1997 SPR Meeting Poster Abstracts: M - Z

The emotional brain: Subdural recordings of ERPs to affective pictures
Jochen Machetanz, Almut Weike, Heino Mohrmann, Babette Spelsberg, Uwe Runge, Michael Gaab, & Alfons Hamm
University of Greifswald
The neurophysiology of emotional processes in animals is becoming
increasingly well understood. Subcortical structures form the key
site within the aversive motivational system. There is recent evidence
showing that picture stimuli can evoke analogues responses in humans
encouraging a comparative approach to elucidate the neurophysiology
of emotions. In the present experiment, cortical responses to emotion
eliciting picture stimuli were recorded subdurally, thus allowing
a better localization of the electrical dipoles that drive evoked
potentials. Subjects were hospitalized patients with implanted sub-dural
electrode grids, being evaluated for surgical treatment of severe
epilepsy. Patients viewed a series of 60 color slides (6 s exposure;
selected from the International Affective Picture System) including
15 pleasant, sexually oriented pictures, 10 pleasant control pictures
(flowers and babies), 10 neutral slides, 15 unpleasant pictures of
violence and aggression, and 10 unpleasant control stimuli (grief
and disgust evoking scenes). Event related potentials were recorded
subdurally from 32 different electrodes. Replicating previous results,
the slides elicited event related potentials and succeeding cortical
slow waves which varied substantially for the different emotional
contents. Preliminary results indicate that the cortical response
waveform to the slide was characterized by a marked late negative
voltage change which was substantially larger for emotional compared
to neutral materials. Interestingly, the largest negative voltage
change was observed for highly arousing erotic pictures. Moreover,
this voltage shift began 200-300 ms after stimulus onset and was sustained
throughout the entire viewing period.

Differentiation of cardiovascular responses to experimental pain or mental stress in humans
Walter Magerl, Joerg Spiegel, & Rolf-Detlef Treede
Johannes-Gutenberg University
Autonomic reflexes to acute experimental pain (sustained cutaneous
vasoconstriction, but only minor cardiac responses) differ from responses
to non-somatosensory stimuli (Magerl et al., Clin. Autonom. Res.
(1995) 346). Somatotopic organization of vasoconstrictor reflexes
induced by noxious stimuli in animals and humans suggest a major contribution
of spinal reflex circuitry (Magerl et al., J. Auton. Nerv. Syst. 57
(1996) 63-72; Sato. et al., Rev. Physiol. Biochem. Pharmacol., 130
(1997) 1-328). Here we investigate cardiac responses and vascular
responses in cutaneous and deep tissue to distinguish pain-related
reflexes from stress responses and to elucidate whether pain-related
autonomic responses fulfill suggested criteria of orienting vs. defense
reflexes (cf. Turpin, Psychophysiology 23 (1986) 1-14). Twelve subjects
were exposed to the stress of mental arithmetic (backward subtraction)
and to sustained painful heat stimuli. Cardiovascular responses were
quantified by heart rate, skin blood flow (laser Doppler flowmetry)
and muscle blood flow (near infrared spectroscopy). Both, stress and
pain caused a strong vasoconstriction in thumb glabrous skin. Vasoconstriction
adapted rapidly under stress, but was sustained throughout the heat
pain stimulation. In forearm muscles, stress caused vasodilatation
and enhanced blood flow, whereas muscle blood flow was only marginally
affected by pain stimuli. Heart rate was constantly elevated during
the stress task, but was slightly lowered during painful heat. In
conclusion, human autonomic reflexes to sustained pain differ markedly
from sustained stress in cardiac and peripheral vascular targets.
Moreover, the autonomic response pattern suggests the elicitation
of orienting, but not defense reflexes. (Supported by the Robert Mueller

Is a frontal positive slow wave in the ERP a consequence of eye movements?
Stefanie Maier1, Gabriele Becker1, Oliver Diedrich2, Ewald Naumann1, & Dieter Bartussek1
1University of Trier, 2University of Tuebingen
In a series of experiments presenting IAPS slides, a frontal positive
slow wave was elicited in the ERP. The functional significance of
this slow wave is still controversial. An experiment was designed
in which the amount of eye movement was varied systematically to exclude
that a frontal positive slow wave is an artifact of eye movements
during the presentation of visual stimuli. Different 6 x 6-character
matrices were generated in which predefined targets (a 2 x 2-character
matrix) had to be detected. The main hypotheses stated that no systematic
influence of eye movements on the frontal electrodes should be observable.
ERPs were recorded from 60 subjects. Two independent variables were
used. Amount of eye movements (less, much) was varied through the
position of target matrices within the background matrices. Task difficulty
(easy, difficult) was varied by use of different characters for the
background matrices. Four experimental groups were examined. Each
subject watched 75 matrices (presented for 6-7.5 sec) in succession
and had to decide whether no, one, or two target matrices were embedded
in the back-ground matrices. EEG was recorded from F3, C3, P3, Fz,
Cz, Pz, F4, C4, and P4 from 460 ms before till 4540 ms after stimulus
onset. The main hypotheses is confirmed: In the time range betwee
1000 and 3500 ms no influence of amount of eye movements on frontal
ERPs is found. However, significant interactions between frontality
and task difficulty are observable: The difficult task condition results
in more positive amplitudes at frontal locations and more negative
amplitudes at parietal locations as compared to the easy task condition.
The results will be discussed with regard to usability of visual stimuli
for ERP experiments. Supported by Deutsche Forschungsgemeinschaft
(grand #BA926/4-4).

Assessment of beliefs about heart rate and the relationship between heart rate beliefs and heartbeat counting
Jennifer Mailloux, Ronnie Dubinski, & Jasper Brener
State University of New York at Stony Brook
Performance on Schandry's heartbeat counting task may be influenced
by both beliefs about heart rate and by the processing of cardiac
sensations. This experiment sought to determine (1) whether subjects
possess accurate beliefs about heart rate, and (2) the extent to which
(i) beliefs about heart rate and (ii) actual heart rate influence
heartbeat counting. Subjects' beliefs about the effects of posture
and exercise upon heart rate were assessed by verbal report and also
by requiring them to adjust an electronic metronome to oscillate at
the believed heart rate. They were then submitted to Schandry's heartbeat
counting task during which actual heart rate was manipulated by changes
in posture and exercise. Thereafter, accuracy of heartbeat detection
was assessed by the Method of Constant Stimuli (MCS) in which subjects
judged the simultaneity of heartbeats and tones presented at various
intervals after the R- wave. Verbally-expressed beliefs were found
to underestimate actual heart rates but were similar to counted heart
rates. However, heart rate beliefs expressed by adjusting a metronome
were similar to actual heart rates. Among good heartbeat detectors
on the MCS task, the number of heartbeats counted in the Schandry
task was significantly correlated with the actual number of heartbeats
recorded, but not with heart rate beliefs. However, among poor heartbeat
detectors, the number of heartbeats counted was significantly correlated
with beliefs but not with actual heart rates. Hence, performance on
Schandry's heartbeat counting task is determined by cardiac sensations
in good heartbeat detectors and by beliefs in poor heartbeat detectors.

Independent component analysis of visual evoked responses during selective visual attention
Scott Makeig1,3, Marissa N. Westerfield1,2, Jeanne Townsend1,2, James W. Covington2, Eric Courchesne1,2, & Terrence J. Sejnowski1,4
1University of California-San Diego, 2Children's Hospital Research Center, 3Naval Health Research Center, 4Salk Institute-La Jolla
Independent Component Analysis (ICA) is a new signal processing techniqu
for decomposing spontaneous or evoked electrophysiological data into
spatially fixed and temporally independent components. ICA allows
comparison of component amplitudes and time courses across related
conditions. Applied simultaneously to target and nontarget responses
in 30 conditions of a visual selective attention experiment (see Westerfield
et al., this session), ICA derived at least four components of the
early visual evoked response which were differently amplitude-modulated
by spatial location and attention without effects on component latency,
but were not affected by the target/nontarget distinction. Other components
accounted for data artifacts in single conditions. Later components
common to several conditions were sensitive to both spatial attention
and target feature. ICA allows quantitative comparison of objectively-derived
and temporally-sparse ERP components and subcomponents across 30 or
more stimulus or task conditions. (Supported by the Office of Naval
Research, Howard Hughes Medical Institute, NINDS NS34155 and NIMH

Psychophysiological consequences of anger ideals
Roberta A. Mancuso & Barbara L. Fredrickson
University of Michigan
Anger ideals are socialized beliefs that people hold about their experience
of anger. Individuals with negative anger ideals are those who believe
that the experience of anger is not appropriate, while people with
positive anger ideals believe that the experience of anger is appropriate.
The present study explored the psychophysiological consequences of
anger ideals. It was hypothesized that people with negative anger
ideals would show significantly different psychophysiological responses
to anger when compared to individuals with positive anger ideals.
Participants were 45 female college students who relived an anger
memory in the laboratory. Anger ideals were measured using a combination
of self-report items, where a higher score indicated negative anger
ideals, and low scores indicated positive anger ideals. Data on participants'
emotional responses were obtained using physiological measures, such
as systolic and diastolic blood pressure, and retrospective self-report
questionnaires. Analysis of the self-report data indicated that there
were no mean differences in the level of anger experienced by people
during the relived anger task. However, individuals with negative
anger ideals showed significantly higher levels of systolic blood
pressure during the relived task (F(2,42) = 29.36, p<.0001) compared
to those with positive anger ideals. Our findings suggest that there
may be negative ramifications for individuals who hold negative anger
ideals. Given that the cardiovascular reactivity associated with negative
emotions has been implicated in the development of cardiovascular
disease, people with negative anger ideals may suffer negative consequences
in their emotional experiences as well as their physical health

Statistical detection tests for signals extracted by Woody filter
Walter S. Marcantoni & Andre Achim
University of Quebec-Montreal
The Woody filter (WF) is an iterative cross-correlational technique
to compound variable onset event-related potentials (ERPs) into a
representative response waveshape. This procedure, however, can produce
plausible waveforms without ERP response in the EEG. We describe and
compare two signal detection methods for WF responses. Both require
that noise remains statistically equivalent under polarity reversal
(excludes background activity having definite polarity, like mu rhythms).
Method-A ranks the variance explained by the template with all trials
in their original polarity among the variances obtained similarly
for all possible subsets of trials inverted in polarity. In Method-B,
individual trials are aligned with the current template at the maximum
absolute value of correlation rather than the maximum correlation;
signal is detected when the signed correlations are asymmetrically
distributed about zero (Wilcoxon signed rank test). Synthetic noise
and signal corresponding to a 32 point steady-state ERP were generated
by a first-order auto- regressive function, with parameter (r) .95
or .80. Nine trials per test allowed permuting all distinct subsets,
for Method-A, rather than sampling them. Neither method exceeded the
nominal false alarm rate in 1000 tests. Three signal intensity levels
were chosen by pre-testing for partial success by both methods. Method-A
was systematically superior to Method-B (by about 15% near 50% detection),
and both operated better at r=.80 than r=.95. Approximately 70% detection
was obtained with Method-A when the signal to noise intensity level
in individual trials was 1.33 at r=.95 and 1.11 at r=.80. Unless processing
time is crucial, method-A is recommended.

The effects of task difficulty on readiness potentials preceding vocalization
Hiroaki Masaki1, Noriyoshi Takasawa2, & Katuo Yamazaki1
1Waseda University, 2National Research Institute of Police Science
The present study was designed to investigate the effects of task
difficulty on the readiness potentials (RPs) preceding vocalization.
The RPs before a nonspeech task (lip rounding), saying a voiced sound
of /gu/ (easy task), and saying a nasal voiced sound of /gu/ specific
to Japanese (difficult task) were recorded from seven right-handed
university students who had already acquired a high skill of pronunciation
in order to be an announcer. They performed 70 trials in each task
arranged with a counterbalanced order. Bipolar EMG recording was made
via Ag/AgCl surface disk electrodes placed over the orbicularis oris
muscle. EEG averaging was triggered by the rectified EMG signal. Althoug
a 100-ms averaged amplitude measure started 900 ms before the EMG
onset did not differ among tasks, significant task effects were observed
in a 100-ms averaged amplitude measure started 200 ms before the EMG.
The later 100-ms averaged amplitude measure was significantly larger
in the difficult task than in the easy and lip rounding. In addition,
the RP amplitude at EMG onset became larger as the task shifted from
lip rounding to easy to difficult. The lateralized readiness potentials
(LRPs) revealed no lateralized activity indicative of left hemispheric
dominance. Results suggest that the late RP component recorded at
Cz is susceptible to changes in task difficulty.

Motor inhibition on the absence of go stimulus in the precuing reaction time task
Hidehiko Matsumoto, Kiichi Kiriyama, & Takashi Morotomi
Hokkaido University
Motor preparation and inhibition were investigated under S1-S2 paradigm,
using the lateralized readiness potential (LRP) as a measure of motor
preparation. Although motor inhibition has been studied by the presentation
of nogo stimulus (Go/Nogo task), the absence of the go stimulus signaled
to inhibit a motor preparation in this study. After that, the subjects
must start preparing the responses again to the prolonged go stimulus.
The task was the precuing reaction time task with two ISIs (1500 or
500 ms). Under 1500ms/80% condition the probability of 1500 ms ISI
was 80 %, and under 500ms/80% condition the probability of 500ms ISI
was 80 %. RT and onset latency of the LRP for 500 ms ISI trial under
the 1500ms/80% condition were both delayed compared to those under
the 500ms/80% condition, reflecting that subjects well expected higher
probable ISI trials. Further, under the 500ms/80% condition the LRP
appeared to the 1500 ms ISI trials at the same latency range as to
the 500 ms ISI trials even though the subjects succeeded in inhibiting
motor responses. In addition, this increase of the LRP was observed
only for trials that was slower than mean RT but not for faster response
trials. These findings suggest that the inhibition of motor preparation,
which occurred in the absence of go stimulus, affects the following
motor response.

Fear or craving: Startle reflex modulation during viewing of food cues in deprived and non-deprived bulimics
Birgit Mauler1, Brunna Tuschen2, & Alfons Hamm3
1Christoph Dornier Foundation for Clinical Psychology, 2University of Marburg, 3University of Greifswald
This study was designed to test whether food cues act as elicitors
of craving in bulimics or alternatively are fear evoking. 32 clinically
diagnosed bulimics and 32 healthy controls served as subjects. Hal
the subjects in each group were instructed not to eat for 16 hours
prior to the experiment. Beta-hydroxic-butyric-acid was analyzed from
the serum to assess, whether subjects followed this instruction. After
taking of the blood sample, subjects viewed eight pleasant, neutral,
unpleasant and food related pictures, respectively. Acoustic startle
probes were administered during and between slide viewing. Afterwards
subjects could select from several food items from a buffet and eat
as much as they wanted. Then subjects viewed a second series of 32
slides which were selected from the same four affective categories
and were carefully matched according to the affective ratings. There
was a robust affect-startle effect for the standard affective categories
which did neither vary for groups nor for slide series. Moreover,
in control subjects startle response magnitudes were significantly
inhibited when elicited during viewing of food stimuli in non-deprived
subjects and slightly enhanced in deprived volunteers. Non-deprived
bulimics, however, showed a substantial potentiation of the startle
eyeblink during viewing of food cues. Deprived bulimics did not show
such startle facilitation. After the consumption of a small amount
of food these patients, however, also showed strong potentiation of
their startle responses during viewing of food stimuli. These data
suggest that food cues might act as fear stimuli in bulimics and that
this fear is reduced by food deprivation.

A P300 (P3) study of target detection, response inhibition, and novelty during auditory and visual sustained attention
Danielle C. McCabe, Janet L. Shucard, & David W. Shucard
State University of New York at Buffalo
Studies in our laboratory that combine electrophysiology with visual
and auditory continuous performance tasks (CPTs) have shown that P300
(P3) amplitude was affected by the different stimulus sequences or
conditions of the CPT. In the present study, a novel stimulus was
included in the CPT to evaluate the differing neurocognitive demands
of target detection, response inhibition and novelty. Parallel forms
of auditory and visual CPTs were presented in a counterbalanced order
to right handed males and females between 18-24 years of age. The
CPTs were equated for number of stimuli (n=450),type of stimuli (letters)
and interstimulus interval (1.7s). Event-related potentials were obtained
from 12 electrode sites. Here, data are reported only for the midline
scalp sites. The results showed similar condition-dependent P3 topographic
relationships between modalities for both target detection (Go) and
response inhibition (No-Go). The Target Detection condition (A-X)
showed greater posterior P3 amplitude, while the response inhibition
or No-Go condition (A- not-X) showed greater central and frontal P3
amplitude in both the auditory and visual modalities. The Novelty
Condition, however, showed a different pattern for visual and auditor
modalities, with greater posterior activation only during visual presentation.
Results are discussed in terms of both modality and condition effects
on P3, and the relationship among target detection, response inhibition
and novelty.

Component overlap and the N400: Methodological issues
Christina M. McCann & Jeffrey M. Clarke
University of North Texas-Denton
Studies involving event-related potentials (ERPs) may encounter problems
with component overlap, where one component reduces the amplitude
of another. For example, decision making has been shown to produce
a large P300, which, in turn, attenuates the N400. This effect should
be reduced by delaying the decision making process. A primed lexical
decision task (LDT: participants identified target stimuli as words
or non-words), and a primed delayed matching task (DMT: participants
identified a second target stimulus as either the same or different
from the first target), were used to investigate component overlap
between the N400 and the P300. The same 3-6 non-word and concrete
word stimuli were visually presented for both tasks and consisted
of prime word + target [80 each of: associated, unassociated, pronounceable
non-word (PNW), and non-pronounceable non-word (NPNW)] pairs. For
the LDT, P300 amplitudes were larger for associated and NPNW conditions,
due to increased ease of processing. An N400 was evoked by unassociated
and PNW conditions, and surprisingly by NPNW as well. P300 amplitudes
were greatly reduced and undifferentiated across all conditions for
the DMT, where task demands changed from immediate (as in the LDT)
to delayed decision making. N400 amplitudes were more negative for
the DMT across all conditions, and were largest for unassociated and
PNW conditions, as predicted. Thus, not only were the P300 and N400
dissociated by condition, but a task dependent pattern also emerged.
These results reflect changes in the N400 by a reduction of P300 amplitude,
a change in processing demands, or perhaps a combination of both.

Anger behavior predicts cardiovascular reactivity in old age but not in middle age
Loren M. McCarter & Robert W. Levenson
University of California-Berkeley
Anger behavior is an hypothesized mediator of the relationship between
trait hostility and cardiovascular health (e.g., hostile individuals
frequently become angry, which increases cardiovascular reactivity
and blood turbulence, damages the endothelium at arterial bifurcations
and advances atherosclerotic plaque formation). Yet, few studies have
measured anger behavior directly (e.g., using behavioral coding of
facial and verbal expressions) while concomitantly measuring cardiovascular
reactivity. We decided to examine the relationship between anger behavior
and cardiovascular reactivity in married couples while they discussed
a problem area in their marriage. We included both middle-age (mean
age = 41.9; n=52) and elderly (mean age = 63.8; n=60) couples to examine
age-related changes in emotional behavior and the cardiovascular system.
Anger behavior (i.e., angry facial and verbal expressions) and cardiovascular
reactivity (i.e., change from baseline in heart rate, pulse transmission
time to the ear and finger, finger pulse amplitude, and finger temperature)
were monitored continuously. Results showed that, when averaged over
the 15 minute discussion period, the frequency of anger behaviors
was associated with increased cardiovascular reactivity for elderly
men (r=.32; p=.015) and women (r=.35; p=.010) but not middle-age men
(r=-.06; p=.695) and women (r=-.007; p=.963). The implications of
these results are that negative emotions could be particularly toxic
for elderly individuals by virtue of their close links with cardiovascular
reactivity, which may reflect age-related changes in the functional
capacity of hemodynamic regulatory mechanisms. Our findings may help
to explain why elderly individuals report greater investment in regulating
their negative emotions than do younger individuals.

Sugar and spice versus snails and puppy dog tails: Gender and temperament in children's emotional responding
Mark H. McManis1,2, Bruce N. Cuthbert2, Margaret M. Bradley2, & Peter J. Lang2
1Harvard University, 2University of Florida
Data from our laboratory has shown that a child's pattern of responses
to affective pictures varies with the age and gender of the child.
The present study explores these relationships more thoroughly, adding
psychophysiological and self-report measures of emotional reactivity
and temperament in children. These measures included resting EEG
(frontal activation asymmetry), heart rate, and self-report measures
of fear and temperament. Children (7 - 10 years old) filled out a
fear survey while parents completed a temperament questionnaire.
Sensors were then placed for recording EEG (F3, Fz, F4, Cz, and Pz),
eye blink, corrugator EMG, heart rate, and skin conductance. Resting
EEG and heart rate were collected during 4 one-minute baseline periods.
Children then viewed 36 pictures varying in affective content; an
acoustic startle probe was delivered during each 6 s picture-viewing
period. There was a significant interaction between gender and temperament
measures. For boys, but not girls, greater left frontal activation
was related to lower reported fears. For girls, but not boys, high
resting heart rate was related to greater reported fears. The relationship
between temperament and startle modulation was also mediated by gende
and age. Unlike girls and adults, boys showed, overall, inhibition
of the startle reflex while viewing unpleasant pictures; however,
greater left hemisphere activation was related to larger startles
during unpleasant pictures and smaller startles during pleasant pictures
in boys. This relation was stronger for younger boys. Girls showed
a weaker relationship. The complexity of the relationship between
emotional responding, temperament, gender and development is discussed.

An event-related brain potential study of PTSD in abused children
W. Brian McPherson, Peggy T. Ackerman, Joe E.O. Newton, and Roscoe A. Dykman
University of Arkansas for Medical Sciences & Arkansas Children's Hospital
Event-related brain potentials (ERPs) were used to assess cortical
reactivity to weak and strong auditory stimuli in children, 8 to 12
years of age, with and without post-traumatic stress disorder (PTSD).
Subjects included 166 physically and/or sexually abused children recruited
from both inpatient and outpatient clinics at Arkansas Children's
Hospital, and 26 non-abused children also recruite d at Arkansas Children's
Hospital. All children were administered the Diagnostic Interview
for Children and Adolescents (Reich, 1990), and the diagnosis of PTSD
was based upon DSM-IIIR criteria. ERPs were recorded from 5 active
scalp sites, including F3, F4, P3, P4, and Cz. Acoustic stimuli consisted
of 240 1KHz tones lasting 500 ms each and occurring at 4 different
intensities (60 each @ 65dB, 80dB, 95dB, &102dB) in a pseudo random
order. Artifact from eye movements were monitored and trials with
eye movements registering over 100 uv were not included in the ERP
averages. Results revealed that the difference between the P2 and
N2 peaks taken across increasing tone intensity differentiated the
children with PTSD from those without PTSD (both abused and non-abused).
An increase in tone intensity normally precipitates an increase in
the P2-N2 difference. All groups showed such an increase, but the
subjects with a PTSD diagnosis showed a significantly steeper incr
ease than those groups without PTSD. Non-abused children and abused
children without PTSD did not differ on this measure. These results
contrast with ERP records of combat veterans with PTSD (Paige, Reed,
Allen & Newton, 1990). Adults with PTSD have shown a smaller than
normal increase in the P2 component with increasing tone intensity
in a similar ERP paradigm. Together these results suggest that the
psychophysiological correlates of PTSD differ between children and

On the nature of recognition memory deficits in patients with cerebral hypoxia
A. Mecklinger, G. Matthes-von Cramom, & D.Y. von Cramon
Max Planck Institute of Cognitive Neuroscience
Hypoxic brain injury due to cardiac arrest may lead to posthypoxic
amnesia. Although certain brain structures like the CA-1 region in
the hippocampus have been shown to be predomimantly vulnerable to
acute hypoxia, little is known about the functional and neurological
aspects of impaired memory functions in chronic hypoxic patients.
This study aims at identifying the subprocesses underlying recognition
memory by means of a combined analysis of event-related potential
(ERP) and performance data. Six chronic hypoxic patients (time since
lesion > 3 months) as well as four groups of six age and gender matched
controls each, performed two versions of a recognition memory tasks
requiring recognition judgments for either object forms or spatial
locations as well as a visual oddball task, imposing negligible memory
requirements. The oddball task evoked P300 components, highly similar
in timing, amplitude and scalp topography for all patients and controls.
For the controls reliable old/new effects (i.e., larger ERP waveforms
for previously studied than for unstudied items) were found between
300 and 600 ms in both recognition tasks. In contrast, for all six
patients investigated these old/new effects were absent or even inverted
in polarity while their recognition performance was well above chance
(about 70% correct judgments). These results indicate that recognition
memory performance is not contingent upon intact ERP old/new effects.
Consistent with the view that old/new effect reflect the retrieval
of explicit memories, the results further suggest that recognition
memory in cerebral hypoxia is not associated with explicit memory
functions (i.e., conscious recollection) but rather relies on intact
implicit memory (i.e., judgment of an items familiarity.

An examination of affective modulation, psychopathy, and negative schizotypy in a non-incarcerated sample
Veronica Y. Mejia1, Eric J. Vanman2, Michael E. Dawson1, Adrian Raine1, & Todd Lencz3
1University of Southern California, 2Texas A & M University, 3The Long Island Jewish Center Hillside Hospital
Startle eyeblink measures were recorded while participants viewed
positive and negative slides from the International Affective Picture
System (IAPS). Slides were probed at 300 ms, 800 ms and 4500 ms with
a 104 dB white noise. Subjects were recruited from temporary employment
agencies located in the metropolitan Los Angeles area. Prior to participation,
these subjects were assessed for psychopathy with the Psychopathy
Checklist (PCL - Screening Version), had completed the Raine Schizotypal
Personality Questionnaire (SPQ), and had an MRI structural brain scan.
Participants were matched on the basis of their PCL Factor 2 scores
(measuring antisocial behavior) and then split into high and low PCL
Factor 1 (measuring emotional detachment) groups. Subjects classified
as high in antisocial behavior and low in emotional detachment showed
the expected significant emotional modulation at the .05 level whereas
subjects scoring high on both factors failed to show this effect
replicating Patrick's psychopath startle pattern (Patrick, Bradley,
& Lang, 1993, Journal of Abnormal Psychology, 82-92) with a non-incarcerated
sample. Additionally, at 300 ms, subjects high in antisociality but
low in emotional detachment displayed significantly more prepulse
inhibition than subjects scoring low on both factors. These subjects
were reexamined on the basis of their SPQ Flattened Affect (Factor
2) scores. In contrast to the pattern defined by the emotional detachment
factor of the PCL, subjects scoring high in affective flatness (SPQ
factor 2) displayed significant emotional modulation of startle at
the .05 level. Subjects scoring low on this factor failed to show
this effect.

P300 hemispheric differences from oddball, verbal, and spatial tasks
Ralf Mertens1 & John Polich2
1University of Arizona, 2The Scripps Research Institute
Event-related brain potentials (ERPs) were elicited in n=32 strongly
right-handed, male adults to determine the degree of P300 hemispheric
asymmetries from centrally presented stimuli. The paradigms included
auditory and visual oddballs; a verbal semantic memory access task
in which subjects categorized nouns as living or nonliving things;
a spatial judgment task in which subjects indicated whether the number
of black squares in a matrix was even or odd. Topographic distributions
for P300 component measures were obtained from 16 electrode sites.
As reported previously, P300 amplitude from the auditory and visual
oddball tasks was generally larger over the right hemisphere frontal/central
sites, with a tendency toward reverse asymmetries for the more parietal
electrodes for the visual task. P300 amplitude for the verbal semantic
task was consistently larger over the left compared to right hemispheres-especially
at the parietal electrode sites; the spatial judgment task produced
larger components over the right hemisphere-especially at the frontal
and central electrode sites. P300 latency tended to be shorter at
electrode sites where amplitude was large, but these effects were
not consistent or robust across task conditions. Taken together, the
findings indicate that P300 amplitude asymmetries can be obtained
even when stimuli are presented centrally (vis ¦ vis unilaterally)
for simple stimulus discrimination as well as hemispherically specialized

The effects of epinephrine administration on cardiovascular function
Elizabeth S. Mezzacappa1, Robert M. Kelsey2, & Edward S. Katkin2
1Columbia-Prebyterian Medical Center, 2State University of New York at Stony Brook
The effects of epinephrine administration on cardiovascular function
were examined in 26 male undergraduates who were given a subcutaneous
injection (.007cc/kg) of either 1:10,000 epinephrine hydrochlorid
or physiological saline. Impedance cardiography and continuous blood
pressure measures were recorded during a 2 minute pre-injection baseline
and in the post-injection period. The differences between ensemble
averaged values for the six 20-second epochs of the baseline and post-injection
periods were calculated for several cardiovascular measures. As compared
to saline controls, epinephrine was associated with the following
significant changes in cardiovascular function (all p<.005): (a) greater
shortening of heart period (eta=.61) and pre-ejection period (eta=.84);
(b) greater increases in cardiac output (eta=.81), stroke volume (eta=.80),
dZ/dt amplitude (eta=.81), and Heather Index (eta=.74); (c) greater
decreases in total peripheral resistance (eta=.75). The effects of
epinephrine on pre-ejection period were due entirely to shortening
of the R to B interval (eta=.84). There was no significant impact
of epinephrine on LVET, or on the intervals comprising it (B to Z
and Z to X), nor was there a significant effect on mean arterial blood
pressure. The results are consistent with the expected effects of
epinephrine on cardiovascular function, and indicate that the R to
B interval may be the best index of sympathetic-adrenomedullary activation
during psychological or high physical stress.

Event-related potentials and pain reports to painful stimuli during hypnosis and disattention
Wolfgang H. R. Miltner1 & Christoph Braun2
1Friedrich Schiller University-Jena, 2Eberhard Karls University-Tuebingen
Hypnosis and disattention represent powerful methods for acute pain
control. Some studies have reported reduced nociceptive reflexes and
pain reports during hypnosis and disattention as compared to baseline
conditions. Similar results were obtained for the N1 and P2 components
of the somatosensory event-related potential (SEP) with smaller amplitudes
during both conditions as compared to baseline. However, results of
experiments on hypnosis did not reveal similar consistent results.
For hypnosis, N1/P2 amplitudes were reported to be smaller than during
baseline in some studies whereas other studies could not find such
differences. Therefore, our present study compared SEP-amplitudes/latencies
of the N1, P2, and P3 components of the SEP and pain reports (PR)
to painful stimuli during baseline, hypnosis and a disattention paradigm
in 13 subjects. Subjects were preselected for successful hypnotic
pain control while exposed to a cold pressor test prior to the experiment.
EEG was recorded from 21 electrodes (10-20 system) referenced to linked
earlobes. Additonally, PRs were recorded for each single stimulus.
Averaged SEPs and PRs were determined for baseline condition, wh ile
subjects were under hypnosis or instructed to dissattend from painful
stimulation. Results show significant smaller PRs during hypnosis
and disattention as compared to baseline with no difference between
non-baseline conditions. N1, P2, and P3 amplitudes at frontal, central,
and parietal electrodes were reduced while subjects disattended t
pain stimulation as compared to baseline. However, during hypnosis
amplitudes and latencies remained unchanged compared to baseline.
Results indicate that different brain mechanisms are involved in hypnosis
and attention/disattention while processing painful stimuli.

Task dependent effects of hypoglycemia on the information processing of type-I diabetic subjects
Wolfgang H. R. Miltner1, Christoph Braun2, & Michael Mann2
1Friedrich Schiller University-Jena, 2Eberhard Karls University-Tuebingen
Studies on hypoglycemia in type-I diabetes subjects reported impaired
cognitive and motor functions during hypoglycemia while others did
not. The present study investigated changes of cognitive and motor
functions at different levels of hypoglycemia by means of event-related
potentials (ERPs) and behavioral measures. In 10 patients, blood glucose
levels were fixed at 100 (level 1), 60 (2), 40 (3), 30 (4) and 100
mg/dl (5) for minimum 30 minutes each using an artificial pancreas.
6 diabetes type-I subjects served as controls with blood glucose levels
maintained constantly at 100 mg/dl. Brain-stem auditory evoked potentials
(BAEPs) to clicks were recorded during blood glucose levels 1 to 5.
Motor functions were tested by a Go/NoGo task at levels 1 to 5 and
visual information processing and choice reaction motor functions
were tested by a Continuous Performance Task (CPT) at levels 1, 3,
and 5. During all tests ERPs were recorded from FZ, CZ, PZ, F3, F4,
C3, C4, P3, P4 and behavioral data on errors were collected. During
hypoglycemia patients and controls did not differ in early and midlatency
BAEPs. No ERP difference between groups was seen at all blood glucose
levels during the Go/NoGo-task. Significant differences were obtained
for the P300 during the CPT: control subjects showed larger P300 amplitudes
while processing target stimuli than patients. Results confirm that
hypoglycemia differentially affects cognitive and motor functions:
In reaction time task processing of auditory stimuli and motor functions
are not affected by hypoglycemia whereas processing o f complex cognitive
stimuli becomes impaired significantly.

The source of the magnetic equivalent of the error-related negativity
Wolfgang H. R. Miltner1, Ulrike Lemke1, Thomas Weiss1, Clay Holroyd2, Marten K. Scheffers2, & Michael G. H. Coles2
1Friedrich Schiller University-Jena, 2University of Illinois at Urbana-Champaign
When subjects make errors in a variety of tasks, a negative event-related
brain potential, the Error-Related Negativity (ERN), can be recorded
from electrodes placed on the scalp. There is now considerable evidence
to suggest that this potential reflects the activity of an error processin
system implemented in frontal brain regions. In the present experiment,
we investigated the neural source of the magnetic equivalent of the
ERN using a magnetoencephalograph (MEG) with 31 channels. Six student
volunteers were required to perform a Go-NoGo task, in which they
had to make a speeded response with their right finger to one tone
but to withhold the response to another tone. Tones were 1000 Hz and
2000 Hz (duration 50 ms) and the inter-tone interval was 700ms. Concurrent
EEG recordings revealed the presence of an ERN when subjects responded
(incorrectly) to the NoGo tone. For each subject, dipole source analyses
of the MEG fields at the time of the ERN peak were conducted using
a realistic head model based on individual MRI recordings. The dipole
solutions were superimposed on the MRI recordings. For the four subjects,
for whom reliable sources could be obtained, the analyses suggested
a frontal source in the region of the anterior cingulate. These MEG
results are consistent with previous source analyses based on EEG
recordings and support our earlier claims that the frontal brain contains
a system for handling errors.

Electrophysiological measures of masked and unmasked repetition priming
Maya Misra, Tatiana Sitnikova, & Phillip J. Holcomb
Tufts University
Misra and Holcomb (Cognitive Neuroscience Society, 1997) presented
participants with prime words shown for 50 ms then masked for 250
ms and followed by targets words presented in the clear. Event-related
potentials (ERPs) to the targets revealed evidence for short interval
masked repetition priming. In a subsequent recognition test, however,
participants were at chance levels in recognition of words that had
been previously masked, indicating that the original processing of
these words was at a shallow or perhaps even a "preconscious" level.
The goal of the current experiment was to see if recognition performance
would improve and ERP priming effects would change in size or form
when the same paradigm was run without masking. Nine participants
viewed words presented sequentially on a computer monitor. Prime words
were displayed for 50 ms and were followed 450 ms later by a target
word (300 ms) that was either the same word (repetition) or a different
word. Participants monitored this stream of words for occasional occurrences
of non-target animal names to which they pressed a button. A subsequent
recognition memory test was given to determine the percentage of primes
and targets remembered. Results suggest that performance on the memory
test was greatly improved for the unmasked primes in this study compared
to the masked primes from our earlier study. In addition, a robust
short interval ERP priming effect was also present which was at least
as large as the effect noted in the masked condition. (Please note
that P. Holcomb is a member of SPR.)

Somatotopic organization of the somatosensory cortex in subjects with congenital limb atrophy, traumatic amputees with phantom limb pain, and healthy controls
Pedro Montoya1, Karin Ritter1, Ellena Huse1, Wolfgang Larbig1, Werner Lutzenberger1, Herta Flor2, & Niels Birbaumer1
1 University of Tuebingen, 2 Humboldt University
Animal and human studies have shown that nerve deafferentiation following
amputation leads to functional reorganization of the primary somatosensory
cortex. The objective of the present study was to examine the somatopic
organization of several body parts in subjects with congenital limb
atrophy, traumatic amputees suffering from phantom limb pain and healthy
controls. The hypothesis was tested that cortical reorganization in
upper-limb amputees is related to phantom limb pain. Somatosensory
evoked potentials were recorded from 60 scalp EEG electrodes during
left- and right-sided tactile stimulation of the lower lip, the thumb
and the first toe. Amputees with phantom limb pain showed a significant
shift (about 24 mm) of the cortical representation of the lip contralateral
to the amputation side towards the ,phantom hand area", resulting
in an asymmetrical representation of the lower lip over the somatosensory
cortex. In contrast, no hemispheric differences were found in the
cortical maps of the first toe of pain amputees. In subjects with
congenital limb atrophy and in healthy controls, no hemispheric difference
was found in their cortical maps. These findings are in agreement
with previous results demonstrating that the cortical reorganization
of the primary somatosensory cortex is related to the presence of
phantom limb pain. Furthermore, our data suggest that these plastic
changes are specifically confined to the cortical representation of
the face area. This work was supported by the Deutsche Forschungsgemeinschaft
(Research group ,Clinical Psychophysiology of pain", Bi 195/24).

Emotional modulation of startle in schizophrenia
Sarah E. Morris, Cindy M. Yee, & Keith H. Nuechterlein
University of California-Los Angeles
Abnormalities in emotional responsiveness have long been considered
key features of schizophrenia. It remains unclear, however, whether
these deficits are best characterized as hypersensitivity to negatively
valenced material, hyposensitivity to pleasantly valenced material,
deficits in the expression of affect, and/or abnormalities in the
experience of emotions. The present study used a startle blink paradigm
to investigate whether schizophrenic patients demonstrate the same
patterns of affective modulation as non-psychiatric subjects. It has
been reliably established that in normal subjects the startle blin
reflex is modulated by the affective valence of the foreground stimuli.
Thus, blink amplitude is potentiated when elicited while viewing unpleasant
images and diminished when elicited during presentation of pleasant
images, relative to neutral conditions (e.g., Bradley et al., 1993,
Psychophysiology, 30, 541-545). In the current study, startle probes
are presented in the prepulse (300 and 800 ms), affect (1300 and 3800
ms) and image offset (6300 and 9800 ms) time periods relative to image
onset. Startle blink amplitude of schizophrenic patients are contrasted
for pleasant, unpleasant, and neutral images. Preliminary analyses
of data from chronic schizophrenic subjects indicates that these patients
show patterns of affective modulation of the startle blink reflex
that are similar to those reported in non-patient populations. The
relationship of affective modulation of startle blink amplitude to
level of negative symptoms within schizophrenic patients will be presented.
Heart rate and skin conductance responses also will be examined.

Auditory oddball paradigm developmental and migraine specific results
Bernhard Mueller1 & Gudrun Sartory2
1University Clinics Essen, 2University of Wuppertal
Migraine is characterized by an increased CNV which may be due to
deviant information processing or motor organization. Children suffering
from migraine (n=30, age 11.6) were compared to healthy children (n=15,
age 12.1) and adults (n=20, age 26.4), with regard to ERP-components
in an auditory oddball paradigm. The standard stimulus (70%) consisted
of a tone of 1200 Hz; the deviant stimulus (15%) was a tone of 400
Hz and required a button press (Go), the other deviant stimulus (NoGo)
was of varying frequencies. Migraine children showed an increased
N1-latency compared with the other groups. Both groups of children
had increased N2-latencies compared with adults. Children's amplitudes
were larger with regard to N2 in the Go condition and smaller with
regard to P3a at C3 and C4 during the NoGo condition and again larger
with regard to P3b in the Go condition at Pz and P3. Slow negative
wave (SNW) amplitudes at frontal sites were also larger in children
in the Go and more so in the NoGo condition. Migraine affects oddball
ERPs during the early stages of information processing. Developmental
effects were most prominent with regard to N2 and SNW at frontal sites
in the NoGo condition and in the parietal P3b in the Go condition.

Dipole modeling of the neural generators of the steady-state visual evoked magnetic field
M. M. Mueller1, W. Teder2, & S. A. Hillyard2
1University of Konstanz, 2University of California
Steady-state visual evoked magnetic fields (SSVEFs) were recorded
in response to an either 6.0, 11.9, or 15.2 Hz flickering light sourc
using a 37-channel magnetometer. The SSVEFs showed a dipolar topography
over the posterior scalp. Equivalent current dipole (ECD) locations
and dipole moments were extracted at the peak for which the root-mean-square
(RMS) field amplitude was maximal. For three subjects, MRI-scans were
obtained to superimpose ECD localizations upon cortical structures.
Results showed a significant effect of frequency on dipole moment
(p < 0.05) and RMS values (p < 0.001), which was mainly due to significantly
higher amplitudes for the 6.0 Hz as compared to the 11.9 and 15.2
Hz responses (p's < 0.05). The ECD localizations showed a significant
effect of stimulus frequency for the medio-lateral (y) and anterior-posterior
(x) axes (ps < 0.05) but not for the inferior-superior (z) axis. Specific
comparisons indicated that the ECD for the 15.2 Hz SSVEF was situated
medially (p < 0.01) and anteriorly (p < 0.01) with respect to the
6.0 Hz ECD and medially with respect to the 11.9 Hz ECD (p < 0.05).
The superimposition upon MRI sections showed ECD localizations of
the 6.0 and 11.9 Hz SSVEF in the posterior occipital cortex near the
calcarine fissure. For the 15.2 Hz SSVEF the ECD was situated anteriorly
in the lingural gyrus, ventral to the calcarine fissure.

Impact of acoustic and vibrotactile prepulse on acoustic and electrical blink reflexes: Startle inhibition and task accuracy results
Becky J. Mussat-Whitlow & Terry D. Blumenthal
Wake Forest University
Four experiments were conducted to evaluate the relationship between
sensory modality, prepulse inhibition, and identification accuracy
on a matching task. Experiments 1 (N = 18) and 2 (N = 18) used an
acoustic startle stimulus, and Experiments 3 (N = 24) and 4 (N = 24)
used an electrical startle stimulus. Acoustic pre/postpulses were
presented in Experiments 1 and 3, and vibrotactile pre/postpulses
were presented in Experiments 2 and 4. The matching task required
subjects to determine whether the postpulse (sound in Experiments
1 and 3; vibration in Experiments 2 and 4) which followed startle
stimulus onset by 500 ms was the same as or different than the prepulse,
which preceded startle stimulus onset by 150 ms. In all four experiments,
the prepulse was found to inhibit startle response amplitude. Experiment
1 (acoustic prepulse-acoustic startle) showed evidence of loudness
assimilation, which may be one way in which startle interrupts prepulse
processing. However, no relationship was found between PPI and matching
task accuracy, contrary to expectations based on Graham's (1975) theory
that PPI is a protective mechanism. The present data suggest that
protection of prepulse processing may vary with the cognitive target
properties of the prepulse. If the prepulse is the stimulus on which
a decision is made, as in a study by Norris and Blumenthal (1996),
PPI and task accuracy will increase together. If the prepulse is simply
stored for comparison with a later stimulus, as in the present study
PPI and task accuracy may not be related.

Spontaneous blink rates predict resting anterior electroencephalographic asymmetry
Lynn A. Myers1, John P. Kline2, John J.B. Allen1, & Gary E. Schwartz1
1The University of Arizona, 2Eastern Washington University
It has been hypothesized that left frontal regions of the brain mediate
behavioral approach (Davidson & Tomarken, 1989), and that dopamine
is related to both behavioral engagement (Depue et al., 1990) and
spontaneous blink rate. Increased levels of dopamine have been hypothesized
to relate to increased blink rate in SAD's patients (Depue et al.,
1990). Left frontal activation has therefore been hypothesized to
relate to spontaneous blink rate (Allen et al, 1993). To test this
hypothesis, EEG was recorded from 61 right-handed participants (27
men). Six 60-second baselines of EEG, 3 eyes open and 3 eyes closed,
were recorded from 19 channels referenced to linked ears. Impedances
were less than 5 Kohm and homogolous channels within 1 Kohm. EEG was
digitized on-line at 256 Hz (band pass 2 to 64 Hz). EEG epochs with
EOG or EMG deflections greater than 50 microvolts in the frontal channels
were rejected. Average alpha (8-13 Hz) power (microvolts squared)
was computed across artifact-free epochs and log transformed to normalize
EEG power. Asymmetry scores were computed (log[right]-log[left]) for
F4-F3, Fp2-Fp1, and F8-F7. Ocular blinks were manually counted and
summed as large, distinct deflections in frontal EEG channels. As
predicted, participants with high blink rates showed relatively greater
left frontal activation than did participants with low blink rates.
Participants who were unable to inhibit rhythmic blink motions during
eyes closed conditions showed the largest effect. The results are
consistent with the hypothesis that relative left frontal activation
is related to dopamine function.

The lateral presentation of emotional faces: An ERP study with subjects scoring high on a bodily complaint scale.
Ewald Naumann, Fridemann Gerhards, Alexander Luerken, Dieter Bartussserk, & Dirk Hellhammer
University of Trier
Previous research suggests, that subjects who are sensitive to bodily
complaints show reduced functional hemispheric lateralisation. The
present study used the lateral presentation of neutral and emotional
faces to find possible cortical asymmetry differences in subjects
scoring low or high (n=20 in each group) on a bodily complaint scale
by means of event related potentials. Two faces were projected in
each visual half field (VHF) for 150 ms (144 trials, one third emotiona
face right, one third emotional face left, one third both faces neutral).
Subjects had to respond with a press on one of three buttons, which
were assigned to the position of the emotional faces (right, left,
none). Contingent on the presentation of faces EEG was measured from
locations F3,C3,P3, Fz,Cz, Pz, F4,C4,P4 from 300 ms prior to stimulus
presentation to 1500 ms post presentation. Averages were computed
for each stimulus class. The reaction time data showed a smaller left
VHF advantage for subjects scoring high on the complaint scale. The
ERP data showed interaction of hemisphere and group for N1 and P2.
Subjects scoring high on the complaint scale had more negative N1-Amplitudes,
but only over the midline and the left hemisphere. P3 Amplitude differed
between the groups for left VHF presentation only over the midline
with no differences for right VHF presentation of the emotional faces.
There seems to be a general difference between these groups in the
early stage of information processing and a difference depending on
the information which is presented to an cortical hemisphere for later
stages of the information process. Supported by the Deutsche Forschungsgemeinschaft,
Forschergruppe Stressvulnerabilitaet und Stressprotektion

Measuring the extension of priming effects during arithmetic fact retrieval
Michael Niedeggen & Frank Roesler
In simple multiplication problems (e.g. 3*8), activation spread which
is driven by the operands will excite the correct solution (24) as
well as table-related false multiples (16, 32). Behaviorallly, response
time is longer for table-related than for table-unrelated errors (e.g.,
27). The two types of errors also find a different expression in an
"arithmetic N400-effect" (Niedeggen, Jost & Roesler, 1996, Psychophysiology,33,
S65). We examined if priming of table-related errors is restricted
to adjacent multiples of the correct solution only. After seeing the
two operands of a multiplication problem, subjects (n=16) had to verify
an offered solution which was correct in 50%, a table-related error
in 25%, or a table-unrelated error in 25% of all trials. Related errors
varied systematically with respect to their "table"-distance to the
correct product (e.g.: 5 * 6 = 24 (small distance), 18 (medium), 12
(large)). Each related error was matched with an unrelated error with
respect to the size of the numerical distance to the correct product.
Event-related potentials were recorded from 61 electrodes. Reaction
times revealed a magnitude effect for both related and unrelated errors:
RT increased with increasing numerical distance between solution and
correct product. In contrast, the amplitude of the arithmetic N400-effect
followed a different trend for related and unrelated errors: For unrelated
errors the N400-amplitude was of constant size for all numerical distances.
For related errors the amplitude was reduced for small and mediu
numerical distances, and of the same size as with unrelated errors
in case of large distances. Thus, N400-amplitude is specifically related
to the associative distance of table-related errors, not to the numerical
distance as such.

Cerebral asymmetry, life stress, and depression: Testing a diathesis/stress model
Jack B. Nitschke, Wendy Heller, Patrick A. Palmieri, & Gregory A. Miller
University of Illinois at Urbana-Champaign
Substantial evidence has been obtained to indicate that regional asymmetries
in brain activity (as inferred from EEG, hemodynamic, lesion, and
behavioral studies) are associated with depression, but there has
been some controversy as to the conditions under which the association
occurs. For example, although a significant number of patients with
left frontal lesions show symptoms of depression, many do not. Similarly,
not all neurophysiological studies report the more common finding
of a reduction in left compared to right frontal activity in depression.
To account for these discrepancies, Davidson (1993) formulated a diathesis-stress
model proposing that frontal asymmetry confers a vulnerability to
depression but that environmental elicitors are necessary to produce
it. As a direct test of this hypothesis, we examined the relationship
between depression, life stress (daily hassles and major life events),
and frontal EEG asymmetry in 80 participants, selected from among
783 undergraduates for extreme scores on self-report measures of depression
and anxiety. Neither depression nor daily hassles alone predicted
frontal asymmetry. However, increased depression, in the presence
of increased daily hassles, was associated with a relative decrease
in left- compared to right-hemisphere activity. In contrast, there
were no significant effects involving major life events. We also obtained
evidence that variance associated with daily hassles influences the
patterns of activity associated with anxiety and with the co-occurrence
of anxiety and depression for anterior and posterior brain regions.
The results favor a diathesis/stress model and further support our
previous work emphasizing the importance of subtypes and comorbidity.

Passive auditory oddball paradigm in childhood epilepsy
Nancy E. Noldy & James Stieben
The Toronto Hospital Research Institute & York University
Impairment of cognitive functioning commonly accompanies childhood
epilepsy. Children with epilepsy can be very difficult to test using
traditional standardized assessment techniques because of coexisting
cognitive and motor disorders. A primary information processing difficult
in these children is in "the brain's capacity to react adaptively
to incoming information" (Binnie, 1990). This preliminary investigation
examines the usefulness of ERPs in measuring cognition in this population.
To investigate how rare stimuli in the environment are spontaneously
processed by this group, a passive auditory oddball paradigm was used.
No task instructions were given. Five cognitively delayed children
with seizures involving the temporal lobes, and five age matched normal
controls were tested. Two conditions of a passive auditory oddball
paradigm were employed: "easy" (common 1000 Hz, rare 2500 Hz) and
"difficult" (common 1000 Hz, rare 1500 Hz). ERPs were recorded from
Fz, C3 and C4, and referred to linked mastoids. The difference waveforms
(rare minus common) were characterized by two negative waves and a
positive wave which separated them. Significant differences were found
only in the "easy" condition. The positive wave was later and smaller
in the epilepsy group (406 ms; 3.5 uV) compared to the normal group
(322 ms; 8.5 uV). The second negative wave was later in the epilepsy
group (552 ms) compared to controls (472 ms). This preliminary study
indicates that the passive auditory oddball paradigm is sensitive
to cognitive impairments in children with seizures involving the temporal
lobes, and further investigation is warranted.

The relationship of the P200 generated at encoding and subsequent retrieval
Amanda O'Donnell and Bruce R. Dunn
The University of West Florida
The relationship between the P200 component of the Event Related Potential
(ERP) generated at encoding and subsequent retrieval was explored
using two memory tasks. The first task was a serial-order task where
participants were instructed to memorize a list of 20 words in serial
order (rote encoding task). The second task was a Bousfield type category
task where participants were instructed to memorize the same list
of 20 words by categories (an elaborative task). There were three
trials in each condition. After the presentation of each list, the
participants were given a free recall sheet and asked to list all
the words recalled from the list. Following the free recall task,
the participants were given a recognition sheet (with distractors)
and asked to check all the words they recalled from the list. Both
the behavioral and ERP data from items that were both recalled and
recognized (retrieved) and neither recalled nor recognized (not retrieved)
were analyzed. Results indicate that there were no P200 amplitude
differences between serial order and category tasks at any brain recording
site. Nor were there any differences in the number of items recalled
and recognized across tasks. However, a greater mean posterior P200
(or possible P3A) amplitude was generated when items were retrieved
compared to when they were not. Our data suggest the P200 indexes
encoding processes that are sensitive to subsequent word retrieval.

The impact of ethanol challenge on psychophysiological and neuroendocrine stress responses in subjects with a familial risk for alcoholism
Robert Olbrich, Bernhard Croissant, Ralf Demmel, & Fred Rist
Central Institute of Mental Health-Mannheim
One of the most interesting approaches in current alcoholism research
refers to ethanol challenge studies carried out with male subjects
who have no personal psychiatric problem but a positive family history
(FH+) regarding an alcoholic father. When exposed to aversive psychological
stimuli in laboratory experiments this group, who is considered to
bear an heightened genetic risk for addictive behavior, is much more
likely to exploit the autonomic stress response dampening effect of
alcohol, when available, than FH-subjects. Extending this line of
research, we are at present conducting an alcohol (and placebo) challenge
study (targeting at a 0.7 per mille blood ethanol level), for which
not only adult sons but also daughters and siblings (both male and
female) of male alcoholics are recruited. Employing RT paradigms,
in addition to aversive stimulation a reward condition is given in
the laboratory experiment. Response assessment includes not only psychophysiological
parameters but also monitoring of neurohormones. In an analysis of
the data of 62 subjects at risk and 39 normal controls which will
be described in the submitted paper, stress dampening effects due
to alcohol could be demonstrated. Interestingly they applied to the
male risk group but not to female FH+ subjects. Dampening phenomena
were observed in the psychophysiological measurement area, i.e. for
heart rate and tonic electrodermal indices as well as in the HPA-axis,
indexed by serum cortisol.

Gender differences in late positive components evoked by human faces
Juan C. Oliver-Rodriguez, Victor S. Johnston, & Guan Zhiquiang
Universidad 'Jaime 1' Castellon & New Mexico State University
The emotional value hypothesis predicts a positive relationship between
P300 amplitude and affective value of a stimulus. It was tested in
the context of facial perception, in conditions where probability
and task relevance of stimuli (the two classical P300-eliciting variables)
were controlled. 22 males and 40 females participated in the study.
Two levels of Hair, Eyes, Mouth, Chin and Proportions served as independent
variables to form 32 male and 32 female faces. The subject task consisted
of simply looking at a face and pay close attention. Attractiveness
ratings on a 5-point scale were postponed to the final experimental
session to eliminate possible response selection confounds. EEG was
recorded at frontal, central and parietal electrode sites only during
stimulus presentation. A positive correlation between average ratings
and average P300 scores to opposite sex faces was observed in both
male (r = 0.40) and in preovulatory (r = 0.41) and postovulatory (
= 0.44) female subjects. Correlations to same sex faces were only
found in postovulatory females (r = 0.61). ERP dissociations in gender
effects undetected by ratings were observed: males showed a much larger
average P300 than females and simple face sex effects occurred only
in female subjects, whose P300s were unexpectedly larger to female
than to male faces. Neither experimental nor real world stimulus probabilities
are plausible explanations for these differences. These results support
the emotional value hypothesis, according to which classical P300
processes reflect an affective evaluation of the stimulus, which in
turn produces context updating.

Working memory capacity and syntactic processes
Patrick J.W. Oor, Herman H.J. Kolk, & Dorothee J. Chwilla
University of Nijmegen
The relation between working memory (WM) capacity and ERPs to sentences
being read has been investigated with manipulations of subjects' capacities,
WM load, or both (e.g., King & Kutas, 1995). This study assessed effects
of WM capacity limits on syntactic processes in a grammaticality judgement
task. WM load was manipulated by comparing Subject Relative (SR) sentences
to syntactically more demanding Object Relatives (OR)(the latter yielded
longer reaction times (RT) and more errors in a RT pre-test [p<.01]).
23 Subjects (9 high, 14 low capacity) read sentences presented word-by-word
(SOA=645ms); EEG was recorded from 13 electrodes. In half of the sentences,
a verb at a critical mid-sentence position disagreed in number with
its subject (violating syntactic constraints), being expected to elicit
a late positivity (i.e., P600). Following each sentence, subjects
indicated its correctness by a delayed response. Reducing the capacity
available, either by increasing sentence complexity, or by comparing
subjects differing in capacity, was predicted to decrease P600 amplitude
and/or increase its latency. To compare effects of a syntactic violation
with those of a semantic violation at the same mid-sentence position,
a semantic judgment task was added (half of the sentences violating
selection restrictions). Consistent with the literature, the syntactic
violations yielded a parietally distributed P600 (p<.01; window: 500-800ms
after verb onset), and the semantic violations elicited a centro-parietally
distributed N400 effect (p<.01; 350-500ms window). Effects of varying
syntactic complexity and subjects' capacities will be discussed.

Aspects of Cook-Medley hostility and hemodynamic reactivity: Effects of gender
Matthew D. Orenstein1, Laura Cousino Klein1, Susan S. Girdler2, & Larry D. Jamner1
1University of California-Irvine, 2University of North Carolina-Chapel Hill
Cardiovascular(CV) reactivity to stress has been linked to the development
of hypertension and coronary heart disease. Recent reports suggest
that subscales of the Cook-Medley Hostility Scale(Ho) may be more
predictive of CV reactivity than the overall Ho scale. The present
study examined the relationship of 5 Ho subscales (cynicism, hostile
affect, hostile attribution, social avoi dance, aggressive responding)
with CV reactivity to 3 laboratory stressors (math, speech preparation,
stroop) in 48 women and 30 men. Overall, men had higher Ho (t(75)=2.74,
p<0.05), aggressive responding (t(74)=2.83, p<0.05), and cynicism
(t(74)=3.12, p<0.05) scores than did women. There were no gender differences
in scores on the other 3 subscales. Ho scores predicted baseline systolic(SBP)
and diastolic(DBP) blood pressure in men but not in women. During
the tasks, Ho scores predicted total peripheral resistance(TPR) and
cardiac output(CO) changes in men but were not related to CV changes
in women. With respect to Ho subscales, hostile affect predicted baseline
SBP and cynicism predicted baseline DBP in men. For women, social
avoidance predicted both baseline SBP and DBP. T he cynicism and hostile
affect subscales predicted changes in stroke volume, pre-ejection
period(PEP), CO, TPR, SBP, and DBP during the tasks for men. For women,
hostile affect predicted SBP reactivity during all tasks and DBP changes
during the math and stroop tasks. Although social avoidance predicted
PEP changes in women during speech preparation, no other Ho subscales
predicted CV changes in women. These results suggest that certain
Ho subscales are better predictors of CV reactivity in men and women
than are overall Ho scores. In addition, cynicism may be a useful
predictor of reactivity in men, whereas hostile affect may be a valuable
predictor of CV reactivity in men and blood pressure changes in women.

The effects of repeated occlusion measurement on blood pressure and pulse rate
Rebecca Palacios-Esquivel, Jeannie B. Concha, & Joe Tomaka
University of Texas at El Paso
We examined the effects of repeated occlusion measurement on blood
pressure and pulse rate during rest and during mental arithmetic stress.
Each participant performed two rest- and-task sequences. We used a
common automated blood pressure monitor (Dinamap, Model 8100) to take
readings during each minute of one rest- and-task sequence (i.e.,
all minutes measurement strategy) and during every other minute for
the other sequence (i.e., odd minutes measurement strategy). We chose
these strategies because both allow for examination of reactivity
trends over short time periods (e.g., 5 min). Each rest and task period
was 5 min long and we counterbalanced the two rest-and-task sequences
for order. We found small but reliable measurement strategy effects
for diastolic blood pressure that interacted with task novelty/task
repetition. Specifically, diastolic pressure task levels (F[1, 30]
= 6.67, p = .015) and reactivity (F[1, 30] = 15.36, p < .001) wer
lower during the all minutes measurement strategy compared to the
odd minutes strategy, particularly during the second of two repeated
tasks. We also found a very small, but similar, trend for systolic
blood pressure, but observed no effects of measurement strategy on
mean arterial pressure or pulse rate. The results of this study have
implications for tracking patterns of blood pressure reactivity over
time, depending on the focus of the research.

Landscape aesthetics, nonverbal memory, and EEG
Russ Parsons1, Louis G. Tassinary2, Daniel Bontempo2, & Eric Vanman2
1University of Illinois, 2Texas A&M University
Evidence suggests that the experience of emotion and the perception
of affect-laden stimuli elicit discriminable patterns of information
processing. The experience of emotion is associated with bilateral
frontal lobe activation, while the perception of affect-laden stimuli
is associated with more posterior, unilateral activation of the right
hemisphere. Aesthetic judgments of landscape scenes, however, do not
clearly elicit processing indicative of the experience of emotion
or the perception of affective information, and thus may constitute
a dissociation between nonsocial evaluative judgments and more socially-oriented
affective processing. We constructed two reduced-information stimulus
sets from 90 photographic quality 16-bit color digital images of forest
landscapes, a 4-bit color set and a contrast-enhanced, edge-sharpened
achromatic set. Separate groups of participants provided either 'liking'
ratings or performed similarly engaging cognitive tasks for half of
the images in the training phase of a recognition experiment. Because
preference ratings for the full color images correlate weakly with
the 4-bit color images (r = .43) and not at all with the achromatic
images (r = .04), we expected the reduced-information sets to elicit
less emotional engagement than the full color images, regardless of
the task, and thus to be less memorable when recognition was tested.
As predicted, the reduced-information sets were liked less, rated
as less complex, and were less well remembered than the full color
images (All ps < .05). This interpretation of greater memorability
for and affective engagement with the full color images was consistent
as well with bilateral EEG activation recorded from frontal (F3, F4)
and temporal (T3, T4) scalp sites.

Event-related potential measures of auditory priming and recognition as a function of lag: Aging and Alzheimer's disease
Julie V. Patterson, Carl Cotman, & Curt Sandman
University of California-Irvi
Event-related potential (ERP) measures of auditory priming and recognition
were compared in young and older controls, and individuals with presumed
Alzheimer's disease at 15 scalp sites. Three lists of words were prepared
for 3 tasks (priming, recognition, target detection). Each list contained:
(1) words repeated immediately (0 lag); (2) words repeated after 5
intervening words (5 lag); (3) words not repeated (fillers); and (4)
infrequent target words not repeated. In the recognition task, subjects
pressed one of two reaction time (RT) buttons to indicate whether
each word was a first ("new") or second ("old") presentation of a
word in the list. In the priming task, subjects indicated whether
each word contained one or two syllables. In target detection, subjects
pressed one RT button to infrequent target words. For each group,
RTs were shorter to old than to new words for both the syllable and
recognition tasks at 0 lag. At 5 lag, this RT advantage was diminished,
but still present, except for the Alzheimer's group during recognition.
Word repetition, task, lag, and topography reflected group differences
in ERPs. ERPs to new compared to old words differed in the region
of N400 and a late positivity for each task and group, even when a
new/old decision was not required. For young and older controls, ERP
differences between new and old words were largest in the recognition
task, and for 0 compared to 5 lag; for the Alzheimer's group, task
effects on the ERP were diminished, largely due to an attenuated amplitude
late positivity to old words during recognition. For old words, ERP
differences between older controls and the Alzheimer's group were
larger and more widespread for the recognition compared to the syllable
and target tasks. The findings suggest that brain processes related
to priming and recognition are differentially affected in Alzheimer's

Cardiovascular reactivity patterns among obese and non-obese adolescents: A two-year longitudinal study
Stephen M. Patterson1, Bruce S. Alpert1, Marie Barnard1, & Dawn K. Wilson2
1The University of Tennessee-Memphis, 2Medical College of Virginia
Obesity poses a significant risk factor for several health conditions
in adolescents that can persist into adulthood. However, little is
known about the influence of obesity on cardiovascular reactivity
among adolescents. This study examined whether cardiovascular reactivity
profiles during acute psychological stress differed among obese (BMI>90%ile)
and non-obese adolescents and whether the patterns were stable over
a two-year period. Systolic (SBP) and diastolic (DBP) blood pressure
and heart rate (HR) reactivity to cold pressor (CP) and a video game(VG)
were assessed during an initial visit and a two-year follow-up visit
in 23 obese and 46 non-obese adolescents (age 9-19 yr.). Blood pressure
and HR were measured during a 5-minute rest period, 1-minute CP, an
a 3-minute video game. This procedure was repeated again at follow-up.
Results of the initial CP task indicated that obese adolescents had
significantly higher resting DBP (p<.0001) and marginally higher SBP
(p's<.09) compared to non-obese adolescents. Non-obese adolescents
exhibited greater SBP reactivity during CP (p's<.05) than obese adolescents.
Follow-up visit CP results also showed higher resting SBP and DBP
among the obese group (p's<.05), and greater SBP and DBP reactivity
among the non-obese group (p's<.01). VG results revealed a similar
cardiovascular pattern during the follow-up visit with resting SBP
being higher among the obese group (p<.03) and SBP reactivity being
greater among the non-obese group (p<.005). Results suggest that there
are possible ceiling effects for cardiovascular reactivity among obese
adolescents and that reactivity profiles among obese adolescents appears
to be stable over time.

Frontal brain asymmetry in panic disorder
Paul Pauli, Georg Wiedemann, Wilhelm Dengler, Y Lyle Bourne Jr.
University of Tuebingen
According to Davidson (1992, Psychological Science, 3, 39-43) an increased
right frontal hemispheric activity reflects negative affect and activation
of an avoidance-withdrawal system. Increased left frontal hemispheric
activity, on the other hand, is associated with positive emotions
and activation of an approach system. The alpha activity at left and
right frontal electrodes seems to be a reliable index of such an asymmetry.
Panic patients are characterized by negative affect and avoidance
behavior. Therefore, we predict that these patients exhibit an increased
right frontal brain activity (reduced alpha power). The EEG of 23
panic patients and 25 control subjects was measured under conditions
of rest and when confronted with panic-relevant (emergency situation),
anxiety relevant (spider), arousing (erotic scenes), or neutral (mushroom)
picture stimuli. Alpha power at left and right frontal (F3, F4) and
parietal (P3, P4) electrodes was analyzed. Only panic patients exhibited
a reduced right frontal alpha power under rest conditions and when
confronted with panic-relevant, anxiety relevant, and arousing picture
stimuli. However, no frontal asymmetry was observable in panic patients
when confronted with a neutral picture stimuli. In addition, the observed
asymmetry in the alpha power of panic patients was restricted to frontal
electrodes. These findings indicate that panic disorder is characterized
by a reduced right frontal alpha activity. This asymmetry may reflect
the preponderance of a negative emotionality and an increased activation
of an avoidance-withdrawal system. Research was supported by the Deutsche
Forschungsgemeinschaft, DFG.

The role of worrisome and relaxing thinking on the heart rate response to fear image
Kate E. Peasley-Miklus & Scott R. Vrana
Purdue University
This study replicated and extended previous research suggesting that
worry inhibits emotional processing of fearful imagery. Female participants
categorized as either victimization fearful (n= 24) or victimization
and speech fearful (n=27) completed 24 trials of worrisome or relaxing
thinking and tone- cued fear imagery. For each trial, subjects engaged
in 30s of relaxing or worrisome (speech- or victimization-related)
thinking and then imagined speech or victimization fear scenes for
15s. Subjects were exposed to 4 trials each of 6 possible thought-imagery
pairings (e.g., speech worry then speech image; speech worry- victimization
image; relax-speech; relax-victimization, etc.). Heart rate and corrugator
EMG were measured during the thinking and imagery periods to estimate
degree of emotional processing of the fear imagery. The two fear groups
did not exhibit heart rate differences during victimization and speech
imagery, replicating previous research showing that heart rate increases
during speech imagery for both fearful and nonfearful people. Consistent
with previous findings, participants had a significantly larger heart
rate increase from thinking to imagery after they engaged in relaxing
as opposed to worrisome thinking. Contrary to earlier findings, however,
heart rate also increased more during speech and victimization thinking
than during relaxing thinking. Thus, it appears that previous research
showing that worry inhibits emotional imagery processing may have
been caused by different heart rate levels during the preceding thinking

Effects of difficulty and ability in a dual-task video game paradigm on attention, physiological responses, performance, and emotion-related appraisal
Anna Pecchinenda1, Arvid Kappas1, & Craig A. Smith2
1Laval University, 2Vanderbilt University
In a study in which task-difficulty and ability were varied, we assessed
changes in physiological activation, behavior, and self-reported appraisals
associated with performing a video game. In addition, a secondary
reaction-time task was used to assess the participants' attentional
resources not allocated to the game. Participants played a sequence
of 10 games while pressing a pedal when one of two tones was presented.
Task difficulty (high vs. low) was manipulated by varying the number
of points to be obtained in each trial. Success probability was high
for the first 5 games for both groups. Starting with game 6, the probability
of succeeding decreased for both groups, albeit much more quickly
for the "high difficulty" group. Changes in RTs to the secondary task,
number of points collected during a trial, participants' self-reporte
appraisals, and changes in SC, HR, skin temperature and EMG were analyzed
as a function of task difficulty and ability. Participants' performance
decreased for both groups during the last 5 games, but especially
for the low ability, high difficulty participants. Starting from game
6, evaluation of difficulty increased while evaluations of certainty
of succeeding and importance of performing well decreased, but only
in the high difficulty groups. In general, RTs to the secondary task
were longer during the last 5 games. Additionally, SC mean decreased
significantly during the first 5 games. Finally, SCR maximum amplitude
was larger and SCR-Rise time was longer during the last 5 games for
low ability participants in the high-difficulty group.

Effects of false heart rate feedback on the schandry and Whitehead heartbeat detection tasks
G. C. Phillips, G. E. Jones, E. J. Rieger, & J. B. Snell
University of Southern Mississippi
Previous research has indicated that performance on heartbeat counting
tasks may be influenced by beliefs about heart rate. To test this
hypothesis, 60 male subjects were administered the heartbeat counting
task developed by Schandry (1981) after viewing fast, slow, or no
heart rate feedback. The "standard" procedure of Schandry was used
and involved three counting periods of 25, 35, and 45 seconds duration.
Subjects were also administered the Whitehead heartbeat perception
task, a signal-detection-type heartbeat awareness task, for comparison
purposes. Subjects were presented 30 Whitehead trials conforming to
our standard implementation. Results indicated that subjects who received
fast or no heart rate feedback performed better on the Schandry task
than subjects who received slow heart rate feedback. False feedback
presentation did not significantly affect the Whitehead task. Also,
subjects who were administered the Whitehead task prior to being administered
the Schandry task demonstrated poorer performance on the Schandry
task than subjects who had not previously been administered the Whitehead
task. Data again confirms that subjects tend to rather substantially
underestimate their actual heart rate on the Schandry task (12.2%
with no feedback), and underestimated heart rate by 28.4% under slow
(false) feedback conditions. Fast (false) feedback appeared to raise
subjects' inappropriately low expectations of heart rate and led to
improvements in Schandry error score. These results suggest that the
Schandry task is more influenced by external variables such as subject
beliefs and expectations about HR, and thus may be a less powerful
method for determining ability to detect individual discrete heartbeats.

Short-term memory for visuospatial and verbal material in the elderly: An ERP study
Natalie A. Phillips and Ron Hoop
Concordia University
Although mild memory deficits are a consequence of aging, deficits
for visuospatial material may be greater than for verbal material.
However, previous research typically compared abstract novel visuospatial
stimuli with verbal material which is highly familiar. Thus, the deficit
may result from the unfamiliarity of visuospatial stimuli rather than
their visuospatial nature per se. We sought to disentangle this by
examining event-related brain potentials (ERPs) recorded during a
fixed-set memory-scanning paradigm which used three stimulus types:
abstract words (Familiar Verbal), non-nameable dot arrays (Unfamiliar
Visuospatial), and pseudowords (Unfamiliar Verbal). Ten young (mean
age = 24 years) and 10 elderly (mean age = 80 years) adults participated.
Memory set sizes of 1, 2, and 4 were used per stimulus type; 100 probes
were presented per set size. Subjects indicated whether or not each
probe was a member of the memory set (50% were). ERPs were recorded
to each probe; the late positive component (LPC; 300-800 msec) was
scored. Elderly subjects had both longer reaction times (RT) and LPC
latencies (p's < .05) relative to young subjects. However, an interesting
dissociation between RT and LPC latency emerged in the older subjects.
While RTs to the visuospatial stimuli were significantly longer compared
to words and pseudowords, LPC latency did not differ between the three
stimulus types. This suggests that, in the elderly, stimulus evaluation
and memory comparison processes are similar regardless of stimulus
type, and that differences in RT across stimulus type are due to processes
occurring after the LPC which may be related to response selection.

EEG, ERPs, and cognitive development
Alonn Pitzer & John Polich
The Scripps Research Institute
EEG (eyes open/closed) and ERP (auditory/visual) measures were obtained
at midline recording sites from n=32 participants who ranged in age
from 6 through 20 years to assess if developmental electrophysiological
changes affect EEG and the P300 component concomitantly. With increasing
age: EEG power decreased for the delta, theta, and early alpha bands;
EEG mean band frequency did not change dramatically except for the
late alpha band, which demonstrated an appreciable increase in frequency.
P300 amplitude increased and peak latency decreased for both auditory
and visual ERP conditions. Correlational analyses for EEG power/P300
amplitude and EEG frequency/P300 latency did not reveal any consistent
or reliable associations for any of the bands or experimental conditions.
However, EEG power in the delta, theta, and early alpha bands was
positively correlated with P300 latency across subject age-a result
that suggests at least some association between EEG and ERP change
with development. Although P300 amplitude scalp distributions were
affected by age, EEG power over the midline electrodes did not correlate
strongly with these changes. The findings suggest that EEG and P300
measures are both modulated by developmental increases in age, but
that EEG developmental changes are indirectly related to P300 component
development in children.

ANS differentiation of emotion across the lifespan
Kirsten M. Poehlmann, David J. Klein, Gary G. Berntson, & John T. Cacioppo
The Ohio State University
The link between physiology and affect has been of tremendous interest
in recent years, with much of the work focusing on the extent to which
specific emotions can be differentiated by autonomic measures. Evidence
for the autonomic differentiation of discrete emotions has been equivocal.
However, indications of this differentiation might be masked by moderating
factors such as age. Because peripheral physiology changes across
the lifespan, age-related changes in physiological function may alter
the autonomic manifestations of emotions. Using age as a possible
moderator variable, the current work uses meta-analytic techniques
to review published research relevant to the autonomic differentiation
of emotions. These analyses revealed some consistent relationships
between autonomic activity and some discrete emotions. A number of
measures differentiated anger from fear; anger was associated with
higher diastolic blood pressure; more nonspecific skin conductance
responses; smaller increases in heart rate, in stroke volume, and
in cardiac output; and larger increases in total peripheral resistance,
facial temperature, and finger pulse volume. These results suggest
that anger appears to act more on the vasculature and less on the
heart than fear. Ancillary analyses using age as a moderator variable
revealed that in general the magnitudes of autonomic responses in
older adults was much smaller than those observed in young adults.
For instance, there appeared to be greater autonomic differentiation
of anger-fear in younger than older participants on the measures of
diastolic blood pressure, facial temperature, finger temperature,
and stroke volume. Analyses also suggested that the autonomic differentiation
across condition may be due to factors other than emotional states
per se.

Relationships between psychophysiology, emotion, and therapeutic technique in a single case psychotherapy for depression
Nnamdi Pole & Robert W. Levenson
University of California-Berkeley
Previous psychophysiological studies of psychotherapy have been criticized
for: (a) lacking substantive theoretical foundation to guide data
interpretation and (b) lacking convincing evidence of the relationship
between psychophysiological data and psychological constructs. This
study tests Control-Mastery Theory, which postulates that clients
succeed in therapy when they are made to feel "safe" and experience
setbacks when they feel in "danger". In an intensive study of a single
case, we advanced three hypotheses: (a) "safety" feelings will be
associated with decreased autonomic arousal, (b) "danger" feelings
will be associated with increased autonomic arousal, and (c) supportive
interventions will be associated with decreased autonomic arousal
and increased "safety" feelings. A 30 year old woman diagnosed with
Major Depressive Disorder received sixteen sessions of Control-Mastery
psychotherapy. During each session, her heart rate, skin conductance,
and general activity level were continuously recorded. Following each
session, she rated her average level of "danger" and "safety" during
the session and the therapist rated the supportiveness of his interventions.
Correlations were computed between client and therapist ratings and
client's mean physiology in each session. Analyses revealed that the
client's "danger" feelings were associated with her greater arousal
(faster heart rate [r = .42], greater skin conductance [r = .31]).
Her "safety" feelings were associated with lower arousal (lower heart
rate [r = -.51]). Supportive therapeutic interventions were associated
with higher client "safety" (r = .59) and lower arousal (lower heart
rate [r = -.61], lower skin conductance [r=-.46]), thus supporting
our three hypotheses.

Orienting responses to structural features of media
Robert F. Potter, Annie Lang, & Paul D. Bolls
Indiana University
In previous studies we have demonstrated that some structural features
of television (e.g., changes in visual scene, changes in program,
and the onset of video graphics) elicit cardiac orienting responses
in attentive television viewers. The goal of this study was (a) to
determine if the change from one camera to another in the same visual
scene (called an edit) elicits orienting responses, and (b) to generalize
this finding to other media by determining if structural features
of other media (e.g. radio and computers) elicit orienting responses
in attentive users. 30 adult subjects viewed 32 headlines appearing
on a computer screen, watched 20 one-minute television vignettes,
and listened to a 12 minute radio show. Heart rate was collected,
time-locked to the media presentation, during each of these tasks.
Cardiac response curves were generated for (a) the ten-seconds following
the onset of the headline on the computer screen, (b) the ten-seconds
following a camera change in the television message, and (c) the te
seconds following identified radio structural features (which included
voice change, content change, and audio special effects). Results
showed significant deceleratory quadratic trends in the heart rate
data indicative of orienting following radio structural features and
television camera changes. However, no significant Time or quadratic
trend effects were found in response to headlines appearing on the
computer screen. This is surprising given that cuts from a picture
to text on television and the appearance of pictures on computers
have both been shown to elicit orienting responses.

Self-regulation of slow cortical potentials and cerebral blood flow
Hubert Preissl1, Niels Birbaumer1,2, Friedemann Pulvermueller1, H.J. Heinze3, H. Scheich, & C. Tempelmann3,4
1University of Tuebingen, 2University of Padova, 3University of Magdeburg, 4Federal Institute for Neurobiology-Magdeburg
8 healthy subjects were trained for up to 20 sessions to change their
slow potentials (SCP) with a biofeedback device at the vertex location.
After training all Ss were able to change even without feedback their
cortical negativity and positivity. Several days after training Ss
cerebral blood flow (CBF) was measured with a 3-Tesla Brucker scanner
with a gradient echo procedure. Ss tried to produce cortical negativity
and positivity contingent upon the presentation of two discriminative
stimuli, letter A for negativity and letter B for positivity. The
session also incorporates a resting condition when the Ss fixated
a simple cross. The CDF was measured in 6 slices with a thickness
of 10 mm and the scans were corrected for movements. The main result
of the analysis was a significantly increased number of activated
voxels prefontrally during negativity compared to positivity. Some
Ss showed also increased CBF in thalamic regions during negativity.
The data indicates that self-regulation of slow cortical potentials
is achieved through local activation of prefrontal attention systems.
Supported by the DFG(German Research Society)

The effect of visual complexity on the dynamics of the EEG
Hubert Preissl1, Werner Lutzenberger1, Friedemann Pulvermueller1, & Niels Birbaumer1,2
1University of Tuebingen, 2 University of Padova
Non-linear measures are useful and relative new techniques for the
characterization of brain dynamics. The assumption that complex tasks
lead to a increased complexity measured with the pointwise dimension,
was investigated by the use of different stimuli , where it is possible
to quantify there complexity. The experiment incorporated two different
visual stimuli. In the first condition people observed the motio
of a simple pendulum, this is in a first approximation a periodic
movement (friction is very low). In the second condition people had
to observe a double pendulum. The motion of the double pendulum is
one of the simplest chaotic movements. EEG was recorded from 29 Ag/AgCl
electrodes with a sampling rate of 500 Hz in a frequency band between
.1 and 100 Hz referenced to Cz. The electrodes were placed according
to the extended 10-20 system . The pointwise dimension and the alpha
power was determined for the two different stimulations conditions.
Statistical analysis of data recorded from all 28 electrodes revealed
a significant effect of the task level, f(1,23) = 7.7 with P<0.01.
The major differences are located at left and right fronto-temporal
regions. The topography of the alpha power is characterized by a decrease
to the more complex stimulus. In this work we could show that the
complexity of an task is directly related to the complexity of the
cortical dynamics. This is clearly seen in the increase of pointwise
dimension during the second task. Supported by the DFG (German Research

Examination of alpha-rhythm nonlinearity using a new type of surrogate data
Walter S. Pritchard1,2 & Cornelis J. Stam3
1R. J. Reynolds Tobacco Company, 2Bowman Gray School of Medicine, 3Leyenburg Hospital
Nonlinearity in experimental time series is often tested by applying
a nonlinear discriminating statistic (NDS) to both original data and
surrogate data that have been linearized by randomizing phase angles
while maintaining a constant periodogram. This "standard" procedure
will incorrectly indicate nonlinearity in sine waves that mismatch
fundamental FFT frequencies. An algorithm that avoids this artifact
is described that, based on the Wold decomposition, produces surrogates
by first subtracting a scaled, best-fitting sine wave from the data,
phase-angle randomizes the residuals, and then adds the sine back
prior to estimating the surrogate NDS. Thirty 8-s alpha-rhythm EEGs
were analyzed using estimated correlation dimension (D2) as the NDS.
For each EEG, a Z score based on 1000 normally distributed surrogates
(surrogate-distribution Jarque-Bera statistic < 1.5) was computed
as [original D2 minus average surrogate D2]/[surrogate-distribution
standard deviation]. The Wold Z scores of EEGs whose dominant alpha
frequency was an FFT mismatch were significantly lower than the comparable
standard Z scores, indicating that alpha EEG, being highly periodic
in nature, is subject to the FFT-mismatch artifact. Examining each
EEG individually, p-values (Bonferroni-corrected for a total of 30
Z tests for each type of surrogate data) indicated that, for the standard
surrogates, three of the EEGs were nonlinear (Z > 3), vs. two for
the Wold surrogates. However, inferential statistical testing indicated
that the mean Z score was highly significantly greater than zero for
both types of surrogates. It is concluded that, as a population, alph
EEGs should be considered "weakly nonlinear".

The emotional experience of anhedonia: Psychophysiology, facial expression, and self report
Katherine M. Putnam1 & John M. Neale2
1Long Island Jewish University Hospital, 2State University of New York at Stony Brook
Results from studies comparing control undergraduates with anhedonic
undergraduates, defined by high scores on the Physical Anhedonia Scale
[PAS;Chapman, Chapman, & Raulin ( 1976). Jrnl Abn Psych, 374-382.]
have found behavioral, phenomenological, and physiological differences
between these two groups. However, the results from studies designed
to measure the affective experience of anhedonic individuals have
been equivocal. In order to further understand the putative affective
deficits in this population, the present study measured self report
[Self Assessment Affect Manikin; Hodes, Cook, & Lang (1985). Psychophys,
545-560], facial expressivity (corrugator and zygomatic EMG), and
psychophysiological variables (startle blink, SCR) in response to
affect eliciting slides selected from the International Affective
Picture System [Lang, Ohman, & Vaitl (1988). Gainesville: Center of
Research in Psychophysiology]. Participants were university students
who scored high on the PAS and thos who scored within normal range.
This study utilized the affective startle blink paradigm validated
by Peter Lang and colleagues as an index of the valence of emotional
experience. Both groups watched slides of positive, negative, and
neutral valence with startle probes presented during slide viewing.
Anhedonic participants reported less pleasant affect in response to
the positive slides and had less corrugator activity in response to
the negative slides than the nonanhedonic control group. No differences
were found between the groups for startle blink or skin conductance
amplitude. These findings indicate that anhedonia may reflect an uncoupling
of different dimensions of affective experience rather than simply
a reduction in affective experience in response to pleasant stimuli.

The effects of social influence on cardiovascular responsiveness in the natural environment
Douglas A. Raynor, Thomas W. Kamarck, Saul S. Shiffman, Philip Ituarte, Leslie Smithline, Hayley Thompson, Jeff Goodie, Jean Paty, Marianne Gnys, & Jon Kassel
University of Pittsburgh
Ambulatory blood pressure (ABP) may have unique prognostic significance;
the psychosocial determinants of ABP have yet to be adequately characterized.
Smith and colleagues (see JPSP, 1996, 70, 1012-1024) showed that social
influence attempts are associated with enhanced cardiovascular respondin
in the laboratory. Interactions involving social influence may also
exert effects on blood pressure during daily life. 120 healthy adults
(54% female, 50% African American) were monitored at 45 minute intervals
over a six-day period, using an automated ambulatory blood pressure
monitor and a hand-held computer, permitting near real-time self-reports
of behavioral states (including "social influence episodes"). An average
of 109 observations were recorded for each participant, and 70% of
these involved social interactions. The effects of social influence
episodes on CV activity were assessed using random effects regression.
In the sample as a whole (within-subjects), social influence episodes
were associated with increases in systolic (SBP; b = 0.44, T = 2.43,
p = .016) and diastolic blood pressure (b = 0.43, T = 2.65, p = .009)
after adjustment for posture and activity, replicating the Smith et
al. findings. However, individuals exhibiting more frequent social
influence attempts (between-subjects) manifested marginally smaller
SBP responses (b= -1.82, T = -1.89, p = .059) during periods of negative
mood, when compared with their low-influence counterparts. Social
influence is associated with short-term increases and longer-term
decreases in cardiovascular activation. Ambulatory assessment methods
allow for the characterization of the within-subjects and between-subjects
effects of psychosocial processes on CV activity in the natural environment.

Brain potentials and food fatigue: ERP evidence of sensory-specific satiety
Martin F. Regan, Marion M. Hetherington, & Linda Pirie
University of Dundee
Sensory-specific satiety (SSS) is a decline in the self- reported
pleasantness of the sensory characteristics of a food, as a function
of eating that food. This study attempted to provide objective psychophysiological
evidence of SSS. Normal-weight, non- dieting males (n=14) participated
in the experiment following an overnight (14 hour) fast. There were
two conditions: eaten food (cheese on cracker), uneaten food (chocolate).
Order of conditions was counterbalanced across subjects. At baseline
(T0), self-reported hunger and desire to eat the two foods were measured
using visual analogue scales. ERPs were recorded from 3 leads (Fz,
Cz, Pz) in response to 40 slide presentations of each food (slide
exposure = 1,200 ms). All measures were repeated immediately (T1)
and 30 minutes (T2) following consumption to satiety of the eaten
food. As a function of eating, hunger decreased (p < .0001) and desire
to eat the eaten food (but not the uneaten food) decreased (p < .0001).
N100 amplitude decreased for the eaten food over time (T0-T1, p <
.03; T0- T2, p < .01), but not for the uneaten food. Decrease in N100
amplitude was significantly correlated with decrease in desire to
eat the eaten food (T0-T1: r = 0.674, p < .01). P200 amplitude was
larger for the eaten than the uneaten food at T2 (p < .001), but did
not change as a function of eating. The results indicate a close correspondenc
between self-report indices of SSS and N100, while P200 appears to
be sensitive to the "task-relevance" of the stimulus.

Task-relevant affect: Effects of stimulus content on P300
Martin F. Regan & David M. Sutherland
University of Dundee
ERPs can be utilized to explore affect within an information processing
framework. This study examined the effects of stimulus fear-relevance,
probability, and task instruction on ERPs in a visual oddball task.
Subjects (n=20, unselected) underwent 4 trial blocks (order randomized)
in a 2 x 2 x 2 repeated measures design. Two slide stimuli were presented
in each block (slide exposure = 800 ms): fear-relevant (FR: rat),
fear-irrelevant (FIR: landscape). Trial blocks differed according
to which stimulus was frequent (80%) or rare (20%), and which required
motor response (Go) or inhibition (NoGo). ERPs were recorded from
3 leads (Fz, Cz, Pz). Self-report measures of stimulus pleasantness
and arousal were taken. P200 and P300 amplitudes were larger for FR
than FIR stimuli (p < .0001), and larger for rare than frequent stimuli
(p < .0001). P300 latency was longer for FR than FIR stimuli, but
only in the rare condition (p < .002). N100 amplitude was unaffected
by the manipulations, but N100 latency was shorter for FR than FIR
stimuli (p < .01). The FR stimulus was rated as less pleasant (p <
.01) and more arousing (p < .05) than the FIR stimulus. Arousal was
greater for rare than for frequent stimuli (p < .01). Response time
was not affected by the manipulations. The results indicate that fear-relevant
stimuli are more "task-relevant" than fear-irrelevant stimuli, and
require greater evaluation time. Task-relevance is discussed in the
context of informational content and memory representation of fear-relevant

Attentional modulation of language and music processing in opera: An event-related potential analysis
Pascaline Regnault & Mireille Besson
C.N.R.S. Marseille, France
The aim of this experiment was to study the effects of orienting
attention only toward the lyrics or only toward the tunes when
listening to songs (excerpts from opera). Specifically, it was
of interest to determine whether the amplitude and/or latency of
the N400 to sung incongruous words and of the LPC to congruous words
sung out of key would be modulated by the attentional demands.
Two hundred excerpts from French opera, sung a capella, were presented
to sixteen professional musicians. The final word of the excerpt wa
either semantically congruous or incongruous and sung either in or out
of key. The experiment was divided in two blocks; in one block, participants
payed attention only to the lyrics and in the other one, only to the tunes.
Results showed that both the N400 and the LPC were strongly influenced
by attention. The amplitude of the N400 to incongruous words was larger
when participants payed attention to the language than when they payed
attention to the music. Furthermore, the amplitude of the LPC
was larger when participants payed attention to the music than when they
payed attention to the language. Neither the latency of the N400 or the
LPC was modified by attention. Taken together these results show that the
semantic aspects of language and the harmonic aspects of music are
processed independently when listening to songs. They also strongly
suggest that language is processed first independently of the direction
of attention.

The palate pulse in particular
Jason E. Reiss & Robert F. Simons
University of Delaware
Emotions research has consistently found psychophysiological measures
that vary along valence and arousal dimensions. For example, heart
rate (HR) and facial electromyogram reliably index valence while skin
conductance is associated with arousal. Another potential emotion-
related physiological measure is pulse wave amplitude. Although this
measure has received some attention in the emotion literature, it
is more commonly featured in studies of stress and cardiovascular
reactivity. A recent study by Goldstein and Edelberg (1997, Psychophysiology,
34, 124-128) reported the development of a method to record pulse
wave amplitude from the palate. Using this new technique, they found
differential emotion-related responding at palate and finger sites
suggesting both sensitivity and specificity of the palate measure.
Between-site differences also suggest that pulse wave measures are
not simply redundant with the more traditional heart-rate measure.
In the present study, twenty-eight subjects were exposed to 6-s color
slide stimuli previously shown to range in both emotional valence
and arousal (Greenwald, Cook, & Lang, 1989, Journal of Psychophysiology,
3, 51-64). Pulse data were recorded from ear and finger reflectance
plethysmographs as well as from the palate with Goldstein & Edelberg's
new oral plethysmograph placed posterior to the upper incisors. Change
in pulse wave amplitude was calculated between two prestimulus pulses
and the average of all pulses that occurred during the viewing period.
Results indicated the usual linear relationship between HR (derived
from the ear pulse) and valence as well as a main effect for valence
(across sites) in the pulse wave data. More importantly, a site X
valence interaction indicated a significant effect for emotion valence
only at the palate and suggest that this measure may have particula
relevance to research in emotion.

Cardiovascular nonlinear dynamics and emotional responses to the Stroop test as a function of color-word interference and processing speed
Patrice Renaud & Jean-Pierre Blondin
University of Montreal
The purpose of this study was to examine the complexity and instability
of heart rate (HR) dynamics in the context of attentional effort expenditure.A
computerized Stroop Color-Word task was used to induce attentional
conflict and modulate attentional effort. Subjects (N=22) had to identify
the color of non-conflictual and conflictual Stroop stimuli. Additionally,
the task was paced by presenting execution signals either 400 ms or
700 ms after the beginning of each trial in order to force responses
of different speeds. Parameters expressing the nonlinear properties
of HR variations during task performance were extracted: Lyapunov
exponents, which express dynamic instability, and the number of hidden
neural units, which is an index of complexity. Dysphoric emotional
states (anxiety, hostility, depression) were also measured. Results
indicated that the faster 400 ms condition elicited higher HR levels
than the slower 700 ms condition (p < .01). On the other hand, there
were no differences in HR levels between the conflictual and non-conflictual
conditions. The conflictual condition, however, was associated with
less complexity, i.e., with a lesser number of hidden units (p < .05).
The conflictual Stroop condition also elicited more dysphoric emotional
reactions than the non-conflictual condition (p < .05). Furthermore,
emotional states were correlated with the number of hidden neural
units (p < .01) and Lyapunov exponents (p < .01). In conclusion, analysis
of HR nonlinear dynamics may provide important additional measures
of psychophysiological adaptations in tasks requiring attentional

Localizing infant covert attention with scalp ERPs
John E. Richards
University of South Carolina
This study examined infant shifts of covert attention, using scalp-recorded
ERP. Ten 20-week-old infants (4.5 months) were tested in a covert
attention shift paradigm. A central stimulus was presented, and two
seconds later a "cue" stimulus was presented in the periphery for
300 ms, and both stimuli were turned off. Following stimulus-onset-asychronies
(SOA) of 450, 875, or 1300 ms, a "target" peripheral stimulus was
presented ipsi- or contra-laterally to the cue stimulus. A no-cu
control conditon was done in which just the target appeared, and a
no-stimulus control was done in which no peripheral stimulus occurred.
EEG was recorded with the 10-20 system, and ERP averages were made
from similar testing trials. Behaviorally, the infants showed "covert
attention shifts". Saccade reaction time to the target stimulus on
ipsilateral trials was facilitated at short SOA's and lengthened (inhibition
of return) at longer SOA's. There was an enhanced "P1 / N1" complex
to the target when it was ipsilateral to the cue, relative to these
ERP components on the contralateral or control trials, and relative
also to the cue ERP's. Presaccadic potentials were affected by the
stimulus presentation conditions. For example, saccadic spike potentials
were larger when the target and cue were on the same side, than when
the target and cue were on opposite sides. These results demonstrate
that scalp-recorded ERP's may help identify brain areas involved in
infants' covert shifts of attention

Total respiratory resistance, cardiac activity, and ventilation during phasic emotional stimulation
Thomas Ritz, Carmen George, & Bernhard Dahme
University of Hamburg
This study was designed to investigate the direct impact of emotional
stimuli on total respiratory resistance (TRR). Sixteen non-asthmatic
volunteers viewed series of happy, neutral and depressing pictures
and self-referent statements. Each series consisted of eight affectively
homogenous stimuli. The individual stimulus was shown for 12 s, followed
by a 12 s imagination period. TRR was measured throughout the stimulus
series using the forced oscillations technique. Time and volume parameters
of the respiratory cycle, maximum heart period (HP), ventilation corrected
respiratory sinus arrhythmia, and facial EMG (corrugator supercilii,
orbicularis oculi, and masseter) were also recorded. Ratings of pleasure
and arousal were obtained for each stimulus. Evidence for an arousal
modulation of TRR changes was observed, with increases during happy
and depressing stimuli and decreases or no changes during neutral
stimuli. In contrast, maximum HP shortening was more pronounced for
happy and depressing than for neutral stimuli. For ventilation, only
respiratory drive (VT/TI) differentiated significantly between affective
categories, with increases during happy stimuli compared to decreases
during neutral and depressing stimuli. For facial EMG, a reliably
differentiation of affective categories was only found in the orbicularis
oculi site. Imagination periods with eyes closed were accompanied
by an additional number of cardiac and ventilatory adjustments. The
results are compatible with earlier research and clinical observations
which reported aggravations of asthmatic symptoms during emotional
arousal. Methodological issues of respiratory measurements in studies
of emotion will be discussed

Frontal changes in auditory ERP waves during chemotherapy and radiotherapy in children treated for acute lymphoblastic leukemia (ALL)
P. Robaey1, S. Kabene1, I. Lamothe1,2, S. Precourt1,2, & A. Moghrabi1
1Ste-Justine Hospital, 2University of Montreal
Therapeutic successes in the treatment of ALL have been accompanied
by late neuropsychological sequelae. ERPs to auditory stimuli were
recorded from 14 electrodes as a novel method to monitor brain acute
functional neurotoxicity. The child was asked to listen passively
to 900 pure tones presented via a headphone to the left ear. Two tones
(85% vs 15%) of different pitch (100 vs 1100 Hz) were presented for
20 ms in a pseudo-random order with an interstimulus interval varying
randomly between 550 and 950 ms. Electrodes were clustered in frontal,
central and parietal regions.23 children (5,58 +/-2,69 years) treated
for ALL and 8 healthy age-matched controls (4,85 +/-2,78 years) were
included in this longitudinal study. An expected P165-N2b-P3a complex
with a fronto-central topography was consistently observed. Immediately
after the diagnosis and more clearly just before brain irradiation
(day 23), the leukemic children showed a significantly smaller negativity
in the 50-150 ms post-stimulus time window over the frontal regions.
During the week following brain irradiation, and more clearly 7-9
weeks later, they exhibited a smaller positivity in the 150-250 ms
time window, also over the frontal region. The same pattern was observed
over the central region, but it was significant only before brain
irradiation and 7-9 weeks after. The latter difference could be accounted
by the presence of a large negative wave instead of the expected positivity
(P3a) in some subjects. These results suggest that ERP could be useful
to assess acute brain irradiation damage and to predict late sequelae.

Larger ERP-positivity in schizophrenics? Electrocortical correlates of learning and recall in a proactive inhibition task in schizophrenic patients and healthy controls
Brigitte Rockstroh, Patricia Roessner, Rudolf Cohen, Michael Wagner
University of Konstanz
Electrocortical and performance correlates of proactive inhibition
(PI) were examined in 15 schizophrenic patients (DSM-III-R diagnosis)
and 15 healthy subjects matched for age, sex and level of education.
Using a modified paired-associate learning task, two series of 30
pairs of words each with low associative relationship were presented
(A - B) on a TV-screen for 1 s separated by 1 s ISIs. In the following
cued recall the subject had to name the associated word "B" upon presentation
of word "A". Subsequently, PI was established by combining 50% former
"A" words with a new associate "C" and mixing these pairs with 1
new word pairs (D - E). In the final cued-re-call impaired recall
of "C"-words (upon "A"-presentation) as compared to (D-)"E-words would
indicate PI. Despite generally poorer performance of patients, a significant
PI-effect was present in both groups. The EEG was recorded from frontal,
central and parietal, midline, left- and right-hemispheric leads using
a DC-amplifier. Encoding induced a broad positivity peaking 400-600
ms after stimulus-onset, which was smaller in patients than controls.
In both groups larger positivity accompanied A-C-encoding relative
to D-E, suggesting positivity a correlate of inhibition. During recall,
presentation of the cue-words induced a markedly larger posterior
positivity in patients than to controls, who exhibited anterior negativity.
Thus, schizophrenics do not always show reduced ERPs. Their larger
positivity may indicate increased processing effort due to deficient
memory search processes (related to anterior negativity) or more pronounced
inhibitory activity to suppress various other associations during
recall due to deficient encoding. Research was supported by the Deutsche

Alexithymia: A deficit in emotional arousal?
Thomas M. Roedema & Robert F. Simons
University of Delaware
Despite its acceptance in the clinical domain, the construct of alexithymia
has received little attention in the psychophysiological literature.
The few studies that have examined alexithymics' physiological responses
to emotion-provoking stimuli are characterized by a heterogeneity
of findings and methods. The present experiment utilized well-validated
and widely- employed measures and methods in order to connect alexithymia
to the larger data base relating emotion experience to simultaneous
physiological changes. Sixty-five subjects (34 alexithymic) were presented
a standardized set of 21 color slides varying in affective content.
During slide presentation, facial EMG, heart rate and skin conductance
were measured. Emotion self- report was obtained by SAM ratings and
by asking subjects to provide adjectives that described the 'feelings'
induced by each slide. As expected, patterned facial-muscle activity
and the acceleratory component of the heart-rate response were sensitive
to variations in valence while skin conductance varied specifically
with arousal. Alexithymic and control subjects did not differ on their
valence ratings nor on the two valence-sensitive physiological measures.
They did differ, however, on the two arousal-related measures. The
range of arousal ratings was greater among controls than among the
alexithymics and this restricted range of perceived arousal in the
alexithymic group was associated with fewer specific SCRs in response
to the slides. Alexithymics also produced fewer emotion words to describe
their personal reactions to the slides and their heart-rate response
to all slides was significantly less deceleratory suggesting that
alexithymics devote less attention to the processing of affectiv

ERP correlates of sentence parsing in a language with varying word order
Frank Roesler1, Thomas Pechmann2, Judith Streb1, Brigitte Roeder1, & Erwin Hennighausen1
1Philipps-University, 1University of Leipzig
Due to explicit case marking by determiners the relative location
of subject, indirect and direct object (S-iO-dO) can vary in German
sentences. However, sentences are more difficult to process, if the
order of the noun phrases deviates from the canonical sequence, i.e.
S-iO-dO. We hypothesized that sentences with a legal but noncanonical
word order impose extra load on working memory resources and that,
therefore, a left-anterior negativity (LAN) should be evoked by the
noun phrases of such sentences. Seventeen subjects were tested with
300 semantically different sentences in which the word order was systematically
varied. Sentences were presented word-by-word with an ISI of 500 ms.
A question on the role assignments had to be answered 5 s after the
presentation of the final word of each sentence. ERPs were recorded
from 18 electrodes. The study revealed that articles which function
as case markers in German evoked a transient, left anterior negativity
whenever they indicated that a noun phrase sequence would not continue
in its canonical order. Considering the fact that the LAN was temporally
restricted to the case marker and that it resolved before the next
noun could be read, it seems unlikely that the effect indicates working-memory
storage as such. Instead, it seems more likely that it is a manifestation
of some preparatory processing step which enables storage of the forthcoming
noun. In addition to the LAN two other parsing related ERP-effects
could be observed at the very end of noncanonical sentences: A posterior
positivity and a posterior negativity which seem to indicate the inhibition
of a currently held sentence frame and, respectively, the computation
of the meaning of the proposition. Supported by the German Research
Foundation (DFG) and the Berlin-Brandenburg Academy of Sciences (BBAW).

Specificity of scaled P300 scalp amplitude distribution associated with deception
J.Peter Rosenfeld, Susan Wu, & Nahid Eftekhari
Northwestern University
Rosenfeld et al. (1996) showed that scaled P300 scalp distributions
(i.e., profiles; McCarthy &Wood,1985) are different in truth-tellers
vs. liars). Within a liar group, the profiles of honest vs deceptive
responses to the same stimuli differ; these differences hold across
paradigms. Are these profile effects specific to self-awareness of
deception, or do they reflect differing cognitive demands of truth-tellin
vs. deception? We here created 2 similar tasks, one involving deception,
the other not. DECEPTION group subjects feigned cognitive deficit
by randomly responding erroneously to 50% of the stimuli on a matching
to sample task involving 3-digit numbers. BACKWARDS group subjects
chose a random 50% of the same numeric stimuli and repeated them backwards.
P300 was recorded in response to test stimuli. Only MATCH stimuli
were to be responded to deceptively or said backwards, as appropriate.
Results presently available are from ANOVAs, with the independent
variables site (Fz,Cz,Pz) and group (DECEPTION vs. BACKWARDS). In
unscaled data, there was a significant interaction of task x site
( p<.02), and a main effect of site( p<.001). The task effect was
not significant. In scaled data, there was also a significant task
x site interaction (p<.05), and within the DECEPTION group, a response
type (honest vs. dishonest) x site interaction was found (p<.009).
In the BACKWARDS group, the parallel interaction(backwards vs fowards)
x site was not significant (p>.76).

Event-related potentials during sequence learning reveal differences in neural representation of explicit and implicit knowledge
Jascha Russeler & Frank Roesler
We recorded event-related potentials (ERPs) while subjects performed
a modified version of a serial reaction time task. Subjects responded
to eight different letters with a finger lift. Two letters each were
mapped onto one of four response fingers. In an eight element repeating
letter sequence two types of deviating stimuli replaced regular letters:
Perceptual deviants changed the perceptual but preserved the response
sequence whereas motor deviants changed both the perceptual and the
response sequence. This enabled us to dissociate perceptual and motor
components of sequence learning. Subjects were categorized postexperimentally
according to verbalizable sequence knowledge as explicit (n=9) or
implicit learners (n=10). Both groups did learn the underlying structural
regularities as reflected in a reaction time (RT) decrease for structured
compared to interspersed random stimulus blocks. ERPs revealed different
effects for both groups: The explicit group showed an enhanced N200-component
for deviant compared to standard letters and an enhanced P300-amplitude
for motor deviants. None of these effects was seen in the group of
implicit learners. The lateralized readiness potentials (LRP), on
the other hand, showed the same pattern of effects for both groups:
A shortening of LRP-onset latency for standard letters and a slight
activation of the incorrect response for motor deviants. These effects
became more expressed in the 2nd half of the experiment. These results
suggest functional differences in the acquisition and the neural representation
of sequence knowledge in subjects exhibiting different degrees of
verbalizable (i.e. "explicit") sequence knowledge. Supported by the
German Research Foundation (DFG) and the Berlin-Brandenburg Academy
of Sciences (BBAW

Application of a nonlinear dynamical analysis to the EEG in emotional states
Chang Su Ryu1, Seunghwan Kim1, Seon Hee Park1, & Min Cheol Whang2
1Electronics & Telecommunications Research, 2Korea Research Institute of Standards & Science Institute
As an application of a nonlinear dynamical analysis to the electroencephalogram
(EEG) in emotional states, we estimate the correlation dimension for
the EEG in positive and negative emotional states. The EEG's are measured
for the positive and the negative auditory stimuli. For an estimation
of the correlation dimension, the time delay for the reconstruction
of the EEG in embedding space is not determined from a single function
such as the autocorrelation function but determined to yield the largest
plateau in the local slope of the correlation integral for each embedding
dimension. In our case, the absence of an abrupt change in magnitudes
of singular values makes the singular-value decomposition useless,
which implies the absence of a dominant process in emotional states.
Our results show that the correlation exponent for the emotional state
is higher than that for the rest state even if the variation between
subjects is considered. We also obtain similar results on T3 and T4
channels, which implies that the left and the right hemispheres equally
participate in the information processing.

The effects of preattentive and attentive modes of extinction in classical conditioning
Sara Saban & Kenneth Hugdahl
University of Bergen
The present experiment investigated the effects of preattentive/ attentive
modes of extinction in classical conditioning to facial emotional
stimuli. The subjects participated in three different experimental
phases. In the first (habituation) phase they were presented with
a 500 ms. angry face for 6 trials. In the second (acquisition) phase,
for half of the subjects the face was paired with an aversive noise
(experimental group) while for the other half of the subjects the
face and the noise presentations were separated by 6-10 sec. intervals
(control group). In the third (extinction) phase, these two groups
were further divided into two subgroups. One subgroup for both the
experimental and control group had the face CS presented for 30 ms,
and immediately masked with a neutral picture. The other two subgroups
had the CS presented for 500 ms. with no mask. The experimental design
would thus allow both within and between group comparisons. Skin conductance
responses (SCRs) were recorded. The results showed that conditioning
occured in the experimental group which was indicated by a significan
difference between responses during habituation and responses during
extinction. There was no effects of conditioning in the control group.
During the extinction phase, a significant conditioning effect was
observed for both the preattentive backward masked and the attentive
modes of extinction for the experimental group. There was no significant
difference between these two modes of extinction for the experimental
group. No conditioning was observed in any of the control subgroups.
The results suggest that conditioned responses were acquired to emotional
stimuli regardless of attentive or preattentive modes of extinction.

Hold your horses: Agonists and antagonists in motor preparation
Dean Sabatinelli, Ilene Schiselman, & W.Keith Berg
University of Florida
Preparation for action has been hypothesized to activate muscles directly
involved in the upcoming response. Past studies of motor preparation
have generally examined agonist muscle activity,comparing foreperiod
muscle tension in a consistently responding limb to tension in a consistently
passive limb. This study examined electromyographic activity in the
forearms, concurrently recording both agonist and antagonist muscle
groups, in order to further examine the pattern of motor activity
during preparation to respond. In a warned reaction time task, 18
subjects held a "gripper" in each hand, responding with the speeded
closure of the cued gripper at the end of a six-second warning interval.
The responding hand varied randomly through the session, such that
in each trial, one arm prepared and responded, while the other remained
passive. Comparing muscle activity early and late in the warning interval,
agonist muscles of both responding and passive arms increased in tension,
reaching significance in the responding arm only (p=.04). Surprisingly,
however, the antagonist muscles of the responding arm changed as well,
showing a large and highly significant increase (p<.001). Antagonist
muscles in the passive arm, however, showed a small but reliable decrease
in tension during the warning interval (p=.002). Therefore, antagonist,
as well as agonist, muscles became activated in the arm preparing
to respond. In fact, the most reliable difference between responding
and passive arms was found in the relative increase or decrease of
activity in the antagonist muscles.

Abnormal ERP activity and selective performance deficit to homographs in schizophrenia
Dean F. Salisbury, Iris A. Fischer, Robert W. McCarley, Paola Mazzoni, & Martha E. Shenton
Harvard Medical School-McLean Hospital
Thought disorder is a cardinal sign of psychosis, and includes such
features as loss of abstraction ability, tangentiality, loose associations
derailment, thought blocking, and overinclusive thinking. The presence
of underlying thought disorder is inferred from bizarre or unusual
speech. A crucial question regarding thought disorder is whether it
reflects a stable dysfunction in thought processes or a transient
phasic problem in executive attention. On language tasks, a random
pattern of errors alternating with correct responses implies attention
abnormalities. In contrast, a differential pattern of errors to different
word types suggests a deficit more fundamental to language and thought.
Homographs, or common English words with multiple meanings (e.g.,
date, bank, toast), are difficult for schizophrenic patients to interpret,
since the ambiguity of the words makes them confusing. On a reading
task where subjects judged the sensibility of sentences containing
homographs, schizophrenic subjects made significantly more errors
to unusual but valid homograph meanings. In addition, the ERPs in
schizophrenic patients to these less common homograph meaning sentence
endings showed significantly more negative N400 activity than controls,
who showed significantly more positive LPC activity. These results
provide evidence for a semantic processing bias in schizophrenia,
and support the hypothesis of an underlying increase in automatic
semantic network activation in schizophrenia based on probability
of meaning usage, with subsequent failure to inhibit meaning selection
with executive context based or strategic activity.

Effect of stimulus intensity, risetime, and duration on the cardiac defense response
J. Vila, M. Sanchez, I. Ramirez, & M. C. Fernandez
University of Granada
We present three studies aimed to examine, respectively, the effect
of stimulus intensity, risetime and duration on the heart rate components
of the defense response to an acoustic stimulus along the 80 s after
stimulus onset. In the first study, stimulus intensity was manipulated
at two levels: 79dB and 109dB (duration and risetime were kept constant
at .5 s and instantaneous risetime). In the second study, stimulus
duration was manipulated at 5 levels: 50, 100, 250, 500, and 1,000
ms (intensity and duration were kept constant at 105dB and 0 risetime).
In the third study, stimulus risetime was manipulated at 5 levels:
0, 24, 48, 96 and 240 ms (intensity and duration were kept constant
at 105dB and 1000 ms. The following results were obtained: (1) The
three studies confirm the description of the response to the first
stimulus presentation as a pattern of heart rate changes with two
accelerative and two decelerative components in alternating order.
(2) Intensity increases the amplitude of the heart rate components
but does not modifies the response pattern. (3) Duration linearly
affects the first acceleration increasing its amplitude but not the
second acceleration which is only elicited by the two longest durations.
(4) Rise time does not affect either the amplitude of the component
nor the response pattern. The implications of the results are discussed
in the context of traditional assumptions concerning differentiation
between orienting, startle and defense.

Phonological expectancy effects on visually presented word recognition and event-related potentials
Nami Sanehira, Jun'ichi Katayama, & Takashi Morotomi
Hokkaido University
The N340 ERP component is readily elicited by phonological processing
of a visually presented homophone in a two-word priming paradigm,
is larger over the left hemisphere, and more evident when the prime-target
SOA was short. The present study assessed phonological expectancy
to the target-modulated N340 by manipulating homophone target stimulus
probability. The prime was always a word and the target was either
homophone word (HW), unrelated word (UW), pseudohomophone nonword
(HNW), or unrelated nonword (UNW), with a constant SOA of 1500 ms.
Prime-target pairs were employed in two probability conditions: For
the low probability conditions, the number of HW, UW, HNW, and UNW
was 30, 90, 30, 90; for the high probability conditions, the numbers
were 90, 30, 90, 30. Participants (n=12) were undergraduate students
(20-31 yrs), who responded with a button press to indicate whether
the target stimulus was a real word or not (lexical decision). In
the low probability condition, N340 elicited by HW stimuli did not
differ from those from UW stimuli, since no expectancy about target
phonology was induced. In the high probability condition, HW stimuli
elicited a smaller N340 than UW stimuli, because the probability conditions
induced an expectancy for the homophone of the prime as the target.
These results suggest that the N340 reflects the processing or the
activation level of phonological representation, which can be affected
by controlled processes such as expectancy.

The electrically-elicited eyeblink is modified by the alteration of visual stimulus parameters
A. J. Sarno1,3, A. J. W. Boelhouwer2, & T. D. Blumenthal1
1Wake Forest University, 2Tilburg University, 3University of Missouri-Columbia
The present experiment was conducted to investigate the effects of
visual pulse intensity (113.8 cd/m2, 433 cd/m2, & 2030 cd/m2) on the
modification of the electrically-elicited blink reflex at ten SOAs
(-80, -60, -40, -20, 0, 20, 40, 60, 80, and 100 ms). A negative SOA
indicates that the electrical blink stimulus preceded the visual pulse,
0 ms SOA signifies a simultaneous presentation, and a positive SOA
indicates that the electrical blink stimulus was presented after the
visual pulse. Results from 20 subjects indicated that increases in
visual pulse intensity did not affect the overall magnitude of facilitatio
or inhibition, but did affect the timing of the modification. Blink
facilitation occurred for the high, medium, and low intensity visual
pulses at SOAs of 0-20, 40-60, and 80-100 ms, respectively. Blink
inhibition occurred for the low and medium intensity visual pulses
at SOAs of 60-100 and 80-100 ms, respectively, with no inhibition
caused by the low intensity visual pulse at any SOA. The data support
the temporal summation hypothesis of sensory input with convergence
of visual and somatosensory (electrical) information. The results
suggest that the neural information produced by the three intensity
visual pulses was transmitted to a brainstem location at different
rates, supporting the conclusion that perceptual lag is a subcortical
process. The facilitation of the eyeblink in the low intensity visual
pulse condition at the same SOA (100 ms) as the inhibition of the
eyeblink in the medium and high intensity visual pulse conditions
supports the proposal that these two mechanisms of modification are

Developmental and migraine-specific aspects of the Bereitschaftspotential
Gudrun Sartory1 & Bernhard Mueller2
1University of Wuppertal, 2University Clinics Essen
Migraine in both adults and children is characterized by an elevated
contingent negative variation (CNV). To shed light on the functional
significance of this finding, the Bereitschaftspotential (BP), thought
to be partly responsible for the late CNV wave, was investigated in
children suffering from migraine (n=30), in healthy age-matched controls
(n=16), and in a group of healthy adults (n=20). The children had
a mean age of 11.7 years (sd=2.0) and the adults a mean age of 26
years (sd=4.3). Methods: Subjects pressed a button 80 times with the
right finger while the motor related potential was registered from
12 electrode sites (10-20 system) referenced to linked ears. EMG from
the right forearm and vertical EOG from the right eye were also recorded.
EEG was recorded with a time constant of 10 s and a low bandpass filter
of 35 Hz and sampled with 200 Hz from 1,700 ms before to 200 ms after
the button press. Data were corrected for EOG blink-artifacts, averaged,
and baseline corrected with regard to the initial 200 ms. Results:
Comparison of groups yielded differences in terms of topography and
amplitude. Adults showed more distinct lateralization with a more
pronounced BP at C3 than did children whose BP was more prominent
in the midfrontocentral region (Cz, Fz). Compared with the other two
groups, migraineurs also showed an elevated BP at these electrode
positions. Healthy control children exhibited a positive deflection
500ms before the button press at Cz and C3 and up to the button press
at F3 and F4, which was not evident in the other two groups. Conclusions:
The BP of healthy children thus appears to be characterized by inhibitory
potentials in the frontal area, whereas that of migraine childre
is more similar to that of adults in shape but less lateralized and
more elevated at mid-line derivations.

Caffeine effects on blood pressure and syncopal reactions in novice female blood donors
Lori A. Sauer, Christopher R. France, & Laura Wilhelm
Ohio University
A major deterrent to the retention of blood donors is the experience
of syncopal reactions (e.g., faintness, dizziness, lightheadedness).
These reactions are associated with a diphasic physiological response:
an initial increase in blood pressure associated with sympathetic
arousal followed by a sudden drop in blood pressure. Due to its potent
pressor effect, caffeine was investigated as a modulator of risk for
syncopal reactions to blood donation. Participants were 41 healthy
female first-time blood donors who were regular caffeine consumers.
On the blood donation day, participants were administered either 0
mg or 250 mg of caffeine prior to donation in a randomized double-blind
fashion. Repeated measures of resting blood pressure and heart rate
were obtained in the laboratory before caffeine administration, 45
min later at the donation site prior to blood donation, and immediately
post-donation. Syncopal reactions were assessed immediately post-donation
using the Blood Donation Reactions Inventory. This inventory is an
11-item scale of common blood donation reaction symptoms. Participants
receiving the 250 mg dose of caffeine reported significantly fewer
syncopal reactions [t(39) = 2.03, p < .05]. Further, symptom scores
were negatively correlated with systolic (r = -.30, p < .05) and diastolic
(r = -.42, p < .05) blood pressure change from pre-caffeine to post-donation,
suggesting that caffeine's pressor effect was associated with fewer
negative donation symptoms. These findings suggest that a moderate
dose of caffeine helps prevent negative reactions to blood donation
and may encourage repeat donation.

Physiological reactivity moderates stress-induced negative mood
Angela Scarpa1, Bruce H. Friedman2, Kadee Smallee3, & Kristen A. Luscher1
1University of Georgia, 2Virginia Polytechnic Institute and State University, 3Eastern Washington University
Independent lines of research have linked increased physiological
reactivity with a) the expression of negative emotions such as depression
and anxiety, or b) the repression of negative emotions such as anger
and hostility, with corresponding implications for psychological and
physical health. This study clarifies these divergent findings by
examining the relationship of sympathetically-mediated cortisol reactivit
and vagally-mediated heart rate variability (HRV) with the expression
of negative mood before and after a stressful task. Self-report mood
ratings (N=53), HRV (N=46), and saliva samples for cortisol (N=15)
were obtained before and after an uncontrollable stressor task. Participants
were assigned to groups based upon their reactivity tendencies in
these physiological measures from pre- to post-task. Individuals who
increased their cortisol levels reported higher tension (p=.016),
frustration (p=.012), anger (p=.033), and sadness (p=.036) both before
and after the stressor, which suggests that chronic levels of negative
mood are associated with sympathetic reactivity. Those who increased
HRV reported higher levels of anxiety (p=.03), confusion (p=.026),
anger (p=.01), and depression (p=.014) only after the stressor, which
indicates that situationally-appropriate mood changes were associated
with parasympathetic modulation of HRV. These results suggest that
the relationship of physiological reactivity to negative mood varies
as a function of the autonomic mediation of the response measure.
Specifically, sympathetic reactivity may reflect a dispositional tendency
towards negative affect, while vagal control of HRV may index the
state-dependent self-regulatory ability to express negativity after
stressful situations. These findings are consistent with the literature
and help integrate discrepant views of mood and physiological reactivity.

Baroreflexes mediate blood pressure and heart rate increases during hypnotically induced negative emotions
Hartmut Schaechinger, Julia Schuerch, & Wolf Langewitz
University of Basel
The investigation of emotion-related physiological changes is hampered
by the difficulty to obtain reliable mood states. Hypnosis may be
used to elicite mood states with high intensity. After two pre-measurement
training sessions hypnosis was induced in 20 healthy volunteers (age:
30±7 yrs., 9 females), and suggestions were given about previously
defined events with a positive (PE) or negative (NE) emotional content
(counter-balanced sequence). Non-invasive blood pressure (BP) (Finapres
system), heart rate (HR), and skin perfusion (laser doppler flowmetry)
were recorded continuously during successive 5 minutes periods. A
43-item questionnaire was used to describe mood changes. Power spectra
of HR and systolic BP variability were calculated by discrete Fourier
transformation. Cross-spectral analyses between systolic BP and inter-beat-interval
length were performed to determine baroreflex sensitivity (BRS) (modulus
transfer function in the frequency range of 0.07 to 0.14 Hz). Systolic
and diastolic BP (5.4±1.8/2.7±1.1 mmHg, p<0.008/p<0.03) and HR (3.0±1.2
bpm, p<0.02) were higher during NE as compared to PE. HR variability
in the high frequency range (0.15-0.40 Hz) was lower (-0.24±0.08 MI**2,
p<0.007) during NE indicating a shifting sympatho-vagal influence
towards less parasympathetic cardiac control. During NE BRS (-3.37±1.2
ms/mmHg, p<0.01) and skin perfusion (-10.2±4.8, p<0.05) were decreased.
Correlation analysis revealed that 'rage' was the most powerful single
item from the questionnaire explaining emotion-related physiological
changes. Conclusions: BP and HR changes during hypnotically induced
negative emotional states are mediated via reductions in BRS and vagal

Psychophysiological work strain during multi-tasking human-computer interaction
Florian P. Schaefer, Ralf Schaefer, & Wolfram Boucsein
University of Wuppertal
Since modern computer operating systems provide multi- tasking features,
long computer processing times do not necessarily force the user to
wait since they can be utilized performing another work step. Although
strain imposed by forced waiting is avoided, additional effort is
needed for an optimal scheduling of parallel tasks, producing new
psychophysiological strain. In the present study 40 students performed
a mocked power plant control center setup procedure. A panel showed
36 displays to be set to particular initial values, some of which
were to be requested from a remote data bank by transmission lines.
During the processing times for those requests being 10s for half
of the subjects and 30s for the other half, the setup procedure could
be either continued or additional requests could be started. As an
additional within subjects factor, feedback on the progress of opened
requests was either not provided or given as static or dynamic display.
EDA, EKG, PVA and respiration were continuously recorded, and keystrokes
and request handling were written on log-files. Subjective ratings
were taken after each feedback condition. Mixed-design ANOVAs for
effects of processing time and feedback type showed that the multi-tasking
features were more often used when the processing times were long,
where subjects also reported more activation and pain symptoms. For
short processing times, both frequency and amplitudes of nonspecific
EDRs were significantly increased, indicating emotional strain. Cardiovascular
parameters did not show significant effects. Compared to our former
results with single- tasking systems, the psychophysiological patterns
seem to be reversed.

Heightened acoustic startle reflex in adult children of alcoholics
Steven L. Schandler, Whitney V. Leach, Isabel NuÒez, & Michael J. Cohen
Department of Veteran Affairs Medical Center-Long Beach
Recent studies of alcohol and information processing indicate that
alcohol disrupts attentional operations. However, rather than onl
the result of alcohol consumption, disrupted attentional operations
appear in children of alcoholics who are at the greatest risk for
developing alcoholism. One hypothesis is that recovering alcoholics
and persons at risk for alcoholism experience hyperarousal of attentional
processing mechanisms, reducing the ability to scan and select relevant
information. The present study employed the startle reflex to directly
examine the attention hyperarousal hypothesis in children of alcoholics,
and, if present, to more precisely evaluate the generalization of
attentional hyperarousal across stimuli of varying intensities. Thirty
healthy nonalcoholic volunteers, aged 22-45 years, served as participants.
Participants were matched on intelligence, years of school completed,
and neurological status. However, one group of 15 participants came
from families in which alcoholism was present in the biological father
and at least one other relative. The second group of 15 participants
had no family history of alcoholism. The startle eye-blink response
was recorded during pseudorandomized 40-msec presentations of a white-noise
stimulus presented at 90, 96, 102, 108, and 114 dB SPL. Compared to
children of nonalcoholics, the startle-blink responses to all acoustic
stimuli were significantly higher for children of alcoholics. The
effect was very robust, with very low variation. The results support
the hypothesis that the information processing deficits displayed
by children of alcoholics reflect hyperarousal of their general attentional

Neural basis of visual perceptual learning in humans
Karen M. Schrank, Murtaza F. Singaporewala, Aliasgar M. Dhoon, & Nancy K. Squires
State University of New York at Stony Brook
Psychophysiological studies of perceptual learning in adults have
indicated that practice at performing a difficult visual discrimination
task can result in long-term changes in the visual system. This follow-up
investigation utilized monocular presentations in order to further
localize the neural area responsible for the visual perceptual learning.
In order to trace the origins of the changes to V1, visual evoked
potentials (VEPs) were recorded in a visual orientation discrimination
task where the subjects learned to differentiate between vertical
and horizontal targets within an array of diagonally oriented distractors.
The stimuli were displayed monocularly for 25 ms, followed by a 20
ms mask, with an SOA of 155 ms. Subjects were given a series of six
to seven blocks of trials in which the stimulus was displayed 75 times.
On blocks 1-4 stimuli were presented to either the left or right eye,
after which presentation was switched to the opposite eye. VEPs were
recorded from the O1, Oz, and O2 sites over the visual cortex utilizing
the International 10-20 System. Electrodes were also placed at FPZ
and the back of the neck for reference and ground, respectively. The
vertical target, which was presented approximately 15% of the tim
was detectable with an accuracy of 70-100% by the third or fourth
trial, thus indicating a learned response. A grand average of subject
VEPs indicate a change in the morphology of the VEP between 50-200
ms after the onset of the stimulus by the fourth trial. Results will
be discussed in terms of the monocular training implications and transference
within the V1 area.

Crossmodal effects in the processing of task-irrelevant location changes
Erich Schroeger & Andreas Widmann
University of Leipzig & University of Munich
Frequent standard light-sound stimuli were distally presented from
a constant position (lateralization either plus or minus 15 ; distance
2 m; duration 100 ms; stimulus-onset asynchrony 700 ms). Infrequently,
a location deviant was interspersed, in which the sound, the light,
or both were emitted from the contralateral position. In the Attend
Auditory condition (AA), the participants had to press a response
button to auditory location deviants, whereas they had to respond
to visual location deviants in the Attend Visual condition (AV). In
AA, hit rates (false alarm rates) were 97.2% (1.9%), and 97.6% (1.5%)
in AV. The event-related potential results revealed enhanced deviance-related
negativities (DRN) in the 100-300 ms range and enhanced positivities
(DRP) in the 300-500 ms range. In the case of the auditory deviants,
the early part of the DRN (100-200 ms) reversed polarity at the mastoids
(nose reference). Most interestingly, the DRN and DRP effects were
even elicited by deviants of the unattended modality, that is, by
task-irrelevant visual deviants in AA and, vice versa, by task-irrelevant
auditory deviants in AV. The attentional influence mainly affected
the late part of the DRN (200-300 ms) and the DRP. In the case of
auditory deviants, the early part of the DRN (100-200 ms) was not
reduced when auditory deviants were task-irrelevant. Results suggest
that task-irrelevant changes in the unattended modality are automatically
processed when task-relevant and task-irrelevant modality are embedded
in the same object.

Quick pics: Emotional pictures and cortical ERPs
Harald Schupp, Bruce Cuthbert, Margaret Bradley, Mark McManis, Charles Hillman, John Cacioppo, & Peter Lang
University of Florida
Previous research has demonstrated a robust effect of affective picture
processing on event-related potentials (ERPs): Pleasant and unpleasant
pictures prompt a larger positive slow wave compared to neutral, wher
all pictures are equiprobable. In contrast, in a modified oddball
paradigm developed by Cacioppo and associates, evaluatively inconsistent
stimuli (e. g. an aversive picture embedded in a series of pleasant
ones or vice versa) elicit a large late positive potential (LPP).
Accordingly, the goal of this study was to investigate ERP modulation
as a function of equiprobable affective stimuli, using the short inter-stimulus
interval and other parameters of the modified oddball paradigm. Stimuli
included 20 pleasant, 20 neutral, and 20 unpleasant pictures selected
from the International Affective Picture System. Pictures were shown
for 1.5 s (3 s ISI) and presented in 30 blocks of 6 images each, with
a random order of picture pleasantness. The task was to evaluate each
picture as either pleasant, neutral, or unpleasant. EEG was recorded
from 9 sites, rereferenced off-line to linked ears. Consistent with
earlier findings, the LPP was more pronounced for pleasant and unpleasant
pictures compared to neutral. This effect emerged around 300 ms. and
lasted approximately 400 ms. Further analysis indicated that pleasant
and unpleasant pictures rated high in arousal elicited larger LPPs
than low-arousing pictures. The results show clearly that both the
intrinsic affective properties of the pictures and the context of
presentation are important factors in modulating ERPs to affective

Digital filtering of brainstem and mid-latency auditory ERPs: Enhanced identification and test-retest reliability
Steven B. Schwarzkopf1, Elna M. Nagasako1, Kirsten M. VanMeenen1, Gregory A. Light2, & Steven M. Silverstein1
1University of Rochester, 2University of California-San Diego
Auditory brainstem responses (ABRs) and mid-latency responses (MLRs)
have proven diagnostic utility in neurologic and audiometric settings.
In contrast, ABR and MLR studies in psychiatric syndromes have not
yielded results consistent enough to for clinical utility. Due to
heterogeneity within psychiatric syndromes, potential importance of
more subtle abnormalities than those found in neurologic/audiometric
settings, and greater difficulty in testing patients with psychiatric
conditions, signal enhancement is critical for this line of research.
Toward this goal, we tested the efficacy of digital filtering techniques
for improving waveform identification and test-retest reliability
of ABRs and MLRs in normals. Twelve subjects were tested twice for
ABRs with a one week inter-session interval. Nine subjects underwent
MLR testing three times with a one week between session interval.
Finite impulse response (FIR) digital filters with varying bandpass
and roll-off characteristics were applied to the averaged waveforms.
"Optimal" bandpass filtering resulted in improved peak detection for
ABRs and MLRs. Measurement was enhanced by eliminating problems with
multiple peaks (high frequency noise) and "merged" waves (low frequenc
distortion). Test-retest reliability was greatly enhanced for MLRs
using a bandpass filter with sharp roll-off at 10 and 50 Hz. For ABRs,
attenuating activity greater than 1,500 Hz enhanced identification
and generally improved reliability. Contrary to expectations, attenuation
of activity less than 300 Hz did not enhance reliability. These results
point to specific filtering approaches that may significantly enhance
ABR and MLR waveform identification as well as the accuracy of amplitude
and latency measures.

ERP correlates of categorical perception: Cross-linguistic variation
Valerie L. Shafer1,2, Bhuvana Karunakaran2,3, Richard G. Schwartz2,3, & Diane Kurtzberg2,3
1Hofstra University, 2Albert Einstein College of Medicine, 3City University of New York
This study investigated categorical perception of native and non-native
phonetic contrasts using behavioral and electrophysiologic (mismatch
negativity [MMN]) measures. A place of articulation continuum based
on Hindi bilabial, dental and retroflex contrasts was created by resynthesizing
edited versions of a naturally produced dental /da/. F2 and F3 transition
starting frequencies varied in 7-step intervals of 100 Hz and 50 Hz,
respectively. Native Hindi speakers divide this continuum into 3 contrasts:
bilabial /ba/, dental /da/ and retroflex /Da/; the dental and retroflex
speech sounds are not contrasted in English. Both the English and
Hindi subjects were required to group the sounds into 3 categories.
Individual variation in categorization and discrimination were seen
in all subjects, and, interestingly, a number of the English speakers
performed similarly to the Hindi subjects. A sharp category boundary
was revealed between /ba/ and /da/ at the same location for most of
the subjects. All subjects, even Hindi, showed large individual differences
for the boundary between dental and retroflex, although for the Hindi
subjects, these categories were more distinct. The amplitude and latency
of MMN were largest and earliest for the stimuli that differed the
most acoustically. However, there was not a linear relationship between
MMN characteristics and absolute acoustic differences across the steps
in the continuum; the individual categorization and discrimination
patterns influenced the magnitude of MMN. Thus, the ERP and discrimination/identification
functions reflect both acoustic differences and individual perception

Light cigarette smokers have lower pain thresholds than non-smokers, but heavy smokers don't!
David Sheffield, B. Todd Granger, Paula L. Biles, Paula F. Miller, W. Maixner, & David S. Sheps
University of North Carolina at Chapel Hill
Although many studies have reported an association between smoking
and pain in patient populations, results of pain threshold studies
have been equivocal. The mechanisms underlying these possible differences
may include blood pressure (BP) stimulation of baroreceptors and increased
beta-endorphin activity. We examined these variables in 19 deprived
smokers and 19 nonsmokers (all men, mean age=31 years). Following
a 15 minute rest period, BP, cotinine levels and thermal pain sensitivity
were determined. Thermal stimuli ranging from 35 to 50oC were delivered
in 0.5oC increments. Subjects stated when they first felt pain (threshold)
and when the temperature was too hot (tolerance). Subjects then rested
for 30 minutes, after which they smoked a cigarette, and BP, cotinine
levels and pain sensitivity were assessed. These procedures were repeated
after 10 minutes, following a second cigarette, and after 30 minutes,
following a final cigarette; nonsmokers did not smoke, but rested.
Based on their baseline cotinine levels, smokers were divided into
light and heavy smoker groups and compared with nonsmokers. ANOVAs
revealed that pain threshold (p<.05) and tolerance (p<.01) differed
across the three groups at different periods. Heavy and non-smokers
did not differ at any time period (all p>.10), whereas light smokers
had lower tolerance and threshold than the other groups at baseline;
these measures increased after the first cigarette. These differences
were mirrored by increases in systolic and diastolic BP (all p<.05),
but not beta-endorphin levels (p>.10). In conclusion, deprived heavy
smokers resembled nonsmokers, whereas light smokers differed. We speculate
that BP stimulation of baroreceptors may mediate these antinociceptive

An event related potential study of spatial abilities in Turner syndrome
Janet L. Shucard, Gregory L. Ciupak, Teresa Quatrin, & David W. Shucard
State University of New York at Buffalo
It has been theorized that the visual-spatial deficits present in
Turner Syndrome (TS) are due to the effects of missing chromosomal
material and to subsequent deficient perinatal/postnatal estrogen
exposure. It has been shown that early exposure to sex steroids in
animals influences the development of spatial abilities, and is also
thought to play a role in the neurocognitive differences between genders
in humans. While deficits in spatial ability are well documented in
TS, little is known about differences in functional brain organization
related to these cognitive deficits. In the present study, it was
hypothesized that ERPs would reflect different patterns of cerebral
functional organization among TS females, and male and female controls
during performance of visually presented spatial and verbal tasks.
Subjects were 28 control boys and girls (14 each) determined to be
prepubertal by Tanner Scale staging, and 14 TS females. ERPs to tone
probe stimuli were obtained from 13 scalp sites while subjects performe
a dot localization, line orientation, 2-D rotation and nonsense word
recognition task. Control boys and girls showed different condition
dependent patterns of lateralization. Boys tended to use more right
hemisphere resources, particularly for the spatial tasks, while girls
tended to use more left resources, particularly for the verbal task.
These effects were strongest at temporal, central and parietal scalp
sites. TS females, however, showed an overall absence of hemispheric
asymmetry during all tasks. The results suggest differences in cortical
organization among the three groups that most likely are established
early in development.

Breast-feeding alters the blood pressure response to the cold pressor task
Elizabeth Sibolboro Mezzacappa1 & Edward S. Katkin2
1Columbia-Presbyterian Medical Center, 2State University of New York at Stony Brook
Previous human and animal investigations have found that breast-feeding
or lactation is associated with decreased self-reports of stress and
decreased hormonal response to stressors. The present study examined
the effects of breast-feeding on the blood pressure response to a
one minute cold pressor task. Blood pressure was continuously monitored
at baseline and during the cold pressor in 13 exclusively breast-feeding
mothers and 8 mothers who had never breast-fed. All mothers had an
infant under one year old. For systolic pressure, breast-feeding mothers
had a mean decrease of 3.00 mmHg (sd = 3.44), while mothers who had
never breast-fed had a mean increase of .38 mmHg (sd =2.07). For diastolic
pressure, breast-feeding mothers had a mean decrease of .92 mmHg (sd
= 1.12), while mothers who had never breast-fed had a mean increase
of .25 mmHg (sd = 0.71). Both differences were significant (p < .05).
Impedance cardiography data showed no differences in cardiac output
measures between the groups, suggesting that the differences in blood
pressure responses were due to differences in total peripheral resistance.
The results indicate that lactation may alter vascular reactivity
to stressors.

Electrophysiological correlates of configural and automaticity effects in perceptual organization
Steven M. Silverstein, Steven B. Schwarzkopf, Geoffrey Nowlis, John Chapman, Scott Nuernberger, Lindsay S. Schenkel, & Robert Chapman.
University of Rochester
The ability to perceive a set of visual stimuli as belonging to a
single object is related to both the prepotent figural goodness of,
and prior exposure to the stimulus set. We previously developed a
pattern recognition task (PRT) in which we assessed (using accuracy
and RT indices) the ability to determine whether a visual patter
was previously seen as a function of both experience and configural
stimulus properties. On each of 600 trials, subjects saw a 2-D pattern
of six asterisks and were asked to determine whether they had seen
the pattern before. Two patterns, one with strong configural properties
and one without, were each repeated 120 times at random intervals
throughout the session. The remaining 360 trials consisted of non-repeating
patterns. Our previous results indicated parallel automaticity curves
for configural and nonconfigural patterns, with consistently superior
performance for configural patterns. In the current study, we recorded
ERPs from several sites during performance of a 480-trial version
of the PRT. At Cz, P300 was of greater amplitude for the configural
repeated pattern than for the nonconfigural repeated pattern, which
was in turn of greater amplitude than the P300 for the non-repeating
patterns. The same hierarchy of amplitudes was also evident between
100 and 200 ms after stimulus delivery. This pattern suggests involvement
of relatively early neural mechanisms in the discrimination of good
from poor form patterns, and repeated from nonrepeated visual stimuli.

Emotion-processing in three systems: The medium and the message
Robert F. Simons, Benjamin H. Detenber, Thomas M. Roedema, & James L. Poston
University of Delaware
It has been shown in a variety of experiments and in several different
laboratories that consistent and specific relationships exist between
two emotion dimensions (valence and arousal) and self-report, physiological
and overt-behavioral responses. Pictures and imagery varying in valence
prompt differential heart-rate and facial muscle activity and significantly
modulate startle-probe magnitude. Likewise, stimuli that vary in perceived
arousal also vary in evoked skin-conductance activity, are recalled
more reliably and are viewed longer under free-operant conditions.
These consistent relationships between stimulus content (the message)
and emotion topography tell only part of the story, however. There
is evidence that stimulus form (the medium) can also influence the
nature of the emotion response, and studies of emotion and media properties
such as screen size and color have been recently reported. The present
experiment focused on another important presentation attribute --
stimulus motion -- and explored its impact on both the subjective
and physiological components of the induced emotion. Thirty-four subjects
were exposed to moving and still versions of stimuli which varied
widely on perceived valence and arousal. Each stimulus was presented
for six seconds while facial-muscle, heart-rate and skin- conductance
activity was monitored. Subjects provided self-reports of valence
and arousal (SAM) shortly after stimulus offset. Consistent with the
literature, facial-muscle and heart-rate change were associated with
stimulus valence; skin conductance was associated with arousal. Thi
relationship held for both moving and still images. The impact of
motion was dramatic. Self-report and physiological data strongly suggested
that motion increased arousal, had little impact on valence and captured
and maintained the subject's attention throughout the presentation

Movement related potentials and rate force development
S. M. Slobounov & W. J. Ray
The Pennsylvania State University
It was generally reported that the amplitude of movement-related potentials
(MRP) are higher before ballistic than before ramp movements. However,
the rate of force development (e.g., peak force of individual trials
over time-to-peak force, Pf/Tp) was not experimentally manipulated
to support this finding. This study investigated whether different
rates of force development influence the spatio-temporal characteristics
of MRPs. Twelve healthy subjects performed the index finger force
production tasks where force levels and time to reach target force
were experimentally manipulated. The force trajectory feedback was
provided from computer screen. A Syn- Amps amplifier was used to record
continuous EEG at sites Fz, F3, F4, Cz, C3, C4, Pz, P3 & P4 referred
to as linked mastoids. Fifty sweeps associated with force production
tasks were averaged following artifact reduction. It was found that
neither force level itself nor time to the target force predicted
the amplitude of the MRPs. Rather, the rate of force development predicted
the peak amplitude of MRPs (r = .65, p < .01). Also, the shape of
MRPs changed as a function of task, with the tendency to grow throughout
the trial until the peak force was attained. Finally, there was the
lack of contralateral preponderance of the MRPs over precentral and
frontal electrode locations. A rather symmetrical distribution of
the slow negative shift with maximum amplitude at Cz electrode location
was observed. This study adds to current knowledge about movement
related potentials in that differences in MRP characteristics could
predict the dynamics properties of forthcoming movement.

Brain-computer interface based on non-orthogonal wavelet transform of EEG
S. Slobounov & R. Tutwiler
The Pennsylvania State University
Several algorithms and methods for automatic classification of non-averaged
single trial EEG as a prerequisite for constructing the brain-computer
interface (BCI) have been developed recently. We have proposed a new
algorithm for BCI based on non-orthogonal wavelet transform for time-frequenc
analysis of EEG associated with various movements. In this algorithm,
the BCI signal analysis stage receives signals from the NeuroScan
system and produces outputs that characterize input signal patterns.
Once the wavelet transform is computed, features are extracted from
the matrix of wavelet coefficients. The original time-frequency map
is adaptively thresholded using energy measures to create a binary
image of the dominant features. The feature vector is used as input
to the pattern recognition system that classify the specific movement
vector associated with the time-frequency footprint. The classification
stage consists of an adaptive learning network. The adaptive learning
network consists of a hierarchical structure that incorporates self-organizing
feature maps into a nodal network structure. Genetic algorithms and
fuzzy logic is applied to improve the system performance. The input/output
relationships tend to be nonlinear, so the system is well suited to
a "self-organizing" system. The output of the classifier consists
of a movement vector used by the visualization system to control the
motion of 3d objects on computer screen. It is hypothesized that the
BCI might be used not only by healthy subjects to move a cursor on
the monitor to the left or right but also as a valuable tool for visualization
of impaired joint(s) restricted movability resulted from various disabilities.

Job strain and ambulatory blood pressure in a healthy community sample: Failure to replicate
Leslie Smithline, Thomas W. Kamarck, Saul S. Shiffman, Hayley Thompson, Jeff Goodie, Jean Paty, John Kassel, & Maryann Gnys
University of Pittsburgh
Stress in the workplace, or "job strain" (defined by Karasek and colleagues
as high psychological demand and low decision latitude), has been
associated with various indices of hypertension and coronary heart
disease. Previous findings have demonstrated that men facing higher
job strain showed elevated ambulatory blood pressure readings at work.
We examined the association between job strain and blood pressure
obtained at work, home, and the clinic, in a normotensive sample.
120 subjects (52.5% women, 50% African-American), ages 23-50, were
assessed with ambulatory blood pressure monitors every 45 minutes
during waking hours over a 6 day period. Following each blood pressure
reading, subjects recorded their location on a palmtop computer. Clinic
blood pressure readings and job strain measures were also obtained.
Job strain was derived from the Job Content Questionnaire in a manner
similar to previous research. Inconsistent with previous findings,
we found that high job strain was not associated with larger differences
between average clinic and work blood pressure. In fact, among men,
job strain was associated with lower SBP and DBP at work, home, and
the clinic (SBP, F=8.94, p<.01; DBP, F=6.71, p=.01). Controlling for
race, body mass index, and age did not account for this unexpected
association. In women, we found that high job strain was associate
with elevated clinic DBP (F=5.21, p<.05), but with no other blood
pressure indices. Explanations for these findings will be explored.

Early temperamental predictors of adolescent behavior and cardiovascular physiology
Nancy Snidman1, Carl E. Schwartz2, & Jerome Kagan1
1Harvard University, 2Harvard Medical School
Previous research has described two behavioral categories of children:
inhibited and uninhibited following initial encounters with unfamiliar
events. These two groups also differ in the magnitude of peripheral,
ANS reactivity to mild stress. The purpose of the present study was
to examine the relation between a classification of behavioral inhibition
in the second year of life and the behavioral and physiological profile
of these children at adolescence. Seventy-nine of the original 106
subjects were available for participation at 13 years. Coded behaviors
included number of smiles and spontaneous comments during a series
of cognitive tasks. Cardiovascular measures included ECG and blood
pressure. Results for physiological measures included: a significant
difference between inhibited and uninhibited children, following the
laboratory tasks, in heart period (F(3,63)=4.51, p< .05), and in systolic
blood pressure between a sitting and standing posture (F(3,75)=6.47,
p<.05). There was also a relation, in males, between a cardiovascular
sympathetic index, comprised of standing heart rate and systolic blood
pressure, and classification: boys originally classified at 2 years
as inhibited had a greater sympathetic index at adolescence than boys
classified as uninhibited (t=2.8(34), p< .01). For females there was
a stronger relationship between concurrent behavior and cardiovascular
measures. Fewer smiles and comments was related to a larger cardiovascular
sympathetic index (r=-.43, p<.01). These data demonstrate that important
behavioral and physiological aspects of the original inhibited temperamental
profile have been preserved and that gender differences are important
to consider.

Evaluation of emotional changes based on the wavelet transform of EEG signals evoked by tactile stimulation
Jin-Hun Sohn1, Jae J. Im2, Dae-Im Kang3, Ji-Eun Kim1, Kyung-Hwa Lee1, & Hyung S. Yeo2
1Chungnam National University, 2Inje University, 3Korea Research Institute of Standards and Science
We are exposed to the various stimuli input from the environment,
which cause emotional changes based on the characteristics of the
stimuli. To find the physical and mental responses to a particula
stimulus, it is necessary to analyze physiological signals which reflect
the emotional changes. Also, it is critical to use appropriate methods
for data analysis to extract meaningful information from rapidly changing
physiological data. The objective of this study was to evaluate emotional
changes evoked by tactile stimulation in EEG signals using wavelet
transform. The experiment was devised with three conditions, control,
tactile stimulation, and subjective rating. Twenty-one subjects were
used to obtain EEG signals while applying five different types of
materials, dry cotton, wet cotton, vinyl, paper, and sand paper. Stimulation
was applied to the left hand of each subject using tactile stimulator.
EEGs were acquired from F3 and F4 electrode location. Wavelet transform
was employed to localize the EEGs both in time and frequency domain.
The results showed that the frequency contents of the EEG signals
appeared at a higher frequency ranges when the stimulation was applied.
Also, spectral energies were distributed to the higher frequency ranges
as the subjects experienced an unpleasant emotion. Moreover, above
results were observed clearly at the electrode site F4. The conclusion
drawn from this study indicates that EEG characterization using time-frequency
energy distribution could be used as a useful baseline for establishing
algorithm which classifies subjects' psychophysiological states. (This
work was supported by a grant 17-01-01 from Korean Ministry of Science
and Technology)

Time-frequency analysis of EEG signals evoked by auditory stimulation for the evaluation of subjective comfort
Jin-Hun Sohn1, Jae J. Im2, Imgap Yi1, Eun-Kyung Ryu1, & Hyung S. Yeo2
1Chungnam National University, 2Inje University
Various external stimuli were continuously input from the environment,
and they cause subjective emotional changes based on the characteristics
of the stimuli. Unfortunately, there are no quantitative results on
the relationship between human sensibility and the characteristics
of physiological signals. The purpose of this study was to quantify
EEG signals evoked by auditory stimulation. The theory was based on
the assumption that the analysis of the variability on the characteristics
of the EEG waveform may provide the significant information regarding
changes in psychological states of the subject. The experiment was
devised with eleven experimental conditions, which were control and
ten different types of auditory stimulation, five positive ones and
five negative ones. Thirty subjects were used to obtain EEGs from
O1, O2, T3, T4, F3, and F4 electrode sites while introducing auditory
stimulation. Multiresolution wavelet transformation was employed to
construct a time-frequency energy distribution map of EEG signals.
As a results, reconstructed signals at the decomposition level revealed
the different energy values on the EEG signals obtained from T4, right
hemisphere of temporal lobe. Also, general patterns of EEG signal
in rest state compare with negative and positive auditory stimulus
were found to have significant differences at p<0.05. This study could
be extended to establish an algorithm which distinguishes subjective
comfort of the subjects exposed to the auditory stimulation. (This
work was supported by a grant 17-01-03 from Korean Ministry of Science
and Technology)

A study of discrete emotions using the International Affective Picture System
Jin-Hun Sohn1, Ae-Reyoung Oh1, Ji-Eun Kim1, Kyung-Hwa Lee1, & Jae J. Im2
1Chungnam National University, 2Inje University
The International Affective Picture System (IAPS) developed by Lang
and colleagues (manual, 1997) is a world-widely adopted tool in studies
relating a variety of physiological indices to emotions induced by
the presentation of standardized pictures of which subjective ratings
are well established in the 3 dimensions of pleasure, arousal, and
dominance. Especially, it has great utility compared to other research
methods containing verbal and/or written reports when used in non-English
speaking countries. In the present study we investigated whether distinctive
EEG characteristics for six discrete emotions can be found using the
IAPS pictures. In a preliminary study we determined 12 pictures out
of IAPS that scored highest subjective ratings in terms of discrete
emotions, i.e., happiness, sadness, fear, anger, disgust, and surprise.
Consequently, each emotion category contains 2 pictures. These pictures
as visual stimuli were randomly given to 40 right-handed Korean college
students (20-26 years old) with 60 sec of exposure time and 30 sec
of inter-stimulus interval for each picture while EEG signals were
recorded from F3, F4, O1, and O2 referenced to linked ears. Fast Fourier
Transformation and Wavelet algorithms were used to analyze the EEG
data. There were significant differences between positive and negative
emotions and also among negative emotions mostly in the right frontal
and occipital cortices . It seems possible to find EEG indices for
discrete emotions using the pictures of IAPS although it was first
developed for dimensional approaches to emotion. Supported by KOSEF

A constitutional type associated with decreased variability in EEG
Jin-Hun Sohn1, Imgap Yi1, & Chul-Joong Kim2
1Chungnam National University, 2Korean Research Institute of Standards and Science
It has been known that there are great individual differences in EEG
characteristics both at resting states and during and after treatment
of certain stimuli. The purpose of the present study was to elucidate
some psychological and constitutional factors such as anxiety level
personality type, and Korean Sasang Chejil constitution type which
is suspected to show distinctive EEG characteristics in subgroups.
Fifty-three right-handed college women(20-24 years old) participated.
Each subject was given 10 random-ordered sequential auditory stimuli
with 30 sec of ISI. Five were pleasant sounds and the other five unpleasant
ones. EEG was recorded from O1, O2, T3, T4, F3, and F4 referenced
to linked ears at resting state and during the entire period of the
stimuli presentation. The most pleasure- and displeasure-inducing
sounds were subjectively decided by each subject and then their corresponding
EEG was analyzed. There appeared significant differences in EEG among
the constitutional types, personality types and anxiety levels respectively
at resting state and also during emotional state evoked by auditory
stimuli. It means that the variability in human EEG signals is due
to all these three factors. Interestingly, there were statistically
significant differences in most recording sites and also in EEG bands
according to the constitutional type rather than to the others. It
is recommended that the Korean Sasang Chejil constitution type should
be also considered in order to make the subject group homogeneous
for EEG researches. Supported by KOSEF #96-0101-02-01-3.

The locus of the grouping effect during overlapping task performance
Werner Sommer, Hartmut Leuthold, Rasha Abdel-Rahman, and Eva-Maria Pfuetze
Humboldt University-Berlin
In overlapping tasks subjects respond to each of two stimuli (S1,
S2) presented at short stimulus onset asynchronies (SOA). Whereas
reaction time to S2 (RT2) usually decreases with increasing SOA, the
SOA effect on RT1 depends on the subject's strategy. When S1 processing
is given priority RT1 is independent of SOA whereas RT1 increases
with SOA, when subjects postpone R1 until R2 is specified (response
grouping). In the present study the lateralized readiness potential
(LRP) - indicating specific response activation - is used to determine
the locus of R1-postponement. If grouping merely concerns postponement
of response execution, SOA should affect RT1 and R1-locked LRP onset,
but not S1-locked LRP onset. However, if both response activation
and execution is postponed, SOA should affect RT1 and S1-locked LRP
onset, but not R1-locked LRP onset. Manual choice reactions were required
to high or low tones (S1) with the left or right index fingers and
to the symbols X or O with the left or right middle finger. S1 was
followed by S2 at SOAs of 100, 400, or 700 ms. Twelve subjects gave
priority to S1 processing, whereas twelve others postponed R1 until
R2 was known (grouping). In addition to the LRPs the vertical and
horizontal EOG was registered. Only trials with correct responses
and free of EOG artifacts were evaluated. Reaction times confirmed
the use of the required strategy because RT1 was independent of SOA
for S1-priority and it increased with SOA under grouping instructions
With S1 priority, both S1- and R1-locked LRP onset were unaffected
by SOA. With grouping instructions there was again no SOA effect on
S1-locked LRPs but now, a clear delay as a function of SOA appeared
for R1-locked LRP onsets. Thus, when subjects group responses in overlapping
tasks, they first select and activate the response to S1 and then
postpone its execution until R2 is specified.

Activation (BAS) and inhibition (BIS) processes during Wisconsin Card Sorting in pre-adolescent children: A heart rate and skin conductance study
Riek Somsen & Maurits van der Molen
University of Amsterdam
Gray has proposed that two important interactive brain systems that
motivate behaviour. The behavioral activation and inhibition brain systems
are crucial regulators of behaviour. Fowles (1988) provided evidence that
Heart Rate acceleration is connected to behavioral activation, while Skin
Conductance responses are elicited during behavioral inhibition. Recently
Iaboni , Douglas, & Ditto (1997) studied reward and extinction responses in
ADHD children. AHDH children did not exhibit Skin Conductance responses to
extinction, suggesting that these children have a weak inhibition system.
The prefrontal cortex plays an important role in the control of higher
cognitive -executive- functions such as strategic planning and inhibition.
Executive behaviour is measured with the Wisconsin Card Sorting test. This
test demands adaptive activation and inhibition of earlier behaviour.
Failure to inhibit an earlier strategy results in perseverative behaviour.
Welsh, Pennington, and Groisser, (1991) showed that in young children a
relatively high proportion of WCST responses is perseverative. This
perseverative response tendency decreases with age. We hypothesized that,
the WCST responses in children may be sensitive to maturational changes in
prefrontal functions. To test this hypothesis the WCST test was presented
to pre- and post adolescent children while measuring Heart Rate en Skin
Conductance responses
The results showed a high variation in perseverative responses in the
pre-adolescent group. A marginal significant interaction in the number and
amplitude of the Skin Conductance responses suggested differential BIS
activity between low and high perseverative children. Fasic Heart Rate
responses were sensitive to negative feedback (HR deceleration) and to rule
searching behaviour (HR acceleration.)

The development of respiratory sinus arrhythmia
Riek Somsen & Maurits van der Molen
University of Amsterdam
The systematic variation in heart rate with respiration, Respiratory Sinus
Arrhythmia, is controlled by centres in the brainstem which, in turn, are
modulated by higher cortical centres. This RSA control mechanism has an
hierachical organization. Higher centres modulate the activity of lower
reflex systems. Evidence comes, from the work of Spyer who showed that
autonomic reflex gating may occur by hypothalamic influences on brainstem
centres that control autonomic reflexes. Higher control centres depress
reflexive RSA activity. During certain sleep stages RSA is relatively high,
while during wakeful stages RSA is highly reduced.
Further evidence comes from developmental studies. In children RSA is much
higher than in adults. This suggests that with maturation the involvement
of higher cortical influence on RSA increases. In an earlier study we
reported that during middle childhood (5 to 12 years) the amplitude of the
RSA spectrum during rest conditions did not change. A second study across a
much broader age range (from 5 years to 12 years to adult) indicated that
the greatest change (reduction) in RSA amplitude occurs during adolescence.
The question how RSA develops during adolescence was further investigate
Different groups of children between 11 and 16 years and adults performed
various rest, and information processing tasks. The results showed that RSA
amplitude during rest and task did not change much until around age 13 and
then decreased. Squared coherence between heart rate and respiration
consistently increased with age, while the peak frequency of RSA decreased
with age. We suggest that the late maturation of specific cortical area's
may influence developmental RSA decrement.

Postural stability of blood pressure and hemodynamic responses to a foot cold pressor test
Thomas W. Spalding, Pedro Ribeiro, Mary E. J. Lott, & Regina T. Richards
University of Maryland at College Park
The postural stability of blood pressure and hemodynamic responses
to a foot cold pressor test (CPT) were assessed in 34 normotensive
men (mean age of 21.9 yrs). Each subject completed a 15-min rest period
and a 2-min CPT in each of two postures (counterbalanced): supine
and standing. Blood pressure (mm Hg), heart period (HP, ms), stroke
volume (SV, ml), cardiac output (CO, L/min), and total peripheral
resistance (TPR, pru) averages were obtained for the last 2 min of
each rest period and the CPTs. SBP and DBP increased during the CPT
in the standing (12.4 and 7.5, respectively) and supine postures (14.5
and 11.5, respectively), p's < .001. DBP reactivity was less in the
standing than in the supine posture (p < .001) but SBP reactivity
was similar across postures. HP and SV decreased during the CPT in
the supine posture (-139.0 and -14.7, respectively; p's < .001) but
did not change in the standing posture (-16.9 and -0.5, respectively).
Both responses were greater in the supine than in the standing posture
(p's < .001). CO did not change in the standing (0.07) or supine (-0.09)
postures. TPR increased during the CPT in the standing (0.12) and
supine (0.18) postures (p's < .001) but the difference in postures
failed to reach significance (p = .09). The results indicate that
posture may partially explain lab-field differences in DBP reactivity
to a cold stressor but cannot explain such differences in SBP reactivity.
Additionally, TPR responses were stable across postures but HP and
SV responses were not.

DSP-4, a selective noradrenergic neurotoxin, reduces the amplitude of peak 2 of the averaged auditory evoked potential in the albino rat
C. M. Specht, M. E. Starick, & D. W. Shucard
State University of New York at Buffalo
The locus coeruleus (LC), which sends noradrenaline (NA)- containing
fibers throughout the neuroaxis, has been widely implicated in modulating
attention, vigilance and arousal. Our laboratory has focussed on fast
habituation (FH), the reduction of auditory evoked potential (AEP)
amplitude across repeated tone pairs, as an index of attention. The
present study was designed to test the hypothesis that NA modulates
FH. Male Sprague-Dawley rats were surgically implanted with permanent
skull electrodes. One week later, they were injected with either a
single dose (50 mg/kg, ip; 20 mg/ml) of DSP-4 (used to selectively
lesion NA neurons of the LC) or an equal volume of vehicle. They were
tested 2 weeks later. Subjects received 60 pairs of identical pure
tone stimuli (1000 Hz, 100 ms, 65 dB SPL). Tones within the pair were
separated by 2 s and tone pairs were separated by approximately 10
s. Resulting AEPs were averaged separately for each tone. The amplitude
of the second positive-going AEP peak (i.e. Peak 2; N1-P2) was determined
for each tone. Results showed that while typical FH occurred in the
control (vehicle) group, the amplitude of Peak 2 was reduced across
both tones in the experimental (DSP-4) group. These findings lend
support to the notion that brain stem NA fibers appear to modulate
FH. Since the nucleus LC projects the only NA-containing fibers to
the neocortex and the hippocampus, it appears likely that these location(s)
may be possible generator(s) of FH. These data provide evidence that
supports our postulate that FH likely reflects attentional processes,
in that a link has been established between FH and a system in the
brain associated with attention.

Absence of auditory evoked potential slow habituation across 60 paired-tone trials in the albino rat
C. M. Specht, M. E. Starick, & D. W. Shucard
State University of New York at Buffalo
When paired tones are presented to humans or animals, the averaged
auditory evoked potentials (AEPs) in response to each tone have comparable
wave forms with corresponding components that have the similar latencies.
Some components, however, have a smaller amplitude in response to
the second tone. This phenomenon is referred to as fast habituation
(FH). Our laboratory has shown that FH occurs in both humans and animals
when paired tones are presented with a 2 s interstimulus interval
and an interpair interval of approximately 10 s. Further studies in
the albino rat have determined that the reduction in amplitude of
Peak 2 (N1-P2) is not accompanied by an increase in single-trial latency
variability across tones. Studies of FH have focused on these change
in amplitude from tone 1 to tone 2 without investigating possible
changes in amplitude within tone but across individual trials. This
issue is important for determining the functional significance of
FH in studies of attention and/or cognition. In the present investigation,
we surgically implanted permanent skull electrodes in 14 male albino
rats. Three weeks later, we presented 60 pairs of tone stimuli (1000
Hz, 100 ms, 65 dB SPL) to each subject and separately averaged the
resulting AEPs into blocks of 20 trials each. Amplitudes for the 3
pairs of AEPs were compared. While AEPs in response to tone 1 had
higher amplitude than those in response to tone 2, there was no reduction
in amplitude across the three blocks for either tone. Further, the
amount of FH (as measured by amplitude difference across tones) did
not change across the three blocks. Thus, FH appears as a reliable
response that is consistent across at least 60 paired tone trials.

ERP measures of interhemispheric interaction in selective attention
Kevin M. Spencer, Marie T. Banich, & Michael G. H. Coles
University of Illinois at Urbana-Champaign
It has been suggested that the right hemisphere is capable of attending
to both sides of space whereas the left hemisphere (LH) attends to
only the right side, which raises the issue of how a single attentional
focus emerges from the interaction of the hemispheres. We investigated
this question in a study using event-related brain potential and overt
response measures recorded from normal individuals, who were divided
into two groups (each N=12) according to their tendency to engage
one hemisphere more than the other regardless of task demands, as
measured by a perceptual bias test. Stimuli were presented simultaneously
in the left and right visual fields (LVF and RVF), and participants
responded to rare and frequent stimuli under Attend-LVF, Attend-RVF,
and Divided Attention conditions. Interhemispheric interaction was
measured by comparing the condition in which one hemisphere received
a rare stimulus and the other a frequent stimulus with the condition
in which both hemispheres received rare stimuli. In the Attend-LVF
condition, an error rate effect was evident for both subject groups,
and the LH-bias group also showed amplitude effects on the N2 and
P2 components. In the Attend-RVF condition, an amplitude effect was
found on the N1 component for the LH-bias group. In the Divided Attention
condition, both groups showed reaction time effects, while the LH-bias
group also demonstrated P300 component amplitude and error rate effects.
The results suggest that interhemispheric interaction affects several
stages of selection, depending upon task requirements, and is modulated
by individual hemispheric biases.

Event-related potential amplitudes in schizophrenia patients during the degraded stimulus continuous performance task: Preliminary results
Scott R. Sponheim1,2 & Kent M. Kodalen1
1Veterans Affairs Medical Center, 2University of Minnesota
Studies have shown schizophrenia patients to have deficient signal/noise
discrimination in identifying degraded stimuli during a continuous
performance task. This study examined the amplitudes of event-related
potentials (ERPs) elicited during the degraded stimulus continuous
performance task to identify ERP abnormalities that may reflect a
disturbance in signal/noise discrimination in schizophrenia patients.
Schizophrenia and control subjects were seated in a sound- attenuated
room and instructed to press a button to indicate when the number
'0' appeared on the screen during a series of 480 numbers. All stimuli
were degraded visually and presented with a duration of 29 msec, an
intertrial interval of 1 second, and target probability of .25. EEG
data were recorded from subjects at all 19 10-20 scalp locations and
C3', C4', TP9, TP10, FT7, FT8, FT9, and FT10. The 27 channels of EEG
data were collected at a sampling rate of 500Hz with a 60Hz notch
filter. Data were corrected off-line for vertical EOG, and blocks
with visually identified high frequency artifacts were excluded (e.g.,
muscle activity). EEG data were digitally filtered with frequency
cut- offs set at 15 Hz (48dB/octave; low-pass) and .5 Hz (48dB/octave;
high-pass). Date were epoched and averaged with amplitude values set
relative to a 100 msec prestimulus baseline. Preliminary results suggest
that schizophrenia patients exhibit diminished P300 amplitudes to
target stimuli compared to normal controls. Data support P300 amplitudes
as being maximally decremented at parietal sites in schizophrenia.

Scalp distribution of EEG frequency abnormalities in schizophrenia: Preliminary results
Scott R. Sponheim1,2 & Kent M. Kodalen1
1Veterans Affairs Medical Center, 2University of Minnesota
Recent studies have shown that schizophrenia is associated with structural
abnormalities in the left superior temporal gyrus and hypoactivation
of the frontal regions of the brain. This study examined the distribution
of low-frequency and alpha band EEG abnormalities in schizophrenia
patients to determine whether the EEG anomalies in schizophrenia are
evident over the left temporal and frontal areas of the brain. Three
minutes of eye-closed resting EEG were recorded from schizophrenia
patients and normal control subjects at all 19 10- 20 scalp locations
and TP7,TP8,TP9,TP10,FT7,FT8,FT9,FT10,AF3,AF4. The 29 channels of
EEG data were collected at a sampling rate of 250Hz with a 60Hz notch
filter. Data were corrected off-line for vertical and horizontal EOG,
and visually inspected for high frequency artifacts (e.g., muscl
activity). EEG data were then divided into 8.2 second epochs and submitted
to Fast Fourier Transform to calculate relative power within delta,
theta, alpha, and beta bands at each scalp location. Preliminary analyses
suggest that schizophrenia patients exhibit augmented low-frequency
power and diminished alpha band power over the frontal regions of
the brain.

LRP topography in blocked vs. mixed stimulus-response compatibility tasks
Johannes E. A. Stauder
University of Montreal
Psychophysiological studies typically address the chronometric aspects
of Lateralized Readiness Potentials (LRPs), but ignore its topography.
The present study points out that LRP topography can provide relevant
information on the mechanisms underlying Stimulus Response Compatibility.
Subjects (n=15) were presented blue or black arrows at the centre
of a monitor in three different tasks: (1) A compatible blocked task
in which the subject was asked press a key with the hand corresponding
to the direction of the arrow, while ignoring its colour; (2) An incompatible
blocked task in which the subject was asked to respond with the hand
opposite to the direction of the arrow, while ignoring its colour;
and (3) A conditional mixed task in which the subject was asked to
respond with the hand corresponding to the direction of the arrow
to one colour and with the hand opposite to the direction of the arrow
to another colour. The EEG was recorded from 30 electrodes across
the scalp. There was a classical RT difference between the compatible
and incompatible instruction of the blocked tasks, but no significant
RT difference between the same conditions in the mixed task. In contrast,
the LRP showed a larger amplitude difference between the compatible
and incompatible condition in the mixed task than in the blocked tasks.
The difference LRPs (compatible minus incompatible condition) for
the blocked versus the conditional instruction were identical for
LRP latency (290 ms), amplitude, and LRP topography. The latter strongly
supports the notion of automatic response activation processes underlying
SRC, with a maximum lateralization at centro-parietal locations.

Brain mapping of bilateral stimulus processing and force control in ADHD and control children
Juerg Steger1, Daniel Brandeis1, Jachen Denoth2, & H.-Ch. Steinhausen1
1University of Zurich, 2Swiss Federal Institutes of Technology
Simultaneous bilateral activation of homologous extremities is typically
slower or less forceful than unilateral activation. These bilateral
interactions apparently result from neuronal inhibition. It has bee
hypothesized that inhibitory processes are reduced in children with
ADHD (attention deficits hyperactivity disorder). We studied such
bilateral interactions in a reaction time task with unilateral and
bilateral GO and NO-GO trials using combined brain mapping of event-related
potentials (32 channel ERPs) and continuous force measurements. Bilateral
interactions were assessed as deviations from the additivity; i.e.
by comparing ERP maps and force curves in the bilateral condition
to the sum of those from the unilateral conditions. Microstates (topographically
stable epochs) of stimulus and response related averages were analyzed
for ADHD (N=12, mean age=11.12 years) and control children (N=12,
mean age=11.08 years). Trials with incorrect or ambiguous force responses
were excluded. Controls had later force onsets in the bilateral than
in the unilateral condition for both hands as hypothesized (p < .01).
The ADHD children showed this delay only for the dominant hand (p
< .05). Stimulus-related ERPs showed microstates with lateral asymmetries
related to the stimulus side during P1/N1 and P300 microstates, and
significant bilateral interactions in P1 and P300 microstates. Response-related
ERPs showed significant lateral asymmetries related to the response
side (hand) in all microstates (ca. +/- 300 ms), and bilateral interactions
around response onset. A condition-specific topographic group difference
for a pre-response microstate suggested some deficits of ADHD in premotor
processing. Supported by the Swiss Nat. Sci. Found. (3100-043790.95).

Pupillary sensitivity and occupational exposure to solvents
Stuart R. Steinhauer, Lisa A. Morrow, Ruth Condray, & Annette Kasparek
University of Pittsburgh School of Medicine/VAMC
Clinical impairment following toxic exposure to organic solvents has
been associated with CNS dysfunction as indicated by autonomic and
electrocortical measures. Similar effects might occur in workers who
are exposed to noxious chemical work environments, but who have not
yet reported significant clinical symptoms. Pupil diameter was measured
as 23 male union painters and 29 controls (3 female) from other unions
(mean ages 39.0 and 36.6, respectively) performed a modified auditory
oddball paradigm. Subjects first performed a Counting task, and then
performed a Choice Reaction task; in both tasks, event probability
was varied. There were no significant differences between groups in
the phasic dilation following stimuli, or in overall pupil diameter.
However, there was a significant interaction of group by task for
prestimulus pupil diameter (p < 0.02). While controls showed the typical
pattern of decreased diameter - habituation - during the second (Choice
Reaction) task, painters showed an increase in diameter during performance
of the Choice Reaction task as compared to Counting. This effect was
greater for those 11 painters who had been working within the previous
66 hours than for those who had not been painting for at least 4 days
The same finding has been reported previously for patients with toxic
exposure to organic solvents. The pupillary findings suggest that
even subclinical chemical exposure may result in a need for greater
focusing of attention, or generation of greater anxiety, when more
demanding task demands are placed upon such individuals.

Efficiency of post-hoc subject selections in psychophysiological emotion research
Gerhard Stemmler & Cornelia A. Pauls
University of Marburg
Emotion research is confronted with the problem of unresponsiveness
of some subjects to experimental emotion inductions. Therefore, post-hoc
subject selections using, e.g., emotional self-reports are recommended
as a "gold standard" before physiological emotion effects are analyzed.
We evaluated different selection criteria according to subject overlap,
statistical power, effect size, and physiological patterns. N=158
subjects were studied in a crossed Emotion (anger vs. fear) x Group
(treatment [TR] vs. control [CO]) design with real-life inductions;
29 somatovisceral variables, emotional self-reports (ESR), emotion
appraisals (EA), and verbal reports of dominant feelings (VR) were
registered. For the fear (anger) induction, 80-93% (84-100%) of subjects
scoring high on fear (anger)-ESR, stating fear (anger)-VR, or reporting
fear (anger)-EA belonged to the treatment group. Statistical power
to detect physiological differences between responders and nonresponders
within selection groups varied markedly (average during anger for
TR-CO: 0.64, ESR: 0.63, VR: 0.48, EA: 0.46), as did effect size (TR-CO:
0.36, ESR: 0.35, VR: 0.23, EA: 0.24). Somatovisceral profile levels
distinguished between responders (higher levels) and nonresponders
(lower levels) within selection groups but not between TR subjects
and responders in ESR, EA, or VR. Fear-anger comparisons of somatovisceral
profile pattern differences between responders and nonresponders were
significant for Group (p=.008), ESR (p=.015), and EA (p=.021), but
not for VR. Profile patterns were not different between TR and responders
in ESR, EA, or VR, nor between CO and nonresponders in ESR, EA, or
VR. It is concluded that post-hoc subject selections in emotion research
do not necessarily pay.

Pubertal influences on sex differences in cardiovascular reactivity
Catherine M. Stoney1, Bruce S. Alpert2, Stephen M. Patterson2, & Sammie Walker3
1Ohio State University, 2University of Tennessee-Memphis, 3West Tennessee Regional Health Office
Several studies have shown sex differences in the patterning of cardiovascula
stress responses among adults. Generally, men show larger BP responses
to stress, while women show similar or larger HR stress responses.
The patterns of these sex differences vary with individual differences
in reproductive hormones, type of stressor employed, ethnicity, and
with the cardiovascular measures which are indexed. Nonetheless, the
majority of studies have identified consistent sex differences in
cardiovascular reactivity among adults. A previous study (Matthews
& Stoney, 1988) showed that sex differences are also apparent among
post-pubertal children, but not pre- pubertal children, when examined
cross-sectionally. To our knowledge, no other investigation has examined
this issue in a longitudinal design. The purpose of this study was
to longitudinally test sex differences in blood pressure and heart
rate reactivity among pre- and post-pubertal children. 347 boys and
girls were tested yearly from the 3rd-12th grades in a standard reactivity
protocol, while HR and BP were measured. Results showed that pre-
pubertal boys and girls have similar SBP reactivity, while post-pubertal
SBP reactivity is greater in boys than girls (Gender x Time p<.001).
The greater DBP reactivity in boys was consistent across pubertal
stage (Gender p<.001), while the HR reactivity decline with puberty
was larger in girls than in boys (p<.001). These data support the
notion that the psychological and physiological changes which occur
during puberty affect cardiovascular reactivity, and are consistent
with the notion that gonadal hormones influence reactivity. Supported
by HL35788, HL48222 and an Ohio State University College of Social
and Behavioral Sciences Small Grant award.

Changes in lipid metabolism during stress
Catherine M. Stoney & Linda Bausserman
Ohio State University & Brown University
Several studies have shown that acute stress generates significant
increases in the atherogenic lipids (low density lipoprotein-c, triglycerides,
total cholesterol) and apolipoproteins (apolipoprotein B). While several
individual difference variables are related to the pattern and magnitude
of the lipid response, the mechanisms responsible for the increases
are unknown. One possible mechanism is that the enzymes responsible
for metabolizing lipids are altered during stress. The purpose of
this study was to test the hypothesis that lipoprotein lipase activity
(LPLA) and hepatic triglyceride lipase activity (HTGLA) are lowered
during acute psychological stress. Heparin-stimulated lipase activity
was measured in plasma. 27 men and women were tested in three separate
sessions over one week. During session 1, a standard acute laboratory
reactivity protocol was initiated, while lipids were measured. During
session 2, heparin-stimulated lipase activity was measured during
a non-stress period. During session 3, heparin-stimulated lipase activity
was measured during an acute stress reactivity protocol; all session
were counterbalanced. As expected, significant increases in the atherogenic
lipids occurred during session 1 (all p's<.05). Analysis of the metabolic
data indicated that total lipase activity was lower (M=25.91) during
stress, relative to rest (M=27.19), with the changes most prominent
for LPLA, the enzyme which hydrolyzes triglycerides and aids in LDL-c
metabolism. These data show that lipase activity is altered during
acute psychological stress, and suggest that the significant elevations
in the atherogenic lipids during stress may be due to diminished metabolic
clearance of lipids. Supported by HL48363, HL44847, and the Ohio State
University General Clinical Research Center #M01-RR0034-37.

Anger expression style and risk for cardiovascular disease
Catherine M. Stoney & Tilmer O. Engebretson
Ohio State University
We previously described a particular style of expressing anger, which
we termed Anger Expression Style (AES). The AES scale determines the
relative dominance of Anger-in and Anger-out. A negative score on
the measure indicates a greater use of nonexpressive behaviors when
angry, while a positive score indicates a greater use of expressive
behaviors when angry. Scores clustering around zero indicate cross-situational
flexibility in expressive behaviors when angry. We have shown that
those who have flexible AES scores have significantly lower atherogenic
lipid concentrations, relative to those who have more inflexible AES
scores (Engebretson & Stoney, International Journal of Behavioral
Medicine, pg. 281-298, 1995). The current study further explores the
relationship of AES with other cardiovascular risk factors. 127 individuals
enrolled in a longitudinal study of the effects of chronic (occupational)
and acute (serial subtraction and speech) stress on lipids were examined.
Results showed greater total cholesterol and LDL-c reactivity to both
chronic and acute stress among those who typically are nonexpressive
when angry, relative to those with a more flexible anger expression
style (all p's<.05). Health behaviors, including dietary and exercise
behaviors, did not discriminate the AES groups. However, those with
the less flexible anger expression styles were significantly more
likely to have a positive family history of CHD (p<.04), were more
likely to be depressed (p<.03), reported higher perceived stressed
(p<.002), and higher negative affect (p<.04). These data suggest that
a rigid style of anger expression is associated with several physiological
and psychological risk factors for CHD.

Cardiac reactivity during appetitive picture processing: Dietary restraint and ambivalence
Werner G. K. Stritzke1, David J. Drobes2, Alan R. Lang3, Christopher J. Patrick3, & Peter J. Lan
1University of Western Australia, 2Medical University of South Carolina, 3Florida State University, 4University of Florida
Researchers have often used a picture viewing paradigm to examine
physiological and subjective reactions to emotionally-provocative
material. Recently, this paradigm has been used to examine reactions
to appetitive cues. We report on research from two studies that examined
cardiovascular reactivity to appetizing pictures of food. In one study,
we compared a group of highly restrained eaters to non-restrained
participants (n=36). In the other study, we compared a group of individuals
who reacted to food pictures with ambivalence (i.e., high approach
and high avoidance ratings) to a control group of nonambivalent responders
(i.e., high approach and low avoidance ratings) (n=32). Food pictures
were embedded within a series of slides ranging in affective valence
from highly pleasant to highly unpleasant. Heart rate change scores
were computed as the mean difference between the 6-s picture viewing
period and the 1-s preslide baseline. Participants characterized by
restraint (Study 1) or ambivalence (Study 2) showed significantly
elevated heart rate responses to food pictures, whereas control participants
in both studies showed a pronounced heart rate deceleration. In contrast,
all groups showed marked heart rate deceleration during viewing of
non-food, appetitive pictures (i.e., erotica). The elevated heart
rate shown by restrained or ambivalent participants suggests that
they may have engaged in more mental activity when presented with
food-related stimuli, perhaps due to the simultaneous activation of
two motivational systems (i.e., appetitive and aversive). Implications
of these findings for recent models of craving will be discussed.

P300, slow wave, and short vs. long inter-stimulus intervals
Daniel Strueber1 & John Polich2
1University of Bremen, 2The Scripps Research Institute
Event-related brain potentials (ERPs) were elicited with visual oddball
and single-stimulus paradigms that employed inter-stimulus intervals
(ISI) of 2.5 or 30 sec. The oddball paradigm presented both target
(.20) and standard stimuli (.80) in a random series; the single-stimulus
paradigm presented a target without a standard stimulus. Subjects
(8F, 8M; M=20.5 yrs) were instructed to respond by pressing mouse
keys whenever a target or standard stimulus was presented in the oddball
task and to press a single key whenever the target was presented in
the single-stimulus paradigm. ERP waveforms were morphologically distinct
for the 2 compared with 30 sec ISI conditions, with multi-peaked slow
wave activity observed for the 30 sec ISI condition of both paradigms.
Mean amplitude values were computed within each of three latency window
that corresponded to the different peaks: 300-500 ms for P300; 500-700
ms for slow wave early (SWE), and 700-900 ms slow wave late (SWL).
P300 mean amplitude was largest in the oddball paradigm for both ISIs,
followed by the single-stimulus 30 sec ISI condition. SWE and SWL
mean amplitudes were largest for the 30 sec ISI conditions in both
paradigms. The findings indicate that relatively long ISIs (30 sec)
induce pronounced slow wave activity, which is not found for typically
short ISIs (2.5 s). It is concluded that P300 and slow wave ERPs are
distinct components, and that SW processes may require longer ISIs
to be observed readily.

Risk for hypertension and hypoalgesia: Paternal versus maternal history
Shannon Suchowiecki & Christopher R. France
Ohio University
Previous studies have demonstrated that individuals at risk for hypertension
report decreased pain when exposed to various noxious stimuli. The
present study examined pain ratings during cold pressor and forearm
ischemia in individuals with and without a parental history of hypertension,
with an emphasis on possible differences as a function of paternal
versus maternal history of hypertension. Participants were 208 healthy
undergraduate men and women. In a two-day protocol, a one minute cold
pressor task was completed on day 1 and a five minute forearm ischemia
task was completed on day 2. Post-task pain ratings were obtained
using the McGill Pain Questionnaire (MPQ). For both cold pressor and
ischemic tasks, separate 2 x 2 ANOVAs of MPQ pain ratings were conducted
for paternal history [2 Paternal History (hypertensive, normotensive)
x 2 Gender (male, female)] and maternal history [2 Maternal History
(hypertensive, normotensive) x 2 Gender (male, female)]. Paternal
history analyses revealed significant main effects for paternal history
and gender, but no significant interactions. Specifically, a paternal
history of hypertension was associated with lower MPQ scores for cold
pressor [F(1,192)=8.5, p<.01] and ischemia [F(1,187)=6.2, p<.01].
Further, men reported lower pain ratings than women for both cold
pressor [F(1,192)=28.0, p<.001] and ischemia [F(1,187)=4.4, p<.05].
In contrast, maternal history analyses failed to reveal significant
main effects of maternal blood pressure history on either cold pressor
or ischemic pain ratings. These findings suggest that previous evidence
of decreased pain perception in offspring of hypertensives may be
restricted offspring of hypertensive fathers.

An investigation of auditory stream segregation using event-related brain potentials
Elyse Sussman1,2, Walter Ritter2, & Herbert G. Vaughan2
1City University of New York, 2Albert Einstein College of Medicine
There is uncertainty concerning the extent to which the perceptual
organization of sound by acoustic properties is a function of attentive
or pre-attentive mechanisms of the auditory system. The mismatch negativity
(MMN), a component of event-related brain potentials which indexes
pre-attentive acoustic processing, was used as a probe to determine
where in the stages of auditory processing the perceptual organization
of high and low tones to streams occurs. A sequence of six different
high (H) and low (L) tones (H1, L1, H2, L2, H3, L3, etc.) were alternated
at a rapid pace, to induce a streaming effect and at a slow pace,
wherein the tones were heard as alternating high and low pitches.
When the tones were perceptually segregated, a pattern of standards
emerged separately in each stream (e.g., H1, H2, H3), with a deviant
pattern occurring infrequently within each stream (e.g., H3, H2, H1).
During the experiment, subjects ignored the stimuli by reading a book.
A MMN was observed to the deviant within a stream, created by the
rapid stimulus presentation, indicating a pre-attentive locus for
the stream segregation effect.

ERP indices of threat detection in a visual oddball task
David M. Sutherland & Martin F. Regan
University of Dundee
Research using emotional stimuli has shown that ERPs are influenced
by arousal, while the effects of valence are equivocal. There is evidence
however that information processing resources are directed towards
threatening stimuli. This study utilized a visual oddball task to
examine the effects of valence on ERPs to task- relevant stimuli.
ERPs were recorded from 7 leads (Fz, Cz, Pz, C3, C4, P3, P4) in response
to positive and negative valence slides (matched for arousal), drawn
from the International Affective Picture System. Slide exposure time
was 2,500 ms. Subjects (n=18) were instructed to detect rare target
stimuli (20%) presented within a sequence of opposite valence frequent
stimuli (80%). There were two conditions: negative target, positive
target. Different slides were used in each condition and condition
order was counterbalanced. Positive slides were rated as more pleasant
than negative slides (p < .001), and there was no difference in arousal.
Response time was not affected by valence. ERP analyses revealed that
N100 amplitude was not influenced by valence. However P200 amplitude
was larger for positive than negative stimuli (p < .05). A late positive
component (LPC: mean latency = 530 ms) displayed larger amplitude
for negative than for positive stimuli, particularly at parietal leads
(p < .01). Both P200 and LPC amplitudes were larger for targets than
for non- targets (p < .001). The LPC valence effect is interpreted
as evidence of a threat detection mechanism, whereby additional processin
resources are allocated to negative stimuli in the environment.

Asymmetry in prefrontal glucose metabolism during appetitive and aversive emotional states: An FDG-PET study
Steven K. Sutton, R. Terry Ward, Christine L. Larson, James E. Holden, Scott B. Perlman, & Richard J. Davidson
University of Wisconsin-Madison
Electrophysiology research has shown greater left anterior activation
during appetitive emotional states, and greater right anterior activation
during aversive emotional states. The specific cortical territories
asymmetrically associated with appetitive and aversive emotion remain
unclear because this approach has limited spatial resolution. Positron
emission tomography (PET) with 18 -fluorodeoxyglucose (FDG) as the
radiotracer was used to address this issue. Eight female volunteers
participated in three sessions. Aversive emotion was generated via
(a) participation in the initial session (e.g., initial FDG injection,
continual blood draws, PET scanning); and (b) a picture presentation
of threatening objects and situations (e.g., mutilated body, poisonous
snake). Pictures of desirable objects or situations one might approach
(e.g., delicious food, puppies) were used to generate appetitive emotion.
The aversive and appetitive pictures were presented during either
the second or third session. Neutral pictures were shown during the
initial session. Acoustic startle eyeblink reflex, corrugator supercilii
activity, and self-report measures showed that aversive emotion was
greater during the initial and aversive picture sessions relative
to the appetitive picture session. Statistical parametric mapping
analyses showed greater metabolic activity in right dorsolateral prefrontal
cortex (BA 9) in these aversive conditions. Asymmetry analyses showed
this difference in activity was significantly greater than in homologous
left prefrontal cortex, where no significant difference was observed.
In the appetitive condition, greater metabolic activity was observed
in left inferior prefrontal cortex (BA 47). Asymmetry analyses showed
this difference in activity was significantly greater than in homologous
right prefrontal cortex, where no significant difference was observed.

Multivariate analyses of sleep hygiene and polysomnographic sleep parameters
Tomoka Takeuchi1, Maki Inugami2, Kaneyoshi Ishihara3, Kazuhiko Fukuda4, Yukari Yamamoto5, & Katuo Yamazaki5
1Japan Society for the Promotion of Science, 2Tokyo Metropolitan Institute for Neuroscience, 3Notre Dome Seishin University, 4Fukushima University, 5Waseda University
The aim of present study was to reduce many aspects of sleep hygiene
to a few factors (Sleep-related Habit (SH) scales) by factor analysis,
and to investigate the relationship between SH scores and physiological
sleep parameters. The "Sleep-related Life Habit Inventory of the Tokyo
Metropolitan Institute for Neuroscience" consists of 60 items evaluating
habits, attitudes, preferences, and self-estimated feelings relating
to sleep hygiene. This inventory was administered to 2408 undergraduates
or graduates. Using factor analysis (varimax rotation) and repeating
item correction, three factors composed of 21 items were extracted.
Factor 1 ("Phase") had nine items corresponding to "regular- irregular"
type(Taub, 1978) and "morningness-eveningness" type (Horn & Ostberg,
1976), Factor 2 ("Quality") had six items corresponding to "good-poor
sleeper" type (Monroe, 1967), and Factor 3 ("Quantity") had six items
corresponding to "long-short sleeper" type (Hartmann et al., 1971).
Internal consistency was confirmed by the first principal component
analysis in each scale. Canonical correlation analysis on 47 subjects
who slept in the laboratory revealed a relationship between three
SH scales and 21 physiological parameters (15 standard sleep measures
plus 6 rectal temperature variables). Results showed that "Phase"
was more closely related to sleep parameters than "Quality" and "Quantity"
scales were. Canonical structure suggested that irregular and delayed
"Phase" and poor "Quality " were related to several polysomnographic
measures: longer sleep latency, shorter and interrupted REM sleep,
decreased number of NREM-REM cycles, increased stage 2 percentage,
and lower rectal temperature. This suggests a strong relationship
between psychometric and physiological sleep measures.

Sex-related differences indexed by P300 (P3) amplitude in visual and auditory sustained attention tasks
Ayda Tekok-Kilic, Jeanine M. Sparks, & David W. Shucard
State University of New York at Buffalo
Previously we studied hemispheric involvement under different letter
sequence conditions within a Continuous Performance Task (CPT-AX).
Results showed that lateralized effects were dependent on the sex
of the subject. In the A-X (Go) condition, males had higher right
hemisphere P3 amplitudes across all leads while females had higher
left hemisphere amplitudes. In the A-not-X (NoGo) condition, males
showed similar amplitudes over both hemispheres while females exhibited
a greater P3 amplitude over the right hemisphere. In order to extend
these findings to the auditory modality, we developed both visual
and auditory CPTs. Twenty male and 20 female college students (18-35
years) participated in the study. Number of stimuli, types of stimuli
(letters) and stimulus sequences were identical for both modalities
of CPT presentation. Subjects were required to respond with both hands
to the letter "X" only when it was preceded by the letter "A". Event-relate
potentials (ERPs) were recorded from midline (Fz, Cz, Pz, Oz) and
8 bilateral scalp sites all referenced to ear leads. Separate ERPs
were obtained for 5 letter sequences in both visual and auditory modalities.
P3 was designated as the maximum positive deflection occurring between
250-550 ms post-stimulus. Results supported part of our previous findings.
For example, in the visual A- X (Go) condition, females showed greater
left parietal activation. Males on the other hand did not show lateralized
effects for any stimulus sequences or conditions within the visual
CPT. Findings are discussed in terms of both modality effects and
sex differences in cognitive processing.

CNV rebound and distraction effects before and after a TM session
Frederick Travis
Maharishi University of Management
Tecce probed attentional contributions to the CNV through trials with
and without a letter recall task in the S1-S2 interval. With letters,
he observed a distraction effect: diminished CNVs and slower RTs.
Without letters, he observed a rebound effect-heightened CNVs-indicating
attention shifting from a divided to a unitary set. The Transcendental
Meditation (TM) technique is described as shifting attention to less
excited levels of thinking. If TM practice does redirect attention,
it may influence rebound and distraction effects. Seventeen subjects
were tested-mean age 20.8 yrs, mean TM practice 4.7 yrs. The protocol
comprised three blocks of trials before and after a 20-min TM session.
The middle block of each set randomly had 15 trials with the letter
recall task and 15 trials without. EEG was recorded with a .01-100
Hz band pass from Fz, Cz, and Pz using an Electrocap, referenced to
linked ears. After eye movement correction and manual artifacting,
data were averaged and the integrated amplitude 250 ms prior to S2
calculated. Before TM practice, significant CNV rebound effects and
CNV- and RT-distraction effects were seen. After TM practice, rebound
was absent, and "distracting" letters facilitated frontal CNVs. Subjects
reported that after their TM session, the letters acted as timing
cues for the S2. This is supported by earlier peaks in their terminal
CNV (1528 ms vs 1607 ms) and similar RTs to comparison blocks. After
TM practice, subjects appeared to process information more in terms
of underlying patterns.

Age-related differences in ERP activity during an old/new recognition memory test with subsequent contextual/familiarity and temporal source judgments
Charlotte T. Trott1, David Friedman1, Walter Ritter2, & Monica Fabiani3
1New York State Psychiatric Institute, 2Albert Einstein College of Medicine, 3University of Missouri
Older adults perform more poorly than younger adults on direct memory
tests, which could be due to a greater deficit in contextual- compared
to familiarity-based processes. Sixteen young (21-28) and 16 older
women (65- 81) studied two lists of sentences (each with 2 nouns)
and were subsequently tested by old/new recognition memory. For nouns
judged old, Remember (contextual) vs. Know (familiarity), and list
(temporal source) judgments were elicited. The groups did not differ
in the percentage of correctly recognized old nouns (hits) or Remember
(R) vs. Know (K) judgments, but the elderly were significantly poorer
at correctly identifying the source of hits. For both young and old
at posterior sites (400-800 ms), hits associated with R, K, and list
judgments elicited more positive ERPs than foils. This effect was
more prolonged and larger for hits associated with R than K judgments.
ERPs for hits with list correct and incorrect showed less differentiation.
The groups differed at prefrontal sites (800-1800 ms) where, for the
young only, hits were associated with greater positivity than foils
regardless of type of subsequent judgment and hits with list incorrect
were more positive than those with list correct. The posterior old/new
effect seen in both age groups may reflect item retrieval. The prefrontal
effects seen only in the young (occurring after the old/new reaction
time decision) may reflect effortful search for and/or retrieval of
source. The data suggest that the remember, know and list judgments
reflect similar, but not identical, underlying neural and cognitive
substrates. The lack of the prefrontal effect in older subjects is
compatible with the hypothesis of a deficit in source memory contingent
upon an age-related change in frontal lobe function.

Culture, physiology, and reported affect: Couples during enjoyable conversations
Jeanne L. Tsai1 & Robert W. Levenson2
1University of Minnesota, 2University of California-Berkeley
How does culture influence our emotional responses? Because cultural
norms regulate social relations, we hypothesize that culture influences
those components of emotion with greater social consequences (reported
affect) more than those with fewer social consequences (physiology).
This hypothesis was supported by our previous studies of Chinese American
and European American dating couples' emotional responses during interactions
that primarily elicited negative affect (i.e., conversations about
a conflict in their relationship). The present study examines whether
this hypothesis generalizes to couples' emotional responses during
an interaction that primarily elicits positive affect. Fourteen Chinese
American and 13 European American couples conversed about something
they enjoyed (e.g., vacations spent together). Continuous measures
of couples' cardiovascular, electrodermal, and somatic activity an
their reports of positive and negative affect experienced during the
conversation were obtained. Our results supported our hypothesis and
replicated our previous findings. Analyses revealed no cultural differences
in physiological responding during the enjoyable conversations. Cultural
differences emerged in reported affect for men. As in our previous
studies, these differences were consistent with ethnographic notions:
Chinese American men moderated and controlled their reported affect
more (i.e., reported less variable and less negative affect) than
European American men. Chinese American and European American women
did not differ in their reported affect. We attribute these gender
differences in the influence of culture on reported affect to the
faster acculturation of Asian women than men to European American
cultural norms (Nguyen & Williams, 1989).

Individual differences in vagal tone predicts cardiovascular recovery from sadness
Michele M. Tugade & Barbara L. Fredrickson
University of Michigan
This study examined whether individual differences in vagal tone,
as indexed by respiratory sinus arrhythmia (RSA), predicts the speed
of recovery from cardiovascular activation generated by sadness. Continuous
measures of cardiovascular responding were collected from 53 female
participants who viewed a film known to elicit sadness. Measures included
RSA, heart rate, finger pulse amplitude, pulse transmission time to
the ear and finger, and systolic and diastolic blood pressure. Low
and High Vagal Tone groups were defined by a median split on mean
baseline RSA. Results indicate that the sad film produced changes
from pre-film baseline levels in finger pulse amplitude, pulse transmission
time to the finger, and systolic and diastolic blood pressure. To
calculate the duration of this cardiovascular reactivity, we determined
the time elapsed from the end of the sad scenes until the cardiovascular
changes induced by the film returned to each subject's own baseline
levels. High Vagal Tone subjects returned to baseline significantly
faster than Low Vagal Tone subjects (24 s. versus 47 s.), t (51) =
-2.00, p = .05. Results suggest that individual differences in vagal
tone may predict the ability to regulate the cardiovascular effects
of negative emotional experiences.

Adaptation of oculomotor resting states depends on autonomic demands
Richard A. Tyrrell1, Melanie A. Pearson1, & Julian F. Thayer2
1Clemson University, 2University of Missouri
Both the heart and the eye muscles receive neural input from the ANS
and we have previously demonstrated that observers who experienced
adaptation of the resting states of visual accommodation (dark focus)
or of binocular vergence (dark vergence) also experienced a parasympathetically
mediated acceleration of heart rate. The present study tested whether
oculomotor adaptation to near work can be influenced by the autonomic
demands inherent in the task. Twenty healthy college students (ten
emmetropes and ten late-onset myopes) performed two visual tasks for
20 minutes. When in the visual search task participants competed for
monetary rewards in a demanding visual search / reaction time task
at 15 cm; when in the reading task participants read a Sherlock Holmes
story at the same near distance. Immediately before and after the
tasks dark focus and dark vergence were assessed. Heart rate, MSD,
and PEP results indicated a relative parasympathetic withdrawal during
the visual search task. The magnitude of oculomotor adaptation varied
across the tasks: In the reading task, significant changes in dark
focus occurred for both groups. In the visual search task, however,
dark focus did not change. A significant interaction indicated that
changes in dark vergence depended on both task and group: while the
emmetropes experienced greater dark vergence changes during the visual
search task, the myopes reacted oppositely. These results indicate
that accommodative adaptation is inhibited during parasympathetic
withdrawal and that dark vergence adaptation depends on both the autonomic
demands of the task and on refractive status.

The influence of social support on age-related differences in cardiovascular function
Bert N. Uchino, Timothy S. Garvey, Julianne Holt-Lunstad, & Darcy Uno
University of Utah
Aging is associated with reliable changes in cardiovascular function.
However, these changes do not appear to be a simple function of chronological
age and researchers have begun to examine factors that appear associated
with "successful aging" (Rowe & Kahn, 1987). Of these factors, social
support may be particularly important due to its epidemiological links
with lower rates of morbidity and mortality (House, Landis, & Umberson,
1988). In this study, we examined the influence of social support
(Cohen et al., 1985) on age-related differences in resting systolic
blood pressure (SBP), diastolic blood pressure (DBP), and heart rate
(HR) in 67 healthy women and men (ages 20 to 70). All analyses were
conducted statistically controlling for gender. Consistent with prior
research, age predicted increased resting SBP (r = .30, p < .02) and
DBP (r = .36, p < .01). More important, regression analyses revealed
that social support moderated age-related differences in both SBP
(B = -.47, p < .05) and DBP (B = -.44, p < .01). Subsequent analyses
revealed that age predicted higher resting SBP (B = .28, p < .05)
and DBP (B = .35, p < .01) for individuals low in social support,
but was unrelated to blood pressure for individuals high in socia
support. These results were unchanged while statistically controlling
for health-related variables such as body mass index, sleep patterns,
exercise, alcohol consumption, and caffeine consumption. Future research
is needed to specify more precise psychosocial mechanisms responsible
for these effects.

Pacing and mental workload effects on the structure and distribution of the P300
Peter Ullsperger1 & Darryl G. Humphrey2
1Federal Institute for Occupational Safety and Health, 2Wichita State University
The goal of the study was to extend earlier efforts aimed at evaluating
how mental workload affects the scalp distribution and component structure
of a P300 elicited by events in a complex dual-task paradigm. Fifteen
subjects performed a mental arithmetic task and a gauge monitoring
task in both single and dual task conditions. The gauge-monitoring
task consisted of six continuously updated gauges. The task was to
reset each gauge as quickly as possible once its cursor had reached
the critical zone. To see the cursor location, the subjects pressed
one of six keys (gauge-view). For the mental arithmetic task, the
center of each gauge served as a display area for the operators and
operands. ERP eliciting events included gauge-view samples and the
onset of the mental arithmetic trials. ERPs were recorded from 30
channels with a Cz reference. Task performance was adversely affected
by adding the dual task requirement. This change was reflected in
the ratings obtained via the NASA-TLX workload scales. The P300s from
presentation of the forced-pace arithmetic trials and from the self-paced
gauge-views both discriminate between single and dual task conditions.
Whereas the P300 amplitude to gauge-views decreased from the single-
to dual-task condition by about 17%, the P300 to presentation of arithmetic
trials decreased by 53% (midline derivations). In addition, the P300
scalp distribution differed between gauge-views and math trial presentations.
For the latter, P300 scalp distribution differed between single and
dual-task conditions. These results stress the sensitivity of ERPs
as an index of mental workload.

Ambulatory monitoring of respiratory activities in real daily life situations
Akio Umezawa1 & Akira Kurohara2
1Fukui University, 2Fukui Pref. Police Headquarters
Respiratory control is the strategy used most frequently in Japanese
people to calm down some stressful conditions in daily life. There
are few studies concerning ambulatory monitoring and self-regulation
of respiration in real daily life situations. The present study, therefore
described the methods for ambulatory monitoring of respiratory movements
by using respiratory inductive plethysmography (RIP) and for tidal
volume (Vt) measurements based on the sum of volume displacement of
rib cage (RC) and abdomen (AB). The RC and the AB volume-motion coefficients
were calculated from the data obtained in three body positions (supine,
sitting, and standing) by using the multiple regression analysis method.
Prior to the ambulatory measurement, seven male and female subjects
aged 22 to 47 yrs participated the calibration procedure comparing
with changes in lung volumes measured by RIP to simultaneous pneumotachograph
(PT) recording in the three body positions. Intrasubject correlation
coefficients between estimated Vt (RIP) and actual Vt (PT) were significant
in all subjects. The data obtained in 6 hours' ambulatory monitoring
in daily life showed that breathing patterns in real daily life situations
are characterized by thoracic-dominant breathing and also showed that
predicted minute ventilations using RIP indicated greater than ones
obtained in both pre- and post- baseline rest values even when desk
work without gross body movements was done. These data suggested that
respiratory activities in real daily life situations tend to be hyperventilated
by psychological tension.

Mismatch Negativity to speech stimuli in language impaired children
Ruth Uwer, Ronald Albrecht, & Waldemar von Suchodoletz
Ludwig-Maximilians-University Munich
The Mismatch Negativity (MMN) is an automatic cortical evoked potential
that reflects the detection of acoustic changes. It provides a tool
to study central auditory processing mechanisms involved in the perception
of simple acoustic stimuli and of speech sounds. Since subjects are
not required to pay attention to the stimuli it might be useful in
assessing language impaired children who often suffer from attention
deficits as well. Auditory event related potentials were recorded
from language impaired children and healthy controls (7-11 years).
All children had normal hearing threshold and normal non-verbal cognitive
skills. The difference between scores for language skills and the
non-verbal IQ was at least 1,5 standard deviations for the language
impaired children. As stimuli we used pure tones (standard 1000 Hz,
175 ms, frequency deviant 1200 Hz, duration deviant 100 ms) and digitized
consonant-vowel- syllables (standard da, deviants ga and ba) to elicit
the MMN. According to Korpilahti and Lang (EEG Journal,1994) we hypothesize
a deficit in the automatic detection of the frequency difference for
the language-impaired children. As well we expect an attenuated MMN
to different speech stimuli in language-impaired. Preliminary analyses
show no differences between patients and control children in the pure
tone condition while the MMN was attenuated in amplitude or absent
in the speech-stimuli condition in language-disordered children. This
supports the hypothesis of a specific processing deficit in language-disordere
children. Perspectives and limitations for clinical applications of
the MMN in children are discussed.

Accessory stimulus and response selection effects on the Lateralized Readiness Potential (LRP)
Fernando Valle-Inclan & Steven A. Hackley
University of La Coruna & University of Missouri-Columbia
A task-irrelevant prepulse, or accessory stimulus, delivered concurrently
with the imperative stimulus in a choice reaction time (RT) task reliably
speeds RT. To determine if the locus of this effect is at the response
selection stage, we compared conditions in which the imperative stimulus
(a single letter) was mapped to either 3 or 9 responses. Letter shape
(S or T) determined hand of response and color determined Go/NoGo
and finger (1 or 4 per hand). In the 9 response condition, for example,
the stimulus required a keypress with one of eight fingers or withholding
of the response. On half of the trials, an 80-dB tone accompanied
the visual stimulus with a lead time of 83 ms. As expected, responses
were faster on accessory (796 ms) than control (835 ms) trials and
for the condition with 3 (648 ms) as opposed to 9 (990 ms) response
alternatives. LRP onset was used to partition mean RT into two time
intervals, one subsuming perceptual and decision processes and one
including only late motoric processes. Number of response alternatives
affected the duration of both intervals, but accessory stimulation
influenced only the first interval. The effect of number of response
alternatives on RT and LRP appeared not to be affected by accessory
stimulation. We conclude that alerting triggered by an accessory stimulus
facilitates RT by influencing some process prior to the response selection
stage. This research was supported by the Ministry of Education and
the regional government of Galicia.

Frontal ERPs vary with cardiac cycle time in response inhibition
G.J.M. van Boxtel12, J.R. Jennings3, M.W. van der Molen2, & C.H.M. Brunia1
1Tilburg University, 2University of Amsterdam, 3University of Pittsburgh
Response inhibition in the stop-signal task has been
associated with the frontal N2 and with cardiac deceleration in
separate studies. Cardiac deceleration is found to be largest when the
stop stimulus is presented in about the middle of the inter beat
interval. A relation between the N2 and cardiac cycle time has nev
been investigated. We recorded ERPs from 28 electrodes, heart rate,
and respiration in a visual stop signal task with a response
probability of 50%. As expected, cardiac slowing was larger on Inhibit
than on Respond trials, and the amount of slowing was largest in the
middle of the cardiac cycle. The N2 had a frontal maximum, and was
also larger on Inhibit than on Respond trials. Trials were selected in
which the interval between the stop stimulus and the N2 was contained
in the time between the R-wave and the R-wave + 450 ms (early), and
compared to trials in which the stimulus to N2 interval was contained
in the time between R+450 ms and the following R-wave (late). The N2
was preceded by a small transient positivity (P2), the peak of which
determined the onset of the N2. This positivity was only present on
Inhibit trials at frontal electrodes, and in that case it was largest
when the stop stimulus was presented late in the cardiac cycle. These
results are discussed in terms of the Lacey hypothesis, which relates
baroreceptor activity (predominantly present in the 450 ms after the
R-wave) to cortical deactivation.

When physiological responses and feelings diverge: Personality moderators of early and late affective startle reflex modulation and self-reports
Eric J. Vanman1, Patricia A. Brennan2, & Michael E. Dawson3
1Texas A&M University, 2Emory University, 3University of Southern California
Previous research has demonstrated both early (less than 500 ms) and
late (greater than 500 ms) affective modulation of the startle eyeblink
reflex. Using the affect-directed attention paradigm, this experiment
examined individual differences in emotional responsivity by recording
the eyeblink reflex while 57 college students viewed affect-laden
pictures. All participants first completed personality measures of
affect intensity, alexithymia, and depression, and then made pleasantnes
ratings of the pictures. As participants viewed the pictures the second
time, startle probes were presented on some trials at either 120,
300, 800, or 4500 ms after slide onset. An alpha level of .05 was
used for all statistical tests. As early as 300 ms, but also at 800
and 4500 ms, blinks elicited during negative slides were larger than
those elicited during positive ones. Negative slides were also rated
as more unpleasant. Moreover, all three personality variables moderated
either the valence ratings and/or the magnitude of startle modification.
High affect intensity was associated with diminished affective modulation
of startle, but more extreme valence ratings. Alexithymia had no effect
on the startle measure, but high alexithymia participants showed more
attenuated valence ratings. Depressed participants exhibited accelerated
(120 ms) affective modulation of startle. The results signify the
importance of measuring both physiological responses and subjective
feelings, as well as their interrelationships, in the study of individual
differences in emotion.

Pupillary response indexes cognitive processing in the visual backward masking task
S. P. Verney, E. Granholm, & D. P. Dionisio
University of California-San Diego, VA San Diego Healthcare System, & San Diego State University
This study investigates task-evoked pupillary responses as an index
of the underlying cognitive mechanisms tapped by the visual backward
masking (VBM) task. This paradigm has been used to investigate early
visual information processing in a great many studies, however, the
mechanisms that are involved have remained at the theoretical level.
In Breitmeyer's popular theory, the amount of processing allocated
to the mask either interrupts or is integrated with processing allocated
to the target. Twenty-three young adults performed a standard target
duration, equal energy, visual backward masking task while pupillary
responses were simultaneously recorded. Inter-stimulus intervals between
the target (two unequal lines) and the mask (two equal lines spatially
overlapping the target) were presented randomly at 17, 33, 50, 100,
300 milliseconds, along with a no-masking condition. Pupillary dilations
appeared to index the extent of cognitive processes used in the VBM
task as dilations were significantly larger in Cognitive Load (making
length judgments) relative to Cognitive No-Load conditions (passively
viewing the stimuli), F(1,22) = 174.79; p<.01. Relative to the no-masking
condition, detection accuracy was significantly lower in all conditions,
except the 300 ms condition, whereas, pupillary responses were also
significantly smaller in the 17 ms condition but were significantly
larger in the 300 ms condition (t = -2.75; p<.05; t = 2.34, p<.05,
respectively). Thus, a greater amount of processing was allocated
in the 300 ms masking condition to achieve a comparable level of performance
accuracy in the no-masking condition. This finding suggests that this
psychophysiological index is able to delineate the amount of processin
allocated to the mask and, therefore, can be used to test current
theoretical models of masking mechanisms.

Psychopathy and physiological response to emotionally evocative sounds
Edelyn Verona1, John J. Curtin1, Gary K. Levenston1, Margaret M. Bradley2, Peter J. Lang2, & Christopher J. Patrick1
1Florida State University, 2University of Florida
Considerable evidence indicates that psychopathic criminals are deviant
in their reactions to affective stimuli, but observed differences
may vary with cuing context and diagnostic symptomatology. This study
investigated physiological reactions to sound depictions of neutral
and emotionally arousing events in four male prisoner groups: (a)
emotionally detached offenders (n = 15), possessing only the core
personality symptoms of psychopathy; (b) antisocial offenders (n =
18), possessing only the behavioral features; (c) psychopaths (n =
18), possessing both; and (d) nonpsychopaths (n = 18), possessing
neither. Nine digitized sounds were presented through headphones via
computer: 3 pleasant (passionate moan, infant laugh, crowd cheer),
3 neutral (toothbrush, toilet flush, chicken cluck), and 3 unpleasant
(assault, baby cry, female scream). Pleasant and unpleasant sounds
were matched in rated arousal, with both exceeding neutral. Each sound
was presented three times, in counterbalanced fashion across trial
blocks, and skin conductance, heart rate, and facial muscle reactions
were recorded. For the sample as a whole, autonomic and facial responses
covaried predictably with the pleasantness and arousal properties
of the sounds. Group differences in patterns of physiological response
were also found. In particular, psychopathy--the emotional detachment
component specifically--was associated with reduced electrodermal
reactivity to emotionally evocative sounds. This study provides further
evidence that abnormalities in affective stimulus processing distinguish
the classic syndrome of psychopathy from criminal deviance per se.
(Supported by NIMH grants MH48657 & MH52384.)

The effects of age, ethnicity, sex, and social context on the physiology of a brief social interaction
Scott R. Vrana & David Rollock
Purdue University
This study examined adolescents' physiological responses during a
brief encounter with an unfamiliar person. Participants were 85 African
American and 90 White 10-17 year-old boys and girls. After a resting
baseline, an unfamiliar adult of the same sex as the subject entered
the room and introduced himself or herself as a research assistant
checked equipment for 30 seconds, and then took the subject's pulse
for 30 seconds. Social context was manipulated by varying whether
the adult was of the same or different ethnic background (Black or
White) as the subject. Entry of the unfamiliar person resulted in
a smile (zygomatic EMG increase), a skin conductance response, and
heart rate acceleration. Heart rate was higher and skin conductance
took longer to habituate when someone from a different ethnic group
entered the room. For both Black and White subjects, blood pressure
was greater when a Black person entered the room. For boys, a more
positive expression (more smiling and less frowning) was seen in response
to someone from a different ethnic background, whereas among girls
a more positive expression was seen to someone from the same ethnic
background. Both frowning and smiling were relatively greater (meaning
less reduction below baseline) when touched by someone from the same
compared to a different ethnic background. This research shows that
there are significant physiological effects of even a very minimal
social interaction, and that group differences and the social context
of the interaction influence this response.

Emotional imagery in adolescents: Effects of sex, ethnicity, and social context
Scott R. Vrana & David Rollock
Purdue University
This study examined self-reported emotion and physiological response
during imagery of emotional situations in younger (10-13 years old)
and older (14-17 years old) adolescents. Eighty-five African American
(44 boys) and 90 White (49 boys) adolescents imagined joy, anger,
fear, and neutral emotional situations. Two scenes were presented
for each emotion: one described interacting with a White person and
one described interacting with a Black person. Subjects imagined each
scene while listening to an audiotaped description for 50 seconds,
continued imagery for 30 seconds, and then rated their imagery along
several dimensions. Both self-report and physiological data show that
adolescents respond to emotional imagery in ways consistent with adults:
Pleasure ratings and zygomaticus EMG activity were greater for joy
than any other type of imagery; pleasure ratings were lowest and corrugator
EMG was highest during anger and fear imagery. Joy, anger, and fear
images evoked report of higher arousal, and greater skin sweat activity,
compared to neutral images. White subjects frowned more than Black
subjects during anger and fear scenes; whereas Black subjects smiled
more than White subjects during the joy scenes, and also smiled during
anger scenes. Subjects both smiled more and frowned more during imagined
interactions with people from a different ethnic group. Overall, the
data show that adolescents can participate reliably in emotional imagery
tasks, and that their physiology and self-reported experience are
sensitive to differences in emotional content and social context i

Dipole analysis of laser evoked brain potentials (LEP)
Thomas Weiss, Jennifer Dillmann, Jens Ehrhardt, Kai Kumpf, & Wolfgang H.R. Miltner
Friedrich Schiller University-Jena
Only a few studies have been attempted to localize the anatomical
generators of human laser evoked brain potentials (LEPs). In the present
study we analyzed LEPs to painful and nonpainful laser stimuli (thulium
YAG laser) in a dense electrode montage using a realistic volume conductor
model. In a first experiment, laser heat stimuli below and above pain
threshold were applied to 16 volunteers. LEPs were recorded from 60
electrodes during stimulation of the left hand. The two major LEP
components were a negative peak at 210 ms (N210) and a positive peak
with a mean latency of 350 ms (P350). In some but not all Ss we found
an earlier negative component (N150). Boundary element method (BEM)
was used for source location of the LEP components. A three-dimensional
head coordinate system was created with a 3D digitizer in order to
overlay the functional LEP data to the MRI. Single equivalent dipole
analysis showed that the sources of the N150 component could be found
in the primary somatosensory cortex contralaterally to the stimulated
hand. The dipole localizations of the later components were not as
clear as for the N150. In a second experiment, laser heat stimuli
were applied to three different locations of the body (face, hand,
foot). Clear differences were detected for latencies of the main peaks
of the LEP. In some volunteers single equivalent dipole analysis showed
sources in the contralateral primary somatosensory cortex. However,
using PET localizations from other labs as constraints we were able
to localize generators for the main components in different brain
areas involved in pain processing. This localizations will be interpreted
in the light of multidimensional approaches to pain. (Supported by
BMBF 01ZZ9602.)

Functional plasticity within the somatosensory and motor cortex in phantom limb patients
Thomas Weiss, Winfried Meissner, Jennifer Dillmann, Ralf Huonker, & Wolfgang H.R. Miltner
Friedrich Schiller University-Jena
Conditions for the development of chronic phantom limb pain are still
a matter of discussion. Recent reports concerning the functional plasticity
of the somatosensory cortex give support for a central nervous genesis
of phantom limb pain. The aim of our study was to provide further
evidence for the central nervous reorganisation within the somatosensory
and the primary motor cortex of patients suffering from phantom limb
pain. Event-related changes of EEG and MEG as well as different behaviora
data were recorded from patients after amputation at different sides
of the upper limb. Psychophysiological examinations were performed
under tactile and laser heat stimulation. Additionally, patients requested
to move different parts of the body (finger, arm, shoulder). The event-related
fields and potentials were analyzed using single moving dipole analysis
in a relastic volume conductor model (boundary element method). A
three-dimensional head coordinate system was created with a 3D digitizer
in order to overlay the results of dipole fitting to the MRI. We found
changes in the primary somatosensory and motor cortex. Cortical representations
of the amputated part of the limb seem to be occupied at least in
part by neighboring cortical representations. Furthermore, it was
found that phantom sensations can be evoked from different sites on
the remaining extremity, face and trunk. We conclude that this reorganization
of the cerebral cortex may represent one reason for the phantom limb
pain. This kind of "use-related" plasticity seems to be an important
mechanism for the adaptability to the changing environment. (Supported
by BMBF 01ZZ9602.)

Effect of parietal damage on P1 modulation
Marissa N. Westerfield1,2, Jeanne Townsend1,2, James W. Covington2, & Eric Courchesne1,2
1University of California-San Diego, 2Laboratory for Research on the Neuroscience of Autism-Children's Hopsital
Previous research has shown that the P1 component of the human event-related
potential (ERP), thought to be generated in extrastriate visual cortex,
is sensitive to the spatial focus of attention. P1 amplitude is largest
following stimuli occurring at an attended location and decreases
as stimulus distance from the attended location increases. Townsend
and Courchesne (1994) showed that the P1 gradient displayed by autistic
subjects with bilateral parietal lobe abnormalities differed from
that displayed by normal subjects. The current study extended these
results by recording ERPs from normal subjects and subjects with parietal
damage due to stroke during a spatial attention task. One condition
required subjects to respond to all stimuli at an attended location;
a second condition required subjects to discriminate between targets
and standards at an attended location. Subjects with parietal abnormalities
displayed a P1 attenuated at (contralesional) unattended locations,
indicating that parietal damage can modulate activity in extrastriate
visual cortex as measured by the P1 component. Additionally, the gradient
displayed by the autistic subjects in the previous study was mimicked
in normal subjects by changing task demands. Normal subjects displayed
a P1 gradient in which changes in amplitude were gradual, but in the
discrimination task showed a P1 gradient similar to that seen in the
parietal-abnormal subjects during the much easier detection task.
(Supported by NINDS NS34155 & NIMH MH36840)

EEG sensitivity of human auditory emotion
Min Cheol Whang, Hee Kwan Cho, & Chul Jung Kim
Korea Research Institute of Standards and Science
The general aim of our research is to improve human- centered product
design and environment by using physiological parameters such as EEG,
ECG or GSR being sensitive to emotional variations such as comfort
or pleasantness. In the present study, EEG was recorded from 21 locations
on the scalp of fifteen undergraduate students during auditory presentations
of five positive and five negative sounds such as running through
a creek, bird singing, car crashing and industrial noise, taken from
a commercial compact disc. Sounds were presented over loudspeakers
for 30 sec with 2 min intervals in random order. To evaluate the stability
of emotions elicited by the sounds, correlations between subjective
ratings performed after each sound and after five sounds were obtained.
EEG was analyzed by the Cadwell Spectrum 32 System. Comparisons between
positive and negative emotional responses were performed by paired
Student's t-tests. The relative power spectra of the delta, theta,
alpha and beta wave bands allowed for a highly sensitive discrimination
between positive and negative emotions.

The postauricular reflex in schizophrenia: A potential measure of impaired suppression
Patricia M. White, Cindy M.Yee, & Keith H. Nuechterlein
University of California-Los Angeles
Previous research has demonstrated that when two clicks are presented
500 ms apart, normal individuals show P50 suppression to the second
click compared to the first, while schizophrenic patients show limited
suppression. Similarly, schizophrenic patients exhibit less prepulse
inhibition of the blink reflex than normal subjects when a startle-eliciting
stimulus is preceded by a weak prestimulus. Both P50 suppression and
prepulse inhibition of the startle blink reflex have been hypothesized
to reflect an impairment of sensorimotor gating in schizophrenia.
The present study examined the postauricular reflex (PAR) as a possible
measure of impaired suppression in schizophrenia using pairs of 90
dB SPL clicks, separated by 500 ms, delivered against a 40 dB SPL
white noise background. PAR responses in 12 male, chronic schizophrenic
patients on a typical antipsychotic medication, fluphenazine decanoate,
were compared with those of 9 male, non-psychiatric comparison subjects.
Analysis of variance revealed a significant group by stimulus interaction.
In the non-psychiatric sample, amplitude of the negative deflection
of the PAR to the second click was reduced compared to the first click.
The patients, in contrast, showed impaired suppression to the secon
stimulus. To examine the potential influence of medication, PAR responses
from 15 recent-onset schizophrenic patients (11 males) on an atypical
antipsychotic medication, risperidone, were compared with PAR responses
from the 12 chronic patients on fluphenazine. Recent-onset patients
on risperidone showed significantly more PAR suppression than the
chronic patients on fluphenazine and were similar in suppression to
the non-psychiatric sample. Implications for future research will
be discussed.

Faster peripheral nerve conduction velocity predicts introversion
John C. Wickett1, Philip A. Vernon2, John D. Brown2, & Ronald Broome2
1University of Ottawa, 2University of Western Ontario
Eysenck's theory of extraversion holds that introverts react more
strongly to stimulation than do extraverts, hence their relatively
greater avoidance of highly stimulating situations. In searching for
a biological manifestation of this effect, it was hypothesized that
introverts would have faster nerve conduction velocity. Forty-two
adult male subjects completed a measure of extraversion (drawn from
the 6FPQ; D. N. Jackson & S. V. Paunonen, in press), and had nerve
conduction velocity measured along the median nerve of the right arm.
A correlation of -.32 (p < .05) was found, indicating that introverts
were more likely to have faster neural speed. Most of this correlation
was carried by the spine to axilla segment of the neural pathway,
with more distal segments showing lower correlations. Corrections
for body size, arm temperature, and IQ did not substantially change
the magnitude of this relationship. This suggests that introverts'
stronger reaction to stimulation is, at least in part, mediated by
their faster neural speed.

Intelligence is predicted by peripheral nervous system conduction velocity, but not by central nervous system conduction latency
John C. Wickett1, Philip A. Vernon2, John D. Brown2, & Ronald Broome2
1University of Ottawa, 2University of Western Ontario
Research on the relationship between intelligence and peripheral nerve
conduction velocity has demonstrated a small positive correlation
between speed and IQ. Research relating intelligence to central nerve
conduction (to percutaneous stimulation) does not appear in the literature,
but neural speed in the central nervous system should be even more
strongly related to greater IQ. This hypothesis was tested in a sample
of 48 healthy adult males. Peripheral nerve conduction velocity was
assessed along the median nerve from cervical spine to wrist, an
central latency was assessed from apex of head (over motor cortex)
to cervical spine. Speed along the axilla to elbow segment of the
median nerve showed the strongest correlation with IQ at .37 (p <
.05), and all velocity measures showed correlations in the predicted
direction. Contrary to prediction, central latency showed a positive,
but nonsignificant, correlation with IQ at .10. The peripheral findings
indicate faster neural speed for higher IQ subjects, but the central
findings indicate if anything longer transmission times for these
same subjects. These findings were interpreted as indicating that
although speed is an important factor underlying intelligence, the
number of synaptic connections in the brain is likely even more important.
The more synapses through which an impulse must travel the longer
will be the transmission time, and if high IQ subjects have more neurons,
and hence more synapses, then the expected correlation with central
latency is as found in this study.

Brain volume and general intelligence in adult male siblings
John C. Wickett1, Philip A. Vernon2, & Donald H. Lee3
1University of Ottawa, 2University of Western Ontario, 3London Health Sciences Centre
Current research on the relationship between brain volume (measured
via MRI) and IQ has indicated a moderate positive correlation of about
.40 between the two variables. The present research expands on this
finding by examining the g-factor, in addition to IQ, and by testing
for the presence of within-family correlations. Sixty-eight right-handed
males (ages 20 to 35 years), including 32 sets of brothers with complete
data, completed the Multidimensional Aptitude Battery, an additional
battery of 12 cognitive tests, and submitted to MR imaging of the
brain. It was found that brain volume and full-scale IQ correlated
.35 (p < .01); g and brain volume correlated slightly higher at .38
(p < .01). Most interestingly, a vector analysis indicated that the
most g-loaded tests also had the highest correlations with brain volume
(r = .59, p < .01). This clearly shows a very close correspondence
between g and that aspect of intellectual ability that is indexed
by brain volume. Finally, and most importantly, there was evidence
for a within-family correlation between brain volume and both IQ and
g (rs = .16 and .24, respectively). These effects were not statistically
significant, but were of the magnitude expected. A within-family correlation
indicates that the relationship between brain volume and intelligence
is almost definitely intrinsic and causal, with pleiotropy being the
most likely explanation.

Links in audio-visual spatial attention
Andreas Widmann, Martin Eimer, & Erich Schrog
University of Munich & University of Leipzig
Ten participants were presented with a cue (100 ms duration) followed
by a target (100 ms) with an SOA of 500 ms. The cues were not informative
with regard to the location of the forthcoming target. There was a
Visual Cue Auditory Target condition (VCAT), in which the cue was
a light signal and the target a burst of white noise, and an Auditory
Cue Visual Target condition (ACVT). Stimuli were presented at a distance
of 2 m from four positions in the horizontal plane revealing an eccentricity
of -30 , -15 , +15 , or +30 . Subjects were instructed to press one
of two buttons depending on the eccentricity of the target. Reaction
times for valid trials (cue and target at the same position) were
shorter than for invalid trials (cue and target at a different position).
In VCAT and in ACVT, an increase of this validity-effect as function
of the spatial distance was observed. In both conditions, ERPs revealed
an enhanced negativity in valid as compared with invalid trials with
a posterior maximum at a latency of about 160 ms (Nd1). As shown in
previous studies, this Nd1 reflects attentional orienting in trial-by-trial
cuing situation. Results suggest that auditory cues (although unpredicitve
with respect to the position of the auditory target) elicited an exogenous
orienting of attention in the visual modality and that unpredicitve
visual cues elicited an orienting of attention in the auditory modality.
This argues for a link in the exogenous orienting between vision and

Concurrent validity of two continuous blood pressure monitors and relationship of heart period variability to blood pressure reactivity
Stefan Wiens, Mariya Romanenko, Sarah Reiff, & Robert M. Kelsey
State University of New York at Stony Brook
The concurrent validity of SBP and DBP measured by Cortronic and Finapres
monitors during rest and vocal mental arithmetic was assessed in 17
undergraduates. Measurements of BP and heart period (HP, measured
from an ECG) were aggregated across four 30-s intervals of rest and
task. The means of the six correlations among the four measurements
for SBP-Cortronic, DBP-Cortronic, SBP-Finapres, and DBP-Finapres were
respectively .88, .98, .75, and .95 for baseline and .91, .96, .80,
and .94 for task. A 2 (BP device) x 2 (rest vs. task) x 4 (time) doubly
multivariate ANOVA of DBP and SBP showed that Finapres yielded higher
levels of DBP and SBP, and that the elevations of DBP and SBP during
the task were greater for Finapres. Changes in DBP accounted for changes
in SBP during the task. The correlation of BP reactivity (task minus
baseline) between monitors was low (r = -.01 for SBP and r = .12 for
DBP). HP variability, which was the mean of HP standard deviations
during baseline, correlated negatively with BP reactivity for Finapres
(r = -.66 for SBP and r = -.51 for DBP) but not for Cortronic (r
-.05 for SBP and r = .30 for DBP). The finding by Brondolo et al.
(1996, Psychophysiology, 33, S25) that heart rate variability during
rest is negatively correlated with blood pressure variability was
replicated, but only for the Finapres. Although Cortronic and Finapres
provide stable measures during rest and vocal mental arithmetic stress,
they lack concurrent validity.

Cardiovascular adjustments to mental stress in healthy 19- to 91-year-old men and women
Frank H. Wilhelm & Paul Grossman
Harvard University
Changes in cardiac and peripheral vascular function with age are likely
to affect hemodynamic responses to mental stress. However, age-related
differences have not been well characterized. Thirty-six women and
37 men (19-91 years) without a history of cardiovascular disease underwent
a standardized stress test protocol. A variety of cardiovascular
responses were assessed by peripheral pulse-contour measurement at
rest, during mental arithmetic, and simulated public speaking. Although
subjects reported being made more anxious by math stress, speech elicited
greater responses in HR (17.3 vs 8.2 bpm) and effective arterial elastance
(EA; ~ventricular afterload), and more negative responses in stroke
volume and left ventricular ejection time (LVET). Higher age was
generally associated with increased responses in systemic vascular
resistance and EA, and with decreased responses in cardiac index.
Stroke volume increased for young and decreased for elderly individuals.
Only during math was higher age related to increased responses in
arterial pressure and rate-pressure-product (~myocardial oxygen demand),
while only during speech LVET responses increased and stroke work
index responses decreased with higher age. Women had greater heart
rate and rate-pressure-product responses to speech than men. We conclude
that in younger individuals the main mechanism for blood pressure
rise during mental stress is increased cardiac output due to HR and
stroke volume increase, whereas in older individuals it is increased
vascular resistance with only moderate increase in cardiac output
(HR increase is compensated by stroke volume decrease). Systemic arterial
pressure responses to math, but not to speech, are more pronounced
in older individuals.

Assessment of heart rate variability under non-stationary conditions: Complex demodulation vs. spectral analysis
Frank H. Wilhelm1, Paul Grossman1, & Walton T. Roth2
1Harvard University, 2Stanford University
Complex demodulation (CDM) has been proposed as a method for the analysis
of high- and low-frequency variabilities of heart rate and blood pressur
under non-stationary conditions. In contrast to power spectral analysis,
CDM provides time-dependent changes in signal amplitude and frequency
on a continuous basis and may yield insights into short-term changes
in autonomic regulation. In particular, CDM may be uniquely suited
for quantifying changes in respiratory sinus arrhythmia (RSA) during
panic attacks. In a simulation analysis we generated R-R interval
time series within a normal physiological range that represented different
typical sources of non-stationarity present during a panic attack.
Sources of non-stationarity included abrupt changes in a) level (from
1000 to 500 ms within 60 sec), b) oscillatory amplitude (from 50 to
10 ms peak to trough), c) oscillatory frequency (from 0.2 to 0.4 Hz),
and d) a combination of the above. In general, CDM-estimated amplitude
and frequency accurately reproduced characteristics of the simulation
data under all conditions, even after substantial noise and a 0.10
Hz oscillation were added. However, during some transitions CDM estimates
fluctuated around the true values for up to 10 s before they stabilized.
Compared to CDM, power spectral analysis results were inaccurate since
they did not allow the disentangling of unique contributions of distinct
amplitudes and frequencies at different time points. We conclude that
CDM may provide a powerful means of continuously assessing time-dependent
changes in RSA during panic attacks. CDM may also hold promise for
a variety of other physiological and environmental conditions where
rapid dynamic changes in autonomic control are likely to occur.

Secretory immunoglobulin A and cardiovascular reactions to mental arithmetic challenge: Effects of task difficulty
Gonneke Willemsen, Christopher Ring, Douglas Carroll, & Sam McKeever
University of Birmingham
Secretory immunoglobulin A (sIgA) in saliva and cardiovascular activity
were recorded in 27 healthy volunteers during mental arithmetic (PASAT)
challenges, constructed as easy, hard, and impossible, and during
preceding pre-task resting baselines. Task difficulty was counterbalanced
across participants and the consistent variations observed in measured
task performance (F(2,52)=1035.97,p<.05), as well as self-ratings
of task difficulty (F(2,52)=78.49,p<.05), confirmed the validity of
the difficulty manipulation. The tasks elicited significant increases
overall in systolic and diastolic blood pressure, heart rate, and
total peripheral resistance (F(1,26)=21.42 to 47.08,p<.05). This pattern
of effects and the absence of significant changes in cardiac output
and pre-ejection period implicate mainly alpha-adrenergic activation.
The mental arithmetic tasks also provoked increases in sIgA concentration
and secretion rate (F(1,26)=13.70 and 16.05,p<.05 respectively). However,
there were no significant variations in cardiovascular or immune reactions
with task difficulty, with the exception of diastolic blood pressure
reactions which were larger to the hard than the easy and impossibl
challenges (F(2,52)=6.35,p<.05). These data provide further confirmation
that sIgA is responsive to the sort of brief laboratory challenges
that elicit sympathetically-mediated cardiovascular reactions; increases
in sIgA would seem to be particularly evident where the pattern of
cardiovascular activation suggests the involvement of the alpha-adrenergic
system. However, the present study failed to find convincing evidence
that sIgA reactions are sensitive to variations in task difficulty.

Effects of stimulus sequence and modality on P300 (P3) in two Continuous Performance Tasks (CPTs)
Jennifer K. Williams, Christine M. Oliver, & David W. Shucard
State University of New York at Buffalo
Prior research from our laboratory has revealed that there are distinct
P300 (P3) scalp topographies associated with target detection (Go)
and response inhibition (No-go) conditions during a continuous performance
task (CPT) of attention. In the present study, P3 amplitude and topography
were investigated during two CPTs that required distinctly different
cognitive operations. These CPTs were designed to have parallel forms
(e.g., identical stimuli) across auditory and visual modalities. The
CPT A-X paradigm required a response to a target letter "X" only if
it was preceded by a warning stimulus, "A". The CPT 123 paradigm required
a response only when the same number was repeated immediately after
it was first presented. Right handed male and female volunteers (17-25
years) were studied. Event-related potentials (ERPs) were obtained
from 12 scalp sites. Data from the midline sites are reported here.
Reaction times and error types were also recorded for each subject.
The data showed the expected P3 topographical relationships for target
detection and response inhibition during the CPT A-X paradigm, irrespective
of stimulus modality. In the CPT 123 paradigm for target detection,
the P3, topography was similar to that for the CPT A-X paradigm (i.e.,
a gradient of P3 amplitude increasing from anterior to posterior sites);
however, there was a clear distinction in P3 topography between the
CPT A-X response inhibition conditions and the CPT 123 nontarget conditions.
In addition, slower reaction time, lower amplitude, and longer latency
P3 potentials were present for the CPT 123 as compared to the CPT
A-X paradigm.

Hypobaric hypoxia effects upon psychophysiological measures and complex task performance
Glenn F. Wilson1, Carolyne Swain2, & Chrysoula Kourtidou-Papadeli3
1Armstrong Laboratory, 2Logicon Technical Services, Inc, 3Wright State University
Cognitive impairment at high altitudes is of vital interest to thos
working in aviation. Few studies have examined the relationship between
task performance and psychophysiological responses during acute high
altitude hypoxia. This study evaluated the psychophysiological correlates
of complex task performance during hypobaric induced hypoxia at altitudes
of 10000, 15000, 20000 and 25000 ft. Ten AF personnel performed a
multi-task battery, consisting of visual monitoring, manual tracking
and auditory communication tasks, under three conditions (hypobaric
normoxia, hypobaric hypoxia, and recovery) at each altitude. EEG,
ECG, EOG and respiratory activity including pO2 and pCO2 were continuously
monitored. For analysis, FFTs of artifact corrected EEG recorded from
19 sites were calculated for 30 s epochs across each condition. The
averaged heart, blink and breath rates, plus breath and blink amplitudes
were also compared. As expected, breath and heart rates increased
significantly at all altitudes and breath amplitude increased at 25,000
ft. The EEG results indicated that theta, alpha, and beta powers increased
at 30-60 s and delta power increased at 90-150 s after oxygen removal
and did not return to normoxic levels for 2-2.5 s after the oxygen
mask was replaced. At 25,000 ft., delta levels were significantly
correlated with tracking errors while at 20,000 ft. theta, alpha,
and beta levels were correlated with tracking RMSE. Furthermore, tracking
performance did not fully recover until 1.5 minute after oxygen was
replaced. However, the visual monitoring and auditory communications
tasks were less sensitive to hypoxia than was the visuo-manual task.

Trauma cues influence EEG spectra in PTSD
Steven H. Woodward, Ned J. Arsenault, Craig Santerre, & Geoff Michell
Psychophysiological approaches to PTSD have focused almost exclusively
upon autonomic measures of sympathetic arousal. We will present an
examination of EEG spectral indices under conditions of exposure to
trauma cues. Twenty-five Vietnam combat veterans undergoing inpatient
treatment for PTSD were presented a series of slides and accompanying
sound tracks, each 30 seconds in duration. Combat-related (18) and
pleasant stimuli (24) were presented in alternating blocks of six.
Classical spectral estimation procedures were applied to EEG recorded
at F3, F4, Cz, and Pz. As EOG artifact contaminated frontal and lower
frequency EEG, initial analyses were limited to high frequency (20-32
Hz) EEG recorded at Pz. Significant main effects of block and stimulus
type were observed. Beta-band power increased over blocks and was
higher for combat-related than for pleasant slides. There was no block
by stimulus type interaction. Preliminary analyses further suggested
that frontal alpha power asymmetry was not reliably influenced by
the presentation of trauma cues, though trends in the data suggested
a shift toward relative right hemisphere activation during the first
set of combat-related slides. We would note that a linked mastoid
reference was utilized. EOG artifact reduction may further sharpen
these results. The patient sample was approximately split between
veterans who were trauma cue "responders" versus "non-responders"
as defined by heart rate accelerations (> 2 BPM) in response to the
first combat-related slides. These groups exhibited no significant
differences in EEG spectral spectral indices under the conditions
of trauma cue exposure. While classical spectral analyses suggest
that the EEG is reactive in response to trauma cues, we may look to
short-time spectral or wavelet methods to better resolve the evolution
of EEG responses to such stimuli in PTSD patients.

Eye fixation related potentials during a computer graphic task
Akihiro Yagi, Yuji Takeda, Masayoshi Nagai, & Eri Sakamaki
Kwansei Gakuin University
We can obtain an ERP associated with the occurrence of fixation pauses
by averaging data starting with the termination of a saccade. This
ERP is referred to as the eye fixation related potential (EFRP). The
purpose of this study was to examine variations of EFRP in a computer
graphic task at a visual display unit. Subjects were seven students
who were beginners of the computer graphic task. They were asked to
make copies of flags of various countries by using a software for
drawing. The task consisted of two sessions. The duration of each
session was one hour. A rest period for 10 min was inserted between
sessions. EEGs (Oz) and EOGs were recorded. Each session was divided
into six periods every 10 min for analysis in time course. EEGs time-locked
to the termination of a saccade were averaged to obtain EFRP for each
period. Before and after the task, subjects was asked to fill in questionnaires
for the feeling of fatigue, mental concentration, the sleepiness scale.
Values in questionnaires after the task were compared with those before
the task. The value of mental concentration increased. Sleepiness
decreased. The values in the feeling of fatigue showed few changes.
Amplitudes of EFRP increased gradually for both sessions as the task
went on. The increases were significant. As the subjects became accustomed
to performing the task, they we re enjoyed and absorbed in the task.
The result suggested that EFFORT might reflect the level of visual
attention to the task.

P50 and clinical symptomatology in recent-onset schizophrenia
Cindy M. Yee, Patricia M. White, Sarah E. Morris, & Keith H. Nuechterlein
University of California-Los Angeles
The P50 component of the ERP has been found to be suppressed in response
to a second stimulus, relative to the response elicited by an initia
stimulus, when two clicks are delivered 500 ms apart. The first click
is believed to activate an inhibitory process that reduces processing
of subsequent stimuli. The second click tests the degree of inhibition.
Schizophrenic patients often fail to exhibit P50 suppression, suggesting
a loss of normal inhibition. Specific factors that influence or mediate
P50 suppression have yet to be fully delineated. Previous research
suggests that in nonpsychiatric subjects, P50 suppression can be disrupted by increased anxiety or heightened muscle activity. We examined, in 17 recent-onset schizophrenic patients, whether P50 might be related
to residual clinical symptom levels when patients are clinically stable.
Symptoms were assessed with the Brief Psychiatric Rating Scale (BPRS).
Increases on the Anxiety-Depression factor were associated with a
larger P50 to the second click (r=.76, p<.001) and an increased P50
ratio score (r=.63, p<.01). Analyses of the items that compose this
factor revealed an association between heightened anxiety and P50
amplitude to the second click (r=.49, p<.05) and the P50 ratio measure
(r=.56, p<.02). Depressive mood was related to P50 amplitude to the
first (r=.48, p<.05) and second (r=.81, p<.001) clicks. Guilt feelings
and somatic concern showed no relationships. Greater range on other
BPRS factors may be necessary to detect significant associations.
It does appear, however, that relationships with affective state
are evident even at a clinically stable point.

HTML Coding by Chris Lovelace, 970819