Altered frontal lobe function suggested by source analysis of event-related potentials in impulsive violent alcoholics

Ina M. Tarkka*,, Jari Karhu1,, Jyrki Kuikka2,, Ari Pääkkönen1,, Kim Bergström2,, Juhani Partanen1, and Jari Tiihonen3,

Department of Neurology and Brain Research and Rehabilitation Center Neuron, Department of
1 Clinical Neurophysiology,
2 Departments of Neurology and Clinical Physiology, Kuopio University Hospital, University of Kuopio and
3 Department of Forensic Psychiatry, Niuvanniemi Hospital, Kuopio, Finland

Received 17 October 2000; in revised form 8 January 2001; accepted 31 January 2001


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Scalp-recorded event-related potentials (ERPs) are sensitive indicators of subtle alterations in cerebral processes. We assessed automatic auditory adaptation and detection of novel stimuli in violent and non-violent alcoholics and normal subjects. Source analysis of ERPs revealed active medial temporal and frontal regions in all subjects. Frontal lobe processed alerting tones in violent alcoholics, whereas non-violent alcoholics and normal subjects processed them in medial temporal brain areas. Detection of deviant tones appeared simultaneously in frontal and temporal areas in violent alcoholics, but sequentially in other groups. These findings imply alterations in arousal and involuntary adaptive processes in cortical networks associated with violent behaviour and alcoholism.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Acute use of ethanol is known to be related to increased aggressive behaviour. Violent behaviour in alcoholics has been linked to altered arousal and presumably to impaired sensory adaptation (Yoshimoto et al., 1992Go; Volavka, 1995Go). Subtle alterations in arousal and in automatic sensory adaptation can be observed in neurophysiological scalp-recorded electrical event-related brain potentials (ERPs). A late auditory cerebral response occurs in humans with a peak latency of 100 ms (known as N100 in electric recordings). This response has been extensively studied and is known to represent the activation of medial temporal auditory and frontal cortex (e.g. Hari et al., 1980; Näätänen and Picton, 1987; Bromm et al., 1992). Furthermore, auditory N100 is a sensitive marker of sensory adaptation (Ritter et al., 1968Go; Fruhstorfer et al., 1970Go). In addition, when unexpected deviant tones are delivered among a string of identical tones, this elicits a mismatch negativity (MMN) observed in the frontocentral electrode sites. MMN presumably reflects the function of an automatic or pre-attentive deviance detector in the auditory system (Näätänen et al., 1978Go; Näätänen and Gaillard, 1983Go). Based on evidence from intracranial and lesion studies, it appears that MMN is generated in the auditory cortex, but that some frontal lobe activity contributes to it (see Alho, 1995).

The cerebral sources of electric and magnetic brain recordings can be estimated using mathematical approaches. Non-invasive 3-dimensional source analysis of EEG and magnetoencephalographic (MEG) data allow both spatial and temporal separation of the major activity of components of ERPs. Thus, active cerebral areas and a specific sequence in their activation patterns can be proposed.

In the present study, auditory ERPs were studied as markers of arousal and adaptation in habitually violent alcoholics, in alcoholics without any history of violent behaviour, and in a group of normal, non-drinking, subjects. The ERPs were elicited using auditory stimuli while the subjects were inattentive. First, identical tones were presented in trains separated by 12 s periods of silence; in the second set, randomly interspersed ‘oddball’ tones were delivered within a continuous string of similar tones. Twelve seconds is considered long enough to allow full recovery of the neural network which was activated by the previous train of tones (Sams et al., 1993Go). Thus, the first tone in each train elicited an arousal. The aim of the present paper was to analyse the sources, as represented by equivalent electrical dipoles, of the auditory ERP wave forms. The source locations, source magnitudes and the temporal characteristics of the source activities between the three groups of subjects were compared.


    SUBJECTS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Subjects
Ten habitually violent alcoholics with poor impulse control (mean age 30.3 years; all males), 10 age-matched normal subjects (mean age 31.3 years; all males), and 10 alcoholics with no history of violent behaviour (mean age 46.7 years; all males) participated in the study. All violent subjects had been committed to an extensive forensic psychiatric examination, because of serious violent offences under the influence of alcohol (for details of offences and diagnosis of violent subjects, see Table 1Go). Examination of the offenders included psychiatric evaluation, a standardized psychological test battery, Structured Clinical Interview for DSM-III-R (SCID), physical examination, electroencephalography (EEG) and observation in the security ward for 4–8 weeks. Violent subjects were abstinent and drug-free for at least 2 months prior to the examination. Violent subjects were diagnosed as typical type 2 alcoholics, six of them early onset. Non-violent alcoholics were recruited from a local rehabilitation centre for alcoholics, where they were treated non-pharmacologically for alcoholism. The abstinence and drug-free period before the recordings varied from 1 to 20 weeks in this group. These subjects had experienced difficulties in their work and social life, but none had committed any criminal offences. All non-violent alcoholics had a previous diagnosis of alcoholism (3039X; Alcoholism; DSM-III-R; American Psychiatric Association, 1987), were interviewed by an experienced psychiatrist (Structured Clinical Interview SCID; Overview and SCID II interview concerning antisocial personality disorder), and also completed the structured interview of the Michigan Alcohol Screening Test (MAST). In association with a related study, a monoamine transporter SPECT measurement, eight violent offenders received once 20 mg of citalopram perorally 2–16 h prior to ERP recordings. The single-dose effect of the 5-HT re-uptake inhibitor on ERP amplitudes was, if anything, towards reducing the ERP amplitudes in those violent subjects. All subjects gave their written informed consent after complete description of the study. The study was approved by the local ethical committee (Kuopio University Hospital).


View this table:
[in this window]
[in a new window]
 
Table 1. Violent alcoholics: anamnestic and diagnostic data
 
Electrophysiological recordings
ERPs were recorded using 19 Ag/AgCl electrodes placed on the scalp according to the International 10–20 System referred to the right mastoid (see Fig. 1aGo for electrode montage). Electro-oculogram was continuously recorded. All signals were amplified and filtered by a Neuroscan Synamps amplifier with a passband of 0.5–50 Hz and digitized at 256 Hz. Unilateral, auditory stimulation was delivered via a small tube inserted in the right ear canal. Intermittent stimulation was 800 Hz, 85 ms tones in trains of four; interstimulus interval (ISI) of 1 s; intertrain interval of 12 s. Recording was transformed off-line to epochs of –100 to 4000 ms relative to the onset of the first stimulus of each train. Data from the second set of stimulation with continuous ISI of 1 s with randomly interspersed ‘oddballs’ (15%; frequency 560 Hz) were transformed to epochs from –100 to 900 ms. Epochs containing eye-movement artefacts were rejected using both automatic (±75 µV) and manual preprocessing. Data were averaged (~40 averages for intermittent stimulation and 300 averages for continuous stimulation for an individual) and digitally filtered with a low pass cut-off frequency at 20 Hz (3 dB point of 24 dB/octave roll-off). No testing was performed between details of the individual drinking history and ERP amplitudes.



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 1. Grand average wave form in violent alcoholics of the first N100, i.e. ‘the arousal’. This is shown in a series of four tones in (a). Topographic instantaneous voltage maps (top row) and current source density maps (bottom row) are shown in (b) from normal subjects (Nor), non-violent alcoholics (Non-viol), and violent alcoholics (Viol). All maps were created for the first in a train N100 peak amplitude at 117 ms.

 
Data analysis
The peak amplitudes of the first and second N100 components of the train stimuli are given in Table 2Go as well as the peak latencies of the standard N100 and the MMN. Non-invasive source analysis of the ERP data, namely the first N100 in a train, the second N100 in a train, the MMN and the standard component of the MMN recording was performed with the Brain Electromagnetic Source Analysis (BESA: Scherg, 1984; Scherg and von Gramon, 1986; Scherg and Picton, 1992; Scherg and Berg, 1996). First, two-dimensional topographic current source density maps were generated for the time window around the peak amplitudes (80–130 ms) of ERPs to identify the differences in the distributions of scalp potentials and to guide the model development. A 4-shell spherical head volume conductor model (with a radius of 85 mm) was used to develop dipole models in order to delineate active brain regions. Spatio-temporal multiple dipole modelling assumes that the equivalent electric dipole represents a small region of active parallel neurons and that the dipole has both a stationary location and orientation but changes its moment, i.e. strength, with time. The source activity represents the dipole moment. It varies in amplitude throughout the analysed window and reflects the temporal variation of the electrical current explained by the dipole. The clearer the main activation pattern of each dipole is temporally separated from the pattern of other dipoles, the more distinct and unique each source is. To achieve this, the model development involves an iterative adjustment of the parameters of each dipole in order to find the best solution minimizing the residual variance between the measured data and the model.


View this table:
[in this window]
[in a new window]
 
Table 2. Peak amplitudes of first and second N100 in a train and their ratio
 
Grand averaged waveforms were created from individual subjects' data and multiple dipole models were independently developed for each of the grand averaged ERPs. Average reference was used. The data for violent alcoholics and normal subjects were drawn from 10 people each, but one recording of the non-violent alcoholics was technically flawed and not included. Based on previous non-invasive modelling of auditory N100, it was known that at least two dipoles, one in each hemisphere, were necessary for a stable model (Hari et al., 1980Go, 1982Go). The modelled window was 300 ms from the stimulus onward for each ERP. Modelling was started with bilateral dipoles and these were fitted for locations and orientations within each hemisphere. Residual variance was still high and thus an additional dipole was included, starting from the centre of the head model. The role of a new dipole was to attract more distant activity in order to refine the temporal activation pattern of the existing ones, but not to overtake them. Because of a relatively small number of recording sites on the scalp it was concluded that three dipoles with their degrees of freedom (x, y, z, {theta} and {phi}) was the maximum number of dipoles in a model for the current data. Throughout the fitting procedure, the same energy, variation and separation constraints were used (20, 20 and 40%, respectively). The constraints add weighting in the algorithm. All comparisons were subjected to Student's t-test (2-tailed) and P < 0.05 was accepted as significant.

As a final step of source analysis, the location coordinates of each dipole were transformed into the coordinates of a stereotaxic brain atlas (Talairach and Tournoux, 1988Go). This transformation was based on the measurements done on the MRI of one healthy male both in BESA and Talairach coordinates. The centre of the BESA coordinate system was placed 16 mm posterior to the vertical line at anterior commissure, and 6 mm inferior to the anterior commissure–posterior commissure line. After transformation we were able to approximate cerebral structures implicated by the models as active.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Grand average wave forms of the first N100, ‘the arousal’ of the violent alcoholics, are shown in Fig. 1aGo as an example of the original data. The montage of the recording electrodes is also seen in Fig. 1aGo. Adaptation of N100 can be expressed as the amplitude ratio between N100 elicited by the first and second tones in a train. N100 responses decreased to almost half in size after the first repetition of auditory input in controls, but not in violent offenders or in alcoholics (see also Table 2Go). Topographic maps were plotted and spatio-temporal multiple dipole models were developed for the grand average wave forms of the four conditions of each group of subjects. Voltage and current source density maps of the first N100 (in the train of four tones) containing ‘the arousal’ are presented in Fig. 1bGo. The three-dipole models of ‘the arousal’ condition of the subject groups are shown in Fig. 2Go.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 2. Spatio-temporal three-dipole models of the first N100, ‘the arousal’, in normal subjects (Nor), non-violent alcoholics (Non-viol), and violent alcoholics (Viol). Each model is shown in posterior and top views. The location of the dipole is at its filled circle and its number is indicated at the end of the orientation arrow. The location differences were not significant.

 
The second N100, the standard N100 and the MMN each were also modelled with three-dipole models. The models were developed independently from each other. The obtained location and orientation coordinates of the dipoles of each of the subject groups are presented in Table 3Go. These dipole locations obtained from models did not significantly differ between the conditions (paired t-test, P < 0.05). The dipole moments were calculated for the window from 81 to 165 ms (residual variances ranged from 9.5 to 15.2%) in the three corresponding sets of N100 data in order to compare activation levels. For the MMN condition, the calculations were made for a longer window (81–242 ms). The total moment, i.e. the sum of the dipole moments in one model, differed between the conditions in the indicated windows, but not within any condition between the subject groups. The total moment in the first N100 was significantly higher than in the second N100. It was also higher than that in the standard tone or in the MMN condition in each subject group. The distribution of the total moment among the three dipoles in each model indicated clear differences in the spatio-temporal processing of auditory information between the subject groups. These moment distributions are presented in Fig. 3Go.


View this table:
[in this window]
[in a new window]
 
Table 3. Location (x, y, z) and orientation ({vartheta}, {phi}) coordinates of the three dipoles (1, 2, 3) in each of the models created using Brain Electromagnetic Source Analysis
 


View larger version (27K):
[in this window]
[in a new window]
 
Fig. 3. Graphs presenting distributions of the total moment between the three dipoles in each model. Moment is represented in µVeff. On the horizontal axis, 1 = first N100 in a train (arousal), 2 = second N100 in a train (adaptation), 3 = standard N100 (control), and 4 = N100 elicited by mismatch negativity (MMN).

 
As explained in the Subjects and methods section, the approximate brain regions implicated by the spatio-temporal models were approximately localized using a stereotaxic atlas. Dipole 1 was located in the left superior temporal gyrus, dipole 2 in the right superior temporal gyrus and dipole 3 in the anterior cingulate gyrus. The temporal lobe dipoles did not locate symmetrically, neither were their moments evenly dis-tributed (see Fig. 3Go). There was a small, non-significant, variation in dipole locations from condition to condition (see Table 3Go). The frontal dipole (accounting for anterior cingulate activity) showed prominent differences between groups. It had a major role in the models of the first and second N100 in violent alcoholics, compared to non-violent alcoholics and normal subjects. This can be observed in Fig. 3Go.

A temporal activation pattern (i.e. a source wave form) of each dipole illustrated the timing of the peak activity of each dipole. In MMN condition, temporal activation patterns in the superior temporal lobe and anterior cingulate area clearly suggested sequential processing of the deviant tones in normal subjects and in non-violent alcoholics. In contrast, violent alcoholics showed simultaneous processing of the deviant tone in these same brain regions, suggesting a functional difference between these groups (for temporal activation patterns see Fig. 4Go).



View larger version (9K):
[in this window]
[in a new window]
 
Fig. 4. Temporal patterns of source activities obtained in the models for mismatch negativity (MMN) for normal subjects (Nor), non-violent alcoholics (Non-viol), and violent alcoholics (Viol). On the left, numbers 1, 2 and 3 correspond to dipole numbers. The vertical bars indicate peak amplitudes. Note that dipoles 2 and 3 are simultaneously active in violent alcoholics, whereas the dipoles peak sequentially (separated by 24 ms) in the other two groups.

 

    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Violent behaviour in alcoholics has been associated with altered levels of arousal (see, e.g. Volavka, 1995) and the present analysis of N100 potentials shows evidence that alcoholics with impulsive violent behaviour have difficulty in automatic optimization of arousal and sensory response to external cues. Moreover, our evidence supports the view that automatic sensory cerebral processes differ between the subtypes of alcoholism and that these functional differences may reside in the frontal lobe.

All auditory ERP data were successfully modelled with three-dipole models. Dipole locations differed neither between the ERP types nor between the subject groups. This was expected, because of the similarity of the given tones and the simplicity of the modelling. Dipole activities (their moments), however, were sensitive measures indicating differences in levels of activity in different brain areas and in the temporal characteristics of the activation patterns. The total moment, i.e. sum of three dipole moments in one model, was higher in the ‘arousal condition’ in each subject group. The distribution of the moment between the temporal and frontal regions indicated more frontal lobe contribution both in the ‘arousal’ and in the second N100 (adaptation condition) in violent alcoholics, than in the other groups. Furthermore, the temporal activation patterns (Fig. 4Go) of the sources of MMN showed simultaneous processes in the frontal and temporal regions only in violent alcoholics. The lack of temporal dissociation between brain regions may have a role in the assumed increased distractibility in these violent subjects.

Adaptive pre-attentive brain mechanisms probably govern the attentive detection of novelty in the surroundings. In violent alcoholics, the altered amount of adaptation in the frontal and temporal brain regions was accompanied by temporal alteration in the frontal lobe activation during deviance detection. This may indicate a continuous inability to inhibit involuntary attention shift to the deviance in the group of violent alcoholics.

Previous studies have shown, for example, that persons exhibiting antisocial behaviour have increased ERP amplitudes (Raine, 1988Go). It has been speculated that the higher amplitudes in these subjects are associated with an increased attention to stimulating events as a consequence of stimulus-seeking or sensation-seeking due to a non-optimal level of arousal. This may also be the case in the current violent subject group. In line with these speculations and the present findings, recent radioligand studies of the current population of violent alcoholics have shown that a strong neurochemical modulator of arousal, dopamine, has lower transporter densities in non-violent alcoholics in the basal ganglia, whereas the densities are slightly higher in violent alcoholics as compared with normal subjects (Tiihonen et al., 1995Go).

Little evidence was found in this study for left hemisphere dysfunction or lateralization abnormalities known to have a role in predisposing to violence (Raine and Scerbo, 1991Go); however, alteration in frontal lobe function was convincing in violent alcoholics. Dysfunction in prefrontal/frontal regions may result in loss of inhibition on phylogenetically older subcortical structures which are thought to facilitate aggression (Weiger and Bear, 1988Go). The current paper presents differences also in automatic sensory processes between the subtypes of alcoholism. This implies that altered frontal lobe function is associated with impulsive violent behaviour and alcoholism. Although alteration in prefrontal/frontal function may be related to violence, it is probably only a biological predisposition factor, requiring environmental, psychological and social factors for the emergence of severe violent behaviour in an individual.


    ACKNOWLEDGEMENTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
This study was supported by the National Alcohol Research Foundation (Finland) and by the Kuopio University Hospital, Kuopio, Finland.


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
* Author to whom correspondence should be addressed at: Brain Research and Rehabilitation Center Neuron, Kortejoki, FIN-71130 Kuopio, Finland. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Alho, K. (1995) Cerebral generators of mismatch negativity (MMN) and its magnetic counterpart (MMNm) elicited by sound changes. Ear and Hearing 16, 38–51.[ISI][Medline]

American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, revised. American Psychiatric Association, Washington, DC.

Bromm, B., Laudahn, R. and Tarkka, I. M. (1992) Late magnetic field components evoked by auditory and pain-inducing stimuli. In Biomagnetism: Clinical Aspects, pp. 265–268. Excerpta Medica, Amsterdam.

Fruhstorfer, H., Soveri, P. and Järvilehto, T. (1970) Short-term habituation of the auditory evoked response in man. Electroencephalography and Clinical Neurophysiology 28, 153–161.[ISI][Medline]

Hari, R., Aittoniemi, K., Järvinen, M. L., Katila, T. and Varpula, T. (1980) Auditory evoked transient and sustained magnetic fields of the human brain. Experimental Brain Research 40, 237–240.[ISI][Medline]

Hari, R., Kaila, K., Katila, T., Tuomisto, T. and Varpula, T. (1982) Interstimulus-interval dependence of the auditory vertex response and its magnetic counterpart: implications for their neural generation. Electroencephalography and Clinical Neurophysiology 54, 561–569.[ISI][Medline]

Näätänen, R. and Gaillard, T. (1983) The mismatch negativity to intensity changes in an auditory stimulus sequence. In EEG Correlates of Information Processing: Theoretical Issues, Gaillard, A. W. K. and Ritter, W. eds, pp. 119–141. Elsevier, Amsterdam.

Näätänen, R. and Picton, T. (1987) The N1 wave of the human electric and magnetic response to sound: a review and analysis of the component structure. Psychophysiology 24, 375–425.[ISI][Medline]

Näätänen, R., Gaillard, A. W. K. and Mäntysalo, S. (1978) Early selective attention effect reinterpreted. Acta Psychologica 42, 313–329.[ISI][Medline]

Raine, A. (1988) Evoked potentials and antisocial behavior. In Biological Contributions to Crime Causation, Moffitt, T, and Mednick, S. eds, pp. 14–39. Kluwer, Dordrecht.

Raine, A. and Scerbo, A. (1991) Biological theories of violence. In Neuropsychology of Aggression, Milner, J. S. ed., pp. 110–138. Kluwer, Boston.

Ritter, W., Vaughan, H. G. and Costa, L. D. (1968) Orienting and habituation to auditory stimuli: a study of short term changes in averaged evoked responses. Electroencephalography and Clinical Neurophysiology 25, 550–556.[ISI][Medline]

Sams, M., Hari, R., Rif, J. et al. (1993) The human auditory sensory memory trace persists about 10 s: neuromagnetic evidence. Journal of Cognitive Neuroscience 5, 363–370.[ISI]

Scherg, M. (1984) Spatio-temporal modeling of early auditory potentials. Revues Laryngologie 105, 163–170.

Scherg, M. and Berg, P. (1996) Brain Electromagnetic Source Analysis, User Manual, V. 2.2. Megis, Munich.

Scherg, M. and Von Cramon, D. (1986) Evoked dipole source potentials of the human auditory cortex. Electroencephalography and Clinical Neurophysiology 65, 44–360.

Scherg, M. and Picton, T. W. (1992) Separation and identification of event-related potential components by brain electric source analysis. Electroencephalography and Clinical Neurophysiology 42 (Suppl.), 24–27.

Talairach, J. A. and Tournoux, P. (1988) Co-planar Stereotaxic Atlas of the Human Brain. Thieme Verlag, New York.

Tiihonen, J., Kuikka, J., Bergström, K. et al. (1995) Altered striatal dopamine re-uptake site densities in habitually violent and non-violent alcoholics. Nature Medicine 1, 654–657.[ISI][Medline]

Volavka, J. (1995) Neurobiology of Violence. American Psychiatric Press, Washington, DC.

Weiger, W. A. and Bear, D. M. (1988) An approach to the neurology of aggression. Journal of Psychiatric Research 22, 85–98.[ISI][Medline]

Yoshimoto, K., McBride, W. J., Lumeng, L. and Li, T.-K. (1992) Alcohol stimulates the release of dopamine and serotonin in the nucleus accumbens. Alcohol 9, 17–22.[ISI][Medline]





This Article
Abstract
FREE Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Request Permissions
Google Scholar
Articles by Tarkka, I. M.
Articles by Tiihonen, J.
PubMed
PubMed Citation
Articles by Tarkka, I. M.
Articles by Tiihonen, J.