Development of the EEG of school-age children and adolescents. I. Analysis of band power

Development in quantitative EEG parameters is studied for a sample of 158 normal children and adolescents aged 6-17 years. This is of interest both for increasing basic knowledge of human neurophysiology and for obtaining age standardized norms, useful in clinical research and applications. After selecting an appropriate epoch and correcting for EOG artifacts, the EEG at 8 derivations was submitted to spectral analysis in order to extract broad-band parameters in absolute and relative power. Change in EEG band power across age was quantified by polynomial regression analysis. This opened automatically the possibility to obtain age-standardized EEG norms. Development was for most EEG parameters non-linear, with more pronounced changes for absolute than for relative power. No sex differences and no pubertal spurt could be identified in contrast to most somatic quantities. A detailed statistical analysis revealed, however, that this might be due to using cross-sectional data. All bands except for alpha 2 decreased in absolute power, whereas the fast bands increased and the slow bands decreased in relative power. Strong evidence was found for a substituting process between theta activity and fast alpha activity.     

Suspense and surprise: on the relationship between expectancies and P3

Few theories of the P3 component have emphasized the distinction between its parietal and frontocentral parts. This study used a new paradigm for testing the predictions that the parietal P3 is evoked by awaited stimuli (suspense) and the frontal P3 by unexpected stimuli (surprise). Subjects had to make simple responses whenever a yellow ring appeared. This signal appeared on the screen within a clock, most frequently when the pointer was at 12 o'clock (every 6 s) but sometimes also at other times. The suspense process should therefore have its minimum shortly after 12 o'clock and then steadily increase until 12 o'clock, and the parietal P3 should accordingly be smallest with stimuli shortly after 12 o'clock, then gradually increase and be largest with 12 o'clock stimuli. Further, the stimuli presented at times other than 12 o'clock should evoke large frontal P3s because they were unexpected. The results confirmed parts of these predictions. A frontocentral and a parietocentral component could indeed be discerned. The frontal P3 was largest with non-12 o'clock stimuli, whereas the parietal P3 was large with all stimuli. The parietal result was not predicted, but these results taken together pose more problems for the usual view, which assumes that the parietal P3 is evoked by unexpected stimuli, than for our assumption that the parietal P3 reflects suspense, and the frontocentral P3 reflects surprise. Generalizing to other paradigms, we assume that different topographies of P3 in different paradigms or in different groups of subjects might be due to different mixtures of these two components.     

Late cognitive event-related potentials in adult Down's syndrome.

Event-related electroencephalogram (EEG) potentials (ERPs) using two different tasks were measured in 14 adults with Down's syndrome (DS; mean age 32 years) without clinically detectable cognitive decline. Two groups, young healthy (YH) and old healthy (OH) adults, served as controls. In the oddball task, DS had prolonged N1 and earlier P2 latencies than the control groups. P3 latency was delayed in comparison to YH. In the PushWait task, P3 latency was later in DS than in YH and OH. In both tasks, DS showed a marked amplitude shift towards positivity overlapping the N1-P2 complex and seemingly also P3: The P3 amplitude evoked by target tones and by "Push" was shifted towards anterior sites resulting in a Cz maximum. Changes of the N1 latency and amplitude in DS may be related to enhanced arousal during stimulus processing, indicating a possible defect of central inhibitory mechanisms. The study suggests that differentiated ERP procedures provide information on adult DS cognition exceeding those given by mere P3 latency measurements. Such procedures may be useful in the evaluation of the cognitive decline due to precocious aging or Alzheimer-type dementia in DS.     

How the self controls its "automatic pilot" when processing subliminal information

Human performance may be primed by information not consciously available. Can such priming become so overwhelming that observers cannot help but act accordingly? In the present study, well-visible stimuli were preceded by whole series of unidentifiable stimuli. These series had strong, additive priming effects on behavior. However, their effect depended on the frequency with which they provided information conflicting to the visible main stimuli. Thus, effects of subliminal priming are under observers' strategic control, with the criterion presumably set as a function of the openly observable error frequency. Electrical brain potentials show that this criterion acts simultaneously at the level of visual discrimination of the primes and at motor activation evoked by the primes, thereby shielding observers from unwanted information.     

Effects of same- and different-modality cues in a Posner task: extinction-type,spatial, and non-spatial deficits after right-hemispheric stroke

The response delay to left target stimuli preceded by right-side cues, first described by Posner et al. [J. Neurosci. 4 (1984) 1863-1874] appears to be a stable marker of right-parietal injury. However, only few studies compared patients' performance to age-matched controls. Furthermore, only few studies compared visual and auditory stimuli in this task. Therefore, two groups of right-hemisphere stroke patients, with and without left visual hemineglect, and a healthy control group were studied in three versions of Posner's paradigm. Visual or auditory target stimuli were presented to the subject's left or right, following a visual or auditory cue by 150 ms. The classical 'extinction-type' effect, an increase in missing responses for right visual cue/left visual target, was specifically observed in neglect patients. In the same condition, an 'extinction-type' response delay was present in patients with neglect and in those without neglect. No such delay occurred in any group when cues were auditory. Specifically in neglect patients, response times were generally longer for left than for right visual targets, regardless of cue side and of cue modality. Response times were generally prolonged in neglect patients regardless of target modality. This suggests that three components impair neglect patients' performance in this paradigm: a non-spatial, supramodal deficit, a global, neglect-type deficit of the contralesional hemi-field, and the extinction-type impairment. The latter two deficits appear to be most marked within the visual domain.     

Comments on the analysis of the P300 components using brain mapping

Recently, Maurer and colleagues reported that the peak amplitude of the P300 was distributed in specific topographical patterns for demented, hebephrenic, and paranoid patients. This evidence is critically discussed. For the hebephrenic and paranoid patients, it is doubtful whether their topographic patterns differed from normal subjects, while for the demented patients the impact of EOG artifacts is not clear. Some recommendations are given how to deal with artifacts before mapping the P300.     

P3-evoking wrong notes: unexpected, awaited, or arousing?

The effect of melodic deviance on P3 was investigated. In one condition, the opening phrases of well-known melodies were presented. Melodies ended in the midst of the phrases, either properly or with a deviant tone, as in the study of Besson and Macar (1987). In the other condition, ending phrases of melodies were presented, with the final tone either having its proper pitch or a deviant one. It was predicted that not only deviant tones of openings would evoke P3s, as in Besson and Macar's study, but also both deviant and normal tones of endings because all these tones unambiguously marked the end of the present epoch. Contrary to prediction, late positive components were evoked by deviant tones only. However, it was in the endings condition only that these components displayed a parietal maximum, as P3s do. It is concluded that two factors were necessary for evoking P3s in this paradigm, one factor that was associated with deviance, the other associated with endings. It is argued that the first factor is arousal, the second factor is subjects' expectation of the closing stimulus.     

Correlating EEG and IQ: a new look at an old problem using computerized EEG parameters

Explorations of the relationship between EEG activity and scores of tests of intelligence have a long history. Results have been rather discrepant and several explanations can be offered for this. Almost all previous studies relied on parameters derived from a visual rather than a computerized analysis of the EEG. Here broad band spectral parameters were mainly used, and in addition the frequency of the dominant peak (at Pz, O1 and O2) between 6.5 and 14.0 Hz. All EEG parameters were standardized for age, as were tests of intelligence. Correlations of EEG with IQ were computed separately for a group of normal and a group of mildly retarded children. As postulated in advance, correlations were substantially higher for the retarded groups. Even more remarkable than the size of the correlations found was their pattern, associated with an interesting interpretation: the sign of the larger correlations was such that children developmentally more advanced in their EEG parameters had on the average higher IQ scores. Moreover, correlations were large for those frequency bands, and to a lesser degree for those derivations where we found a high developmental relevance in a previous investigation for the age range considered. Some disappointing or even discrepant results in earlier studies can be explained in the light of the results given here.     

Dimensional overlap between arrows as cueing stimuli and responses?. Evidence from contra-ipsilateral differences in EEG potentials

In the S1-S2 interval, 400 ms after an arrow as S1, an EEG-potential difference occurs between scalp sites contralateral and ipsilateral to arrow direction. Eimer [J. Exp. Psychol. Hum. Percept. Perform. 21 (1995) 837-854] interpreted this difference as a sign of automatic activation of the manual response, due to dimensional overlap of arrows and responses. However, according to Kornblum et al.'s [Psychol. Rev. 97 (1990) 253-270] notion of dimensional overlap, responses can only be automatically primed if they are included in the response set. Therefore, participants of the present study had to respond to S2 in separate blocks either by key-press, as in Eimer [J. Exp. Psychol. Hum. Percept. Perform. 21 (1995) 837-854], or by making saccades. In addition, contra-ipsilateral differences were recorded not only from central positions, overlying the hand-motor area, but across the whole scalp. Contralateral negativity at 400 ms after S1 was indeed found over the hand-motor area in the 'hand blocks'. However, this 'L-400' (=lateralization at 400 ms) was generally as large in the 'eye' blocks as in the 'hand' blocks. Therefore, L-400 does not reflect automatic activation of manual responses in the sense of Kornblum et al. [Psychol. Rev. 97 (1990) 253-270]. Further, its topographical maximum was more anterior than the hand-motor-related negativity that preceded the manual response ('LRP') with its maximum at central sites. Therefore, L-400 probably does not originate in the hand-motor cortex. Rather, it may be related to activity of the lateral premotor cortex found in fMRI studies of spatial orienting. The present EEG study extends these studies by delimiting the time period of this activity, suggesting that it reflects encoding of the spatial properties of the arrow for action.