Is cortical distribution of spectral power a stable individual characteristic?

Share Embed


Descripción

ARTICLE IN PRESS

INTPSY-09950; No of Pages 11

International Journal of Psychophysiology xxx (2008) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Psychophysiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j p s yc h o

1

Is cortical distribution of spectral power a stable individual characteristic?

2

Gennady G. Knyazev ⁎

3

State Research Institute of Physiology, Siberian Branch of the Russian Academy of Medical Sciences, Timakova str., 4, Novosibirsk, 630117, Russia

i n f o

a b s t r a c t

Article history: Received 5 August 2008 Received in revised form 11 November 2008 Accepted 11 November 2008 Available online xxxx

PR

General understanding in EEG research is that cortical distribution of spectral power varies as a function of time, frequency, state, and experimental condition. There are findings, however, which show that individualspecific patterns of cortical spectral power distribution could be amazingly stable, at least in some ^ experimental conditions. In this study two different experimental datasets were used to analyze stability and variability of individual pattern of cortical spectral power distribution across time, experimental conditions, and frequency bands. First experiment consisted of presentation of pictures of emotional facial expressions. Second experiment was an auditory stop-signal task. In both experiments a number of psychometric ^ measures were obtained from each participant. It has been shown that in spite of high short-term variability, ^ individual-specific patterns of cortical spectral power distribution are remarkably stable across frequency ^ bands, long periods of time, and experimental conditions. These patterns are related to state and trait participant's characteristics. The antero-posterior spectral power gradient emerged as the most prominent ^ ^ feature associated with important personality dimensions. Relatively higher oscillatory activity in the frontal cortical region relates to female gender and Behavioral Inhibition tendencies. Relatively higher activity at posterior sites is associated with Extraversion. Significant differences in event-related spectral perturbations ^ upon presentation of emotionally loaded stimuli were found between high and low antero-posterior gradient ^ participants. These data show that cortical distribution of oscillatory activity may be seen as a relatively stable individual characteristic. Enhanced or diminished oscillatory activity of some cortical regions, such as the prefrontal cortex, may play an important role in organization of human behavior. © 2008 Published by Elsevier B.V.

CT

ED

Keywords: Functional EEG topography Anxiety Emotional facial expressions Stop-signal paradigm ^ ^ Event-related spectral perturbations ^

27

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

Recent studies emphasize the importance of taking into account individual differences in human electroencephalogram (EEG) patterns (Basile et al., 2006). Twin studies indicate that a considerable part of inter-individual variation in EEG variables could be attributed to genes ^ (van Beijsterveldt and van Baal, 2002). Thus, the heritability estimates are about 79% for EEG alpha power (somewhat lower for other frequency bands), 81% for alpha peak frequency, 60% for P300 amplitude and EEG coherence measures (Anokhin et al., 2001; van Beijsterveldt et al., 1998; van Beijsterveldt and van Baal, 2002). It has been also shown that about 60% of the variance in the temporal structure of amplitude fluctuations in spontaneous EEG oscillations could be attributed to genetic factors (Linkenkaer-Hansen et al., 2007). ^ Another important parameter, which has been extensively studied since multi-channel EEG registration has become a usual practice in ^ EEG research, is the cortical distribution of spectral power in different frequency bands. It is well established that different cortical areas participate in different functional processes. On the other hand, it is

UN CO R

30 31

1. Introduction

RE

26 28

Q1 29

OO

a r t i c l e 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

F

4

⁎ Tel.: +7 383 333 48 65; fax: +7 383 332 42 54. E-mail address: [email protected].

increasingly becoming clear that oscillations may play a crucial role in integration of brain activity (Buzsaki and Draguhn, 2004; Salinas and Sejnowski, 2001; Singer, 1999). Growing evidence suggests that different levels of cerebral integration mediated by spatial and temporal synchrony over multiple frequency bands could play a key role in the emergence of percepts, memories, emotions, thoughts, and actions (Cantero and Atienza, 2005; Nunez, 2000; Varela et al., 2001). Therefore, distribution across the cortex of spectral power in different frequency bands may relate to the distribution of functional activity. Numerous observations confirm that this distribution is very variable, and general understanding is that it depends on time, frequency, state, and experimental condition. One line of research concerned with temporal variation in cortical spectral power distribution is the study of EEG microstates, which may correspond to basic building blocks of human information processing. Each microstate class is described by topography, mean duration, frequency of occurrence and percentage analysis time occupied. The microstate variables show a lawful, complex evolution with age, suggesting the existence of biologically predetermined top-down processes that bias brain electric activity to ^ functional states appropriate for age-specific learning and behavior ^ (Koenig et al., 2002). Significant differences in microstate variables between neuroleptic-naive, first-episode, acute schizophrenics and ^ ^ matched controls have also been found (Koenig et al., 1999). It could be

0167-8760/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.ijpsycho.2008.11.004

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

ARTICLE IN PRESS

79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135

2. Methods

149

2.1. Subjects

150

OF

77 78

The first study sample included 40 healthy, right-handed partici^ pants (21 females) aged 17 to 32 years with normal or corrected to ^ normal vision. The second study sample included 51 healthy, righthanded participants (35 females) aged 18 to 30 years with normal ^ ^ hearing. In both studies all participants were university students. All participants received a sum equivalent to about 5% of the living wage for participation. All applicable participant protection guidelines and regulations were followed in the conduct of the research in accordance with the Declaration of Helsinki. All participants gave informed consent and the study was approved by the Institute of Physiology ethical committee.

RO

76

136 137 138 139 140 141 142 143 144 145 146 147 148

151 152 153 154 155 156 157 158 159 160 161

2.2. Instruments

162

In the first study, the following psychometric measures were obtained: Behavioral Inhibition (BIS) and Behavioral Activation (BAS) were measured by a short form of the Gray–Wilson Personality ^ Questionnaire (GWPQ, Slobodskaya et al., 2003). Eysenckian personality dimensions were measured by a short form of the Eysenck Personality Profiler (EPP, Eysenck et al., 2000; Knyazev et al., 2004a). This inventory contains 3 scales for each major personality dimension, plus a Lie scale. The 3 traits selected for Extraversion are Sociability, Activity and Assertiveness. The 3 traits selected for Neuroticism are Anxiety, Inferiority, and Unhappiness. The 3 traits selected for Psychoticism are Risk-taking, Impulsiveness, and Irresponsibility. The Big-Five ^ ^ Factors (Extraversion, Neuroticism, Conscientiousness, Agreeableness, and Intellect) were measured by the International Personality Item Pool 100 Big-Five Factor Markers (Goldberg, 2001; Knyazev et al., ^ 2008b). State and Trait Anxiety were measured by the Spielberger's ^ State Trait Anxiety Inventory (STAI, Spielberger et al., 1970; Hanin, 1989). Aggressiveness was measured by the Buss–Perry aggression ^ scales (BPAS, Buss, 1992; Knyazev et al., 2008b). It consisted of four subscales: Anger, Physical Aggression, Hostility, and Verbal Aggression. In the second study we used the GWPQ (Wilson et al., 1989; Slobodskaya et al., 2001), the BIS/BAS scales (Carver and White, 1994; Knyazev et al., 2004c), the EPP and the STAI.

163

2.3. Experimental procedures

185

As stimulation in the first study we used an ensemble of the photographs presented by Ekman and Friesen (1976). We selected 30 photographs, specifically, 5 different females and 5 different males with 3 different facial expressions (angry, happy, and neutral). The pictures were presented black and white (17 × 17 cm) and displayed on ^ a screen at a distance of 120 cm from the participants. The participants ^ sat in a soundproof and dimly illuminated room. After about 4 min of ^ spontaneous EEG registration with eyes open (first baseline) they completed the State Anxiety scale and a brief questionnaire describing

186

TE

74 75

tone, respectively, but they had to refrain from pressing any button if a click was presented after the tone (for a detailed description see Knyazev et al., 2008c). In both studies a number of psychometric measures of state and trait participants' characteristics were obtained. ^ These measures were chosen so that to obtain a broad description of personality traits from the perspective of most influential contemporary personality theories, such as Eysenckian three-factor model, ^ the Big Five model, and the Gray's Reinforcement Sensitivity Theory, ^ and characteristics of the participant's state during the experiment. ^ So, this study was largely exploratory. The main aim was to explore the stability and variability of individual-specific patterns of cortical ^ spectral power distribution and their relation to trait and state participant's characteristics. ^

RE C

72 73

expected that in main-stream population, these processes may show ^ substantial inter-individual variability. Actually, genetically predeter^ mined traits could be viewed as individual predisposition to experience particular states. This implies that there could be substantial inter-individual differences in the mean duration and ^ frequency of occurrence of specific microstates that are characterized by specific patterns of cortical spectral power distribution. When averaged over time data are analyzed, this may reveal itself in substantial and relatively stable individual-specific patterns of cortical ^ spectral power distribution. A general tendency in psychophysiological research to focus on group measures, an approach which emphasizes the similarities between participants, has largely precluded a systematic investigation of individual differences in cortical spectral power distribution. There are findings, however, which show that individual-specific patterns of ^ cortical spectral power distribution could be amazingly stable, at least in some experimental conditions (Basile et al., 2007; De Gennaro et al., 2005; Tinguely et al., 2006). It has been shown, for example, that a particular topographic distribution of the EEG spectral power along the antero-posterior cortical axis, like a “fingerprint,” distinguishes ^ ^ ^ each individual during non-REM sleep. This individual EEG-trait was ^ ^ substantially invariant across six consecutive nights characterized by large experimentally induced changes of sleep architecture. It has been hypothesized that this EEG invariance can be related to individual differences in genetically determined functional brain anatomy, rather than to sleep-dependent mechanisms (De Gennaro ^ et al., 2005). Similarly, Tinguely et al. (2006) in their study of functional EEG topography in sleep and waking have found that, although EEG spectral power is modulated by state and sleep pressure, basic topographic features appear to be state-independent. Moreover, a ^ high degree of intra-individual correspondence of the spectral power ^ maps was observed across the studied states. Recently two studies by Basile et al. (2006) have made an ^ important contribution to the study of individual differences in the functional EEG topography. In the first study, this group has shown that the inter-individual variability in reactions to even simple stimuli ^ is very large, making the use of averaged-across-participants^ ^ measures meaningless because these measures may reflect the most ^ strong reactions of some individuals and obscure reactions of other individuals, which could be quite different from these averaged representations. In their second study, Basile et al. (2007) have shown ^ that during a simple visual task the generators of task-induced ^ oscillations are largely the same individual-specific sets of cortical ^ areas active during the pre-stimulus baseline. It has been concluded ^ that attention-related electrical cortical activity is highly individual^ specific, and possibly, to a great extent, already established during ^ mere resting wakefulness. These studies raise a number of basic questions. 1. How stable is the pattern of cortical distribution of spectral power? 2. If there indeed were a stable pattern of cortical distribution, would it be similar for different frequency bands? 3. Is there indeed an “EEG-fingerprint,” ^ ^ ^ which distinguishes each individual? 4. Could this pattern of cortical distribution be decomposed into relatively stable (individual-specific) ^ and more variable (task-, or condition-related) parts? 5. Is the ^ ^ individual-specific pattern of cortical distribution related to state, ^ trait, and biological participant's characteristics? ^ These research questions are related to very basic brain processes and it is expected that these processes would be independent of the kind of experimental task. Therefore, we used data from two different studies with very different experimental tasks. In the first study, participants were presented with emotionally loaded visual stimuli (emotional facial expressions) and asked to evaluate the presented stimuli in terms of their emotional content (for a detailed description of this study see Knyazev et al., 2008a). In the second study, a stopsignal experimental paradigm was used. Participants were asked to ^ press a left or right button upon presentation of a high- or low-pitched ^ ^

UN CO R

70 71

G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

DP

2

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184

187 188 189 190 191 192 193 194

ARTICLE IN PRESS G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232

2.4. EEG recording

234

EEG was recorded using a 32-channel PC-based system via silver– ^ ^ silver chloride electrodes. A mid-forehead electrode was the ground. ^ ^ The signals were amplified with a multichannel biosignal amplifier EEG8 manufactured by Contact Precision Instruments Inc., Boston MA, with bandpass 0.05–70 Hz, − 6 dB/octave and continuously digitized at ^^ ^ ^^ 300 Hz. Notch filter for removing 50 Hz power supply artifacts was not ^ ^ used, because in this study only frequencies below 30 Hz were ^ analyzed. The electrodes were placed at 30 head sites according to the International 10–20 system and referred to linked-mastoids. The ^ ^ horizontal and vertical EOG was registered simultaneously. EEG data were artifact-corrected using Independent Components Analysis via ^ EEGLAB toolbox (http://www.sccn.ucsd.edu/eeglab/) with additional visual rejection of artifact-contaminated data off-line. ^ ^

239 240 241 242 243 244 245 246

UN CO R

237 238

RE

233

235 236

247

2.5. Psychophysiological data reduction

248

2.5.1. Baseline, ITI, and PSP spectral power For within-epoch spectral power and variability analysis, time– ^ frequency representations were calculated using Morlet wavelets with ^ the number of cycles linearly increasing from a minimum of 1 cycle for ^ 2.3 Hz to a maximum of 10 cycles for 45 Hz (EEGLAB toolbox). This ^ ^ ^ method allows obtaining better frequency resolution at higher frequencies than a conventional wavelet approach that uses constant cycle length. This method is also better matched to the linear scale which is used to visualize frequencies (Delorme and Makeig, 2004).

249 250 251 252 253 254 255 256

2.5.2. Event-related spectral perturbations (ERSP) ^ To assess event-related changes in spectral power, ERSP were ^ calculated using timef function of EEGLAB toolbox (http://www.sccn. ucsd.edu/eeglab/). The ERSP (Makeig, 1993) shows mean log eventlocked deviations from baseline-mean power at each frequency. The ^ ^ mean value of the spectral power in a particular frequency during the 1000 ms prior to stimulus presentation was considered as the baseline ^ level and was subtracted from the E(t, f) after stimulus onset, where E denotes spectral power in frequency f at time point t. Method of ERSP calculation realized in EEGLAB toolbox is described in Delorme and Makeig (2004). Time-frequency representations were calculated by ^ Morlet wavelets as described in the previous section.

271 272

F

202 203

OO

201

257 258

PR

199 200

The obtained frequency spacing was 0.29 Hz; time resolution was ^ 6.7 ms. For between-trial analysis, power spectral density was ^ ^ calculated for artifact-free baseline, ITI, and PSP epochs using Welch's ^ ^ averaged modified periodogram method of spectral estimation through a Hamming window. The power spectral density estimates were log-transformed (using the base 10 logarithm) in order to ^ normalize their distribution. Spectral estimates were averaged within delta (1–3.8 Hz), theta (4.1–6.7 Hz), alpha (7.0–11.7 Hz), and beta ^ ^ ^ ^ ^ ^ (12.0–30.2 Hz) bands. Since there is evidence that different alpha ^ ^ frequency sub-bands are differently sensitive to cognitive task ^ demands (see e.g. Klimesch, 1999), at the first stage of the analysis we used narrow frequency bands. Since all frequencies apart from beta behaved very similarly, we later opted for choosing fewer bands so as to limit the number of reported correlations.

2.6. Data analyses and statistics

2.6.1. Analysis of stability and variability of the pattern of cortical spectral power distribution Pearson correlation was used as a measure of similarity between two cortical spectral power distributions. For statistical comparison of correlation coefficients they were transformed into Fisher's z using ^ the formula z = 0.5[ln(1 + r) − ln(1 − r)], where ln is the natural loga^ ^ ^ rithm. For the sake of clarity, however, non-transformed coefficients ^ are reported throughout the Results section. Our first aim was to evaluate the degree of correspondence between cortical distributions of spectral power at different time intervals, different experimental conditions (baseline, ITI, and PSP), different frequency bands, and different participants. First, different time intervals within each experimental condition were analyzed. The shortest time interval (6.7 ms) was the interval between adjacent time points in the time^ frequency decomposition of each epoch data obtained by means of ^ wavelet transform. For each frequency band and each epoch of every participant, a matrix of correlations between cortical distributions of spectral power in each time point was calculated and averaged. Averaging these values across all epochs produced a measure of shortterm stability of the pattern of cortical distribution for each frequency ^ band of every participant. At the next temporal level, each trial's data ^ were converted into frequency domain using Welch's method, and ^ inter-trial correlations of cortical spectral power distribution were ^ calculated and averaged for each participant, each condition, and each frequency band separately. Finally, all belonging to each condition epochs were randomly divided into two halves and spectral data were averaged within these halves. Correlation between cortical distributions of spectral power in these halves was calculated. Similarly, correlations between patterns of cortical spectral power distribution in different conditions, different frequency bands and different participants were calculated. To analyze a link of stability of the pattern of cortical spectral power distribution across time, experimental conditions and frequency bands with psychometric variables, all above described correlation coefficients were transformed into Fisher's z, averaged across the conditions, and entered into Maximum ^ Likelihood factor analysis with Varimax rotation. The choice of

ED

197 198

their feelings during the baseline registration. Factor analysis of these questionnaire items revealed several dimensions including arousal level (aroused vs. relaxed), vigilance (paying attention to outside sounds, skin irritants, and so on vs. being absorbed in thoughts), mood and emotion (positive vs. negative), and verbalization (a tendency to think by words vs. non-verbal images). Before the experimental ^ procedure the participants were instructed to evaluate emotional expression of each presented face on an analog scale ranging from − 100 (very hostile) to 100 (very friendly). First, a fixation cross ^ appeared at the center of the screen for 1 s. Then a face picture was ^^ presented for 4 s, which was followed by presentation of the ^ ^ evaluative scale. During the 4 s of test interval the participants were ^^ watching the presented face trying to comprehend its emotional message; after presentation of the evaluative scale they had opportunity to express their understanding of this message. Angry, happy, and neutral faces were delivered randomly, and inter-stimulus^ interval randomly varied between 4 and 7 s to avoid temporally ^ ^ conditioned responses. The number of face stimulations was 120 for each participant, including 40 faces of each category. After face presentation spontaneous EEG was recorded for 2 min (second ^^ ^ baseline). In this study, besides the first and the second baseline, the inter-trial interval (ITI, 1000 ms prior to the fixation cross presentation) ^ ^ and the post-stimulus period (PSP, 1000 ms after the face picture ^ ^ presentation) were analyzed. In the second study, after the above described baseline recording, the participants were presented with 30 trials of a discrimination task in which they had to press left or right button after presentation of, respectively, a high- (2000 Hz) or low- (1000 Hz) pitched tone, 75 dB ^ ^ ^ ^ ^ in intensity, 100 ms in duration, delivered binaurally. In subsequent 60 ^ trials of a stop-signal task they had to refrain from pressing any button ^ if a click was presented after the tone. The high- and low-pitched ^ ^ tones alternated randomly and the inter-stimulus interval randomly ^ varied between 3 and 5 s. The click was presented in on average 25% of ^^ trials. The interval between the tone and the click randomly varied between 200 and 800 ms with a mean of 500 and SD 50 ms (for more ^ ^ details see Knyazev et al., 2008c). In this study only reactions to the tone were analyzed. 1500 ms prior to the tone presentation were used ^ as ITI and 1500 ms after the tone presentation – as PSP. ^ ^

CT

195 196

3

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

259 260 261 262 263 264 265 266 267 268 269 270

273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319

ARTICLE IN PRESS

329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383

2.6.2. Principal components analysis of the within- and between-subject ^ ^ ^ variance Principal component analysis (PCA) was used to reveal most prominent patterns of cortical spectral power distribution which account for most within- and between-subject variance. PCA was ^ ^ invented in 1901 by Karl Pearson (1901) and has been extensively used as a tool in exploratory analysis of both between-subject data, when ^ cases are assumed to be independent, and within-subject data, when ^ “cases” represent repeated measures obtained in the same subject in ^ ^ different time periods. In the latter case PCA was frequently used to deal with complex data obtained from a large number of electrodes or sensors and to reveal a natural clustering of cortical sites (see e.g., Pourtois et al., 2008 for a recent review). PCA applied in the temporal domain would specifically make each successive component account for as much as possible of the variance uncorrelated with previously determined components, in contrast to recently developed alternative linear decomposition method, Independent Component Analysis (ICA), which seeks maximally independent sources (Delorme and Makeig, 2004). Since our goal here was to evaluate the temporal stability of components, which account for most of the within-subject ^ variance, we chose PCA instead of ICA. For analysis of the within-subject variance PCA was applied to each ^ participant's concatenated epochs pertaining to one condition (i.e., ^ first baseline, ITI, PSP, and second baseline). To evaluate stability of a component across conditions, absolute correlation coefficients (without taking into account the sign of correlation) between loadings of this component obtained in different conditions were calculated and averaged across the conditions. For analysis of the between-subject ^ variance PCA was applied to the sample of subjects. Participants were treated as ‘cases’ and cortical sites as variables. The analysis was ^ ^ conducted for each frequency band separately. Each participant's ^ spectral estimates were converted into z-scores across the channels ^ prior to analysis. That was done in order to remove inter-individual ^ differences in total spectral power, because these differences mostly depend on such factors as skull thickness and skin conductance (e.g., Davidson et al., 2000; Leissner et al., 1970).

2.6.4. Analysis of associations between EEG and psychometric variables As a preliminary step, correlations between EEG and psychometric variables were calculated. Personality scales that showed significant correlations with an EEG variable were further entered into a stepwise multiple regression analysis as predictors of these variable scores.

390 391

2.6.5. Analysis of the effect of the antero-posterior spectral power ^ gradient on ERSP To study the effects of the antero-posterior spectral power gradient ^ on ERSP, a median split was applied which divided the sample into two equal groups (with low and high antero-posterior spectral power ^ gradient values). EEG responses to stimuli were statistically compared in these two groups using statcond function of EEGLAB toolbox (http:// www.sccn.ucsd.edu/eeglab/) with non-parametric statistical testing ^ based on data re-sampling. Correction for multiple comparisons was ^ done using the False Discovery Rate (FDR) method (Holm, 1979). For more details on statistical methods used see Knyazev et al. (2008a).

395 396

3. Results

406

OF

327 328

RO

326

384 385

2.6.3. Calculation of the antero-posterior gradient ^ To obtain a comparable measure of the antero-posterior spectral ^ power and variability gradient, the baseline spectral power or variability data were first averaged across all frequency bands and converted into z-scores across the channels. Next, the channel data ^ were averaged along “latitudinal” dimensions, thus obtaining an ^ ^ antero-posterior vector of seven variables: (1) Fp1, Fp2; (2) F7, F3, Fz, ^

386 387 388 389

392 393 394

397 398 399 400 401 402 403 404 405

3.1. Stability of the pattern of cortical spectral power distribution

407

Study 1. Table 1 shows mean (SD) correlations between cortical distributions of spectral power in adjacent time intervals within one epoch, in different epochs, and in two randomly selected halves. For all bands and conditions average correlations were 0.23 for intraepoch data, 0.53 for inter-epoch data, and 0.89 for between-halves ^ ^ ^ data. Next, correlations between patterns of cortical spectral power distribution in different conditions were calculated. Correlations between cortical spectral power distributions during ITI and PSP approached unity (average = 0.985). For other comparisons they were around 0.77 for delta and theta, around 0.82 for alpha, and around 0.63 for beta. Correlations between cortical distributions of spectral power in different frequency bands were about 0.80 for delta, theta, and alpha, and about 0.50 for correlations of these three bands with beta. Average between-subjects correlations were around 0.55 for ^ delta, theta, and alpha and around 0.17 for beta.

408 409

TE

324 325

F4, F8; (3) FT7, FC3, FCz, FC4, FT8; (4) T7, C3, Cz, C4, T8; (5) TP7, CP3, CPz, CP4, TP8; (6) P7, P3, Pz, P4, P8; (7) O1, Oz, O2. Then, differences between all but the last variable and the next one in the anteroposterior vector were calculated and accumulated. High values of the ^ resulting measure reflected relative predominance of anterior over posterior cortical sites in all frequency bands.

RE C

322 323

appropriate factor solution was guided by scree- plot analysis. ^ Associations of obtained factors' scores with psychometric variables ^ were analyzed using stepwise multiple regression method. In such a way associations between psychometric variables and general stability of the cortical spectral power distribution were evaluated. However, this method does not show how psychometric variables are associated with spectral power variability at particular cortical sites. To address this issue, each participant's first baseline data were time– ^ frequency decomposed using wavelet transform and averaged within ^ delta, theta, alpha, and beta frequency bands. For each single epoch, standard deviations along the time dimension were calculated on the spectral power estimates, which were preliminarily transformed into z-scores across the channels. For each participant these measures ^ were averaged across the trials and later used for analysis of their association with psychometric variables. To answer the question as to whether individual-specific pattern of ^ cortical spectral power distribution is related to the distribution of reactivity in the experimental task, for each participant correlations between cortical distribution of ITI spectral power and cortical distribution of absolute (log test – log ITI) and relative ([log test – log ^ ^ ITI]/log ITI) reactivity were calculated and averaged across the trials.

UN CO R

320 321

G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

DP

4

410 411 412 413 414 415 416 417 418 419 420 421 422 423

Table 1 t1:1 Intra-epoch, inter-epoch, and between-halves correlations (SD) averaged across all subjects t1:2 Delta Theta Alpha Beta t1:3 Intra-epoch First baseline Second baseline ITI PSP

0.16 (0.05) 0.15 (0.04) 0.33 (0.03) 0.33 (0.03)

0.15 (0.05) 0.14 (0.04) 0.25 (0.04) 0.24 (0.03)

0.16 (0.05) 0.17 (0.05) 0.27 (0.05) 0.22 (0.05)

0.22 (0.08) 0.22 (0.08) 0.33 (0.06) 0.32 (0.08)

t1:5 t1:6 t1:7 t1:8 t1:9

Inter-epoch First baseline Second baseline ITI PSP

0.46 (0.12) 0.60 (0.11) 0.27 (0.08) 0.33 (0.09)

0.52 (0.13) 0.69 (0.10) 0.31 (0.09) 0.38 (0.10)

0.65 (0.15) 0.68 (0.14) 0.45 (0.12) 0.45 (0.13)

0.62 (0.17) 0.79 (0.12) 0.60 (0.14) 0.61 (0.15)

t1:11 t1:12 t1:13 t1:14 t1:15

Between halves First baseline Second baseline ITI PSP

0.85 0.95 0.86 0.80

0.88 0.96 0.90 0.84

0.90 0.96 0.94 0.86

0.87 (0.20) 0.91 (0.10) 0.88 (0.11) 0.86 (0.12)

t1:17 t1:18 t1:19 t1:20

(0.25) (0.04) (0.12) (0.13)

(0.20) (0.04) (0.06) (0.11)

(0.19) (0.03) (0.07) (0.10)

ITI – Inter-trial interval; PSP – Post-stimulus period.

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

t1:21

ARTICLE IN PRESS

435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467

Fig. 2. Amount of the within-subject variance explained by the first 10 principal components along with the mean size of the between-condition correlations averaged across all participants.

OO

433 434

3.2. Principal components analysis of the within-subject variance ^ ^

PR

431 432

In order to reveal what part of the within-subject spectral power ^ variance is relatively stable across experimental conditions, PCA was applied to each participant's concatenated epochs pertaining to one ^ condition. Correlations between respective component loadings obtained in different conditions were calculated and averaged. Fig. 2 shows the averaged across all participants amount of the withinsubject variance explained by the first ten principal components, ^ along with the mean size of the between-condition correlations for ^ the Study 1 data. It is clear that the first two components, which together explain about 70% of the within-subject variance, are highly stable across the ^ conditions (mean between-condition correlation for these compo^ nents was 0.9). For the Study 2 data the first two components together explained about 71% of the within-subject variance and mean ^ between-condition correlation for these components was 0.93. ^

ED

430

CT

428 429

RE

426 427

Next, for each participant, correlations between cortical distribution of ITI spectral power and cortical distribution of absolute and relative reactivity were calculated and averaged across the trials. For absolute reactivity, mean (SD) correlations were: − 0.02 (0.28), − 0.17 ^ ^ (0.32), − 0.40 (0.26), and − 0.37 (0.36) for delta, theta, alpha, and beta, ^ ^ respectively. For relative reactivity they were: − 0.13 (0.36), − 0.04 ^ ^ (0.36), 0.01 (0.32), and − 0.01 (0.39) for delta, theta, alpha, and beta, ^ respectively. Study 2. The same analysis was performed for the second study ^ data. Fig. 1 shows averaged across all participants, bands, and conditions intra-epoch, inter-epoch, and between-halves correlations. ^ ^ ^ The latter were above 0.9 for all bands and conditions. Average correlations (SD) between ITI and PSP cortical spectral power distributions were 0.87 (0.11) for delta, 0.91 (0.05) for theta, 0.90 (0.08) for alpha, and 0.97 (0.04) for beta. Correlations of ITI and PSP with baseline cortical spectral power distribution were about 0.85 (0.15) for delta, 0.86 (0.15) for theta, 0.84 (0.15) for alpha, and 0.47 (0.31) for beta. Average across participants and conditions correlation among cortical spectral power distributions in delta, theta, and alpha bands was 0.81. Correlations of these three bands with beta, however, were higher during baseline (0.73), than during ITI (0.20) and PSP (0.13). Average between-subject correlations were around 0.65 for ^ delta, theta, and alpha, and around 0.27 for beta. For absolute reactivity, mean (SD) correlations were: − 0.55 ^ (0.05), − 0.52 (0.07), − 0.59 (0.05), and − 0.42 (0.10) for delta, theta, ^ ^ ^ alpha, and beta, respectively. For relative reactivity they were: − 0.05 ^ (0.13), 0.04 (0.06), 0.04 (0.05), and 0.10 (0.15) for delta, theta, alpha, and beta, respectively. Thus, main findings of the first study were reproduced in the ^ second study fairly well. These findings could be summarized as ^ follows. In each participant, cortical spectral power distribution varies strongly in short time intervals, but this variation is around some relatively stable pattern. This pattern is similar for all but beta frequency bands and little changes in different experimental conditions. These patterns are substantially different in different participants (average between- subject correlation is only 0.6). ^ Negative association between ITI spectral power and absolute reactivity means that cortical sites with lower ITI spectral power tend to show higher absolute spectral power increase (or lower absolute spectral power decrease) after the stimulus presentation. However, cortical distribution of the relative spectral power change (which is usually used for calculation of the classical event-related ^ desynchronization) does not depend on the cortical distribution of ITI spectral power.

UN CO R

424 425

5

F

G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

Fig. 1. Intra-epoch, inter-epoch, and between-halves correlations averaged across all ^ ^ ^ participants, bands, and conditions (Study 2).

468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483

3.3. Between-subject patterns of cortical spectral power distribution ^

484

To reveal natural clustering of cortical sites in a sample of participants, PCA was applied to the first and the second study baseline spectral power measures for each frequency band separately. Prior to that, each participant's spectral estimates were converted into z-scores ^ ^ across the channels. Fig. 3 shows cortical projection of the first two components' loadings. In each frequency band these two components ^ together accounted for about 40% of the between-subject variance. ^ Although the topography of components' loadings somewhat ^ differs in the two studies, different variants of the antero-posterior ^ gradient in cortical spectral power distribution constitute its most prominent feature.

485 486 487 488 489 490 491 492 493 494 495

3.4. Association of cortical spectral power and variability distribution 496 with psychometric variables 497 3.4.1. Personality and stability of cortical spectral power distribution Study 1. The above described intra-epoch, inter-epochs, between^ ^ conditions, and between-bands correlation coefficients were trans^ ^ formed into Fisher's z and entered into Maximum Likelihood factor ^ analysis with extraction of just one factor, which showed positive loadings from all correlation coefficients (all loadings N0.3) and explained 38% of variance. This factor scores correlated negatively with Activity, Risk Taking, Irresponsibility, and Physical Aggression.

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

498 499 500 501 502 503 504 505

ARTICLE IN PRESS G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

RE C

TE

DP

RO

OF

6

Fig. 3. Factor loadings of the first two components of the principal components analysis applied to the first and the second study spectral power measures for each frequency band separately. At each cortical site, positive loadings are plotted in red and negative loadings are plotted in blue. Green color marks sites where the loadings are near zero. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526

Study 2. The same factor analysis was conducted on the Study 2 data. The first factor showed positive loadings from all correlation coefficients (all loadings N0.28) and explained 36% of variance. This factor scores correlated negatively with Assertiveness, Irresponsibility, and Psychoticism, and positively with GWPQ and Carver and White's BIS. Thus, in ^ both studies facets of Extraversion and Psychoticism were associated with higher variability of cortical spectral power distribution. Additionally in the Study 2 Behavioral Inhibition was associated with a more stable pattern of cortical spectral power distribution.

UN CO R

506

3.4.2. Association of the antero-posterior gradient of spectral power ^ variability with psychometric variables Study 1. The antero-posterior gradient of spectral power variability ^ correlated positively with all facets of Behavioral Inhibition and Neuroticism and negatively with Extraversion. It also correlated positively with being vigilant and perception of facial expressions as more hostile. Introversion (β = 0.55, p = 0.001) and being vigilant (β = 0.33, p = 0.026) together explained 45% of its variance. Study 2. In the second study, the antero-posterior gradient ^ correlated positively with female gender, feeling tense and vigilant during baseline recording, GWPQ Active Avoidance, and Carver and White's BIS. In the regression analysis Gender (β = 0.39, p = 0.043) ^

emerged as the sole predictor explaining 15% of variance. Thus, in both studies higher spectral power variability in the frontal cortical region was positively associated with Behavioral Inhibition and similar behavioral tendencies.

527

3.4.3. Association of the antero-posterior spectral power gradient ^ (APSPG) with psychometric variables Fig. 4 shows cortical distribution of correlation coefficients between the antero-posterior gradient and spectral power estimates ^ in the four frequency bands for the Study 1 and Study 2 data. Fig. 5 shows cortical distributions of spectral power in the four frequency bands in typical low- and high-antero-posterior gradient ^ ^ ^ participants. It is clear that in all frequency bands high values on the anteroposterior gradient scale were associated with higher spectral power in ^ the frontal and lower spectral power in the posterior cortical area. Study 1. APSPG correlated positively with Trait Anxiety, Inferiority, Behavioral Inhibition, and being vigilant during the experiment. It correlated negatively with Sociability, Extraversion, and Intellect. Stepwise multiple regression analysis showed that Extraversion (β = − 0.59, p = 0.001) and being vigilant (β = 0.34, p = 0.039) ^ together explained 54% of its variance.

531 532

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

528 529 530

533 534 535 536 537 538 539 540 541 542 543 544 545 546 547

ARTICLE IN PRESS 7

OO

F

G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

555 556 557 558 559 560 561 562 563

ED

553 554

responses to emotional facial expressions were statistically compared in these two groups. Event-related spectral perturbations (ERSP) were ^ calculated and statistical comparison was performed for each time– frequency point using the False Discovery Rate (FDR) correction for ^ multiple comparisons. Fig. 6 shows ERSP in low and high anteroposterior gradient participants during presentation of angry and happy ^ faces. All areas with no significant between-group differences are ^ zeroed out and plotted in green. In areas where between-group ^ differences are significant, ERSP values averaged across respective group participants are plotted. Cortical maps at the top of the figure show cortical distribution of most pronounced effects. First of all it should be noted that between-group differences are most ^ pronounced during presentation of angry faces. They are less pronounced during presentation of happy faces and have not been revealed during presentation of neutral faces. Thus, it could be concluded that only emotionally loaded stimuli evoke different reactions in the two groups of participants. Both for angry and for happy faces high antero-posterior ^

3.4.4. APSPG and EEG responses to stimuli Study 1. To study effects of the APSPG on EEG responses to stimuli, a median split was applied, which divided the sample into two equal parts. Independent Samples T-test showed that the two groups ^ significantly differed on Sociability (T = 2.09, p = 0.045), Inferiority (T = − 2.41, p = 0.022), Neuroticism (T = − 2.18, p = 0.037), and Behavioral ^ ^ ^ ^ Inhibition (T = − 2.38, p = 0.023). Sociability was higher in low-, whereas ^ ^ ^ all other variables were higher in high APSPG participants. EEG

CT

552

RE

550 551

Study 2. For the second study data, APSPG correlated positively with female gender, verbalization, Active Avoidance, and Carver and White's BIS. Verbalization (β = 0.48, p = 0.006) and female gender ^ (β = 0.47, p = 0.039) together predicted 40% of its variance. Thus, higher frontal and lower posterior spectral power in all frequency bands was positively associated with female gender and facets of Behavioral Inhibition and Neuroticism, and was negatively associated with Extraversion.

UN CO R

548 549

PR

Fig. 4. Correlations (across subjects) of the antero-posterior gradient with spectral power in the four frequency bands. At each cortical site, positive correlation coefficients are plotted in red and negative correlation coefficients are plotted in blue. Green color marks sites where the correlation is near zero. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. Cortical distributions of spectral power z-scores (across the channels) in the four frequency bands in typical low- and high-antero-posterior gradient participants.

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580

ARTICLE IN PRESS G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

TE

DP

RO

OF

8

583 584 585 586 587 588 589 590

gradient participants show strong desynchronization, whereas low anteroposterior gradient participants show weak desynchronization or synchro^ nization in upper alpha and beta frequency bands. For angry faces, these differential reactions could be already seen between 100 and 200 ms ^ after stimulus presentation and extend till 400 ms. They are mostly ^ pronounced in the central and posterior cortical regions. Study 2. When Study 2 data were analyzed, similar differences ^ between high and low APSPG participants were noted, but they failed to be significant at the chosen level of FDR correction for multiple comparisons.

UN CO R

581 582

RE C

Fig. 6. Event-related spectral perturbations in low and high antero-posterior gradient participants during presentation of angry and happy faces. All areas with no significant betweengroup differences are zeroed out and plotted in green. In areas where between-group differences are significant, ERSP values averaged across respective group participants are plotted. Cortical maps at the top of the figure show cortical distribution of most pronounced effects. a-p grad = antero-posterior gradient. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

591

4. Discussion

592

Main findings of this study, which have been reproduced in two experimental samples, are listed in Table 2. They could be summarized as follows. Cortical distribution of spectral power is highly variable in short time intervals, but this variation is around some individuallyspecific pattern, which is very stable over long time periods. Basile ^ et al. (2007) used independent component analysis of averaged across the trials data and showed that the first component is very similar for ITI and PSP. In this study, it has been shown that even for nonaveraged data the first two PCA components, which account for ^ overwhelming part of the within-subject spectral power variance, are ^ very stable in different experimental conditions and time periods.

593 594 595 596 597 598 599 600 601 602

Most interestingly, these individual-specific patterns of cortical 603 ^ spectral power distribution appear to be very similar for delta, theta, 604 and alpha, but not for beta band. Delta, theta, and alpha oscillations 605

t2:1

Table 2 Summary of findings Finding High short-term CSPD variability High long-term CSPD stability High between-condition CSPD stability High between-band CSPD stability About 70% of the within-subject variance is stable across conditions Extraversion and Psychoticism are associated with higher CSPD variability Behavioral Inhibition is associated with higher spectral power variability in the frontal cortical region APSPG is the most prominent feature of the between-subject CSPD variability Behavioral Inhibition is associated with higher frontal and lower posterior spectral power in all frequency bands Emotionally loaded stimuli evoke stronger alpha and beta desynchronization in high- than in low-APSPG participants

t2:2 Study 1 Study 2 t2:3 + + + + + +

+ + + + + +

t2:4 t2:5 t2:6 t2:7 t2:8 t2:9

+

+

t2:10

+

+

t2:11

+

+

t2:12

+

t2:13

The plus sign (+) designates statistically significant findings. CSPD – Cortical Spectral Power Distribution; APSPG – Antero-Posterior Spectral Power t2:14 Gradient. t2:15

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

ARTICLE IN PRESS G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671

F

617 618

OO

615 616

PR

613 614

ED

612

experience particular states. Keeping in mind that all psychophysiological data are obtained in a very specific environment of psychophysiological laboratory, one may suggest that attitude to this environment may be a pivotal component of the participant's state ^ during the recording. No wonder then that anxiety most frequently emerges as the most consistent correlate of different EEG variables. In this study, the antero-posterior gradient showed most consis^ tent associations with psychometric variables. Higher spectral power and higher temporal variability in the frontal relative to the posterior cortical region was positively related to female gender, Anxiety, Neuroticism, and Behavioral Inhibition, and was negatively related to Extraversion. As for stability of the individual pattern of cortical spectral power distribution in general, it was positively associated with Behavioral Inhibition, and negatively with Extraversion and Psychoticism. What interpretation could be provided for these findings? Does higher spectral power and variability of oscillations across the whole spectrum in a particular cortical region imply this region is more, or less functionally active? For example, does higher spectral power and variability of oscillations in the frontal cortical region in anxious individuals imply anxiety is associated with higher, or lower frontal activity? Recent findings and theoretical advances pave the way to understanding higher oscillatory activity as a correlate of higher functional activity (Cantero and Atienza, 2005; Nunez, 2000; Varela et al., 2001). The present study data show that experimental stimuli evoke only minor changes in cortical spectral power distribution, implying that linked with oscillatory activity internal states and processes are much more important for establishing the current cortical oscillatory pattern. It is known that in the course of evolution, the prefrontal cortex (PFC) grows disproportionately more than other cortical regions (Fuster, 1997). Deficits in frontal lobe function have been proposed as a mechanism underlying deficits in regulatory control, such as, e.g., ADHD symptomatology (Arnsten and Li, 2005; Bush et al., 2005; Niedermeyer and Naidu, 1998). Moreover, patients with orbitomedial prefrontal lesions exhibit inordinate impulsiveness, irritability, hyperactivity, and poor control of instincts (Fuster, 1997). Existing theories of biological roots of personality emphasize the role of motivational and emotional brain circuits located in the brainstem and the limbic system, but somewhat underestimate the role of PFC. Recent accounts, however, acknowledge the role of PFC in behavioral inhibition (McNaughton and Corr, 2004). Thus, increased frontal activity may reflect increased activity of inhibitory mechanisms, whereas diminished frontal activity may evidence disinhibition of acting-out ^ behavioral tendencies. Moreover, it seems tempting to link the observed pattern of associations with the ‘posterior’ and ‘anterior’ attentional systems (Posner ^ ^ ^ ^ and Raichle,1994). The posterior attentional system is a reactive network involved in orienting or shifting attention from one location to another. It is referred to as an ‘involuntary’ attentional system. The act of orienting ^ ^ arises from three component operations: attention must disengage from the attended location, move to a new location, and then engage the new location (Matthews et al., 2000). Studies of performance in patients suffering from regional brain damage indicate that the disengage operations are accomplished through circuits within the parietal cortex. These neuropsychological findings are supported by converging studies using ERP and PET recordings (e.g. Corbetta et al., 1993). Studies on attention shift from a threatening cue to a target in the uncued location suggest that the anxiety-related attentional bias may ^ not involve moving toward and engaging sources of threat, but rather, difficulties in disengaging attention and thus shifting away from threat (Matthews et al., 2000). The association of anxiety with relatively lower oscillatory activity within the parietal cortex may be linked with difficulties in disengaging and shifting attention away from what is perceived as a potential threat. Noteworthy, along with personality variables, verbalization and being vigilant during an

CT

610 611

RE

608 609

have been associated with global processing modes which span relatively large cortical regions. Global modes have been hypothesized to serve the purpose of integration across diverse cortical sites by synchronizing coherent activity and phase coupling across widely spatially distributed neural assemblies (Nunez, 1995). Oscillations of beta and gamma ranges, or local EEG modes, are higher in frequency, lower in amplitude, and distributed over a more limited topographic area. They could be considered as elementary signals of the brain, functionally related to diverse brain processes (Schurmann et al., 1999). From this point of view, higher variability of cortical beta spectral power distribution does not seem surprising. But why the three global processing mode oscillatory systems, which in a number of studies have shown different and frequently opposite patterns of results, show such similarity of cortical spectral power distribution? In many respects alpha could be opposed to slow oscillations of delta and theta ranges. Their different behavioral and functional correlates have been repeatedly noted (e.g., Klimesch, 1999; Knyazev, 2007). It has been even suggested that there is a reciprocal relationship between alpha and delta oscillatory systems (Knyazev and Slobodskaya, 2003; Robinson, 2001). It appears that individualspecific pattern of cortical distribution of oscillatory activity and ^ reciprocal relationships between different frequency bands are two different dimensions. Whereas the former reflects general level of oscillatory activity in different cortical regions, the latter relates to specific between-band relationships which could be observed within ^ each region. The most important question is what these individual-specific ^ patterns of spectral power distribution mean in psychological or biological terms. For example, can we suggest that “attention-related ^ ^ electrical cortical activity is… to a great extent already established during mere resting wakefulness” (Basile et al., 2007)? Our data show ^ that cortical spectral power distribution during ITI correlates negatively with cortical distribution of absolute reactivity, which is in line with abundant evidence showing that pre-stimulus baseline ^ spectral power is inversely related to the response magnitude (Basar, 1998, 1999; Klimesch, 1999). These correlations are substantially stronger in the second than in the first study. This could be explained by the fact that in the second study ITI spectral power was calculated for the time period directly preceding the stimulus presentation, whereas in the first study it was calculated for the time period which was 1 s earlier. ^^ Cortical distribution of relative reactivity, which is most frequently used for measuring the response magnitude (Pfurtscheller and Aranibar, 1977), was not related to the ITI cortical spectral power distribution, implying that individual differences in cortical localization of the response to stimuli may not depend on individual differences in pre-stimulus spectral power distribution. However, ^ the intensity and even direction (synchronization vs. desynchronization) of the response to at least some kinds of stimuli do depend on individual differences in pre-stimulus spectral power distribution ^ (see Fig. 6). What sort of individual differences could be associated with individual-specific patterns of spectral power distribution? Could they ^ be related to stable genetically determined traits, or reflect more variable characteristics, such as moods and other states? Our data show a number of meaningful associations between these patterns and both state and trait variables, but only future longitudinal studies may show as to how stable these patterns are. Basile et al. (2007) in one subject replicated the experiment after one month and the same topography of baseline activity was observed. On the other hand, they note that in three participants, who participated in previous experiments four and six years before, very different topographies of the resting activity were observed. It has been suggested previously that observed associations between EEG spectral power variables and personality traits are mediated by states (Knyazev et al., 2004b), because personality traits could be seen as a predisposition to

UN CO R

606 625 607

9

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737

ARTICLE IN PRESS

747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803

OF

745 746

RO

744

804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826

Acknowledgements

827

This study was supported by grants of the Russian Foundation for Basic Research (RFBR) No. 08-06-00016-a and Russian Humanitarian ^ ^ ^ ^ Foundation (RFH) 07-06-00016-a. I am grateful to Luis Basile whose ^ ^ ^ ideas and findings have inspired me to perform this analysis. I am also grateful to my colleagues Andrey Bocharov, Evgenij Levin, Alexander Savostyanov, and Jaroslav Slobodskoj-Plusnin for data collection. ^

828

TE

742 743

differences in cortical power distribution and other EEG variables, should be exercised. Summing up, this study's findings emphasize the existence of ^ some fundamental regularity in cortical distribution of oscillatory activity. In spite of its high short-term variability, individual-specific ^ ^ patterns of this distribution are remarkably stable across frequency bands, long periods of time, and experimental conditions. These patterns of cortical activity probably reflect some properties that are associated with relatively stable individual differences. Future longitudinal studies may reveal whether these properties manifest traitlike features, or are related to more labile conditions, such as moods. ^ But even if these studies show that the individual-specific patterns ^ may change in the course of several days, months, or years, and thus reflect some less stable features than genetically predetermined traits, the present study findings show that these features are at least indirectly related to personality. The antero-posterior gradient ^ emerged as the most prominent feature of cortical spectral power distribution, which is associated with important personality dimensions. Relatively higher oscillatory activity of the frontal cortical region relates to Behavioral Inhibition tendencies and feelings of anxiety and vigilance during the experiment. Relatively higher activity of posterior regions is associated with extraverted tendencies and disinhibited traits.

References

829 830 831 832 833

834

Aftanas, L.I., Koshkarov, V.I., Pokrovskaja, V.L., Lotova, N.V., Mordvintsev, Y.N., 1996. Pre- 835 ^ and post-stimulus processes in affective task and event-related desynchronization: 836 ^ ^ do they discriminate anxiety coping styles? Int. J. Psychophysiol. 24, 197–212. 837 Anokhin, A.P., van Baal, G.C., van Beijsterveldt, C.E., de Geus, E.J., Grant, J., Boomsma, D.I., 838 2001. Genetic correlation between the P300 event-related brain potential and the 839 ^ EEG power spectrum. Behav. Genet. 31, 545–554. 840 Arnsten, A.F., Li, B.M., 2005. Neurobiology of executive functions: catecholamine 841 influences on prefrontal cortical functions. Biol. Psychiatry 57, 1377–1384. 842 Basar, E., 1998. Brain function and oscillations. I. Brain oscillations. Principles and 843 844 Approaches. Springer, Berlin, Heidelberg. ^ Basar, E., 1999. Brain function and oscillations. II. Integrative brain function. 845 Neurophysiology and cognitive processes. Springer, Berlin, Heidelberg. 846 Basile, L.F., Brunetti, E.P., Pereira Jr., J.F., Ballester, G., Amaro Jr., E., Anghinah, R., Ribeiro, 847 P., Piedade, R., Gattaz, W.F., 2006. Complex slow potential generators in a simplified 848 attention paradigm. Int. J. Psychophysiol. 61, 149–157. 849 Basile, L.F.H., Anghinah, R., Ribeiro, P., Ramos, R.T., Piedade, R., Ballester, G., Brunetti, E.P., 850 2007. Interindividual variability in EEG correlates of attention and limits of 851 functional mapping. Int. J. Psychophysiol. 65, 238–251. 852 Botwinick, M.M., Braver, T.S., Barch, D.M., Carter, C.S., Cohen, J.D., 2001. Psychol. Rev. 108, 853 624–652. 854 Bush, G., Luu, P., Posner, M.I., 2000. Cognitive and emotional influences in anterior 855 cingulate cortex. Trends Cogn. Sci. 4, 215–222. 856 Bush, G., Valera, E.M., Seidman, L.J., 2005. Functional neuroimaging of attention-deficit/ 857 ^ hyperactivity disorder: a review and suggested future directions. Biol. Psychiatry 57, 858 1273–1284. 859 Buss, A.H., 1992. The Aggression Questionnaire. J. Pers. Soc. Psychol. 63, 452–459. 860 Buzsaki, G., Draguhn, A., 2004. Neuronal oscillations in cortical networks. Science 304, 861 1926–1929. 862 Cantero, J.L., Atienza, M., 2005. The role of neural synchronization in the emergence of 863 cognition across the wake-sleep cycle. Rev. Neurosci. 16, 69–83. 864 ^ Carver, C.S., White, T.L., 1994. Behavioural inhibition, behavioural activation and 865 affective responses to impending reward and punishment: the BIS/BAS scales. 866 ^ J. Pers. Soc. Psychol. 67, 319–333.

RE C

740 741

experiment emerged as significant predictors of the APSPG. On the other hand, acting-out behavioral tendencies are bound to be ^ associated with ease of attention shifting and, as a consequence, a more labile pattern of cortical activation, which is reflected in relatively higher oscillatory activity within the parietal cortex and lower general stability of cortical spectral power distribution. The anterior attentional system is referred to as a ‘voluntary’ ^ ^ attentional system because it provides voluntary control of attention. This attentional network involves the anterior cingulate and lateral prefrontal areas and is activated strongly in situations that entail attentional control, such as when there is a conflict situation (Botwinick et al., 2001; Bush et al., 2000; Fan et al., 2003). Attentional control and conflict resolution constitute the core of the contemporary understanding of anxiety and behavioral inhibition (Gray and McNaughton, 2000). Hence, enhanced activity of the anterior attentional system could be expected in anxious individuals in the surroundings of a psychophysiological laboratory. This enhanced activity results in increased attention to impending events and higher event-related activation which is apparent in higher alpha and beta ^ desynchronization observed in anxious individuals in response to arousing emotionally loaded stimuli. Higher event-related alpha band desynchronization in anxious ^ individuals has been already reported previously (Aftanas et al., 1996; Knyazev et al., 2006, 2008a,c). The between-group differences in ^ reactivity of high and low antero-posterior gradient participants ^ essentially replicate findings previously reported for comparison of high and low anxiety subjects. The pattern of these differences imply that they are not merely a mechanical result of higher or lower prestimulus spectral power but reflect more specific differences in ^ processing of emotional information: (1) they were much more salient in response to emotional than to neutral stimuli (actually the latter failed to evoke significantly different reactions in the two groups of participants); (2) among emotional stimuli, negative emotional cues (angry faces) evoked more different reactions than positive ones (happy faces); (3) although the studied APSPG was uniformly distributed across all frequencies, the observed differences in reactivity were specifically related to alpha and beta bands; (4) higher event-related alpha and beta desynchronization in high antero^ posterior gradient participants was observed in the central and ^ posterior cortical regions (see Fig. 6) and not in the frontal region. It could be concluded therefore that observed between-group differ^ ences in reactions to emotional cues corroborate the observed associations of antero-posterior gradient with psychometric variables ^ confirming that high antero-posterior gradient relates to anxiety and ^ similar psychological constructs. Thus, data obtained in this study show that the antero-posterior ^ gradient of cortical spectral power distribution could be seen as an individual measure of the strength of inhibitory control and voluntary attention, which in the environment of a psychophysiological laboratory could be associated positively with personality traits of Anxiety and Behavioral Inhibition, and negatively with Extraversion. It should be noted that psychometric variables have explained about half of inter-individual antero-posterior gradient variance. Given ^ ^ inevitable estimation errors, this should be considered as a large proportion of variance. We deliberately restricted our analysis to the study of APSPG because it appears to be the most prominent source of inter-individual variability. However, some other patterns of cortical ^ spectral power distribution, such as inter-hemispheric asymmetry, or ^ central vs. peripheral cortical areas, also merit attention and should be investigated in future studies. It should be noted also that due to the exploratory nature of the present investigation, its' findings should be ^ treated with caution until they are reproduced in some other study. If confirmed, the findings from this study should impact the choice of design in electroencephalographic studies. Data obtained by means of across-subject averaging should be treated with caution and more ^ person-oriented approach, which takes into account individual ^

UN CO R

738 739

G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

DP

10

Corbetta, M., Miezin, F.M., Shulman, G.L., Petersen, S.E., 1993. A PET study of visuospatial attention. J. Neurosci. 13, 1202–1226. Davidson, R.J., Jackson, D.C., Larson, C.L., 2000. Human electroencephalography, In: Cacioppo, J.T., Tassinary, L.G., Berntson, G.G. (Eds.), Handbook of Psychophysiology, 2nd ed. Cambridge University Press, USA, pp. 27–52. De Gennaro, L., Ferrara, M., Vecchio, F., Curcio, G., Bertini, M., 2005. An electroencephalographic fingerprint of human sleep. NeuroImage 26, 114–122.

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

867 868 869 870 871 872 873 874

ARTICLE IN PRESS G.G. Knyazev / International Journal of Psychophysiology xxx (2008) xxx–xxx

ED

PR

OO

F

Makeig, S., 1993. Auditory event-related dynamics of the EEG spectrum and effects of ^ exposure to tones. Electroencephalogr. Clin. Neurophysiol. 86, 283–293. Matthews, G., Derryberry, D., Siegle, G.J., 2000. Personality and emotion: cognitive science perspectives. In: Hampson, S.E. (Ed.), Advances in Personality Psychology. ^ ^ Psychology Press Ltd, East Sussex, UK, pp. 199–237. McNaughton, N., Corr, P.J., 2004. A two-dimensional neuropsychology of defense: fear/ ^ anxiety and defensive distance. Neurosci. Biobehav. Rev. 28, 285–305. Niedermeyer, E., Naidu, S.B., 1998. Attention deficit disorder and frontal motor cortex disconnection. Clin. Electroencephalogr. 28, 130–136. Nunez, P.L., 1995. Neocortical Dynamics and Human EEG Rhythms. Oxford University ^ ^ ^ Press, New York. Nunez, P.L., 2000. Toward a quantitative description of large-scale neocortical dynamic ^ function and EEG. Behav. Brain Sci. 23, 371–398. Pearson, K., 1901. On lines and planes of closest fit to systems of points in space. Philos. Mag. 2, 559–572. Pfurtscheller, G., Aranibar, A., 1977. Event-related cortical desynchronization detected ^ by power measurement of scalp EEG. Electroencephalogr. Clin. Neurophysiol. 42, 817–826. Posner, M.I., Raichle, M.E., 1994. Images of Mind. Scientific American Library, New York. ^ ^ Pourtois, G., Delplanque, S., Michel, C., Vuilleumier, P., 2008. Beyond conventional event-related brain potential (ERP): exploring the time-course of visual emotion ^ ^ processing using topographic and principal component analyses. Brain Topogr. 4, 265–277. Robinson, D.L., 2001. How brain arousal systems determine different temperament types and the major dimensions of personality. Pers. Individ. Differ. 31, 1233–1259. Salinas, E., Sejnowski, T.J., 2001. Correlated neuronal activity and the flow of neural information. Nat. Rev., Neurosci. 2, 539–550. Schurmann, M., Demiralp, T., Basar-Eroglu, C., Basar, E., 1999. Selectively distributed ^ gamma-band responses studied in cortex, reticular formation, hippocampus, and ^ cerebellum. In: Basar, E. (Ed.), Brain Function and Oscillations. II. Integrative Brain ^ ^ ^ Function. Neurophysiology and Cognitive Processes. Springer, Berlin, Heidelberg, ^ ^ ^ pp. 61–67. Singer, W., 1999. Neuronal synchrony: a versatile code for the definition of relations? Neuron 24, 49–65. Slobodskaya, H.R., Knyazev, G.G., Safronova, M.V., Wilson, G.D., 2001. Reactions of Russian adolescents to reward and punishment: a cross-cultural study of the Gray– ^ Wilson Personality Questionnaire. Pers. Individ. Differ. 30, 1211–1224. ^ Slobodskaya, H.R., Knyazev, G.G., Safronova, M.V., Wilson, G.D., 2003. Development of a short form of the Gray–Wilson personality questionnaire: its use in measuring ^ personality and adjustment among Russian adolescents. Pers. Individ. Differ. 35, 1049–1059. Spielberger, C.D., Gorsuch, R.L., Lushene, R.E., 1970. Manual for the State-Trait Anxiety ^ Inventory. Consulting Psychologists Press, Palo Alto, CA. Tinguely, G., Finelli, L.A., Landolt, H.P., Borbély, A.A., Achermann, P., 2006. Functional EEG topography in sleep and waking: state-dependent and state-independent ^ ^ features. NeuroImage 32, 283–292. van Beijsterveldt, C.E., van Baal, G.C., 2002. Twin and family studies of the human electroencephalogram: a review and a meta-analysis. Biol. Psychol. 61, 111–138. ^ van Beijsterveldt, C.E., Molenaar, P.C., de Geus, E.J., Boomsma, D.I., 1998. Genetic and environmental influences on EEG coherence. Behav. Genet. 28, 443–453. Varela, F., Lachaux, J.P., Rodriguez, E., Martinerie, J., 2001. The brainweb: phase synchronization and large-scale integration. Nat. Rev., Neurosci. 2, 229–239. ^ Wilson, G.D., Barrett, P.T., Gray, J.A., 1989. Human reactions to reward and punishment: a ^ questionnaire examination of Gray's personality theory. Br. J. Psychol. 80, 509–515.

CT

Koenig, T., Prichep, L., Lehmann, D., Sosa, P.V., Braeker, E., Kleinlogel, H., Isenhart, R., John, E.R., 2002. Millisecond by millisecond, year by year: normative EEG microstates and developmental stages. NeuroImage 16, 41–48. Leissner, P., Lindholm, L.E., Petersen, I., 1970. Alpha amplitude dependence on skull thickness as measured by ultrasound technique. Electroencephalogr. Clin. Neurophysiol. 29, 392–399. Linkenkaer-Hansen, K., Smit, D.J.A., Barkil, A., van Beijsterveldt, T.E.M., Brussaard, A.B., ^ Boomsma, D.I., van Ooyen, A., de Geus, E.J.C., 2007. Genetic contributions to longrange temporal correlations in ongoing oscillations. J. Neurosci. 27, 13882–13889. ^

RE

987

Delorme, A., Makeig, S., 2004. EEGLAB: an open source toolbox for analysis of singletrial EEG dynamics including independent component analysis. J. Neurosci. ^ Methods 134, 9–21. Ekman, P., Friesen, W.V., 1976. Pictures of Facial Affect. Consulting Psychologist Press, Palo Alto. Eysenck, H.J., Wilson, G.D., Jackson, C.J., 2000. Eysenck Personality Profiler Short V6. PsiPress, Worthing, UK. ^ Fan, J., Flombaum, J.I., McCandliss, B.D., Thomas, K.M., Posner, M.I., 2003. Cognitive and brain consequences of conflict. NeuroImage 18, 42–45. Fuster, J.M., 1997. The Prefrontal Cortex: Anatomy, Physiology and Neuropsychology of ^ ^ the Frontal Lobe, 3rd ed. Raven, New York. Goldberg, L.R., 2001. International Personality Item Pool. http://ipip.ori.org/. Gray, J.A., McNaughton, N., 2000. The Neuropsychology of Anxiety: An Enquiry into the ^ ^ ^ Functions of the Septo-hippocampal System, 2nd ed. Oxford University Press, New ^ ^ ^ ^ York. ^ Hanin, Y.L., 1989. Cross-cultural perspectives of the individual differences diagnostic. ^ Vopr. Psikhologii 4, 118–125. Holm, S., 1979. A simple sequentially rejective multiple test procedure. Scand. J. Statist. 6, 65–70. Klimesch, W., 1999. EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brains Res. Rev. 29, 169–195. Knyazev, G.G., 2007. Motivation, emotion, and their inhibitory control mirrored in brain oscillations. Neurosci. Biobehav. Rev. 31, 377–395. Knyazev, G.G., Slobodskaya, H.R., 2003. Personality trait of behavioral inhibition is associated with oscillatory systems reciprocal relationships. Int. J. Psychophysiol. 48, 247–261. Knyazev, G.G., Belopolsky, V.I., Bodunov, M.V., Wilson, G.D., 2004a. The factor structure of the Eysenck personality profiler in Russia. Pers. Individ. Differ. 37, 1681–1692. Knyazev, G.G., Savostyanov, A.N., Levin, E.A., 2004b. Alpha oscillations as a correlate of trait anxiety. Int. J. Psychophysiol. 53, 147–160. Knyazev, G.G., Slobodskaya, H.R., Wilson, G.D., 2004c. Comparison of construct validity of the Gray–Wilson personality questionnaire and the BIS/BAS scales. Pers. Individ. ^ Differ. 37, 1565–1582. Knyazev, G.G., Savostyanov, A.N., Levin, E.A., 2006. Alpha synchronization and anxiety: implications for inhibition vs. alertness hypotheses. Int. J. Psychophysiol. 59, 151–158. Knyazev, G.G., Bocharov, A.V., Levin, E.A., Savostyanov, A.N., Slobodskoj-Plusnin, J.Y., ^ 2008a. Anxiety and oscillatory responses to emotional facial expressions. Brain Res. 1227, 174–188. Knyazev, G.G., Bocharov, A.V., Slobodskaya, H.R., Ryabichenko, T.I., 2008b. Personalitylinked biases in perception of emotional facial expressions. Pers. Individ. Differ. 44 ^ (5), 1093–1104. Knyazev, G.G., Levin, E.A., Savostyanov, A.N., 2008c. Impulsivity, anxiety, and individual differences in evoked and induced brain oscillations. Int. J. Psychophysiol. 68, 242–254. Koenig, T., Lehmann, D., Merlo, M.C., Kochi, K., Hell, D., Koukkou, M., 1999. A deviant EEG brain microstate in acute, neuroleptic-naive schizophrenics at rest. Eur. Arch. ^ Psychiatry Clin. Neurosci. 249, 205–211.

UN CO R

875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931

11

Please cite this article as: Knyazev, G.G., Is cortical distribution of spectral power a stable individual characteristic? Int. J. Psychophysiol. (2008), doi:10.1016/j.ijpsycho.2008.11.004

932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986

Lihat lebih banyak...

Comentarios

Copyright © 2017 DATOSPDF Inc.