been identified betweenERP latencies and age in normal children in previousstudies.1,8,15Our study showed an age effect of P3 latency in normal controls, but not in children with ADHD, whichparalleled the results of other studies in children with othercognitive defects.33ERPs have shown an age difference inselective attention disorders.5,6We also identified theeffect of age on P3 amplitude, with significantly loweramplitude in the central area in children older than 10years of age, but not in those younger than 9 years old.Other studies have also found a significant difference in P3waveform in older children (8-12 years) with ADHD, but notin younger children (<7 years).

A significantly increased N2 latency in the parietal regionof ADHD children was identified in the present study. N2waveforms are believed to be related to signal detectionand discrimination.5,34Inconsistent findings in N2 latencybetween ADHD and normal controls have been reported,and the results may be related to the age of the testsubjects.5A number of reports have suggested a decrease inN2 amplitude in the frontal and parietal areas in ADHDpatients compared to controls,13,35whereas others have notfound such a difference.27,36Our study showed a decreasedamplitude of N2 in children with ADHD compared to normalcontrols, but did not reach statistical significance.

P3 in ERPs has been used as a predictor of the responseto treatment with CNS stimulants such as methylphenidateand atomoxetine. Administration of methylphenidatenormalizes ERP indices, P3 amplitude, and latencies inchildren with ADHD.22,37Recently, ERP studies combinedwith functional magnetic resonance imaging have shownalterations in the frontal striatal and parietal lobe functionand its modulation during response inhibition followingadministration of methylphenidate in children withADHD37,38Hemodynamic deficits in the right middle frontalgyrus and right anterior superior temporal gyrus in functional magnetic resonance imaging have been associatedwith lower P3 in patients with ADHD.

In conclusion, children with ADHD show a significantincrease in P3 latency, with a reduction in the amplitudecompared to normal age-matched controls. A longerlatency in P3 suggests slower processing of attentiontasks, and lower P3 amplitude is interpreted as disruptionin the inhibitory control of children with ADHD. Theseresults suggest that children diagnosed with functionalpsychiatric disorder such as ADHD may have physiologicallyrelated neurocognitive abnormality. However, an importantlimitation of this study was that ADHD with comorbiddisorders was not investigated. The change in P3 andN2 waveforms are also linked to other diseases with attention and inhibition disturbances such as mental retardation,autism, speechelanguage disorders, learning disability, andother behavioral and cognitive abnormalities. Futurestudies should involve analysis of detailed comorbidbehavior or learning disorder. It should be noted that ERPstudy may be used as an adjunctive neurophysiologicalreference for the cognitive abnormality in ADHD childrenbut not as a replacement for clinical diagnosis of ADHD.

Acknowledgments

This research was supported by an internal grant fromCheng-Hsin General Hospital (CHGH 98-49). The authorsthank Dr. William Tao-Hsin Tung for his statistical analysis,Ms. Chin-Yi Huang and Ms. Pei-Chun Lai at the EMG laboratory of Cathay General Hospital for their technicalsupport, and Ms. Janet Hsin for her assistance.