70 (1992)3-13

A C T A 0 P H T H A L M 0 L O G IC A

REVIEW

Dyslexia; ophthalmological aspects 1991 Gunnar Lennerstrand and Jan Ygge Department of Ophthalmology, Karolinska Institute, Huddinge University Hospital, Huddinge, Sweden.

Abstract. Dyslexia or specific reading disability is an important and frequent handicap amicting 5-10%of the population. It is basically a disorder of acquisition of written language, probably due to a poorly developed phonological awarness, which in turn may be neurologically related. As anatomical correlates symmetry of the planum temporale and ectopies in the cerebral cortex have been suggested. Functional correlates are discovered with brain electrical mapping and stimulation of brain structures during neuro surgery. From an ophthalmological point of view there are no relations between dyslexia and ocular problems including refractive errors and accommodation, problems of binocular control and stereopsis, eye dominance instability etc. However, contrast sensitivity seems to be reduced in dyslexics for the middle range of spatial frequencies, which may be related to impaired function of the ‘transient’ visual system. With regard to eye movements, there is no firm experimental proof for any disturbances in dyslexia, be it with the different movement systems or in the best movement direction. ‘Backward saccades’ or regressions are typical not only for dyslexic reading but in all types of reading when comprehension is poor. Although there is no treatment for dyslexia itself that can be based on ophthalmological findings, the ophthalmologist must after careful examination discover and treat any ocular, orthoptic or neuroophthalmological problem that may make reading difficult for the dyslexic child. The ophthalmologist must explain to the child and the parents that dyslexia usually has no ophthalmological or visual cause but is a disability with a neurobiological background, still unknown, in which the only efficient treatment is within the area of pedagogy.

United Nations declared 1990 as ‘The Year of Literacy’. During these manifestations it was also recognized that a large part of the population of the I*

developed world, where analphabetism is practically eradicated, still cannot read and comprehend a text as well as one would expect. Many of these individuals, who are of normal intelligence and have received ordinary teaching, have a serious handicap in the form of a learning disability called developmental dyslexia. The problem is a large one since this group has been estimated to consist of 5-10%of the population depending on the criteria one chooses. Most research workers in the field have considered the handicap as a desorder of language acquisition, specifically of written language, and not related to e.g. visual or oculomotor defects (Council Report, J A M 1989). A neural substrate has been suspected for a long time but it is only in recent years that neurobiological observations have accumulated to give indications on which neural michanisms might be affected (Duane 1989; von Euler et al. 1989; Whyte 1989, Pennington & Porter 1991). It must also be emphasized that dyslexia can be compensated for by special teaching but that the condition has to be recognized as early as possible for best results. With regard to visual and oculomotor functions it is generally accepted by researchers in dyslexia that the deviations from normal in these areas, if they exist, probably are secondary to the basic problem of language functions (Beauchamp 1990). In t h i s survey a very brief report w i l l be made on some important recent findings on the neuroanatomy, neurophysiology, neuropsychology and genetics of dyslexia, in order to put the ophthalmological aspects in the right perspective. There have been many reports on oculomotor and visual

3

functions in dyslexia in recent years, mainly in the optometric literature but also in ophthalmology as will be shown in the following.

Recent neurobiological research Neuroanatomical changes

The neurological correlates of dyslexia are most likely congenital abnormalities of brain development, more specifically in the language related areas (Galaburda 1988, Sherman et al. 1989). Two different anomalies have been shown in post-mortem studies of brains of dyslexics. The first was the absence of the normal asymmetry between left and right hemispheres with regard to the volume of the planum temporale. Normally, the left planum is the largest, but in the dyslexic brains the right and left sides were equally large, and larger than in the right planum of normals. This symmetry has been confirmed also in living dyslexics with magnetic resonance tomography (MRT) of their brains (Duara et al. 1991). The symmetry seems most closely linked with the phonological dif€iculties that characterize a subgroup of dyslexics (Larsen et al. 1990). A second brain anomaly is the occurrence of a relatively large number of ectopies (Sherman et al. 1989),which originate from misdirected migration of neurons during embryonal brain development (Goldman-Rakic& Rakic 1984).It is possible that symmetry of planum temporale and ectopies reflect the same disturbance during ontogeny and that immunological factors active in specific phases of development is the common denominator (Sherman et al. 1989). Neurophysiological studies

The electropysiological correlates have been examined with.brain electrical activity mapping (BEAM) during test of language and non-language functions (Duffy and McAnulty 1985).The largest differences between dyslexics and normals were seen in the language areas of the parietal and temporal lobes, but they were also found in the frontal regions which are active in the planning and sequential transformation of different behavioural tasks including reading. It is also interesting that the anomalous BEAM patterns characteristic of dyslexic schoolchildren, can be seen already at preschool age in children who later become dyslexics.

4

The cortical organisation of language function has been studied in the awake human during neurosurgical operations under local anesthesia (Ojemann 1989).Patients with specific reading difficulties had significantly reduced areas for language functions, and these areas were located differently than the same areas in normal readers. It is tempting to suggest that the variability in the BEAM patterns could correspond to the differences in language area localisation demonstrated during neurosurgery, and that the neural basis might be the small ectopic developmental defects demonstrated in the brains of dyslexics. Psycholinguistic aspects

Does dyslexia represent defects of visual perception or is it a symptom of dysfunction in storage and retrieval of linguistic information? Some recent findings of psycholiinguistic research might shed light on this question. Reversals of letters is very common in children learning to write and may represent a transition form before the ability to write has become firmly established. In tests of decoding from left to right and from right to left, normal readers and dyslexics showed the same performance. It is therefore likely that the difficulties dyslectics have in maintaining the proper directionality or reading eye movements is a symptom of the reading disorder rather than the cause of the disorder (Vellutino 1987). Decoding is a fundamental function in reading and can be made through two separate routes, either phonologically or orthographically (Vellutino 1987). The former is used when we read unknown but regularly spelled words and the latter when we meet irregularly spelled words, such as abbreviations. The proficient reader has access to both mechanisms which probably are used concomittantly. Dyslexics may have problems which are larger with regard to one route than the other. It is possible that individuals with orthographic dyslexia have greater problems with visual perception than those with phonological dyslexia. Continuing with the phonological type of dyslexia, it seems firmly established that a large part of the dyslexic children have a poorly developed phonological awareness, which means that they cannot perceive the words as separated into phonems, the smallest unit of the speech sounds (Lundberg and Hoien 1989).The phonems are often coarticulated in speech, and young children have difficulties in

imagining and manipulating these abstract concepts, and in transferring them from the spoken to the written language. There is a strong correlation between early phonological awareness and success in learning to read. Phonological exercises during the pre-school year can improve not only the general sense of language but also have a more long-lastingpositive effect on the ability to learn to read and write in school (Lindberg et al. 1988). Genetics

It is known already from the early studies by Hallgren (1950)that dyslexia is much more common in some families than in others. Newer genetic research including studies of twins and coupling analysis been reported recently (Pennington & Porter 1991). It has been confirmed that the most common subtypes of dyslexia are genetically determined (De Fries et al. 1987), but that the genetic component is most prominent in dyslexia with phonological dysfunction and less so in orthographic subtypes (Olsson et al. 1989). It has been suggested that dyslexia of the phonological type is coupled to an aberration on chromosome 15. Neuropsychiatric aspects

There seems to exist a strong correlation between dyslexia and a number of childpsychiatric problems, manifested particularly as disturbances of social behaviour. It is likely that dyslexia is a predisposing factor for the development of these behavioural abberations and not the reverse. It has been shown in a large longitudinal study that children with disturbances of concentration and motor-perceptual development have prominent delays also in the development of reading abilities. More than 213 of these children must be regarded as dyslexics in comparison with about 8%of children in a control group (Gillberg& Gillberg 1983).A biological CNS-disturbance has been hypothesised, leading to perceptual and cognitive deviations that manifest themselves as motor-perceptual problems and/or dyslexia. The child may feel different from his peers and this in turn can lead to psychiatric reactions of different types, depending on the personality of the child, the family situation, social structure, attitudes in school etc. According to this idea perceptual-motor problems and dyslexia are parallel phenomena and specific treatment for one of them does not affect the other. For example, it has been shown in controlled

studies that exercises of motor skills improve performance in this area, but not reading ability (Gillberg 1983). It is therfore inappropriate to suggest motor excercises including eye movements as a cure for dyslexia.

Visual dysfunction and dyslexia In vision as in other sensory functions one may discern differet levels of dysfimtion. The defect may be in the eye itself, manifested for example as a refractive error, abnormality of the optic media, retina, or the optic nerve. In this context it is important to note that many children with very poor vision due to eye diseases can learn to read quite well with the use of visual aids. The next levels of disability are in visual perception, e.g. reduced visual acuity and/or contrast sensitivity, reduced motion or colour detection or disturbances of binocular vision. The highest level of disability would be represented by memory defects and difficulties in storage and retrieval of visual information. As mentioned before, the generalized cognitive disabilities, combined with mental retardation, should be separated from the condition of specific reading disability or dyslexia (Beauchamp 1990). An analysis of the level of visual problems in relation to dyslexia will be attempted in the following. As mentioned earlier, one should possibly separate dyslexia into subtypes based on the phonological and orthographic deficiencies, even if mixed forms probably are common. However, a subdivison of the patient material has been done in very few studies of visual functions in dyslexia, which may explain some of the often contradictory results. While there has been great interest among optometrists and orthoptists to study visual functions including binocularity in relation to reading ability, there are relatively few ophthalmological reports in this field. Actually, three of the most extensive and best controlled ophthalmological studies during the last decades have been performed in Scandinavia. They include the studies in Denmark at the end of the 60's by Norn and coworkers (Norn et al. 1969; Hamrnerberg & Norn 1972),in Norway at the end of the ~O'S, reported by Aasved (1987, 1988), and an ongoing study in Sweden (Ygge et al. 1991). The Danish and the Swedish studies have used matched control groups. The Danish study included dyslexic child5

ren and children with normal reading ability in grade 3-8 of the same age and sex but not from the same school, while both groups of children in the Swedish study were pupils of grade 3 during testing, and matched pairs were compared from the same class. Occular abnormalitiesand refractive errors In general, ocular disease does not seem to affect the ability of children to learn to read. An example is a child operated for bilateral congenital cataracts at a few months of age and fitted with contact lenses. This child has lost the ability to accommodate and as a compensation needs to wear near correction. Another example is a child with congenital nystagmus where the fixaton instability does not influence the ability to learn to read. In terms of acquired pathology, brain damage would seem to be the only condition to cause reading difficulties of a kind similar to dyslexia,particularly if areas that serve visual cognition are affected. However, other mental functions are usually affected as well. With regard to refractive errors, there are some indications in the optometric literature of a preponderance of hypermetropic refraction among children with learning disabilities (see review by Evans & Drasdo 1990) but the criteria for selection of dyslexics and control children were often doubtful. In the three ophthalmological studies from Scandinavia, refraction was determined with retinoscopy in cycloplegia. No correlation could be found between reading performance and any specific type of refractive error (Norn et al. 1969;Aasved 1987,1988;Ygge et al. 1991).However, dyslexic children were wearing glasses twice as often as normal readers, which probably reflects an attitude towards correction even of small and probably insignificant rehctive errors in the dyslexics (Norn et al. 1969). In a longitudinal study on a population of children followed regularly since birth, no difference in reading scores was found between myopic and hyperopic children at ages 7 and 11 years (Williams et al. 1988). Artificial hyperopia by means of minus lenses in adult normal readers did reduce the reading performance with regard to speed of reqding but not with respect to accuracy or understanding (Garzia et al. 1989). There is no evidence in the ophthalmological studies that accommodation is different in the two groups as has been claimed in the optometric literature (Evans & Drasdo 1990).

6

In conclusion, defects in vision due to peripheral factors in the eye itself, or in the primary visual pathways are not correlated with dyslexia. Visual acuity and contrast sensitivity There is no conclusive evidence in the previous literature for the idea that dyslexics have reduced visual acuity for distance or for near with proper spectacle correction (Evans & Drasdo 1990). However, in the on-going study of Swedish children, a statisticallysignificant difference in acuity for both distance and near was seen between dyslexic children and normal readers at the age of 9 years (Ygge et al. 1991). All normal readers had acuities of 1.0 or 1.3 (20/20 or 20/15) while some dyslexics had acuities also at 0.8 and lower (20/25). The functional significance of this small difference is unclear. It might reflect that the visual acuity development is slower in children with dyslexia than in normal readers, and in that case the difference would have disappeared at the follow-up in 2-3 years. Tests of accommodation width did not reveal any difference between the groups, but the dynamics and the endurance of accommodation could not be adequately examined, and differences between groups in these respects might have influenced the results on acuity for near. It has been claimed that the contrast sensitivity function provides a more complete description of form vision than visual acuity, which is only one point on the sensitivity curve (Campbell 1974; Lundh 1983). In this test contrast thresholds are determined with patterns, usually of a sine wave luminosity prophile, at different sizes or spatial frequencies. It has been known for some time that some dyslexics show reduced contrast sensitivity in the range of lower spatial frequencies (Lovegrove et al. 1982; Breitmeyer 1989). Recently it was discovered that some dyslexics have reduced contrast detection also in the high spatial frequency range corresponding to the difference found in visual acuity mentioned above (Ygge et al. 1991).These findings have been correlated with the abnormalities in visuo-spatialperception that have been demonstrated in dyslexic individuals (Solman & May 1990).The research area has recently been reviewed from many aspects, and the present thinking is that dyslexia involvesa dysfunction or imbalance between the so-called sustained and transient subsystems of vision (Breitmeyer 1989; Lovegrove et al. 1990). The sustained system is mainly rep-

resented in foveal vision and participating in visual discrimination and colour vision, while the transient system subserves mainly parafoveal and peripheral vision and transmits information of movement and spatial location. It should be noted that the physiological concept of sustained and transient visual systems seems to have its neuroanatomical correlate in the X- and Y-cell systems (Stone 1983) and the parvo- and magnocellular channels of visual neurons from the retina over the lateral geniculatebody up to the visual cortex (Livingstone & Hubel 1988). With respect to contrast sensitivity, the sustained system is considered to be responsible for the perception of contrast in the range of higher spatial frequencies, and the transient system for the middle and lower ranges. The results of contrast testing in dyslexics would indicate a reduced sensitivity for contrast in their transient system (Lovegrove et al. 1982).Some worker have gone so far as to suggest that the mismatch between the sustained and the transient systems is an important part of the reading problems (Breitmeyer 1989; Lovegrove 1990; Lovegrove et al. 1990).Their idea is that sustained activity from one fixation normally is eradicated by suppression through saccadic inhibition from the transient system, activated by the saccadic movement between fixations. If the inhibition is weaker than normal, sustained activity from one fixation will remain into the next one and interfere with the acquisition of new information. This would occur in dyslexics with their subnormal transient system function. Additional support for the idea of an imbalance between transient and sustained systems in dyslexia derives from studies of visual integration time and lateral masking between center and parafoveal areas (Williams et al. 1989,1990;Winters et al. 1989;Geiger & Lettvin 1987; Klein et al. 1990). It has also been the basis of two kinds of remediation for reading disability that is presently being tested. The first implies filtering of higher spatial frequencies to reduce sustained channel activity and make transient system inhibition relatively more effective, and the other word-by-word reading within a window, cutting out the parafoveal information from a deficient transient system in dyslexics (Lovegrove 1990). An imbalance between the transient and the sustained system functions could thus explain the reduced average contrast sensitivity found in the

lower range of spatial frequencies in a group of dyslexics. The reduction of contrast sensitivity in the high frequency range is also reflected in a decrease of visual acuity (Ygge et al. 1991), but this finding has not been previously been reported and its physiological meaning is unknown at present. However, it is not possible to state if these abnormalities at the level of perception represent the primary defect in dyslexia or if they are parallel phenomena originating from cognitive dysfunction. It would also be of interest to know whether the contrast sensitivity defects are more common in dyslexics with orthographic reading problems that in dyslexia of the phonological type. In this context, reading with colored lenses might be mentioned, although such therapy onginally was not directed towards imblances in contrast sensitivity. However, no beneficial effect on reading abilities have been proven by the use of colored lenses (Solan & Richman 1990). Binocular functions

In the optometric literature there are claims of a correlation between slight abnormalities of binocular function and reading problems (Brod & Hamilton 1973; Bishop et d. 1979; Rosner & Rosner 1987),but in most of these studies the selection of experimental and control groups has been inadequate. The conclusion of a recent review on the subject was that no clear relation could be demonstrated between convergence insuf€iciency, fusional defects and poor stereopsis, on one hand, and dyslexia on the other (Evand & Drasdo 1990). Neither did fixational disparity seem to be related to dyslexia (Mohindra et al. 1976). The results of the three Scandinavian studies referred to earlier also reject any relation between binocular dysfunction and disabilities of reading morn et al. 1969; Aasved 1988; Ygge et al. 1991). In the Danish and Norwegian studies a few children were treated for poor binocular control, and they obtained improvement with respect to binocularity but not in reading skills (Norn et al. 1969; Aasved 1988). However, studies related to visuo-spatial dysfunction in dyslexia have implicated more subtile abnormalities of binocular coordination as a contributing factor to the reading problems (Stein & Fowler 1984; Stein 1989). This is of importance since such an approach may offer treatment for dyslexia within the ophthalmological sphere. Many dyslexics describe that the text seem to be

7

moving around, the letters jumping and changing places. These symptoms have been interpreted by some workers to indicate that dyslexics suffer from an inability to accurately coordinate oculomotor and retinal information on eye position (Stein & Fowler 1984). This incoordination is particularly disturbing when the eyes are converged as in reading at close distance. The integration of the visual and oculomotor information is presumed to take place in the right parietal cortex and patients with lesions within this area are known to exhibit similar symptoms (Stein 1989). Normal readers avoid the confusion by developing a 'leading' (dominant) eye that they always use for reading. For investigation of eye dominance Stein & Fowler (1989) have used the so-calledDunlop test (Dunlop et al. 1973). They reported that 50 out of 80 ten-year-old children with 'visual' dyslexia have an unstable eye dominance, whereas all but one pupil in the matched control group exhibited a stable eye dominance. However, while Bigelow & McKenzie (1985) could confirm the results of Stein and co-workerson ocular dominance, Newman et al. (1985) and Aasved (1988) could not discriminate between dyslexic and control children with the Dunlop test and neither has this been possible in our study of Swedish children (Ygge et al. 1991).In an another study, Fowler et al. (1988) have found that vergence eye movements to small fixational objects were impaired in children with learning disabilities, but this could not be confirmed by Ygge et al. (1991). Thus, at present there is no clear physiological evidence for a connection between dyslexia and instability of eye dominance or poor vergence control. In an attempt to improve the binocular functions of dyslexics, Stein et al. (1987, 1988) have used monocular occlusion during reading complemented by orthoptic exercises and shown that within 6 months 2/3 of the children with previously unstable eye dominancewill stabilizein eye dominance and improve their vergence functions. These findings have been confirmed by Masters (1988). The most controversial part of the studies by Stein and co-workers is their proposal that the stabilization of the binocular functions in dyslexia occurs in conjunction with an improvement in reading. With monocular occlusion for 6 months they report a 12-month improvement in reading age, whereas children without treatment only showed a change related to age. However, Bishop (1989)has disputed the statistical methods of these studies 8

and on recalculation of the findings could not find evidence for any effects of occlusion on reading ability. At the moment there is no unequivocal evidence for beneficial effect of monocular reading and orthoptic exercises on reading ability in dyslexia. This is not to say that these methods may not work in individual cases, but the pathophysiological basis is still lacking. Strabismus

Several earlier studies have not described any relationship between general orthoptic findings and dyslexia as reviewed by Evans 8c Drasdo (1989). Aasved did not find a significantly higher frequency of strabismus in his study. Nor did Bishop et al. (1979)in a study of 147 eight and a half yearold-children. In fact, a number of investigations have shown a low prevalence of dyslexia in subjects who lack binocular vision, namely those with some kind of strabismus (Norn et al. 1969; Benton et al. 1965; Park & Barry 1984).

Eye movements in normal reading Reading is the most complex of all motor skills and therefore requires a good ability to coordinate different types of eye movements. The first description of the eye movements during reading was made by the French ophthalmologist Emile Javal in 1879. He reported that the line of sight instead of making a smooth movement along the text line moves in small steps, i.e. saccades, progressivly from left to right. Between the saccades, the eyes are steady in fixation. Frequently during reading a backward saccade is released, i.e. the regression, which is thought to represent a verification of a part of the text just read for the comprehensionof the text. Fixations in reading

It is during the fixations that the visual information is percieved. The eyes are steady in fixation about 90%of the reading time. About 85%of the fixations occur after rightward saccades and about 15%occur after leftward saccades, i.e. regressions (Rayner 1978; McConkie & Zola 1989). Several especially shorter words are not furated during reading, whereas longer words are sometimesread with two fixations, one at the beginning of the word and the other at the end (Dunn-Rankin 1978; Rayner 1979). The actual position of the fixation is de-

pendent on where the preceding fixation was made (Rayner 1976,1986)and how wide a ‘perceptual span’ the reader can use. The duration of the fixation varies much, not only between readers but also for the individual reader (McConkie & Zola 1989).Thus, the duration of fixations increase with the difficulty of the text. The mean duration of the fixations is estimated to about 180 ms at normal reading but, both long durations up to 450 ms as well as short fixations of 30-40 ms occur (McConkie & Zola 1989). The fact that there occurs furations that are shorter than 100 ms during reading is quite astonishing. It is known that the time interval for the nerve impulse to pass from the retina to the CNS and there elicit a new saccade is about 90 ms. These short fixation durations must therefore be so short that the text that the eyes are fixating is not yet percieved when the CNS programs a new saccade. Therefore, some saccades during reading are thought to be initiated by some internal command which is not dependent on the fixation (McConkieet al. 1988). Saccades during reading

Saccadiceye movements are the most frequent and the most rapid motion in the entire human motor system. These eye movements take place within a fraction of a second and at velocities of up to 700 degreeslsec. The function of the saccade during reading is to bring a new region of text onto the fovea. During the short saccade the visual perception is considered to be almost entirely suppressed. However, a considerable controversy about this saccadic suppression has been generated (cf. Mitrani et al. 1973, 1975; Matin 1974; Volkman 1976) but the general conclusion today is that no significant perception occurs during the saccade. The mean length of a saccade during reading is about 2 degrees, corresponding to about 8 letters of text at a normal reading distance (McConkie & Zola 1989). However, large individual differences exist between normal readers. The length of saccades is known to have large effects on reading speed (Kowler & Anton 1987) and is influenced by the ability to recognize letters and letter combinations outside the most central visual field. Kowler & Anton (1987) have shown that changing the order or orientation of letters in an otherwise familiar text gives rise to reading with very small saccades, where the fixations were placed on all or every other letter. The length of the saccade is further

dependent on the difficulty of the text, i.e. the more complicated the text, the shorter the saccades. The length of the word to the left of the fmated one affects the length of the saccade.A longer following word gives rise to a shorter saccade (Dunn-Rankin 1978; Rayner & Pollatsek 1987). Regressions during reading

The regressions during reading are shorter than the forward saccades and have a mean length of about 1 degree, corresponding to about 4 letters of text. In normal reading about 15%of the saccades are made up of regressions (McConkie & Zola 1989). Regressions are thought to have a grammatic and syntactic function in that a few preceding key words in the text, such as verbs and important conjunctions are verified. The regressions are thus of importance for the comprehension of the text and their number increase with the difficulty of the text (Rayner & Pollatsek 1987;Jones & Stark 1983).

Eye movements in beginners of reading A beginner of reading uses considerably more time to read a text line than a good reader. This depends on four factors (Rayner 1986;Kowler & Martins 1982; Bryant 1988): i) The duration of the fixations are longer. In the beginner the fixation duration is usually about 300-450 msec, i.e. more than twice the time of the average reader. ii) The beginner also makes many more fixations while reading a text line, sometimes a factor of 2 or 3. iii) The length of the saccade is shorter in the beginner than in the good adult reader. The mean saccadeis about 1 degree, i.e. about four letters of text. iv) The most obvious difference between the good reader and the beginner is the great number of regressions in the beginner. More than 25%of the fixations in the beginner occur after regressions. The developmentof the reading skill accelerates during the first years of reading up to about 10 years of age. The number and duration of the fixations diminish as the child achieves a greater vocabulary. With a greater vocabulary the child can use more of the parafoveal information so that he can use fewer fixations and thus make longer saccades. At 10 years of age most normal children have achieved about 213 of the adult normal reading speed (Rayner 1986). Further improvement of the reading skill is dependent on a reduction of the 9

number of the regressions (McConkie & Zola 1989).

Eye movements in dyslexic reading Generally, the eye movement patterns of dyslexics can be descibed as very similar to the eye movements in children beginning to read. This could imply that dyslexics, as well as beginners of reading, have difficultiesin decoding the words and integrating the text to understand the content. The deviations from the normal eye movements that are observed in dyslexic reading would in that case only mirror the language problems the dyslexics have with the text. Kowler & Anton (1987) have shown that normal readers show the same kind of defect eye movements when individual letters in a text are turned 180 degrees, are in changed position or reversed. In this case the normal readers also reported that they had reading problems similar to that reported by dyslexics, i.e. difficulties with the comprehensionof the text due to the difXculties in the decoding precess (Inhoff et al. 1988). The fixation duration in dyslexic reading has been shown to be longer than in normal reading. It has been speculated that this could indicate that the dyslexis reader needs more time to comprehend a word or letter combination. However, dyslexics have no difficulty in comprehending words and letter combinations when they are presented tachistoscopically with very short durations (Zangwill & Blackmore 1972). During recent years there have been reports in the literature that the defective eye movements that occur in dyslexics while reading could be the cause and not the result of the dyslexia. Pavlidis presented the theory that dyslexic children could suffer from a disability of sequential tracking performance and his results (1981a,b, 1985a,b) indicated that the defective eye movements seen in dyslexia could be the key to the condition. Pavlidis has even postulated that saccadic testing in the form of simple visual-tracking tasks of following sequentially illuminating light sources could be an early indicator of a developing dyslexia. Naturally these results had a large impact on people engaged in education. Pavlidis results also had some support in earlier studies of reading-related tasks not requiring comprehension (Elterman 1980; Griffin 1974).Black et al. (1984a,b)investigated extensively the saccadic parameters in 28 dyslexics and com10

pared them with 31 normally reading children. They showed that the eye movement performance including saccadic latency, velocity, accuracy and acceleration was not distinguishably different for poor readers and normal readers. However, they make a comment that reading requires higher processing and therefore their results regarding saccades are not directly applicable to reading. Several research groups have tried to reproduce the findings of Pavlidis (Brown et al. 1983a,b;Olsson et al. 1983;Stanley et al. 1983a,b).They have all come to the result that dyslexics and normal patients exhibit the same kind of eye movement patterns while doing saccadic tracking movements. Thus, one must assume that Pavlidis dyslexic patients may have had some kind of eye movement abnormality, possibly related to their reading handicap but not necessarily the cause of it. The weight of recent evidence clearly points towards the conclusion that oculomotor control of dyslexic children on sequential saccadic tasks is similar to that of normal children. Therefore, a visual test based on searchingfor abnormalities in sequence tracking will not be diagnostic in early detection of dyslexia. Backward reading

It is a well known fact that dyslexics make many more regressions while reading than the normal reader. Very often these backward regressions are longer than the preceding forward saccade and sometimes a series of regressions can be seen in the dyslexic reading. This has been termed ‘reversed staircase pattern’ and is thought to be a characteristic feature of dyslexic reading (Zangwill & Blakemore 1972). It could even indicate that dyslexics have a tendency to try to read backwards. Many dyslexics have also described that they see the text reversed or the letters in a backward order. There are indications that some dyslexics show almost normal reading eye movements while holding the text upsidedown while reading. They also report that the comprehension of the text was much better and also that the reading speed was greater with this kind of reading (Ciuffreda et al. 1972, 1985).

Slow pursuit movementsand reading It is well known that reading is not dependent on the slow pursuit system. However, Adler-Grinberg

& Stark (1978)reported that dyslexic children more frequently showed abnormalities in a smooth pursuit task than normally reading children. The defect in the smooth pursuit movements they observed was a breakdown of the ordinary ramp gain to a saccadic pursuit or ‘cog-wheeling‘,i.e. intruding saccades in the pursuit movement. In order to investigate these movements further, Black et al. (1984~)made a thorough study of horisontal eye movements associated with visual tracking of a smoothly moving (5 deg/sec) target. They found that essentially all children exhibited small saccades superimposed on the smooth pursuit movements. The most significant difference in pursuit performances between the groups were the number of large amplitude saccades.Accordingto these authors this would suggest that poor readers have difficultiesin maintaining fixation on a moving target. The detailed numerical analysis revealed that about 25% of poor readers showed an abnormal increase in the saccadic component of the smooth pursuit movements. It was further proposed that testing eye movements during tracking of smoothly moving targets at low velocity, might be a useful test for early detection of poor-reading children. However, statistical evaluation of the data is necessary since there are considerablevariations fi-omsubject even among normal readers.

References Aasved H (1987): Ophthalmological status of school children with dyslexia. Eye 1: 61-68. Aasved H (1988): 0yenundersekelser. Bergenprojektet III. Studier av barn med dysleksi og andre laerevansker. H-J Gjessing, HD Nygaard and R Solheim. Universitetsforlaget,Oslo. Adler-Grinberg D & Stark L (1978): Eye movements, scanpaths, and dyslexia. Am J Optom Physiol Opt 55: 557-570. Beauchamp G (1990):Visual correlates of dyslexia and related learning disabilities Pediatr Ann 19(5):334-341. Benton C D, McCann J W & Larsen M (1965): Dyslexia and dominance.J Pediatr Ophthalmol2: 53-57. Bigelow E R & McKenzie B (1985): Unstable ocular dominance and reading ability. Perception 14: 329-335. Bishop D V M, Jancey C & Steel A McP (1979): Orthoptic stasus and reading disability. Cortex 15: 659-666. Bishop D V (1989): Unfixed reference, monocular occlusion, and developmental dyslexia- a critique. Br J Ophthalmol 73: 209-215.

Black J L, Collins D W, de RoachJ N & Zubrick S (1984a): Dyslexia: saccadic eye movements. Percept Mot Skills 58: 903-910. Black J L, Collins D W, de RoachJ N & Zubrick S (198413): A detailed study of sequential saccadic eye movements for normal- and poor-reading children. Percept Mot Skills 59: 423-434. Black J L, Collins D W, de Roach J N & Zubrick S (1984~): Smooth pursuit eye movements in normal and dyslexic children. Percept Mot Skills 59: 91-100. Breitmeyer B G (1989):A visually based deficit in specific reading disability. Irish J Psycho1 1 0 534-541. Brod N & Hamilton D (1973): Binocularity and Reading. J Learn Disab 6: 574-576. Brown B, Haegerstrom-Portnoy G, Adams A J et al. (1983a): Predictive eye movements do not disciminate between dyslexic and control children. Neuropsychologia 21: 121-128. Brown B, Haegerstrom-PortnoyG, Yingling C D, Herron J, Galin D & Marcus M (1983b): Tracking eye movements are normal in dyslexia children. Am J Optom Physiol Opt 60: 376-383. Bryant N R (1988): Children’s eye movements and the learning of words. Doctoral dissertation. Univ Ill, Urbana, Illinois. Campell F W (1974):The transmission of spatial information through the visual system. In: Schmidt F 0 & Worden F S (eds).The Neurosciences Third Study Program 95-103 Cambridge, Massachusetts:The M.I.T. Press. Ciuffreda K J, Bahill A T, Kenyon R V & Stark L (1972): Eye movements during reading: case studies. Am J Optom Physiol Opt 53: 521-530. Ciuffreda K J, Kenyon R V & Stark L (1985): Eye movements during reading: further case reports. Am J Optom Physiol Opt 62: 844-852. De FriesJ C, Fulker D W & La Buda M C (1987):Evidence for a genetic aetiology in reading disability of twins. Nature 329: 537-539. Duane D (1989): Neurobiological correlates of learning disorders. J Am Acad Child Adolesc Psychiatry 28(3): 314-318. Duara R, Kushch A, Gross-Glenn K, Baker W, Jallad B, Pascal S, Loewenstein D, SheldonJ, Rabin L, Levin B & Lubs H (1991): Neuroanatomic differences between dyslexic and normal readers on magnetic imaging scans. Arch Neurol48: 410-416. Duffy F H 8c McAnulty G B (1985):Brain electrical activity mapping (BEAM):The search for a physiologicalsignature of dyslexia. In: Duffy F H, Geschwind N; eds. Dyslexia. A Neuroscientific Approach to Clinical Evaluation. Little, Brown and Company, Boston 105-122. Dunlop D B, Dunlop P & Fenelon B (1973): Vision-laterality analysis in children with reading disability: The results of new techniques of examination. Cortex 9: 227-236. Dunn-Rankin P (1978): The visual characteristics of words. Scientific American 238: 122-130.

11

Dyslexia. JAMA, 1989; 261 (15): 2236-2239. Elterman RD, Abel L A, Daroff R B, Dell’Osso L F & Bornstein J L (1980): Eye movement patterns in dyslexic children. J Learn Disabd 13: 16-21. Evans B J W & Drasdo N (1990): Review of ophthalmic factors in dyslexia. Opthal Physiol Opt 10: 123-132. Fowler M S, Ridell P M & SteinJ F (1988):The effect of varying vergence speed and target size on the amplitude of vergence eye movements. Br Orthopt J 45: 49-55. Galaburda A M (1988): The pathogenesis of childhood dyslexia. In: Plum F (ed). Language, communication and the brain. New York, Raven Press 127-137. Garzia R P, Nicholson S B, Gaines C S, Murphy M A, Kramer A & Potts J (1989): Effects of nearpoint visual stress on psycholinguistic processing in reading. J Am Optom Assoc 6 0 38-44. Geiger G & Lettvin J Y (1987): Pheripheral vision in persons with dyslexia. N Engl J Med. 316 1238-1243. Gillberg I C & Gillberg C (1983):Three-year follow-up at age 10 of children with minor neurodevelopmental disorders. Behavioural problems. Dev Med Child Neurol 25: 438-449. Gillberg I C (1987): Deficits in attention, motor control and perception: follow-up from preschool to the early teens. Thesis. Uppsala University, Sweden. Goldman-Rakic P & Rakic P (1984): Experimentally modified convolutional patterns in nonhuman primates: Possible relevance of connections to cerebral dominance in humans. In: Geschwind N & Galaburna A M (eds).Cerebral Dominance: The Biological Foundations. Harvard University Press, Cambridge MA, USA 179-194. Griffin D C, Walton H N & Ives V (1974): Saccades as related to reading disorders.J Learn Disabil7 52-58. Hallgren B (1950): Specific dyslexia (‘congenital wordblindness’): a clinical and genetic study. Acta Psychiatrica et Neurologica, suppl, No 65: 1-287. HammerbergE & Norn M S (1972):Defective dissociation of accommodation and convergence in dyslectic children. Acta Ophthalmol (Copenh) 50: 651-654. Inhoff A W, Pollatsek A, Posner M I & Rayner K (1988): Covert attention and eye movements during reading. QJ Exp Psychol: in press. Jones A & Stark L W (1983): Abnormal patterns of normal eye movements in specific dyslexia. Eye movenments in reading, perceptual and language processes 481-498. Klein R, Berry G, Briand K, DEntremont B & Farmer M (1990):Letter indifcation declines with increasing retinal eccentricity at the same rate for normal and dyslexic readers. Perception & Psychophysics, 47(6): 601606. Kowler E & Martins A J (1982): Eye movements of preschool children. Science 215: 997-998. Kowler E & Apton S (1987): Reading twisted text: Implications for the role of saccades Vision Res 27: 45-60.

12

Larsen J P, Hoien T, Lundberg I & Odegaard H (1990): MRI evaluation of the size and symmetry of the planum temporale in adolescents with developmental dyslexia. Brain Lang 39: 289-301. Livingstone M & Hubel D (1988): Segregation of form, color movement, and depth Anatomy, physiology, and perception. Science 240: 740-749. Lovegrove W, Martin F, Bowling A, Badcock D & Paxton S (1982):Contrast sensitivityfunctions and specificreading disability. Neuropsychologia 20: 309-315. Lovegrove W (1989): The visual deficit hypothesis. Nature, theory and treatment. In: Singh N & Beale I (eds). Current Perspectives in learning disabilities. Springer Verlag, New York. Lovegrove W, Garzia R P & Nicholson, S B (1990): Experimental evidence for a transient system deficit in specificreading disanbility.J Am Optom Assoc 61: 137146. Lundberg I & Hoien T (1989): Phonemic deficits - a core symptom of developmental dyslexia? Irish J Psychol 10: 579-592. Lundberg I, FrostJ & Petersen 0 P (1988):Long term effects of a preschool training program in phonetical awareness. Reading Research Quarterly 2 8 263-284. Lundh B L (1983):Clinical contrast sensitivity. LinkBping University Medical Dissertations No 144. Masters M C (1988): Orthopic management of visual dyslexia. Br Orthopt J 45: 49-55. Matin E (1974):Saccadic suppression:a review. In: Monty R A & SendersJ A (eds).Eye movements and a analysis. Psycho1 Bull 81: 899-917. McConkie G W & Zola D (1989):Some characteristics of readers’ eye movements.In: von Euler C, Lundberg I & Lennerstrand G (eds). Brain and reading. London: Macmillan Press 369-381. Mitrani L, Yakimoff N & Mateef S (1973): Saccadic suppression in the presence of a structured background. Vision Res 13: 517-521. Mitrani L, Radil-Weiss T, Yakimoff N, Mateef S & Bozkov V (1975): Various background pattern effect on saccadic suppression. Act Nerv Super (Prahe) 17: 161-164. Mohindra I, Scheiman M M & Scheiman M T (1976):Fixation disparity and learing disabilities. Br J Physiol Optics 30: 128-131. Newman S P, Wadsworth J F, Archer R & Hockley R (1985): Ocular dominance, reading, and spelling ability in schoolchildren.BrJ Ophthalmol69: 228-232. Norn M S, Rindziunski E & Skydsgaard H (1969): Ophthalmologic and orthoptic examinations of dyslectics. Acta Ophthalmol (Copenh)47: 147-160. Ojemann G A (1989): Some brain mechanisms for reading. In: von Euler C, Lundberg I & Lennerstrand G (eds).Brain and Reading. Macmillan Press 47-59. Olsson R, Wise B, Conners F, Rack J & Fulker D (1989): Specific deficits in component reading and language skills: Genetic and environmental influences. J Learn Disab 22: 339-348.

Olsson R K, Kliegl R & Davidson B J (1983):Dyslexic and normal readers’ eye movements. J Exper Psychol 9: 816-825. Park G E & Barry C (1984):Eye maturations and reading difficulties. J Educ Psychol 12: 535-545. Pavlidis G T (1981a):Do eye movements hold the key to dyslexia? Neuropsychol 19: 57-64. Pavlidis G T (1981b):Sequencing, eye movements and the early objective diagnosis of dyslexia. In: Pavlidis G T & Miles T R (eds).Dyslexia research and its application to education. London: J Wiley & Sons Ltd 99-163. Pavlidis G T (1985a):Eye movements in dyslexia: their diagnostic significance. J Learn Disabil 18: 42-50. Pavlidis G T (1985b):Eye movement differences between dyslexics, normal, and retarded readers while sequentially fixating digits. Am J Optom Physiol Opt 62: 820832. Pennington B & Porter D (1991): Proceedings of the XVI International Rodin Scientific Conference: ‘Genetic and Neurologica Influences on Dyslexia’, in Boulder, Colorado, USA, September 19-22, 1990. Special issue of Journal of Reading and Writing. Rayner K (1976): What guides a reader’s eye movements? Vision Res 16: 829-837. Rayner K (1978): Eye movements in reading and information processing. Psychol Bull 85: 618-660. Rayner K (1979):Eye guidance in reading: Fixation locations within words. Perception 8: 21-30. Rayner K (1986): Eye movements and the perceptual span in beginning and skilled readers. J Exp Child Psychol 41: 211-236. Rayner K & Pollatsek A (1987): Eye movements in reading: A tutorial review. In: Coltheart M (ed). Attention and performance 12: The psychology of reading, London: Erlbaum. Rosner J & Rosner J (1987):Comparison of visual characteristics in children with and without learning difficulties. Am J of Optom & Physiol Optics 64(7): 531-533. Sherman G F, Rosen G D & Galaburda A M (1989):Neuroanatomical findings in developmental dyslexia. In: von Euler C, Lundberg I & Lennerstrand G (eds). Brain and Reading. Macmillan Press 3-15. Solan H A & Richman J (1990): Irlen lenses: a critical appraisal. J Am Optom Assoc 61: 789-796. Solman R T & May J G (1990): Spatial localization discrepancies: A visual deficiency in poor readers. Am J Psycho1 103(2):243-263. Stanley G, Smith G A & Howell E A (1983a): Eye-movements and sequential tracking in dyslexic and control children. Br J Psychol 7 4 181-187. Stanley G, Smith G A & Howell E A (198313):Eye-movements in dyslexic children: comments on Pavlidis reply. Br J Psychol 74: 195-197. Stein J F & Fowler S (1984): A physiologica theory of visual dyslexia. In: Rose F (ed). ‘Advances in Neurology. Progress in Aphasiology’, Raven Press, New York 42: 233-246.

Stein J F & Fowler S (1985):Effect of monocular occlusion on visuomotor perception and reading in dyslexic children. Lancet 2: 69-73. Stein J F, Ridell P M & Fowler M S (1987):Fine binocular control in dyslexic children. Eye 1:433-438. Stein J F, Ridell P M & Fowler S (1988): Disordered vergence control in dyslexic children. Br J Ophthalmol 72(3): 162-166. Stein J (1989): Visuospatial perception and reading problems. Irish J Psychol 10: 521-533. Stone J (1983): Parallel Processing in the Visual System. New York Plenum Press. Vellutino F (1987): Dyslexia. Scientific American 256(3): 20-27. Whyte J (1989): Dyslexia: Current research issues. Irish J. Psycho1 10: 465-677. Williams S M, Sanderson G F, Share D L & Silva P A (1988): Refractive error, IQ and reading ability: A longitudinal study from age seven to 11. Develop Med and Child Neurology 30: 735-742. Williams M C, Molinet K & LeCluyse K (1989): Visual masking as a measure of temporal processing in normal and disabled readers. Clin Vision Sci 4(2): 137-144. Williams M C, LeCluyse K & Bologna N (1990): Masking by light as a measure of visual integration time in normal and disabled readers. Clin Vision Sci 5(4):335-343. Winters R L, Patterson R & Shuntz W (1989): Visual persistence and adult dyslexia.J Learn Disabil22: 641-645. Volkman F C (1976): Saccadic suppression: a brief review. In: Monty RA & Senders J A (eds). Eye movements and psychological processes. Hillsdale: Lawrence Erblaum ASS.

von Euler C, Lundberg I & Lennerstrand G (eds) (1989): Brain and Reading. Structural and functional anomalies in developmental dyslexia with special reference to hemispheric interactions, memory functions, linguestic progress and visual analysis in reading. Macmillan Press, London. Ygge J, Lennerstrand G, Axelsson I, Rydberg A, Wijecoon S & Pettersson B M (1992):Visual functions and oculomotor functions in a Swedish population of 9-year-old dyslexics and normal readers. A preliminary report. Acta Ophthalmol (in press). Zangwill 0 L & Blakemore C (1972):Dyslexia: Reversal of eye-movements during reading. Neuropsychologia 10: 371-373.

Received on September 26th, 1991. Author’s address:

Gunnar Lennerstrand, MD, and Jan Ygge, MD, Department of Ophthalmology, Karolinska Institute, Huddinge University Hospital, S-141 86 Huddinge, Sweden.

13