B~'mn & Development, 15 (1993) 173-177

03~7-7604/93/$06.00 ,~'? 1993 Elsevier Science Publishers B.V. All rights reserved

BRADEV 00017

Original Articles

Long latency event-related potentials (P300) in gifted children

Florence Martin a, MD, Eric Delpont a, MD, Georges Suisse a, MD,

Christian Richelme b, MD and Claude Dolisi a, MD
aLaboratoire d'Explorations Fonctionnelles du Systdme Nerveux, Hdpital Pasteur, Nice, France
and bService de p{diatrie, Hdpital de Cimiez, Nice, France

Received 7 September 1992; accepted 8 December 1992

In children, P3 latency decreases with increasing age. This decrease could be linked with the maturation of cognitive processes. According to this
hypothesis, event-related potentials P3 were recording in gifted children to research an electrophysiological correlation with the mental precocity.
Auditory long latency event-related potentials were recorded in 10 gifted children (lQs over 140) and 23 control subjects. The part of variance
related to age was extracted by comparing deviations from regression line as a function of age. Stimulus-evoked NI component latency was not
statistically different in the two populations. Event-related P3 component latency was significantly shorter in the gifted children at Cz (P<0.05).
Inter-peak interval N1 P3 was significantly shorter at all three recording sites (P<0.01 at Fz and CZ, P<0.02 at Pz). These results suggest a
relationship between the P3 component and cognitive ability in children.

Key words: P3 (P300); Auditory event-related potential; Cognitive potential; Child development; Gifted child; Cognitive ability

INTRODUCTION MATERIALS AND M E T H O D S

The exact neuropsychological signification of long la-
tency event-related potentials has remained unclear
since Sutton's publication in 1965 [1]. Many authors
have postulated that these potentials are related to cog-
nitive processes [2M].
Numerous studies on children have revealed a de-
crease in the latency of the main positive component
(P3) with increasing age [5-10]. This could be related
to maturational phenomena concerning cognitive pro-
cesses. Two studies demonstrating longer latencies in
mentally retarded children appear consistent with this
hypothesis [11,12].
This study was designed to investigate the possible
relationship between cognitive ability and P3 in gifted
children compared with a control group, rather than in
mentally retarded children.
Correspondence address: Dr. Florence Martin, Laboratoire d'Ex-
plorations Fonctionnelles du Syst+me Nerveux, H6pital Pasteur, 30
Voie Romaine, BP 069, 06002 Nice c6dex, France.
Subjects
Ten gifted children (7 boys aged 4.4-6.9 years and 3
girls aged 5.4-7.7 years) all had intelligence quotients
(IQs) over 140, as measured by the Wechsler Intelli-
gence Scale for Children (WISC) for subjects over 4
years of age, and by the Terman-Merrill Scale for sub-
jects younger than 4 years.
The control children (8 boys aged 5.5-11.2 years and
15 girls aged 3.9-12.2 years) were recruited from chil-
dren whose performance was adequate in their school
tasks.
Recording procedures
Long latency event-related potentials were obtained
by binaural auditory stimulation with an oddball para-
digm. The standard tone pip (80%) of 1,000 Hz and the
target tone (20%) of 2,000 Hz were presented through
earphones. The tones were presented in a random series
at a rate of 0.7/s. Tones were delivered at 60 dB above
each individual's sensation threshold, and had 10 ms
rise/fall and 50 ms plateau time. Potentials were recor-

Brain & Development, Vol 15, No 3, 1993 173

 at Fz, Cz and Pz (international 10/20 system) referred
NI N2 to linked mastoids (A1 A2) with an Evomatic 8000
system (Dantec, DK-2740 Skovlunde, Denmark). The
Fz impedance of all the electrodes was less than 5 kf2.
P2 P~
The bandpass was 0.5 50 Hz. Children were t:amiliar-
ized with the tone pips before recording was started.
They were asked to detect the rare tone pips and to
Cz keep a running mental count of them or to move a
P3 finger gently.
P2
Signals exceeding ÷80 /LV were rejected on-line to
NI N2 eliminate ocular and muscular artifacts. Potentials for
both target and non-target stimuli were averaged sepa-
P2 P3
rately. Event-related potentials for the target event were
IOUV [ , computed from the first 20 artifact-free epochs. Trials
lOOms were repeated to ensure reproductibility.
Fig l. Auditory event-related potentials (boy, 10-years old, control P3 (or P300) was defined as the maximum positive
group). Solid lines, target tone; dashed lines, standard tone. component between 250 and 700 ms after stimulus pre-
sentation on the curves elicited by target stimuli (Fig 1).
ded with silver chloride electrodes placed on the scalp P3 latency was measured at the peak. P3 amplitude was

Fz
Cz Pz

1lo
• • B

g 200
II B i ~ D

[]
100 '
z

• i • i i i
i i i - , , i
i - i , i - , ,

6oo
t m

400

• i • i • i - i , i

ul il

il B
400

~- aoo

2O0

• , , i , i - i * J • i • , - , • , • ,
4 6 8 10 4 6 8 10 4 6 8 10
AGE (years) AGE (years) AGE (years)

Fig 2. N1 and P3 latencies, and IPI N1 P3 as a function of age, in the two groups m: gifted children; []: controls. Regression lines are calculated
for the entire population. N1 latency at Fz, y 287.4123- 13.3135x, r - 0 . 5 6 ; Cz, y 294.0726- 14.3378x, r - 0 . 5 8 ; Pz, y = 301.2654- 15.1077x, r
-0.6~.P3latencyatFz`y-7~8.~853-33.~349x`r-~.7~;Cz~y-693.~62-3~.38~3x~r-~.67:Pz~y 704.4375-32.6909x, r 0.72.1P1 N1 P 3 a t
Fz, y - 4 2 0 , 7 7 3 - 19.8215x. r - 0 . 5 7 : Cz, ) ' - 3 9 9 . 0 3 3 6 - 1 7 . 0 4 2 5 x , r 0.52: Pz, y - 388.2352-16.0547x, r - 0 . 5 4 .

174 Brain & Development. Vol 15. No 3, 1993

measured peak-to-peak from the negative component
just before P3 (N2).
In order to eliminate the possible influence of stimu-
lus-linked evoked potential latency on P3 latency, N1
latency and inter-peak interval (IPI) N1-P3 were stu-
died. N1 was defined as the maximum negative peak
between 75 and 150 ms after stimulus presentation.
A. N1
ms 0
-20
Data analysis
The relationships between age and latency, and be-
tween age and amplitude, were first assessed in the en-
tire population (gifted and control children). Inter-
group comparisons were then performed.
When a correlation exists between age and latency or
amplitude, that part of variance related to age must be
extracted before two populations can be compared. For
this purpose, a theoretical value was calculated for each
child based on age, and the linear regression calculated
for the entire population. The difference between the
measured value and the theoretical value was then cal-
culated for each child. Finally, mean values obtained
for each population were compared. Differences be-
tween groups were analyzed with Student's t-test for
independent samples.

-40
RESULTS

-60 Effect of age in the entire population

B. P3
Latencies
t"20 A significant negative correlation (P<0.01) was ob-
served between the variability of N1 and P3 latencies,
ms 0 and age in the entire population at all three recording
sites. Likewise, IPI N1-P3 decreased significantly with
-20 increasing age (P<0.01). Fig 2 shows the N1 latency,
P3 latency and IPI N l-P3 values, and the linear regres-
-40 sion at the three sites. There was no significant differ-
ence between girls and boys.
-60
P3 amplitude
C.N1 -P3 No significant correlation was found between P3 am-
+ 20 plitude and age at any of the three sites.

ms 0 " - - Comparison between groups (Fig 3)

-20
-40
-6O
Fig 3. Differences between measured data and theoretical values:
mean values for each population at the three recording sites. Control
children: Fz I , Cz [], Pz []; gifted children, Fz [], Cz f-l, Pz I . A:
N1 latency. The mean deviations are positive in the gifted children
and negative in the controls, without a significant difference. B: P3
latency. The mean deviations are negative in the gifted children and
positive in the control, with a significant difference at Cz. C: IPI N1
P3. The mean deviations are negative in the gifted children and posi-
tive in the controls. The mean values are significantly different at all
three sites. *P < 0.05, **P < 0.02, "**P < 0.01.
Latencies
Owing to the significant correlation between NI la-
tency, P3 latency, IPI N1-P3 and age, that part of var-
iance related to age was extracted, as explained above,
by calculating the differences between the measured
latencies and the theoretical latencies, and comparing
the mean values for the two groups.
NI latency. No significant difference was observed at
any of the three sites. However, the mean deviations
were positive in the gifted children and negative in the
control group.
P3 Latency. The mean values were significantly dif-
ferent at Cz (P < 0.05). The mean deviations were nega-
tive in the gifted children and positive in the control

Brain & Development, Vol 15. No 3, 1993 175

group. At Fz and Pz, the mean deviations were also
negative in the gifted children and positive in the con-
trol group, but the differences were not statistically sig-
nificant.
IPI N1-P3. The mean values were significantly differ-
ent at all three recording sites (P<0.01 at Fz and Cz,
P < 0 . 0 2 at Pz). The mean deviations were negative in
the gifted children and positive in the control group.
P3 amplitude
The mean amplitudes were greater in the gifted chil-
dren at all three sites, but the differences were not sta-
tistically significant.
DISCUSSION
Long latency event-related potentials ('cognitive' po-
tentials) have been related to mental processes implica-
ted in the cognition of sensory information. In particu-
lar, several studies have suggested that the main wave
(P300 or P3) may occur in a context of updating, a
process of updating of expectancies, or a process of
context closure [3,4]. In adults, Goodin et al. [13] and
several other authors [14-17] reported longer P3 latency
in dementia. However, the sensitivity in these studies
was extremely variable, and many factors, such as or-
ganic factors of varying etiologies, could have inter-
fered with the results [18]. In children, Rothenberger
and Baschek [11] observed longer P3 latencies in a
group of mentally retarded children. The difference be-
tween this group and a group of normal children was
statistically significant at the 2.5% level. A second
study in children [12] revealed a tendency for longer
latencies with decreasing IQ, but latencies varied signif-
icantly with child IQ only at Pz and for non-targets.
Nevertheless, organic factors could have also influ-
enced the results in these studies.
The purpose of the present study was to test the hy-
pothesis of a relationship between cognitive ability and
long latency event-related components, in gifted chil-
dren rather than in mentally retarded subjects. Stimu-
lus-linked N I component latency was assessed in order
to exclude its possible influence on P3 latency.
All studies to date on long latency event-related po-
tentials in normal children [5 10] have revealed a sig-
nificant relationship between P3 latency and age. How-
ever, the results concerning the rate of decrease were
varied, ranging from 3.6 ms/year [8] to 19.3 ms/year
[10] at Pz. This variation may be due to differences in
task difficulty or in populations. For instance, the num-
ber of young children (under 10-years old) was often
rather small and the greatest rate of decrease was ob-
] 76
served in the study which included the most such young
children [10]. As far as amplitude was concerned,
Goodin et al. [5] reported an increase with increasing
age, although the correlation was not statistically sig-
nificant. The few investigators who have studied ma-
turation of the stimulus-evoked N I component with
age [5, 9] failed to find a significant correlation.
The present study demonstrated a decrease in audi-
tory-evoked P3 latency with increasing age in the en-
tire population (control and gifted children), but the
rate of decrease (37.7 ms/year at Pz) was greater than
previously reported. We also found a significant de-
crease in N1 latency with increasing age. The greater
rate of decrease in P3 latency may be due to the fact
that our study included more young children ( < 1 0
years) than earlier studies, a possibility suggested by
Pearce et al. [10]. Moreover, this study also included
younger children (starting at 3.9-years old). The signifi-
cant decrease in N I latency with age might also be due
to these two facts. Our results concerning amplitude
were similar to those of earlier studies. No significant
correlation was observed.
While the size of the gifted children population was
comparatively srnall, owing to difficulties in recruit-
ment, the group was homogeneous. All the gifted chil-
dren were under 8 years of age and they had all been
tested before the age of 6.
The two populations were not age matched. That part
of variance related to age was therefore extracted by
comparing the deviations between the measured laten-
cies and the theoretical latencies (calculated from the
age of each child and the linear regression for the en-
tire population). N1 latency deviations were not statis-
tically different, whereas P3 latency deviations were
statistically different (P<0.05) at Cz. The mean devia-
tions were negative in gifted children and positive in the
control group, meaning that latencies were significantly
shorter in the gifted children than in the control group
at Cz. When rater-peak interval N I P3 was compared,
deviations became statistically different at all three re-
cording sites (P<0.01 at Fz and Cz, P < 0 . 0 2 at Pz).
The mean deviations were negative in gifted children
and positive m the control group, meaning that IPI
N1 P3 were significantly shorter in the gifted children
at all three sites. These results suggest that passage
from N I to P3 might occur faster in gifted children.
Moreover, since IPI N1 P3 decreases with increasing
age, mental precocity in the gifted children might re-
flect an electrophysiological precocity concerning the
interval, NI P3.
Although there was no significant difference in P3
amplitude, the mean values were higher in the gifted
Brain & Development, Vol 15, No 3, 1993
children. This could be due to the fact that amplitude
and latency are generally linked parameters: greater
amplitudes correspond to shorter latencies [19].
The present study concerning gifted children seems to
support the hypothesis of a relationship between cogni-
tive ability and P3 latency in children. P3 latency might
be related to mental maturation, but further studies with
a greater number of subjects are required to identify any
relationship between IQ and P3 latency in children.
REFERENCES
1. Sutton S, Braben M, Zubin J, John ER. Evoked potentials corre-
lates of stimulus uncertainty. Science 1965; 150: 1187-8.
2. Donchin E. Surprise!... Surprise? Psychophysiology 1981; 18:493
513.
3. Verleger R. Event-related potentials and cognition: a critique of
the context updating hypothesis and an alternative interpretation
of P3. Behav Brain Sci 1988; 11:343 56.
4. Donchin E, Coles MGH. Precommentary: is P300 component a
manifestation of context updating? Behav Brain Sci 1988; i!:
357 74.
5. Goodin DS, Squires KC, Henderson BH, Starr A. Age-related
variations in evoked potentials to auditory stimuli in normal hu-
man subjects. Electroencephalogr Clin Neurophysiol 1978: 44:
447 58.
6. Courchesne E. Neurophysiological correlates of cognitive devel-
opment: changes in long-latency event-related potentials from
childhood to adulthood. Electroencephalogr Clin Neurophysiol
1978; 45:468 82.
7. Polich J, Howard L, Starr A. Effects of age on the P300 compo-
nent of the event-related potentials from auditory stimuli: peak
definition, variation and measurement. J Gerontol 1985; 40:721
6.
8. Finley WW, Faux SF, Hutcheson J, Amstutz. Long-latency
event-related potentials in evaluation of cognitive function in
children. Neurology 1985; 35:323 7.
Brain & Development, Vol 15, No 3, 1993
9. Martin L, Barajas J J, Fernandez R, Tortes E. Auditory event-
related potentials in well-characterized groups of children. Elec-
troencephalogr Clin Neurophysiol 1988; 71: 375-81.
10. Pearce JW, Crowell DH, Tokioka A, Pachero GP. Childhood
developmental changes in the auditory P300. J Child Neurol
1989: 4: 100-6.
11. Rothenberger A, Baschek V. Children with different cognitive
abilities. In: Rothenberger A, ed. Event-related potentials in
children. Amsterdam: Elsevier 1982:317--24.
12. Baumann S, Otto D, Robinson G, Schroeder S, Barton C. The
relationship of late positive ERPs, age, intelligence and lead ab-
sorption in socioeconomically disadvantaged children. In: John-
son Jr R, Rohrbaugh JW, Parasuraman R, eds. Current trends in
ERP research (Suppl No 40 to Electroencephalogr Clin Neurophy-
siol). Amsterdam: Elsevier, 1987:617 23.
13. Goodin DS, Squires KC, Starr A. Long latency event-related
components of the auditory evoked potentials in dementia.
Brain 1978; 101: 635~48.
14. Squires KC, Chippendale TJ, Wrege KS, Goodin DS, Starr A.
Electrophysiological assessment of mental function in aging and
dementia. In: Poon L, ed. Aging in the 1980s. Washington DC:
American Psychological Association, 1980: 125-34.
15. Pfefferbaum A, Ford JM, Wenegrat BG, Roth WT, Kopell BS.
Clinical application of the P3 component of event-related poten-
tials. II. Dementia, depression and schizophrenia. Electroence-
phalogr Clin Neurophysiol 1984; 59: 104-24.
16. Gordon E, Kraiuhin C, Harris A, Meares R, Howson A. The
differential diagnosis of dementia using P300 latency. Biol
Psychiatry 1986; 21: 1123-32.
17. Patterson JV, Michalewski H J, Starr A. Latency variability of the
components of auditory event-related potentials to infrequent
stimuli in aging, Alzheimer-type dementia, and depression. Elec-
troencephalogr Clin Neurophysiol 1988; 71: 45~60.
18. Pfefferbaum A, Ford JM, Kraemer HC. Clinical utility of long
latency 'cognitive' event-related potentials (P3): the cons. Electro-
encephalogr Clin Neurophysiol 1990; 76: 6-12.
19. Polich J. Normal variation of P300 from auditory stimuli. Elec-
troencephalogr Clin Neurophysiol 1986; 65: 23640.
177