International Journal of Neuroscience

ISSN: 0020-7454 (Print) 1543-5245 (Online) Journal homepage: http://www.tandfonline.com/loi/ines20

Visual Evoked Potentials, Attention and Mnemonic Abilities in Children Jorge Bernal, Jacqueline Becker, Thalia Harmony, Mario Rodriguez, Alfonso Reyes, Erzsebet Marosi, Thalia Fernandez & Vicente Guerrero To cite this article: Jorge Bernal, Jacqueline Becker, Thalia Harmony, Mario Rodriguez, Alfonso Reyes, Erzsebet Marosi, Thalia Fernandez & Vicente Guerrero (1992) Visual Evoked Potentials, Attention and Mnemonic Abilities in Children, International Journal of Neuroscience, 66:1-2, 45-51, DOI: 10.3109/00207459208999788 To link to this article: http://dx.doi.org/10.3109/00207459208999788

Published online: 07 Jul 2009.

Submit your article to this journal

View related articles

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ines20 Download by: [Universität Osnabrueck]

Date: 14 March 2016, At: 22:03

0 1992 Gordon and Breach Science Publishers S . A . Printed in the United States of America

Intern. J . Neuroscience, 1992, Vol. 66. pp. 45-51 Reprints available directly from the publisher

Photocopying permitted by license only

VISUAL EVOKED POTENTIALS, ATTENTION AND MNEMONIC ABILITIES IN CHILDREN JORGE BERNAL, JACQUELINE BECKER, THALIA HARMONY, MARIO RODRIGUEZ, ALFONSO REYES, ERZSEBET MAROSI, THALIA FERNANDEZ and VICENTE GUERRERO National University of Mexico

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

(Received January 29, 1992)

We analyzed the correlation between attention and mnemonic processes and different visual evoked potential (VEP) parameters. A group of 34 children between 9 and 13 years old was studied. VEPs were recorded in C3, C4, P3, P4, 0 1 , 0 2 , T 5 and T 6 with linked ear lobes as reference. Two different types of stimuli were used: flash and checkerboard pattern. The power of VEPs was calculated as the sum of the square amplitude values for different time epochs. Correlation coefficients between left and right homologous VEPs were also computed. A visual selective attention task divided into 5 items of increasing difficulty and the Stemberg paradigm were applied. The performance was automatically evaluated by the computer, giving the number of correct responses (NCR) and other measures of performance. Correlation coefficients between VEP parameters and the scores obtained in the performance of tasks were calculated. It was observed that power in P3, P4, T5, and T6 and the correlation coefficients between central, parietal and temporal VEPs were positively correlated with NCR of both tasks. However, power in 01 and 0 2 was negatively correlated with NCR. Keywords: Visual evoked potentials, attention, memory, cognition.

The possibility that the characteristics of evoked potentials (EPs) may reflect neural substrates of intellectual capacity has been suggested in many reports. Chalke and Ertl (1965) were the first to show that subjects with higher intelligence quotients (IQs) had shorter latencies of later EP components (Ertl & Schafer, 1969). Visual Evoked Potentials (VEPs) have shown differences between dull and bright children (Rhodes et al., 1969). Gucker (1973) also reported a high correlation between intelligence test percentile scores and the latency to the third zero crossing of the VEP recorded in C4 in a group of youngsters. Shagass et al. (1981) demonstrated differences in amplitudes in VEPs between high and low IQ nonpatients using t tests. In the same paper the authors described significant correlations between Raven Scores and VEP amplitudes in different time epochs. More recently, Robinson (1989), using a more sophisticated technique to analyze VEP frequency-response, demonstrated a complex nonlinear relationship between full-scale WAIS IQ and frequency. The above-mentioned results were obtained in subjects with gross differences in IQ scales. We were interested in analyzing the relationships between VEPs and intellectual abilities in children with normal IQs. For this reason we decided to develop some computerized psychophysiological tests which measure different type of abilities. In this paper we analyzed the correlation between performance on tests demanding attention and mnemonic processes and different VEP parameters. We chose This project was supported in part by grant number IN205689 from DGAPA, UNAM. The authors wish to acknowledge Sandor John for the correction of the English version of the manuscript. Reprint requests to: Jorge Bemal, ENEP-UNAM-IZTACALA, Apartado 3 14, 54090 Tlalnepantla, Estado de Mexico, Mexico. 45

46

J. BERNAL et al.

not to select the measurement of peak latencies and amplitudes of VEPs, for two basic reasons: 1 ) Some of the peaks were not observable in many children and 2) It was very difficult to decide whether, in comparing a number of different waveforms, we were seeing the same component at different latencies, or different components. Therefore, we computed the power in different time epochs and correlation coefficients between left and right homologous VEPs, as these parameters have been shown to be sensitive to the presence of brain damage (Harmony et al. 1973; Harmony, 1984) and were useful in classifying children according to their pedagogical evaluation (Bernal et al., 1991).

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

MATERlAL AND METHODS A group of 34 children between 9 and 13 years old was studied: 16 male and 18 female. All children had a normal pediatric examination, no behavioral disturbances, no neurological symptoms, and normal IQ (Wechsler Intelligence Scale for Children). None of the children was medicated at the time of recording. Academic performance was evaluated according to a special test assessing reading and writing which we have previously described (Harmony et al., 1990). Examinees were classified according to their performance relative to the expected level for their age and school grade. The classification was intended to reflect the amount of special help or training the children needed in order to reach the expected performance. Four degrees were assigned, with the following meaning: 1) Good: Performance adequate for age and degree, no extra help needed (n = 12, age 10.5-13.2, X = 11.8); 2) Regular: Performance below level with minor difficulties which might be overcome with little or no extra help ( n = 13 age 9.1-12.4 X = 11.2); 3) Poor: Performance below level: the child was unable to overcome the difficulties without special help ( n = 7); 4) Very poor: The child had severe difficulties and special education was recommended; examinee should not continue in regular school (n = 2). Children from groups 1 and 2 had no history of academic failure in school. As only two children were in the fourth group, they were included in the third group for statistical analysis (age 11.2-13.2 X = 12.1).

Recording Subjects were awake and sitting in a comfortable chair in a dimly lit room without acoustic isolation. VEPs were recorded in C3, C4, P3, P4, 0 1 , 0 2 , T5 and T6, with linked ear lobes as reference (10-20 International System). The amplifier bandwidth was set between .5 and 30 Hz. EOG was recorded in the supraorbital ridge versus the external canthus of the right eye. Averaged VEPs to 100 stimuli were computed on-line utilizing a program for rejecting epochs that contained eye movement artifacts. The averaging sweep lasted 512 ms, with 128 points sampled at each 4 ms. Visual stimuli were presented for 40 ms on a video terminal with a random interstimulus interval of between 600 and 1000 ms. Two different types of stimuli were used: diffuse screen illumination (flash) and checkerboard pattern. Each square on the checkerboard subtended an angle of 10’. Recording and stimulation were camed out using a MEDICID 03. VEP analysis All averaged evoked responses were recomputed using as isopotential line the mean amplitude value. Power was then calculated as the sum of the square amplitude values for the following four intervals: 0-52 ms, 52-100 ms, 100-252 ms and 252-512 ms.

VEPs AND VISUAL ATTENTION

47

Correlation coefficients between left and right homologous VEPs were calculated for the same four intervals.

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

Cognitive Tasks Visual selective attention This test was divided into 5 items, of increasing difficulty. The subject had to select one digit (target) and press the space bar of the computer each time the target appeared in the display. Thirty digits in random order were presented each two seconds. Ten of the thirty digits were targets. If the subject responded correctly 3 or more times the test continued with the subsequent item. If not, the test was automatically terminated. In the first item, the digits were presented in the center of the video display. In the second item, two digits were presented simultaneously: one in the center and another in the left upper corner. In the third item, 3 digits were displayed simultaneously: one in the center and 2 in the left upper and right lower comer. In the fourth, 4 digits were presented simultaneously in the center and 3 comers. In the fifth item in the center and the 4 comer of the display. The performance was automatically evaluated by the computer, giving the number of correct responses (NCR) for each item, the number of false alarms and the number of targets to which the subject did not respond.

Memory tusk The Sternberg paradigm (1966) was selected. A five digit number (memory set) was presented at the center of the video display. The subject had to memorize this number and press the space bar when ready. Then a random series of 10 elements was displayed digit by digit and the subject had to make a rapid decision whether the last digit was a member of the memory set, pressing the letter “S” if the digit belonged to the memory set and the letter “N” if the digit did not belong. The following parameters were evaluated: the time elapsed between the presentation of the number and the subject pressing the space bar (retention time RT), the number of correct (NCR) and incorrect responses and the average reaction time (ART) to the presentation of each digit. This task was repeated 5 times. Statistical analysis Our purpose was to analyze whether some relationship existed between VEP parameters and the scores obtained in the performance of the tasks. For this reason we computed the correlation coefficients between both types of variables. However, in order to know whether the presence of such relationships might be due to other factors, such as an age effect both in the performance and in VEPs, we also computed the correlation coefficient between age and the remaining variables. Another aspect that may have affected the relationship between VEP parameters and performance scores was the fact that the children studied belonged to three different categories of academic performance. In a previous paper on a larger group of children, Bernal et a]. (1991) demonstrated that VEP power was associated with the subjects belonging to these groups. Therefore, what was of interest was determining whether there existed some significant differences in the performance of attention and memory tasks between these groups. For this reason, ANOVAs between academic categories were calculated for each score in each item.

RESULTS Correlation with Age VEPs to flash Significant correlations were observed only in T5 and T6. Power values: power increased with age in T5 (interval 100-252 p < .04) and T6 (intervals

48

J. BERNAL et al.

52-loop < .03; 100-252p < .003; 252-512p < .01). The correlation coefficients between homologous pairs of VEPs were not related to age. VEPs to pattern stimuli A significant correlation with age was found in 01 (interval 100-252 p < .02), 0 2 (interval 252-512 p < .04) and T6 (intervals 52-100 p < .008; 100-252 p < .007; 252-512 p < .05). NO age effect was found in the correlation coefficients between homologous left and right VEPs. Visual attention task A very significant correlation was found between the NCR for items 1 ( p < .02), 2 (p < .004),3 ( p < .0004)and 5 (p < ,007) and age.

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

Memory task

No significant correlations with age were found.

ANOVAs between Academic Categories In order to determine whether there were differences in the performance of cognitive tasks between groups, ANOVAs were computed for each score of the performance in each item. Tukey’s multiple comparison test was used in order to evaluate the significance between groups. The single statistical difference ( p < .05) observed was: Children classified as “good” in the educational evaluation had higher NCR than the group with “poor” evaluation in the fourth item of the attention test and in the first item of the memory task. Correlation CoefJicients between Performance and VEP Variables Visual attention task The parameter of the attention task that had higher correlations with VEP variables was the NCR, and for this reason we shall describe only the results obtained with it. For the first item no significant correlation was found between NCR and VEP variables. In the second item higher NCRs were correlated with higher values of power in T5 for flash (interval 52-100 p < .05). In the third item a negative correlation between power and 0 1 (interval 0-50 p < .05) and NCR was observed. Many significant correlations were found for item 4: Higher NCRs were correlated with higher values of the correlation coefficient between VEPs in central areas to flash (interval 0-512 p < .001; interval 0-25Op < .01 and interval 52-100 p < .05) and between patterned VEPs in central areas (interval 0-52 p < .05), parietal regions (interval 100-252 p < .05) and temporal leads (intervals 0250 and 0-512 p < .Ol). Also higher NCRs were correlated with more power in T5 of flash VEPs (intervals 52-100 and 100-252 p < .05). In the fifth item the NCR were correlated with higher values of the correlation coefficient of patterned VEPs in central regions (interval 0-52 p < .05) and with higher values of power of patterned VEPs in the right parietal lead (interval 100-252 p < .05). In Fig 1 we have summarized the significant correlations observed between NCR in the attention test and VEP variables. Memory Task. In this task the NCR also had more significant correlations with VEPs than any other score. In the first item higher values of NCR were associated with higher values of the correlation coefficient between flash VEPs in parietal areas (intervals 0-512 p < .05) and more power of patterned VEPs in P3 (interval 252512 p < .05). In the second item the NCR were correlated with higher values of power in flash VEPs in P3 (interval 0-5- p < . O l ) and T5 (interval 0-5- p < .05) and lower values of power in 01 (interval 52-100 p < .01) in patterned VEPs. In

49

VEPs AND VISUAL ATTENTION

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

AlTENTION TASK CORRELATION WITH NCR

Flash VEP

Pattern

VEP

FIGURE I Places where VEPs were recorded are shown. An arrow pointing upwards means a positive correlation with NCR, negative when the arrow points downwards. The arrows connecting symmetrical leads illustrate the leads where the correlation coefficient showed a significant correlation with the NCR. Flush: power in T5 and Correlation coefficient between central leads were positively correlated with NCR, while power in 01 was negatively correlated. Pattern: power in P4 and T6 and the correlation coefficients between central, parietal and temporal VEPs were positively correlated with NCR.

the flash item the NCR was correlated with lower values of power in 0 2 (intervals 52-100 and 100-252 p < .05) in VEPs to flash and lower values of power in 0 1 (interval 0-52 p < .003) and 0 2 (interval 0-52 p < .05) in pattern VEPs. Higher correlation coefficients between T5 and T6 patterned VEPs were correlated with higher NCR (interval 0-50 p < .05). In the fourth item the NCR was related to higher values of the correlation coefficient between C3 and C4 flash VEPs (interval 0-52 p < .05). In the last item the NCR was associated with higher values of power in P3 (intervals 100-252 p < .05), T5 (interval 252-512p < .05) and T6 (intervals 100-252 p < .05), and lower values of power in 0 2 (interval 52-100 p < .05) in VEPs to flash. Power of patterned VEPs in T5 (interval 252-512 p < .Ol) and T6 (interval 100-252 p < .05) was related to the NCR in the fifth item. In Figure 2 we have summarized the significant correlations found between NCR in the memory task and power and correlation coefficients of VEPs.

DISCUSSION Age effects on VEPs in the sample of children studied were very peculiar, since we found only positive correlations between age and VEP power. This stands in contradiction with previous results (Dustman et al. 1977). These anomalous results might be due to the distribution of age according to academic categories. Although no significant differences between age categories were found, there was a tendency to

50

J . BERNAL et al.

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

MEMORY TASK CORRELATION WITH NCR

Flash VEP

Pattern

VEP

FIGURE 2 Flash: power in T5, P3 and T6 and correlation coefficients between central and parietal VEPs were positively correlated to NCR in the memory task, while power in 0 2 was negatively correlated Pattern: power in T5, P3 and T6 and the correlation coefficient between temporal VEPs were positively correlated with NCR while power in 01 and 0 2 were negatively correlated.

find older ages in group 3. Children in this group could be considered as clearly learning-disabled, and in these children immature responses might be observed (Harmony, 1989). However, for our purpose, that is, to determine whether there is a relationship between VEPs and cognitive tasks, the result obtained could not be explained by an age effect, since occipital VEPs power in particular was negatively correlated with performance. Correlation coefficients between left and right homologous VEPs were not related to age, but there was a clear relationship between a higher NCR- both in the attention and in the mnemonic task- and a higher correlation coefficient between C3 and C4, P3 and P4 and T5 and T6. We consider this a very interesting finding since more symmetric VEPs on waveform would suggest more powerful synchronizing mechanisms, as well as interhemispheric connections. Such characteristics seem to be related to better performance of both tasks. Bernal et al. (1991) have described a relationship between academic performance and VEP power. Children with a good evaluation were characterized by more power in T5 and P4, both to flash and pattern stimuli, while children with a poor evaluation were characterized by more power in T6 and less power in P4 of VEPs to flash. The results obtained in this paper contrasted sharply with such previous results. The fact that comparisons in the performance of attention and memory tasks between educational categories showed only one significant value for each task suggests that there is no relation between the educational categories and the performance of such tasks. Therefore, it would not be surprising to find a different pattern of relationships between these tasks and VEPs. Children with a higher NCR in the attention task were those with more power in T5 and less power in 0 1 of flash VEPs, and more power in P4 and T6 of patterned

Downloaded by [Universität Osnabrueck] at 22:03 14 March 2016

VEPs AND VISUAL ATTENTION

51

VEPs. A better performance in the memory task was also related to more power in T5, P3 and T6 both to flash and patterned stimuli, and less power in 0 1 and 0 2 . Age effect in this sample could explain the relationship between a better performance and higher values of power in T5 and T6. In relation to parietal VEPs a depression of these responses has been related to reading disability (Conners, 1971 ; Preston et al, 1974). Functional abnormality of the parietal areas could be related to a deficit in visual selective attention (Mesulam, 1981, Posner, 1989). It is more difficult to explain the inverse correlation between power of occipital VEPs and performance. While we cannot understand this result, we must accept it, since it was consistently observed in repeated trials and different tasks. Another interesting finding was the observation that in the attention task there were more abundant relationships between performance and VEPs for a particular level of difficulty. In item 1 , when the task was very easy no correlations were found. The same result was reported between performance and EEG spectral parameters (Harmony et al., in press), and we consider it important for future designs of psychophysiological tasks exploring electrophysiological parameters.

REFERENCES Bernal, J . , Harmony, T . , Marosi E., Becker, J., Reyes A . , Rodriguez, M ., Hinojosa, G. & Rocha, C . (199 1). Correlation between visual evoked responses and an educational evaluation. Neurosciences, I , 25-39. Chalko, F. C. R. & Ertl, J. P. (1965). Evoked potentials and intelligence. Life Science, 4. 1319-1322. Conners, C . K. (1971). Cortical visual evoked response in children with learning disorders. Psychophysiology, 7, 418-428. Dustman, R. E., Schenkenberg, T . , Lewis, E. G . & Beck, E. C . (1977). The cerebral evoked potential: Life span changes and twin studies. In J. E. Desrnedt (Ed.) Visual eiloked potentials in man. Oxford: Clarendon Press, 363-377. Ertl, J. P. & Schafer, E. W. P. (1969). Brain response correlates of psychometric intelligence. Nature, 223, 42 1-422. Gucker, D. K. (1973). Correlating visual evoked potentials with psychometric intelligence, variation in technique. Perceptual and Motor Skills, 37, 189-190. Harmony, T . (1984). Neurometric assessment of brain dysfunction in neurological patients. Hillsdale, New Jersey: Erlbaum. Harmony, T . ( 1989). Psychophysiological evaluation of children’s neuropsychological disorders. In C. R. Reynolds & E . Flertcher-Janzen (Eds.) Handbook of clinical child neuropsychology, New York: Plenum, 265-290. Otero, G., Fernindez, G . & Valdks, P. (1973). Symmetry of the visual Harmony, T., Ricardo, .I., potentials in normal subjects. E1ei~troencephalography& clinical Neurophysiology 35. 237-240. Harmony, T., Hinojosa, G., Marosi, E . , Becker, J . , Fernandez, T., Rodriguez, M . , Reyes, A . & Rocha C . (1990). Correlation between EEG spectral parameters and an educational evaluation. International Journal of Nerrrosrienre. 54, 147-1 55. Mesulam, M. M . (1981). A cortical network for directed attentional unilateral neglect. Annals of Neurology. 10, 309-325. Posner, M. I. (1984). Selective attention and the storage of information In G. Lynch, J. L. McGaugh and N. M. Weinberger (Eds.) Neurobiology of learning and memory, New York: Guilford Press, 89-101. Preston, M . S . , Guthrie, J . T . & Childs, B. (1974). Visual evoked responses in normal and disabled readers. Psychophysiology, I I , 4.52-4.57. Rhodes, L. E., Dustman, R . E. & Beck, E. C . (1969). The visual evoked response: a comparison of bright and dull children. Electroencephalogruph~& clinical Neurophysiology, 27, 364-372. Robinson, D. L. (1989). The neurophysiological bases of high IQ. International & Journal of Neuroscience, 46, 209-234. Shagass, C. H . , Roemer, R . A., Straumanis, J . J. & Josiassen, R. C . (1981). Intelligence as a factor in evoked potential studies of psychopathology: I. Comparison of low and high IQ subjects. Biological Psychiatry, 16, 1007-1030.

Visual evoked potentials, attention and mnemonic abilities in children.

We analyzed the correlation between attention and mnemonic processes and different visual evoked potential (VEP) parameters. A group of 34 children be...
621KB Sizes 0 Downloads 0 Views