|
|
| |||||||||||||||||||||||||||||||||||||||
Background The possibility that reading disability in children is associated with visual problems is in dispute. We sought to test the existence of this association by using electrophysiologic techniques to measure the processing of visual information in the magnicellular and parvicellular visual pathways of the brain.
Methods Visual evoked potentials were measured with scalp electrodes in children 8 to 11 years old who were normal readers and in those with reading disability. The potentials were measured for targets with low (0.5 cycle per degree of visual angle) and high (4.5 cycles per degree) spatial frequency, surrounded by either a steady background or a uniform-field flickering 12 times per second. A flickering field normally reduces the amplitude and increases the latency of a transient potential evoked by a low-spatial-frequency target, which preferentially excites the magnicellular visual pathway, but has little effect on the response to a high-spatial-frequency target.
Results With a steady background, the latencies of the early components (N1 and P1) of the visual evoked potentials were longer in the reading-disabled children than in the normal readers when the low-spatial-frequency target was used, but not when the high-spatial-frequency target was used. In normal readers, the flickering background increased the latency and reduced the amplitude of the early components, whereas in the reading-disabled children only the amplitude was affected. No differences were observed in either group with the high-spatial-frequency target.
Conclusions The pattern of results suggests that the response of the magnicellular visual pathway is slowed in reading-disabled children, who do not, however, have a general slowing of the visual response. The possibility that there is a cause-and-effect relation between these findings and reading disability will require further study.
For years it has been argued that visual deficits are not responsible for reading problems in children5. For example, the corrected visual acuity in children with reading disability is often in the normal range. However, reading is a dynamic process that requires proper timing of eye movements for information to be acquired distinctly and sequentially from successive fixations. It has been demonstrated that children with reading disability have various abnormalities in the temporal processing of visual information. For example, such children cannot follow rapidly flickering lights,6 and a perceived image persists longer after the disappearance of the visual stimulus7,8,9,10,11. They also have difficulty with temporal-order judgments12 and other deficits of temporal processing13,14,15. That reading disability is an expression of the disruptive effects of a temporal processing deficit has been proposed by Breitmeyer16.
A temporal deficit in reading ability that was detected by psychophysical testing was recently confirmed in an electrophysiologic study17 in which disabled readers had visual evoked potentials with abnormally long latencies, but only in conditions of low contrast or those in which there were rapid changes of the stimulus. It was concluded that the temporal deficits were due to a defective magnicellular (large-celled) visual pathway, since this pathway preferentially responds at higher temporal rates and lower contrasts18,19. An abnormality in the magnicellular pathway was confirmed anatomically by the demonstration that cells in the magnicellular layers of the dorsal lateral geniculate nucleus of the brains of reading-disabled persons were smaller than those in normal readers17.
Taken together, the psychophysical, electrophysiologic, and anatomical evidence supports the hypothesis that the magnicellular pathway is defective in reading-disabled subjects. The purpose of this study was to test magnicellular-pathway function in such persons by the use of flicker masking20. Visual evoked potentials were compared in two conditions: a low-contrast, low-spatial-frequency (coarsely patterned) target surrounded by a steady, homogeneous field, and a target in which the homogeneous field flickered at a sufficiently high rate to excite primarily the magnicellular pathway. In human beings and cats, a flickering field reduces the amplitude and increases the latency of the transient evoked potential elicited by a target with low spatial frequency. These effects of the flickering field have permitted the contribution of the parvicellular (small-celled) pathway to the evoked potential to be isolated21,22. The large flickering field engages the magnicellular pathway and masks its response to the target stimulus. Since the magnicellular pathway preferentially responds to a target of low contrast and low spatial frequency, removing its contribution to the evoked potential attenuates the overall amplitude of the potential and shifts its latency toward the slower parvicellular pathway. The flickering field has little effect on target stimuli with higher spatial frequencies (finely patterned targets), since such stimuli primarily excite the parvicellular pathway, which is less responsive than the magnicellular pathway to a large flickering field modulating quickly.
If the magnicellular pathway in reading-disabled children is defective, it would be expected that the flickering field would affect the visual evoked potentials of normal readers and reading-disabled subjects differently. If the temporal response of the defective magnicellular pathway is slowed in reading-disabled subjects, as suggested by Livingstone et al.,17 then there will be little effect, if any, of the flickering field on the evoked potentials of those subjects when targets of either low or high spatial frequencies are used. The results indicate that reading-disabled subjects have a pattern of effects of flicker masking that is predictably different from that of normal readers, which is symptomatic of a slower magnicellular pathway.
Methods
Subjects
Visual evoked potentials were recorded in two groups of children 8 to 11 years of age. The reading-disabled group contained eight children (mean age, 10.64 years) with reading levels one to two years below their current grade levels, but with overall scores for mathematics and listening comprehension at or above the grade levels. The control group was composed of 13 age-matched normal readers (mean age, 10.76 years). None of the children in the reading-disabled group had a history of sensory or cognitive impairment, emotional disturbances, neurologic dysfunction, or environmental factors that might underlie the reading disability, and none were receiving prescription medication for attention-deficit disorder. The subjects were selected by school personnel because they met the above criteria. The school district was chosen because the socioeconomic status of its families ranged from upper lower class to upper middle class. The reading-disabled group attended regular classes, with remedial reading provided.
Each subject underwent a complete optometric examination several weeks before the experimental session. All had normal visual acuity. Each child used appropriate refractive correction during the recording of visual evoked potentials.
Stimuli
The target and background stimuli were generated on a display monitor (Tektronix 608, P-31 phosphor) that subtended 17.2 by 14.3 degrees of visual angle at the viewing distance of 40 cm. The stimulus patterns were created by a Picasso Image Synthesizer (Innisfree). The mean luminance of the display was 16.0 candelas per meter squared. The target stimuli were sine-wave gratings (in which the luminance changes spatially in a sinusoidal fashion) with either 0.5 cycle per degree of visual angle or 4.5 cycles per degree at 0.1 contrast, presented for 250 msec. These targets appeared as alternating, vertical bars of light and dark with blurred edges. The low-spatial-frequency target had relatively wide bars (60 minutes of visual angle each), and the high-spatial-frequency target had relatively narrow bars (6.7 minutes of visual angle each). The gratings were displayed as a circular target with a diameter equivalent to 5 degrees of visual angle, surrounded by a uniform field that covered the remainder of the display. The uniform-field background had the same mean luminance as the target. A small fixation mark was placed in the center of the target.
In the steady-background conditions, the target was surrounded by a steady, uniform field, whereas in the uniform-field-flicker conditions, the uniform field flickered 12 times per second at a contrast of 0.70. The uniform-field flicker covered the entire display except the 5-degree area at the center, within which the target was presented. The uniform field flickered continuously throughout the trial and was not fixed in time to the presentation of the target. The target gratings contained either 0.5 or 4.5 cycles per degree of visual angle. Hence, there were four combinations of background and target serving as stimulus conditions: a steady background and a target with 0.5 cycle per degree, a steady background and a target with 4.5 cycles per degree, a uniform-field-flicker background and a target with 0.5 cycle per degree, and a uniform-field-flicker background and a target with 4.5 cycles per degree. Visual evoked potentials were collected from all subjects during exposure to each of the four stimulus conditions.
Recording Procedures
Electroencephalographic activity was recorded from a scalp electrode placed on the midline 1 to 2 cm above the inion, amplified 10,000 times, and band-pass filtered between 0.1 and 100 Hz. The reference electrode was placed along the midline at the top of the head, and the ground electrode was placed along the midline at the top of the forehead. Impedance readings were taken from each electrode. Stimulus presentation, on-line digitization of electroencephalographic activity, and data storage and analysis were controlled by an Apple IIe computer23. Each trial lasted 503 msec, and the sampling rate was 509 Hz. Each visual evoked potential was the average of 100 trials. Approximately one second elapsed between consecutive trials.
Experimental Procedure
Each session lasted less than 45 minutes, including the time required to fit the electrodes. Visual evoked potentials were always collected first for the target with 0.5 cycle per degree, since large, distinctive potentials were typically obtained with these targets. The sequence in which the background conditions appeared was allowed to vary randomly among subjects. For the last two trials, the targets used were those with 4.5 cycles per degree, and the two background conditions were presented in random order.
Each subject was placed in a comfortable head-and-chin rest and instructed to fixate on a small dot in the center of the stimulus display. The subject pushed a button to start and continued as long as fixation was maintained. Most subjects completed the trial with only one or two interruptions.
Results
Visual evoked potentials were obtained from each of the 21 observers in each of the four stimulus conditions without knowledge of the reading levels of the subjects. The N1, P1, and N2 components were measured (Figure 1). Later components were not clearly discernible. The amplitudes of the N1-P1 and P1-N2 complexes and the latencies of the N1 and P1 components were measured. Mean (±SE) values are shown.
|
Steady Background
Composite visual evoked potentials and data summaries are shown in Figure 2 for normal readers and the reading-disabled group with a steady background and target gratings of 0.5 cycle per degree. The amplitude of the N1-P1 complex was similar for both groups. However, the amplitude of the P1-N2 complex was significantly lower for the reading-disabled group (t = 2.985; P = 0.009, two-tailed; df = 9). The latencies of both the N1 (t = 1.836; P = 0.04, one-tailed; df = 9) and the P1 components (t = 1.641; P = 0.059, one-tailed; df = 9) were significantly longer for the reading-disabled group.
|
|
It has been shown in adults that the introduction of uniform-field flicker alters the visual evoked potential elicited by low-spatial-frequency targets by reducing the amplitude and increasing the latency of the N1-P1 complex21,22. These effects of uniform-field flicker were clearly evident in the normal readers.
Composite visual evoked potentials and data summaries for normal readers are shown in Figure 4 under the steady-background and uniform-field-flicker conditions with target gratings of 0.5 cycle per degree. The amplitudes of the N1-P1 (paired t = 3.171; P = 0.004, one-tailed; df = 12), and P1-N2 complexes (paired t = 3.992; P<0.002, two-tailed; df = 12) were smaller in the uniform-field-flicker as compared with the steady-background condition. The latencies of the N1 (paired t = 4.796; P<0.001, one-tailed; df = 12) and the P1 components (paired t = 1.982; P = 0.035, one-tailed; df = 12) were longer in the uniform-field-flicker conditions.
|
|
|
There were two differences between the normal readers and the reading-disabled subjects with regard to the visual evoked potentials recorded in the steady-background condition. First, the latencies of the N1 and P1 components were longer for the reading-disabled subjects than for the normal readers. Second, although the amplitude of the N1-P1 complex did not differ between the two groups, the amplitude of the subsequent P1-N2 component was unexpectedly smaller in the reading-disabled group. These differences occurred only with low-spatial-frequency targets, not with high-spatial-frequency targets. This finding is consistent with a recent report by May et al.,24 who found that the responses to low spatial frequencies and not to high frequencies distinguish the visual evoked potentials of normal readers from those of reading-disabled children.
The critical comparison of this study was that between the normal readers and the reading-disabled group in the uniform-field-flicker conditions. The flickering field had the expected effects on the visual evoked potentials of normal readers: an attenuation in the amplitude of the visual evoked potential, increased latencies of the N1 and P1 components, and more evident effects of the flickering field with targets at a low rather than a high spatial frequency. In the reading-disabled group, the flickering field attenuated the amplitude of the visual evoked potentials elicited by the target, but it did not increase the latency of the visual evoked potentials. The attenuation of these potentials in the uniform-field-flicker conditions was evident only for targets with low spatial frequency; there were no effects of the flickering field for targets with high spatial frequency.
Differences between the normal readers and the reading-disabled subjects cannot be attributed to factors related to attention, motivation, or both, since these factors would have influenced the visual evoked potentials in both background conditions. Moreover, no systematic differences were observed between the normal readers and the reading-disabled subjects when high-spatial-frequency targets were used. The differences in the visual evoked potential, therefore, are most likely due to a sensory deficit in the visual pathway of the reading-disabled group.
The results of the present study are consistent with reports that the magnicellular pathway is defective in reading-disabled subjects25,26,27,28,29. They also demonstrate that the deficiency involves the temporal characteristics of the magnicellular-pathway response. It was shown in the uniform-field-flicker condition for the low-spatial-frequency target that the amplitude of the visual evoked potential was attenuated by the flickering field in both normal readers and reading-disabled subjects. The attenuation of the visual evoked potential in the uniform-field-flicker condition for both groups suggests that the flickering field masked the response of a mechanism that responds to low spatial frequencies modulated at a high temporal rate, which is presumably the magnicellular pathway. The observation that the effects of the flickering field increased the latency of the visual evoked potential of normal readers, but not reading-disabled subjects, suggests that the magnicellular pathway in the reading-disabled subjects has an abnormally long latency of response. It follows that the latency of the visual evoked potential to low-spatial-frequency targets is predominantly determined by a slowed magnicellular pathway in the reading-disabled subjects, whereas in normal readers the latency of the visual evoked potential is dominated by a faster magnicellular-pathway response. This slowing of the response was also observed when the visual evoked potentials in the steady-background conditions were compared in the normal readers and the reading-disabled subjects. The N1 and P1 components were longer in the visual evoked potentials of the reading-disabled group than in the normal readers.
The latency of the visual evoked potential in the reading-disabled subjects was within normal limits for high-spatial-frequency targets, indicating that the delayed response of these subjects was not due to a general slowing of the visual response, but rather was restricted to the magnicellular pathway. A similar conclusion was reached by Livingstone et al.17 to explain the observation that longer latencies were observed in the reading-disabled group for low-contrast targets but not for high-contrast targets (the magnicellular pathway responds better to lower contrast than the parvicellular pathway)30,31,32.
The delayed response of the magnicellular pathway may also explain the reduction in the amplitude of the P1-N2 component. If one assumes that the visual evoked potential represents the combined responses of the magnicellular and parvicellular pathways, a change in the temporal relation between the two responses would distort the composite response. This could explain the abnormal shape of the wave form of the visual evoked potential in the reading-disabled subjects that was produced by the reduction in amplitude of the P1-N2 component.
Taken together, the evidence in this study supports the hypotheses that the magnicellular pathway is defective in reading-disabled children and that the deficiency involves a slowing in the response of this pathway. The challenge now is to determine whether temporal deficiencies of the magnicellular pathway are causally related to the reading problems experienced by reading-disabled people16,33,34 and, if so, whether stimuli can be manipulated to compensate for the temporal deficiency14. It is possible that a defect in the magnicellular pathway creates a timing disorder that precludes the rapid and smooth integration of detailed visual information necessary for efficient reading.
Supported by a grant from the American Foundation of Vision Awareness.
We are indebted to the City of St. Charles School District, St. Charles, Missouri, for assistance.
Source Information
From the School of Optometry, University of Missouri-St. Louis, 8001 Natural Bridge Rd., St. Louis, MO 63121, where reprint requests should be addressed to Dr. Lehmkuhle.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Defective Visual Pathway in Reading-Disabled Children
Victor J. D., Lehmkuhle S., Garzia R. P.
Extract |
Full Text
N Engl J Med 1993;
329:579, Aug 19, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
