www.elsevier.com/locate/pnpbp

Progress in Neuro-Psychopharmacology & B

Electrophysiological substrates of impulsiveness: potential effects

on aggressive behavior

Rebecca J. Houstona,*, Matthew S. Stanfordb

aResearch Institute on Addictions, State University of New York at Buffalo, 1021 Main Street, Buffalo, NY 14203, United StatesbDepartment of Psychology and Neuroscience, Baylor University, Waco, TX 76798, United States

Accepted 19 November 2004

Available online 28 December 2004

Abstract

Previous investigations attempting to examine impulsiveness as a personality construct are likely confounded with a high incidence of

aggressive and antisocial behavior. The present study assessed electroencephalographic activity at rest and during photic stimulation in two

groups: (1) an impulsive group (n=10) scoring high on the Barratt Impulsiveness Scale (BIS-11) and reporting no indication of impulsive

aggressive behavior; and (2) a non-aggressive control group (n=14) scoring within the normal range on the BIS-11. All subjects completed a

brief battery of personality measures related to impulsivity and aggression. Resting EEG was recorded at 9 electrode sites. Photic stimulation

was administered at three frequency levels. The primary findings were consistently lower frontal delta and theta activity in the impulsive

group as well as a different topographical pattern of beta activity between the groups. These differences appeared to be independent of photic

stimulation. Personality analyses indicated significantly greater hostility and lifetime history of aggression in the impulsive group. Taken

together, the personality and EEG results suggest some similarity between the present impulsive group and research on groups regularly

exhibiting premeditated aggression. These results provide unique insight into the construct of impulsivity and its role in the expression of

specific subtypes of aggressive behavior.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Aggression; EEG; Impulsivity

1. Introduction

Impulsivity is generally considered a tendency to respond

quickly to a given stimulus, without deliberation and

evaluation of consequences (Buss and Plomin, 1975).

Moeller et al. (2001) have described it as ba predisposition

toward rapid, unplanned reactions to internal or external

stimuli without regard to the negative consequences of these

reactions to themselves or others.Q As impulsiveness is often

0278-5846/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.pnpbp.2004.11.016

Abbreviations: BIS-11, Barratt Impulsiveness Scale; BDHI, Buss-

Durkee Hostility Inventory; BPAQ, Buss-Perry Aggression Questionnaire;

EEG, electroencephalogram; I7, Impulsiveness Questionnaire; LHAQ,

Lifetime History of Aggression Questionnaire; PSW, positive slow wave;

VEP, visual evoked potential.

* Corresponding author. Tel.: +1 716 887 2579; fax: +1 716 887 2477.

E-mail address: rhouston@ria.buffalo.edu (R.J. Houston).

considered a trait personality feature, there have been

numerous attempts to explain and understand this construct

from a biological perspective (Barratt, 1985; Eysenck, 1981;

Gray, 1987). One of the most popular theories regarding

impulsivity is the notion of physiological underarousal

(Barratt, 1985; Eysenck, 1967, 1981; Eysenck and Eysenck,

1985; Gray, 1987; Raine, 1993). This is related to Hebb’s

(1955) concept of an boptimal level of arousal.Q The

hypothesis states that arousal level and sensory intake are

related in such a way that an optimal level of arousal is

maintained. When the arousal level falls below the

optimum, stimulation seeking behavior increases and serves

to raise arousal back up to the optimum. Thus, at rest,

impulsive individuals function at a lower level of physio-

logical arousal, which, according to Hebb (1955), is a

subjectively unpleasant state. As a result, these individuals

iological Psychiatry 29 (2005) 305–313

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313306

engage in impulsive behavior in order to raise that arousal to

a more optimal level (Barratt, 1985). The notion of

physiological underarousal has also been applied to other

samples often characterized by impulsiveness, such as

aggressive, antisocial, and psychopathic individuals (Convit

et al., 1991; Fishbein et al., 1989; Raine, 1993).

Several studies have demonstrated greater EEG slowing

in individuals scoring high on self-report measures of

impulsivity (Barratt and Patton, 1983; O’Gorman and

Lloyd, 1987; Stenberg, 1992), and most have interpreted

these findings as evidence of low resting cortical arousal in

impulsive individuals (Barratt, 1985). In addition, event-

related potential studies demonstrating reduced P3 ampli-

tudes in impulsive aggressive college student (Gerstle et al.,

1998; Mathias and Stanford, 1999) and prisoner (Barratt et

al., 1997) samples have also contributed to the notion of

physiological underarousal related to impulsivity. However,

it should be noted that studies of individuals characterized

by chronic premeditated aggression, who also score high on

self-report measures of impulsivity, have indicated no

significant differences in P3 amplitude (Barratt et al.,

1997; Stanford et al., 2003). Thus, there appears to be

some discrepancy as to what may be driving the neuro-

physiological findings with regard to impulsivity. Finally,

the augmenting–reducing paradigm, an evoked potential

indication of cortical reactivity to increasing stimulus

intensity, has repeatedly demonstrated significant augment-

ing in impulsive (Barratt et al., 1987; Carrillo de la Pena and

Barratt, 1993), sensation seeking (Zuckerman, 1991),

extraverted (Stenberg et al., 1990), and impulsive aggressive

(Houston and Stanford, 2001) subjects. These results have

been interpreted within the stimulation seeking framework,

but may also suggest greater physiological reactivity in

impulsive subjects.

Studies using autonomic measures have also suggested

low resting arousal in various samples characterized by high

levels of impulsivity including antisocial and delinquent

youth (Farrington, 1997; Gatzke-Kopp et al., 2002),

aggressive (Scarpa and Raine, 1997) and psychopathic

individuals (Hare, 1978; Lykken, 1995), and behavior

disordered children (Lahey et al., 1995; Raine and Jones,

1987). However, a recent study examining heart rate in

college students recruited solely on their self-reported

impulsivity status also demonstrated low resting heart rate

in subjects scoring high in impulsiveness (Mathias and

Stanford, 2003). This investigation also reported greater

initial reactivity upon stimulation in high impulsive subjects

during an arousal challenge task. Consequently there

appears to be evidence from both electrocortical and

autonomic indices supporting the notions of low resting

arousal and greater physiological reactivity in impulsive

individuals.

As noted, physiological underarousal is a popular theory

for explaining the expression of aggressive and violent

behavior as well (Convit et al., 1991; Drake et al., 1992;

Fishbein et al., 1989; Gatzke-Kopp et al., 2001; Houston

and Stanford, 2001). This is not surprising as impulsivity

and aggression are behavioral constructs that are difficult to

separate. As a result, many of the aforementioned con-

clusions on impulsivity were drawn from studies using

samples in which high levels of aggressive behavior were

anticipated. Conversely, the few psychophysiological stud-

ies that have examined subjects based on impulsiveness did

not account for potential mediating effects of aggression or

related factors (Barratt and Patton, 1983; O’Gorman and

Lloyd, 1987; Mathias and Stanford, 2003; Stenberg, 1992).

The current study presents a different approach in the

psychophysiological assessment of impulsivity. In order to

address the issues of physiological underarousal and

reactivity associated with impulsiveness and how it might

relate to aggressive behavior, the authors examined electro-

encephalographic activity at rest and during a photic

stimulation task in subjects scoring high on a self-report

measure of impulsivity (BIS-11; Patton et al., 1995) and

reporting no indication of impulsive aggressive behavior.

Because neurobiological deficits have been repeatedly

demonstrated in impulsive aggressive samples, the authors

chose to specifically exclude this subtype of aggressive

behavior rather than aggression in general. This approach

may be useful in attempting to separate impulsivity as a

personality construct from a specific form of aggressive

behavior associated with a lack of behavioral control.

Furthermore, the criteria for chronic impulsive aggressive

behavior have been validated in several previous studies in

prisoner, community, and college student samples, whereas

determining cutoffs for generalized aggression would be

somewhat arbitrary.

2. Methods

2.1. Subjects

Subjects (n=24) were recruited from undergraduate

courses at a mid-sized southern university as part of a

multiple-study screening process. A two page questionnaire

containing the Barratt Impulsiveness Scale (BIS-11; Patton

et al., 1995), a set of questions concerning impulsive

aggressive criteria (e.g. over the past 6 months, have you

had episodes where you were enraged or angry in such a

way that you felt it was excessive or inappropriate to the

situation; Houston and Stanford, 2001 for details on

impulsive aggression criteria in this type of sample) and

other demographic information was distributed in these

courses.

The criteria for inclusion in the impulsive group were the

following: (i) a total score of z74 on the BIS-11 (greater

than 1 S.D. above published norms; Patton et al., 1995) and

(ii) negative responses to all impulsive aggression criteria,

which included 0 outbursts in the last 6 months and scoring

b8 on the Irritability subscale of the Buss–Durkee Hostility

Inventory (BDHI; Buss and Durkee, 1957). Normal control

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313 307

subjects were drawn from those subjects whose BIS-11

scores were within 1 S.D. above or below the standardized

mean and also responded negatively to all impulsive

aggression criteria (including a score of b8 on the BDHI

Irritability subscale). In addition, all recruited subjects were

right-handed.

Of the 2118 students who completed the screening

questionnaire, 57 met criteria for the impulsive group while

161 were classified as potential control subjects. Subjects

who met some but not all criteria or did not provide

adequate contact information were removed from further

use in this investigation. Subjects who met criteria were

solicited via telephone to participate in the remainder of the

study. Potential subjects were asked a few brief questions

over the phone in order to exclude individuals that were

taking psychoactive medication or had a history of head

injury or seizure. Of the 57 impulsive subjects, 12 refused

participation and 21 reported significant head injury, seizure

activity, or regular psychoactive medication. We were

unsuccessful in contacting 5 of the potential subjects. As a

result, 19 subjects agreed to participate. Seven subjects did

not show up for their appointments and subsequent attempts

to reschedule were unsuccessful.

Participating subjects underwent approximately 2 h of

personality and psychophysiological assessment. Upon

arrival, each subject was asked to read and sign a consent

form. Then each subject was interviewed about psychiatric,

substance abuse, and medical history. Subjects reporting any

Axis I psychiatric diagnoses or significant neurological/

medical problems were excluded from further participation

(n=2 from the impulsive group). Thus, data were collected

for the following groups: impulsive group (n=10, 4M/6F;

mean age 20.3; S.D. 1.2; mean years of education 13.0, S.D.

0.8) and normal control group (n=14, 9M/5F; mean age

19.4, S.D. 1.0; mean years of education 13.1, S.D. 1.0).

2.2. Personality assessment

Subjects were administered the following personality

instruments:

1) I7 Impulsiveness Questionnaire (I7; Eysenck et al.,

1985); the I7 Impulsiveness Questionnaire is a 54-item

questionnaire assessing impulsivity, venturesomeness,

and empathy.

2) Buss–Perry Aggression Questionnaire (BPAQ; Buss

and Perry, 1992); The Buss–Perry Aggression Ques-

tionnaire is a 29-item instrument consisting of four

subscales: Physical Aggression, Verbal Aggression,

Anger, and Hostility. The subscale scores can also be

summed for a total aggression score.

3) Lifetime History of Aggression Questionnaire (LHAQ;

Coccaro et al., 1997); the Lifetime History of Aggres-

sion Questionnaire is administered in an interview

format and assesses incidence of aggressive behavior

from age 13 through adulthood. The LHAQ is

composed of three subscales: the aggression subscale,

the consequences/antisocial behavior subscale, and the

self directed aggression subscale. The aggression

subscale pertains to incidences of aggression such as

verbal and physical fights, throwing/breaking objects,

and childhood temper tantrums. The consequences/

antisocial subscale pertains to the individual’s history of

disciplinary action at school and work as well as any

arrests, convictions, and illegal activities, regardless of

aggressive content. The self directed aggression sub-

scale pertains to acts of self-mutilation and suicide

attempts.

2.3. Electrophysiological assessment

Each subject was seated in a comfortable chair in a sound

and light attenuated room. The scalp was prepared by

application of rubbing alcohol and a mildly abrasive gel

(NuPrep). An electrocap was fitted onto each subject and

EEG data were collected for 9 electrode sites (F3, Fz, F4,

C3, Cz, C4, P3, Pz, and P4). Electrodes were referenced to

linked ears. An eye blink electrode was affixed below the

left eye to allow for differentiation of artifact and removal of

data collected during eye blinks. Impedance for each

electrode was less than 5 Kohms. Filter bandpass was set

at 0.5 and 35 Hz. Five minutes of resting EEG was recorded

with eyes closed.

Selection of EEG segments for analysis was done by

visual inspection, and segments containing artifact (i.e. eye

movements, eye blinks, etc.) were excluded. The process of

eliminating data was performed blind to subject group

assignment. For the resting EEG condition, the average

length of the combined artifact-free segments was 243.21 s.

Signal amplification was conducted using the NE-3 ampli-

fier and the sampling rate was 128 Hz. Artifact-free

segments were subjected to spectral analysis by a Fast

Fourier Transform. Absolute EEG power was calculated by

collapsing the frequency spectrum into EEG frequency

bands of delta (1–4 Hz), theta (5–7 Hz), alpha (8–12 Hz),

and beta (13–20 Hz) (Davidson et al., 2000).

2.4. Photic stimulation

The subject was instructed to focus on the center of a

photostimulator positioned approximately 25 cm from the

subject’s eyelids. The photic stimulation was a red flicker

with flash intensity of 1.40 kcd/m2. The stimulus frequen-

cies used were 5, 10, and 15 Hz. The stimulus train at each

frequency was applied for 30 s with a 10 s time interval. All

subjects were instructed to remain awake and keep their

eyes closed throughout the testing period. EEG power

during photic stimulation was analyzed using the procedures

described above for the resting EEG condition. Average

lengths of the combined artifact-free segments for each

stimulus frequency were as follows: 5 Hz, 23.48 s; 10 Hz,

23.95 s; 15 Hz, 24.03 s.

Fig. 1. Group comparison of delta power at frontal sites. Analyses indicated

significantly lower delta power in the impulsive group for the frontal region

[ F(1,18)=5.31, p=.033].

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313308

2.5. Statistical analyses

All analyses were conducted using SPSS (Version 11.0)

statistical package for Windows 98. The probability of Type

I error was set at 0.05 for all omnibus tests. The Geisser–

Greenhouse (Geisser and Greenhouse, 1958) conservative F

test was utilized as a correction in all cases to guard against

violations of the sphericity assumption. For all follow-up

comparisons the probability of Type I error was set at 0.10

and the Bonferroni inequality was used to control for Type I

error inflation (Stevens, 2002). For the personality meas-

ures, univariate ANOVAs were utilized to examine group

differences with all subscale and total scores serving as

dependent variables. Prior to analysis, power–spectral

densities were transformed to their natural logarithmic

equivalents (Gasser et al., 1982). For analyses of EEG data,

the stimulation condition (resting, 5, 10, and 15 Hz), scalp

region (frontal, central, and parietal), and electrode site

served as within subject factors and group served as the

between subjects factor.

3. Results

3.1. Demographic and personality measures

Analyses of demographic and personality data indicated

no significant differences in age or education between the

two groups. Personality data for each group are located in

Table 1. Analysis of the LHAQ indicated a significant group

difference on all subscales and the total score. Impulsive

subjects scored significantly higher on the aggression

subscale [ F(1,22)=20.85, p=b.001, partial g2=0.49,

Table 1

Comparison of personality measures

Impulsive group

(n=10)

Control group

(n=14)

M S.D. M S.D.

Lifetime History of

Aggression Questionnaire

Aggression** 5.20 2.25 1.57 1.65

Consequences/antisocial

behavior*

2.10 1.97 0.50 0.85

Self directed aggression* 0.40 0.70 0.00 0.00

Total lifetime history of

aggression**

7.70 3.56 2.07 2.05

Buss–Perry Aggression

Questionnaire

Physical aggression 19.20 4.80 17.71 4.29

Verbal aggression 13.90 2.88 13.21 2.83

Anger 14.00 3.02 11.64 2.68

Hostility* 18.10 3.98 14.00 4.87

Total aggression 65.20 8.60 56.57 11.44

I7Impulsiveness** 13.00 2.62 6.00 3.42

* pb05.

** pb01.

Power=0.99], the consequences/antisocial behavior subscale

[F(1,22)=7.40, p=.012, partial g2=0.25, Power=0.74], the

self directed aggression subscale [F(1,22)=4.67, p=.042,

partial g2=0.18, Power=0.54], and the total lifetime history

of aggression score [F(1,22)=24.05, p=b.001, partial

g2=0.52, Power=0.99]. On the BPAQ, impulsive subjects

scored significantly higher than normal controls on the

Hostility subscale [F(1,22)=4.78, p=.040, partial g2=0.18,

Power=0.55]. Higher impulsivity was confirmed in the

impulsive subjects by significantly higher scores on the

Impulsiveness subscale of the I7 [F(1,22)=29.39, p=b.001,

partial g2=0.57, Power=0.99]. There were no other signifi-

cant group differences on personality measures.

3.2. EEG measures

3.2.1. Delta power

Repeated measures analysis in the delta frequency band

indicated a significant region�group interaction [F(2,32)=

7.28, p=.008, partial g2=0.31, Power=0.81], which resulted

in a significant simple effect of lower frontal delta power in

impulsive subjects as compared to control subjects

[F(1,18)=5.31, p=.033, partial g2=0.23, Power=0.59; Fig.1]. There were no significant simple effects for the central or

parietal scalp regions.

3.2.2. Theta power

Repeated measures analysis of theta power revealed a

significant region�group interaction [F(2,32)=9.83, p=.002,

partial g2=0.38, Power=0.92]. Analysis of simple effects

indicated significantly lower theta power in the impulsive

group for the frontal region [F(1,18)=6.46, p=.020, partial

g2=0.26, Power=0.67; Fig. 2]. There were no significant

simple effects for the central or parietal scalp regions.

Analysis of theta power also indicated a significant main

effect for stimulus condition [F(3,48)=5.22, p=.021, partial

g2=0.25, Power=0.69; Table 2]. Further analysis indicated a

significant linear trend [F(1,16)=37.44, p=b.001, partial

Fig. 2. Group comparison of theta power at frontal sites. Analyses indicated

significantly lower theta power in the impulsive group for the frontal region

[ F(1,18)=6.46, p=.020].

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313 309

g2=0.70, Power=1.00] in theta power across stimulation

conditions. Main comparisons revealed significantly greater

theta power in the resting condition compared to the 10 Hz

[F(1,16)=56.88, pb.001, partial g2=0.78, Power=1.00] and

15 Hz [F(1,16)=26.75, p=b.001, partial g2=0.63, Power=0.99] conditions.

3.2.3. Alpha power

Repeated measures analyses in the alpha frequency

indicated a significant stimulus condition�region�group

interaction [ F(6,96)=2.64, p=.047, partial g2=0.14,

Power=0.68]. Follow-up comparisons revealed significantly

lower frontal alpha power in the impulsive subjects at the 10

Hz stimulus condition [F(1,19)=5.80, p=.026, partial

g2=0.23, Power=0.63].

3.2.4. Beta power

Analysis of beta power revealed a significant region-�group interaction [F(2,32)=5.07, p=.031, partial g2=0.24,Power=0.62]. Follow-up comparisons indicated different

regional patterns of beta power between the two groups

(Fig. 3). In the impulsive subjects, beta power increased in

Table 2

EEG power across stimulation conditions

Delta Thetaa Alpha Betab

M S.D. M S.D. M S.D. M S.D.

Rest 4.95 0.71 2.95 0.82 4.11 1.19 2.69 0.80

5 Hz 4.63 1.04 2.78 0.80 3.67 0.95 3.00 0.80

10 Hz 4.69 0.74 2.71 0.82 4.02 1.00 3.04 0.79

15 Hz 4.69 0.74 2.72 0.76 3.64 1.03 3.44 0.91

Table contains natural log power values.a Significant linear trend for stimulus condition [ F(1,16)=37.44,

pb.001] indicating a decrease in theta power with increasing stimulation

frequency.b Significant linear trend for stimulus condition [ F(1,16)=19.82,

pb.001] indicating an increase in beta power with increasing stimulation

frequency.

an anterior to posterior fashion with frontal beta power

being lowest and different from that at both central

[F(1,7)=13.88, p=.007, partial g2=0.67, Power=0.89] and

parietal [ F (1,7)=22.07, p=.002, partial g2=0.76,Power=0.98] regions and central beta power being lower

than parietal beta power [F(1,7)=21.35, p=.002, partial

g2=0.75, Power=0.98]. In control subjects, frontal

[F(1,9)=7.08, p=.026, partial g2=0.44, Power=0.66] and

central [F(1,9)=9.08, p=.015, partial g2=0.50, Power=0.77]

beta power were lower than that at the parietal region, but

were not significantly different from each other. In an

attempt to better understand this group difference in the

topographical distribution of beta activity, correlational

analyses were conducted between mean beta power for

each scalp region (frontal, central, and parietal) and scores

on the personality measures. A significant negative relation-

ship was indicated for frontal beta power and the LHA

consequences/antisocial behavior subscale (r=�0.452). No

significant relationships were revealed for beta power and

personality scores in the central or parietal regions.

Analyses in the beta frequency band also indicated a

significant stimulus condition�region interaction [F(6,96)=

2.92, p=.041, partial g2=0.15, Power=0.68]. This interaction

was ordinal in nature, therefore the main effect for stimulus

condition was further interpreted. For stimulus condition, a

significant linear trend indicated a fairly consistent increase

in beta power as stimulation frequency increased [F(1,16)=

19.82, pb.001, partial g2=0.55, Power=0.99; Table 2]. Main

comparisons proved beta power to be highest for the 15 Hz

condition as compared to the 10 Hz [F(1,16)=22.95, pb.001,

partial g2=0.59, Power=0.99], 5 Hz [F(1,16)=24.23, pb.001,partial g2=0.60, Power=0.99], and resting conditions

[F(1,16)=20.98, pb.001, partial g2=0.57, Power=0.99].

Accordingly, beta power exhibited at the 10 Hz [F(1,16)=

7.62, p=.014, partial g2=0.35, Power=0.78] and 5 Hz

[F(1,16)=8.51, p=.010, partial g2=0.32, Power=0.74] con-

ditions was significantly higher than that exhibited during the

resting condition. Main effects for scalp region were not

Fig. 3. Regional comparison of beta power. Analyses indicated different

regional patterns of beta power between the two groups.

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313310

explored in light of the aforementioned group�region

interaction.

4. Discussion

The present study yielded four primary findings for the

impulsive group: (1) a greater lifetime incidence of

aggressive behavior; (2) higher BPAQ hostility scores; (3)

lower frontal delta and theta activity; and (4) a different

regional pattern of beta activation.

4.1. Impulsivity and aggression

With regard to the greater lifetime history of aggressive

behavior in the impulsive group, it is important to acknowl-

edge that aggressive behavior, in conjunction with high

impulsivity, is common throughout the clinical and empiri-

cal literature regardless of the predominant type of

aggressive behavior (Lee and Coccaro, 2001). Furthermore,

impulsivity is a defining characteristic in several research

groups associated with aggressive behavior that is not

necessarily impulsive or affective in nature such as

premeditated aggression (Barratt et al., 1997; Stanford et

al., 2003) and psychopathy (Cleckley, 1976; Hare, 1993).

As a result, the underlying question is whether impul-

siveness, as a personality trait, differentially influences the

expression of different types of aggressive behavior.

Unfortunately, the data presented at this time cannot provide

an unequivocal answer, but it does allow for some intriguing

suggestions. For example, the impulsive subjects did not

differ from normal controls on most subscales of the BPAQ,

such as physical and verbal aggression, thus corroborating

the ability to control their behavior with respect to

aggression. The higher hostility scores, combined with a

greater lifetime history of aggressive acts, may imply a

personality profile similar to that of the antisocial individual,

who is more likely to commit controlled, premeditated acts

of aggression (Stanford et al., 2003). Interestingly, the

majority of the aggressive acts reported by impulsive

subjects on the LHAQ lacked an affective or reactive

component. This observation is purely anecdotal and cannot

be considered empirical evidence; it is, however, highly

consistent with the greater hostility scores in the impulsive

group.

4.2. Electrocortical findings: interpretation and implications

The findings of lower frontal delta and theta activity in

the impulsive subjects are particularly striking considering

the large body of work demonstrating increased slow wave

activity in impulsive and aggressive samples (Barratt and

Patton, 1983; Convit et al., 1991; Gatzke-Kopp et al., 2001;

O’Gorman and Lloyd, 1987; Stenberg, 1992). Of course,

these previous investigations did not examine the constructs

of impulsivity and impulsive aggression independently.

Historically, decreases in the delta and theta frequency

bands have been associated with greater cortical activation.

Suppressed theta has been repeatedly associated with

hypervigilance and attention (Beatty et al., 1974; Beatty

and O’Hanlon, 1980; Valentino et al., 1993). Decreases in

delta activity in response to nicotine administration have

also been interpreted as evidence of greater cortical activity

and alertness (Knott and Harr, 1996; Pritchard, 1991). Thus,

the present findings could potentially be interpreted as

evidence of greater alertness or attention in this group as

compared to controls. Moreover, group discrepancies in the

frontal region are consistent with a multitude of previous

studies indicating frontal lobe dysfunction in association

with impulsive behavior (Fuster, 1997).

The implication of hypervigilance or enhanced attention

in the current impulsive group is remarkably similar to one

of the more popular theories of physiological function in

psychopath samples (Blackburn, 1979; Raine, 1993). This

research contends that psychopaths exhibit better processing

of relevant stimuli or enhanced attention as indexed by

larger P3 event-related potential and positive slow wave

(PSW) amplitudes (Flor et al., 2002; Raine and Venables,

1988; Raine, 1989, 1993). Given that previous research has

indicated a negative correlation between P3 amplitude and

EEG power in the lower frequency bands (Barratt, 1985),

the present findings could suggest a similarity between our

impulsive group and a psychopath-like sample. As noted,

psychopaths are more likely to commit controlled, premedi-

tated acts of aggression (Williamson et al., 1987; Wood-

worth and Porter, 2002). Thus, it is possible that the current

impulsive group may represent a type of bpremeditated

aggressor,Q possibly similar to a psychopath sample.

In addition, theta rhythms have been linked with various

aspects of cognitive and memory function (Basar, 1998;

Basar et al., 1998). Basar et al. (2001) have concluded that

bspontaneous theta activity of the frontal cortex is a factor

influencing the amplitudes of frontal VEPsQ (p. 198). As

visual event-related potential (VEPs) deficits have been

associated with impulsiveness (Barratt et al., 1987; Carrillo

de la Pena, 1992; Carrillo de la Pena and Barratt, 1993), the

lower frontal theta activity in the current impulsive group

could have important implications in terms of abnormal

cognitive function in these individuals. Future research

might incorporate methodology better designed to inves-

tigate the relationship between these EEG oscillations and

cognitive function in an attempt to further disentangle the

underlying substrates of impulsive and aggressive behavior.

With regard to the group differences in regional beta

activation, correlational analyses indicated a significant

relationship between frontal beta power and the LHA

consequences/antisocial behavior subscale. This subscale

directly assesses behaviors that are often exhibited by

antisocial individuals (i.e. rule violations, arrests) (Coccaro

et al., 1997). Thus, these results are consistent with research

demonstrating less frontal beta reactivity in relation to

antisocial behavior (Finn et al., 2000). Again, this could

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313 311

provide further support for a similarity between the

impulsive group in the current study and an antisocial or

premeditated profile.

As for the photic stimulation, this task did not

demonstrate greater reactivity in the impulsive group. Given

that EEG power, as measured in the present study, is a

relatively non-specific measure of cortical function, it is

possible that measuring it in response to such stimulation

may be ineffective in determining reactivity differences,

especially in a high functioning sample. Previous studies

using a similar paradigm have demonstrated a diminished

capacity to respond in disorders such as schizophrenia (Jin

et al., 2000) and Alzheimer’s disease (Wada et al., 1998).

Therefore, another influential factor in using this paradigm

may be the necessity of more severe pathology.

4.3. Study limitations

The present results should be interpreted with some

limitations in mind. This study is based on a small sample

size, and despite fairly robust effects, replication of these

results could benefit substantially from a larger sample.

This is particularly true as this approach is different from

previous EEG studies of impulsiveness. The two groups

examined here also differed with respect to gender; the

impulsive group contained a higher ratio of female

subjects. It is unlikely that this significantly impacted the

present results, particularly since recent work has demon-

strated lower theta power in males as compared to females

(Clarke et al., 2001). However, it would be more

appropriate to have gender-matched groups. Furthermore,

although we have interpreted these results as potential

evidence of similarity between the present impulsive group

and antisocial or psychopathic groups, the current study

did not employ any direct measures of psychopathy or

related features aside from impulsivity and aggression.

Continuing work in this area would benefit from a more

extensive battery of personality assessments. For example,

a more extensive examination of psychopathy, criminal

history, and other relevant antisocial behaviors would

prove useful in disentangling these constructs. Finally,

future EEG work in this area should also explore the use of

narrow or individually adjusted frequency bands (Kli-

mesch, 1999) to determine whether differences in cortical

activation in these groups may be restricted to a more

specific frequency.

5. Conclusions

The current findings do raise the question as to the exact

role of impulsivity in the expression of different types of

aggressive behavior. A key element in this distinction is the

notion of behavioral control, particularly in the face of

emotion. Individuals high in impulsivity may be unable to

control certain behavioral impulses, but this phenomenon is

not necessarily related to emotion. Therefore, while some

individuals (i.e. antisocials or premeditated aggressors) are

high in impulsiveness, they can exhibit violent behavior of a

controlled nature, unrelated to their emotional state. In

contrast, impulsive aggressive individuals clearly display a

lack of behavioral control specifically when angered,

agitated or upset. The screening questions utilized in this

study explicitly referred to a loss of behavioral control when

enraged, thus excluding those potential subjects who

endorsed such behavior control problems. As a result, the

significantly higher lifetime history of aggression and

hostility scores in the present impulsive group, coupled

with the suggestion of enhanced attention or vigilance in

frontal areas reinforces the suggestion of similarity between

this group and premeditated or antisocial aggressive groups.

Taken as a whole, these results provide intriguing insight

into the role of impulsivity as a personality construct, its

underlying physiological substrates and its relationship with

specific types of aggressive behavior.

Acknowledgement

This research was supported by the Dreyfus Health

Foundation, The Rogosin Institute, New York Hospital-

Cornell Medical Center.

References

Barratt, E.S., 1985. Impulsiveness subtraits: arousal and information

processing. In: Spence, J.T., Izard, C.E. (Eds.), Motivation, emotion,

and personality. Elsevier Science Publishers, New York, pp. 137–143.

Barratt, E.S., Patton, J.H., 1983. Impulsivity: cognitive, behavioral and

psychophysiological correlates. In: Zuckerman, M. (Ed.), Biological

bases of sensation seeking, impulsivity, and anxiety. Lawrence Ear-

lbaum Associates, Hillsdale, NJ, pp. 77–122.

Barratt, E.S., Pritchard, W.S., Faulk, D.M., Brandt, M.E., 1987. The

relationship between impulsiveness subtraits, trait anxiety, and visual

N100 augmenting/reducing: a topographic analysis. Pers. Individ.

Differ. 8, 43–51.

Barratt, E.S., Stanford, M.S., Kent, T.A., Felthous, A., 1997. Neuro-

psychological and cognitive psychophysiological substrates of impul-

sive aggression. Biol. Psychiatry 41, 1045–1061.

Basar, E., 1998. Brain function and oscillations: integrative brain function,

vol. 2. Springer Verlag, Berlin.

Basar, E., Rahn, E., Demiralp, T., Schqrmann, M., 1998. Spontaneous

EEG theta activity controls frontal visual evoked potential ampli-

tudes. Electroencephalogr. Clin. Neurophysiol. Evoked Potentials 108,

101–109.

Basar, E., Schqrmann, M., Sakowitz, O., 2001. The selectively distributed

theta system: functions. Int. J. Psychophysiol. 39, 197–212.

Beatty, J., O’Hanlon, J., 1980. Operant control of posterior theta rhythm

and vigilance performance: repeated treatments and transfer of training.

In: Birbaumer, N., Kimmel, H. (Eds.), Biofeedback and self-regulation.

Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 247–258.

Beatty, J., Greenberg, A., Deibler, W.P., O’Hanlon, J.F., 1974. Operant

control of occipital theta rhythm affects performance in a radar

monitoring task. Science 183, 871–873.

Blackburn, R., 1979. Cortical and autonomic arousal in primary and

secondary psychopaths. Psychophysiology 16, 143–150.

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313312

Buss, A.H., Durkee, A., 1957. An inventory for assessing different kinds of

hostility. J. Consult. Psychol. 21, 343–349.

Buss, A.H., Perry, M., 1992. The aggression questionnaire. J. Pers. Soc.

Psychol. 63, 452–459.

Buss, A.H., Plomin, R., 1975. A temperament theory of personality

development. John Wiley & Sons, New York.

Carrillo de la Pena, M.T., 1992. ERP augmenting/reducing and sensation

seeking: a critical review. Int. J. Psychophysiol. 12, 211–220.

Carrillo de la Pena, M.T., Barratt, E.S., 1993. Impulsivity and ERP

augmenting-reducing. Pers. Individ. Differ. 15, 25–32.

Clarke, A.R., Barry, R.J., McCarthy, R., Selikowitz, M., 2001. Age and

sex effects in the EEG: development of the normal child. Clin.

Neurophysiol. 112, 806–814.

Cleckley, H., 1976. The mask of sanity, 5th ed. Mosby, St. Louis, MO.

Coccaro, E.F., Berman, M.E., Kavoussi, R.J., 1997. Assessment of life

history aggression: development and psychometric characteristics.

Psychiatry Res. 73, 147–157.

Convit, A., Czober, P., Volavka, J., 1991. Lateralized abnormality in the

EEG of persistently violent psychiatric inpatients. Biol. Psychiatry 3,

121–138.

Davidson, R.J., Jackson, D.C., Larson, C.L., 2000. Human electro-

encephalography. In: Cacioppo, J.T., Tassinary, L.G., Berntson, G.G.

(Eds.), Handbook of psychophysiology 2nd ed. Cambridge University

Press, New York, pp. 27–52.

Drake, M.E., Hietter, S.A., Pakalnis, A., 1992. EEG and evoked potentials

in episodic-dyscontrol syndrome. Neuropsychobiology 26, 125–128.

Eysenck, H.J., 1967. The biological basis of personality. Charles C.

Thomas, Springfield, IL.

Eysenck, H.J., 1981. General features of the model. In: Eysenck, H.J. (Ed.),

A model for personality. Springer-Verlag, New York.

Eysenck, HJ., Eysenck, M.W., 1985. Personality and individual differences:

a natural science approach. Plenum Press, New York.

Eysenck, S.B.G., Pearson, P.R., Easting, G., Allsopp, J.F., 1985.

Impulsiveness and venturesomeness: their position in a dimensional

system of personality description. Pers. Individ. Differ. 6, 613–619.

Farrington, D.P., 1997. The relationship between low resting heart rate

and violence. In: Raine, A., Brennan, P.A., Farrington, D.P., Mednick,

S.A. (Eds.), Biosocial bases of violence. Plenum Press, New York,

pp. 89–105.

Finn, P.R., Ramsey, S.E., Earleywine, M., 2000. Frontal EEG response to

threat, aggressive traits and a family history of alcoholism: a

preliminary study. J. Stud. Alcohol 61, 38–45.

Fishbein, D.H., Herning, R.I., Pickworth, W.B., Haertzen, C.A., Hickey,

J.E., Jaffe, J.H., 1989. EEG and brainstem auditory evoked response

potentials in adult male drug abusers with self-reported histories of

aggressive behavior. Biol. Psychiatry 26, 595–611.

Flor, H., Birbaumer, N., Hermann, C., Ziegler, S., Patrick, C., 2002.

Aversive Pavlovian conditioning in psychopaths: peripheral and central

correlates. Psychophysiology 39, 505–518.

Fuster, J.M., 1997. The prefrontal cortex: anatomy, physiology and

neuropsychology of the frontal lobe. 3rd ed. Raven, New York.

Gasser, T., Bacher, P., Mocks, J., 1982. Transformation towards normal

distribution of broad spectral parameters of the EEG. Electroencepha-

logr. Clin. Neurophysiol. 53, 119–124.

Gatzke-Kopp, L.M., Raine, A., Buchsbaum, M., LaCasse, L., 2001.

Temporal lobe deficits in murderers: EEG findings undetected by

PET. J. Neuropsychiatry Clin. Neurosci. 13, 486–491.

Gatzke-Kopp, L.M., Raine, A., Loeber, R., Stouthamer-Loeber, M.,

Steinhauer, S., 2002. Serious delinquent behavior, sensation

seeking and electrodermal arousal. J. Abnorm. Child Psychol.

30, 477–486.

Geisser, S., Greenhouse, S.W., 1958. An extension of Box’s results on the

use of the F distribution in multivariate analysis. Ann. Math. Stat. 29,

885–891.

Gerstle, J.E., Mathias, C.W., Stanford, M.S., 1998. Auditory P300 and self-

reported impulsive aggression. Prog. Neuro-Psychopharmacol. Biol.

Psychiatry 22, 575–583.

Gray, J.A., 1987. Perspectives on anxiety and impulsivity: a commentary.

J. Res. Pers. 21, 493–509.

Hare, R.D., 1978. Electrodermal and cardiovascular correlates of psychop-

athy. In: Hare, R.D., Schalling, D. (Eds.), Psychopathic behavior:

approaches to research. Wiley, New York, pp. 107–144.

Hare, R.D., 1993. Without conscience: the disturbing world of psychopaths

among us. Pocket Books, New York.

Hebb, D.O., 1955. Drives and the CNS (conceptual nervous system).

Psychol. Rev. 62, 243–254.

Houston, R.J., Stanford, M.S., 2001. Mid-latency evoked potential in self-

reported impulsive aggression. Int. J. Psychophysiol. 40, 1–15.

Jin, Y., Castellanos Jr., A., Solis Jr., E.R., Potkin, S.G., 2000. EEG resonant

responses in schizophrenia: a photic driving study with improved

harmonic resolution. Schizophr. Res. 44, 213–220.

Klimesch, W., 1999. EEG alpha and theta oscillations reflect cognitive and

memory performance: a review and analysis. Brain Res. Rev. 29,

169–195.

Knott, V.J., Harr, A., 1996. Assessing the topographic EEG changes

associated with aging and acute/long-term effects of smoking. Neuro-

psychobiology 33, 210–222.

Lahey, B.B., McBurnett, K., Loeber, R., Hart, E.L., 1995. Psychobiology.

In: Sholevar, G. (Ed.), Conduct disorders in children and adolescents.

Ameri. Psychiat. Press, Washington, pp. 27–44.

Lee, R., Coccaro, E., 2001. The neuropsychopharmacology of criminality

and aggression. Can. J. Psychiatry 46, 35–44.

Lykken, D., 1995. The antisocial personalities. Lawrence Erlbaum,

Hillsdale, NJ.

Mathias, C.W., Stanford, M.S., 1999. P300 under standard and surprise

conditions in self-reported impulsive aggression. Prog. Neuro-Psycho-

pharmacol. Biol. Psychiatry 23, 1037–1051.

Mathias, C.W., Stanford, M.S., 2003. Impulsiveness and arousal: heart rate

under conditions of rest and challenge in healthy males. Pers. Individ.

Differ. 35, 355–371.

Moeller, F.G., Barratt, E.S., Dougherty, D.M., Schmitz, J.M., Swann,

A.C., 2001. Psychiatric aspects of impulsivity. Am. J. Psychiatr. 158,

1783–1793.

O’Gorman, J.G., Lloyd, J.E.M., 1987. Extraversion, impulsiveness, and

EEG alpha activity. Pers. Individ. Differ. 8, 169–174.

Patton, J.H., Stanford, M.S., Barratt, E.S., 1995. Factor structure of the

Barratt Impulsiveness Scale. J. Clin. Psychol. 51, 768–774.

Pritchard, W.S., 1991. Electroencephalographic effects of cigarette

smoking. Psychopharmacology 104, 485–490.

Raine, A., 1989. Evoked potentials and psychopathy. Int. J. Psychophysiol.

8, 1–16.

Raine, A., 1993. The psychopathology of crime: criminal behavior as a

clinical disorder. Academic Press, San Diego.

Raine, A., Jones, F., 1987. Attention, autonomic arousal and personality

in behaviorally disordered children. J. Abnorm. Child Psychol. 14,

583–599.

Raine, A., Venables, P.H., 1988. Enhanced P3 evoked potentials and

longer P3 recovery times in psychopaths. Psychophysiology 25,

30–38.

Scarpa, A., Raine, A., 1997. Psychophysiology of anger and violent

behavior. Psychiatr. Clin. North Am. 20, 375–394.

Stanford, M.S., Houston, R.J., Villemarette-Pittman, N.R., Greve, K.W.,

2003. Premeditated aggression: clinical assessment and cognitive

psychophysiology. Pers. Individ. Differ. 34, 773–781.

Stenberg, G., 1992. Personality and the EEG: arousal and emotional

arousability. Pers. Individ. Differ. 13, 1097–1113.

Stenberg, G., Rosen, I., Risenberg, J., 1990. Attention and personality in

augmenting/reducing of visual evoked potentials. Pers. Individ. Differ.

11, 1243–1254.

Stevens, J., 2002. Applied multivariate statistics for the social sciences.

Lawrence Erlbaum Assoc, Mahwah, NJ.

Valentino, D.A., Arruda, J.E., Gold, S.M., 1993. Comparison of QEEG and

response accuracy in good vs. poorer performers during a vigilance task.

Int. J. Psychophysiol. 15, 123–134.

R.J. Houston, M.S. Stanford / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 305–313 313

Wada, Y., Nanbu, Y., Kikuchi, M., Koshino, Y., Hashimoto, T., Yamaguchi,

N., 1998. Abnormal functional connectivity in Alzheimer’s disease:

intrahemispheric EEG coherence during rest and photic stimulation.

Eur. Arch. Psychiatry Clin. Neurosci. 248, 203–208.

Williamson, S., Hare, R.D., Wong, S., 1987. Violence: criminal psycho-

paths and their victims. Can. J. Behav. Sci. 19, 454–462.

Woodworth, M., Porter, S., 2002. In cold blood: characteristics of criminal

homicides as a function of psychopathy. J. Abnorm. Psychology 111,

436–445.

Zuckerman, M., 1991. Psychobiology of personality. Cambridge University

Press, New York.