Dyslexia: a review of two theories


Krithika Nandakumar
School of Optometry
University of Waterloo
200, University avenue West
Waterloo, ON. N2L 3G1
E-mail: k2nandak@uwaterloo.ca


Optometrists will frequently see patients, who may have a diagnosis or a suspected diagnosis of dyslexia (specific reading disorder) and will need to manage and counsel such patients. There are many propounded theories on the cause(s) of dyslexia. Although most professionals in this area consider that dyslexia is chiefly a linguistic disorder, the possibility of a visual component is contentious. This article is a selective review of two commonly discussed theories that suggest a visual component in dyslexia; the magnocellular deficit theory and Meares-Irlen syndrome.

Dyslexia or reading disorder is defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) as ‘reading achievement (that is, reading accuracy, speed or comprehension as measured by individually administered standardised tests) that falls substantially below that expected given the individual's chronological age, measured intelligence and age-appropriate education. The disturbance in reading significantly interferes with academic achievement … If a sensory deficit is present, the reading difficulties are in excess of those usually associated with it.’ According to the British Dyslexia Association (BDA), dyslexia is a ‘combination of abilities and difficulties that affect the learning process in one or more of reading, spelling and writing. Accompanying weaknesses may be identified in areas of speed of processing, short-term memory, organisation, sequencing, spoken language and motor skills. There may be difficulties with auditory and/or visual perception. It is particularly related to mastering and using written language, which may include alphabetic, numeric and musical notation. Dyslexia can occur despite normal intellectual ability and educational opportunity. It is constitutional in origin, part of one's make-up and independent of socio-economic status or language background.’1 Although the BDA's definition does not explicitly state the cause of dyslexia, it reflects an important assumption about its cause. It suggests that dyslexia is caused by a general mechanism, which affects language, perception and motor skills. It is a persistent condition. Commonly it is seen that these children fall significantly behind in reading/literacy skills compared to the performance that would be expected from age and IQ2,3 irrespective of language.4

There is a distinction made here between children who have a specific reading disorder (dyslexia) and those with reading difficulties or non-specific reading difficulty. Non-specific reading difficulty may occur from a wide range of causes such as ADHD (attention deficit hyperactivity disorder), lower verbal IQ, poor role models, socio-economic factors and sensory deficits (for example, poor hearing),3,5,6 therefore, these factors must be determined and corrected or treated (as far as is possible) before a true dyslexia can be diagnosed.

In the definition of dyslexia, there is some consensus that dyslexia describes differences in the way in which the brain processes information but this does not imply any abnormality, disease or defect, however, experts are not agreed on its underlying causes and components. Many consider that a phonological deficit is the cause of dyslexia and that there is no significant visual cause.3,7–10 In recent years, a number of researchers have suggested that the auditory and visual difficulties in dyslexic individuals are caused by abnormal magnocellular pathway development.11–13 Another commonly quoted theory is that dyslexia is sometimes associated with a visual condition called Meares-Irlen syndrome/visual stress (MISVS).14–16

In this article, we will review the evidence for and against these two theories of the cause of dyslexia, which involve vision and therefore, are of interest and concern to optometrists.


The lateral geniculate nucleus (LGN) comprises six layers, arranged in two major divisions. The upper division, comprising four layers of smaller cells, is the parvocellular layer, and the lower division has two layers of larger cells and is called the magnocellular layer.12,17 The parvocellular (or sustained) system receives input from cones and therefore mainly the central retina. It runs from retinal ganglion cells to the smaller cell bodies of the four dorsal layers of the LGN. It projects via the visual cortex and the ventral or ‘what’ stream to the infero-temporal cortex and mediates colour vision and detection of fine spatial details. The magnocellular pathway receives input from both rods and cones across the retina and extends from ganglion cells in the retina to the two ventral layers of the LGN. This system is more sensitive to lower contrast and higher temporal frequency but less sensitive to higher spatial frequency than the parvocellular system. Most recently, another distinctive layer of neurons has been found in the LGN, the K-layer or the koniocellular layer.18 The function of the K-cells or the konio cells is still debated and a subject for research, though they are believed to be involved in at least some form of colour vision. Parvo cells constitute approximately 80 per cent of the cells in the LGN nucleus, magnocellular cells about 10 per cent and koniocellular cells about 10 per cent.17 Thus, the three streams flow separately from the retina to the cortex and their output is a blend to form a unified perception.


This is a selective review of articles searched from Google scholar and PubMED using the search key words in Table 1. The works of certain authors were then traced backwards and forwards in time and were reviewed.

Table 1. Search words used for this selective review
Primary search wordSecondary search word
Magnocellular pathway deficitDyslexia
Coloured overlays
Reading disability
Reading disabilityMeares-Irlen syndrome
Precision tints
Coloured overlaysScotopic sensitivity syndrome
Specific reading disability
Randomised control

Deficit of the magnocellular (or transient) system

It has been suggested that as many as 75 per cent of dyslexics have a defect in their main magnocellular pathway.19 Evidence in favour of a magnocellular association comes from several approaches. We shall review a few methods that have been used to study this association.

Psychophysical studies have measured contrast sensitivity at different spatial and temporal frequencies.20–22 For example, if deficits are found in people with reading disabilities for high temporal and low spatial frequencies (but no deficit for low temporal and high spatial frequencies) this would be seen as showing a deficit in the magnocellular pathway.23,24 Other studies have used motion perception, speed matching or speed discrimination tasks,21,25,26 however, there is the question of internal consistency of measures of M pathway, that is, are measures of M-pathway function actually measuring the same function? A few studies using different methods to assess/detect the M-pathway function show some consistency in identifying dyslexics. Talcott and colleagues13 assessed magnocellular sensitivity by coherent motion perception and critical flicker fusion frequency and found a deficient in a majority of adult dyslexics. Their results showed that dyslexics were less sensitive than controls both for coherent motion in random dot kinetic stimuli and for detecting high frequency flicker (CFF) and these measures showed a moderate correlation with one another (r = -0.494, p < 0.05). Evans, Drasdo and Richards24 investigated the use of two different paradigms supposedly to measure transient visual function, namely, low spatial frequency contrast sensitivity and flicker detection on the same subject group and reported them to be weakly correlated (r = 0.265), indicating that the variables did not assess the same function.

VEP studies also confirm a deficit in dyslexic individuals.22,27 Some studies have suggested that the presence of a magnocellular deficit occurs in certain types of dyslexia and not others.11 Neuro-imaging has also shown that cells in the magnocellular layers of the LGN of dyslexic subjects are smaller than cells from brains of normal readers.9,22 There is evidence from fMRI studies of brains of some dyslexics, showing that there are strong relationships between the M-pathway, visual motion perception and reading ability28 and that the brains of some have subtle neurological abnormalities in the so-called magnocellular pathways. Studies on colour have also implicated the role of the M-pathway in text perception.29

Many studies do not support the magnocellular deficit theory. Just as many (if not more) psychophysical studies of contrast sensitivity find that the contrast sensitivity deficits do not fit the predicted pattern. When researchers have selected stimulus conditions so that only (or mainly), the M-system is activated, they have not been able to unequivocally demonstrate a deficit on the part of dyslexic readers.30,31 Dyslexic subjects were less sensitive at the higher spatial frequencies, but showed no difference from control participants at lower spatial frequencies.30,32–34

Thus, we see that the evidence of an association is inconclusive at best. Even if a clear link is found, it does not necessarily imply causality; both dyslexia and an M-pathway deficit may be related to another, more fundamental cause, that is, they are non-causal correlates.24 Evans, Drasdo and Richards24 studied similar links in a group of dyslexic and non-dyslexic individuals, who were subjected to a simulated reading visual search task (SRVST), measurement of flicker thresholds and contrast sensitivity. The dyslexics showed poorer performance in both SRVST and flicker thresholds and an apparent lower sensitivity for low spatial frequencies. There were, however, low, mostly non-significant correlations between the flicker thresholds and contrast sensitivity and there was no significant correlation between flicker thresholds and SRVST. Therefore, it seems that the two measures of ‘transient function’ are not measuring the same function and Evans, Drasdo and richards24 argue that there is little evidence for a transient deficit being a causal factor of impaired reading, although it may still be found to be a correlate. Instead the SRVST deficit was discussed in terms of impaired visual sequential memory (VSM).

Several authors have suggested theories that explain how a magnocellular deficit might cause dyslexia. The saccadic suppression theory,35 which is now discredited, was one of the initial theories and suggested that the sustained system inhibited the transient system as a result of saccadic suppression. If this were correct, it would be expected that it would influence many activities in addition to reading. Presumably, any activities that involve eye movements from one scene to another would be affected by the lack of inhibition of the parvocellular mechanism. Another explanation (transient pathway deficit) is that the M-pathway modulates visual attention in some way thus causing impaired reading.12,36–39 Yet another suggestion is that defects in rapid processing may not be limited to only the visual pathways but to auditory processing as well.40

Thus, it seems that there is not enough evidence to fully validate this theory and more studies are required to help us understand the magnocellular and parvocellular pathways and their interaction and functions in the process of reading. Currently, there is little treatment based on this theory, although Solan and associates41 suggested training for the magnocellular system. There is a more recent suggestion of the possibility of a link between the use of coloured filters for reading and the magnocellular deficit theory, which moves us into the consideration of Meares-Irlen syndrome.


The first published paper on this syndrome was in 1980 by Olive Meares,42 in which she described symptoms of eyestrain, spatial distortions and headache, experienced while reading and which can be alleviated by using coloured overlays or coloured filters. Educational psychologist Helen Irlen later described similar symptoms, while she was working with adult learners in the early 1980s and which were also found to be more prevalent in people with dyslexia. She coined the terms ‘Irlen syndrome’ and ‘scotopic sensitivity syndrome (SSS)’ but there was little published explanation, assessment or treatment protocols for this perceptual disorder, until it was described in her book, Reading by the Colors.15 The premise of Irlen's method was that SSS, which causes perceptual distortion, could be an obstacle to reading and learning. Coloured overlays and lenses can be a remedial intervention in such cases. According to Irlen, 12 to 14 per cent of the population suffers from Irlen syndrome and this figure rises to 46 per cent of those diagnosed with dyslexia, attention deficit disorders and learning difficulties.43 Some good readers may have similar symptoms, which may be associated with a history or family history of migraine.44 Individuals with MISVS are described as seeing letters moving on the page, blurring or forming strange patterns. In less severe cases, the misperceptions are said not to occur or to be suppressed. These symptoms generally get worse the longer a person tries to read or do other visually intensive activities. Bright lights, fluorescent lights or glossy paper are described as often making the symptoms worse, as they increase the problem of persistent images.

MISVS is thought to manifest itself most strongly when reading words or music because of the repetitive patterns on the page. When the eyes scan across the page, the patterns of words on the page and persistent images will jumble in a manner that is difficult for the brain to interpret properly.

Irlen developed two intervention methods:

  • 1the use of coloured transparencies or overlays (to improve reading).
  • 2tinted lenses (to improve visual perception of the environment, including reading).

It has been suggested that these interventions may also be helpful for individuals who experience other related problems, such as faulty depth perception or night driving difficulties.45 In the Irlen method, the individual is assessed with a wide array of coloured filters, singly and in combination, to find the most suitable colour. The colour is selected by subjective preference.46,47 Wilkins and co-workers48 developed a colorimeter for this purpose.

The theoretical basis of the SSS is not particularly clear, although initially Irlen hypothesised that the visual perceptual disorder is due to the excessive sensitivity of the retina to particular frequencies of the visual spectrum, and this varies with different individuals. At a similar time, Wilkins and co-workers48 were studying visual discomfort in relation to headaches and used the term ‘cortical hyperexcitability’. According to Wilkins's hypothesis49‘the perceptual distortions of text occur because a spread of excitation causes visual neurones to fire inappropriately. The degree of susceptibility to distortions increases with, and therefore reflects, the degree of cortical hyperexcitability, which is said to be non-uniform.’ Wilkins hypothesises that the tints help to redistribute the cortical excitation that occurs in response to a visual stimulus. For people with MISVS, the brain is described as expending more energy in processing the images than is required by most people, resulting in headaches, eyestrain and/or fatigue. We also see that the choice of the term ‘scotopic’ is perhaps not appropriate to describe the syndrome, as the phenomenon has more to do with glare and sensitivity at photopic levels.

Evidence for MISVS

Two approaches to the study of MISVS have been used. First, if people with the syndrome have different visual sensitivities, presumably some differences in visual function should be measurable with the application of the appropriate test. Second, as interventions have been described for MISVS, the validity of the theory can be studied by looking at the efficacy of the intervention and the ideal method to do this would be by using randomised clinical trials (RDT). We will consider the evidence from both these approaches.


Numerous studies describe the visual correlates of MISVS. Evans and co-workers50 found certain ocular motor factors to be correlates of dyslexia, namely, reduced vergence, accommodative amplitudes and stereo-acuity. They suggested that pattern-glare may be involved in the mechanism of MISVS. Two studies of school children showed that more than one-third chose to use an overlay and read more quickly with it than without. In this study,50 numerous optometric measures were compared for those who continued to use overlays and those who did not. The only functions that were significantly different between the groups were binocular amplitude of accommodation, distance fusional reserve break points and aligning prism in the near Mallet fixation disparity test. In their statistics, there was no consideration of multiple comparisons but they concluded that binocular and accommodative anomalies did not appear to be the underlying mechanism for benefit from coloured lenses in most cases.51 The binocular and accommodative anomalies were thought to be separate factors (correlates), contributing to the reading difficulty. Using psychophysical measures, Simmers and colleagues52 found no significant difference in visual function between the subjects with and without MISVS. Similarly, Lopez and associates53 reported no significant relationships between preferred overlay colour and any of the visual functions that they measured, however, there was a significant tendency for subjects with higher levels of the syndrome to be candidates for vision therapy and to derive greater benefit from the coloured overlays. No significant relationships were found between academic classification and degree of MISVS.51

Several studies have shown an increased prevalence of pattern glare among those with MISVS.16,24,54,55 In one study comparing dyslexics and controls using coloured overlays, prevalence of MISVS was similar in both groups, except that it was slightly higher in dyslexics (odds ratio for a criterion of greater than five per cent: 2.6, 95% confidence limits 0.9 to 7.3). They conclude that MISVS and dyslexia are separate entities that are to be detected and treated in different ways and that dyslexic children benefit more from overlays than non-dyslexics.56


There have been numerous studies using both coloured overlays and precision tinted lenses on the effects of reading speed. According to Wilkins, Sihra and Myers,57 the effects of colour on reading fluency are specific and different for each individual but the precise tint for an overlay is not necessarily the same as that for tinted spectacles.58 Wilkins, Sihra and Smith59 studied the precision of the tints and the number of tints required to determine the exact tint for a patient and propose that the number of trial tints needs to exceed 1,000. Wilkins60 reported that five per cent of children in mainstream education read at least 25 per cent more quickly with an overlay, provided they have chosen the colour. In two studies described by Scott and co-workers,51 more than one-third of the children chose to use an overlay and read more quickly with it than without. Even though these subjects may require vision therapy, they may also benefit from the use of a coloured filter, that is, the MISVS and the binocular and accommodative anomalies were separate factors contributing to the reading difficulty.

Chan and Robinson61 studied the effect of tinted lenses in children with reading disability. Although they found improvements with the lenses, this could not be attributed to the intervention by tinted lenses alone, as the disabled readers were not randomly assigned to the tinted lenses. Alternatively, Winter62 reported that the school records of standardised tests for children who had been prescribed tinted lenses did not reveal any beneficial changes in academic performance after the lenses were prescribed. Menacker and colleagues63 did not find improvements in reading speed for non-specific coloured lenses. The long-term benefits of these tints for specific learning disability have been studied and more than 70 per cent of those who were prescribed precision tinted lenses were wearing them daily 18 months after being prescribed.64

The studies described so far were not randomised clinical trials and therefore cannot give reliable information of whether any measured improvements are due to the tint or a placebo effect. In clinical management, a placebo effect from the use of a tint can be positive. Children are given a reason for their difficulty with reading, that is, they have a visual problem rather than a lack of intelligence. This can give children a fresh start and a new belief that they have the ability to learn to read.65 In research studies of the effectiveness of tints, it is important to eliminate any placebo effect by proper randomisation. The trial should also be double-blinded, that is, neither the participant nor the experimenter who takes the measurements should know whether the participant has been assigned to the intervention or the control group. This is difficult to do with tints. Wilkins and co-workers66 claim that the subjects were not aware of which was the true tint and which was the control.

There are a few randomised clinical trials evaluating coloured lenses. A randomised (but not masked) pilot study by Blaskey and associates67 investigated the effectiveness of Irlen filters for improving comfort and reading performance compared to traditional optometric intervention. The subjects were randomly placed in a coloured filter treatment group, a vision therapy treatment group or a control group. The subjects in the coloured filter group did not show significant gains in reading rate, word recognition or comprehension. Wilkins and co-workers66 undertook a double-masked placebo-controlled trial, in which tints similar to the tint identified by the colorimeter were used as controls. Half the children with MISVS type symptoms were randomly assigned to wear the control tint for one month followed by the true tint, while the other half wore the tints in the reverse order. They were asked to record their symptoms during the whole period. A number of children recorded significantly fewer symptoms while wearing the true tints.

Robinson and Foreman68 also used a double-masked, placebo-controlled crossover experimental design, with subjects being assessed over a period of 20 months. There was a significant improvement for all groups in reading accuracy over the period and the intervention group did not improve at a significantly different rate from the control group. They explained the failure to find a significant improvement for the experimental groups over the control group, despite subjects’ reports of improved print clarity, as being related to linguistic difficulties that were also present, that is, they had difficulty reading aloud. Analysing data from their previous study, Robinson and Foreman69 reported the effect on reading speed, accuracy and comprehension as well as perception of academic ability. The treatment groups (blue and true tints) increased at a significantly greater rate than the control group (no tint) in reading comprehension and there was a further increase when those with the blue tint crossed over to the correct tint. This shows a strong placebo effect but also an effect of the correct tint. Speed of reading and reading accuracy did not show such clear trends. In another study by Bouldoukian, Wilkins and Evans,70 the effect of individually prescribed coloured overlays on the rate of reading was investigated. The subjects carried out the Wilkins Rate of Reading Test (which assesses visual rather than linguistic factors) under two conditions, namely, with their chosen coloured overlay and with a control filter. Steps were taken to ensure that a strong placebo effect was associated with the control overlay (that is, the subjects were convinced that the control overlay might be as good as the true tint) and, when asked which they preferred, subjects were not significantly more likely to prefer their coloured overlay than the control filter. Nonetheless, the rate of reading was significantly faster with the correct coloured overlay than with the control. To conclude from these randomised clinical trials, there may be some genuine effect (which goes beyond a placebo effect) of the correct tint (two out of three groups of researchers finding a significant effect).

The mechanism of MISVS

The mechanism by which these filters work is still controversial. Pattern glare is more commonly observed in dyslexics than normals54 and in people with MISVS, migraine or seizures.44,70,71 In one experiment, Evans and associates54 used a placebo-controlled paradigm to investigate the effect of pattern glare and coloured overlays on performance at a simulated reading visual search task. The results showed a border-line significant effect to support the conclusion that coloured filters improve reading through ameliorating pattern glare. They suggest that the tints help by lowering the activity in certain sets of neurons tuned for wavelength.

Evans, Patel and Wilkins72 found that a group with visually sensitive migraine was more prone to pattern glare (p = 0.004) than a control group and pattern glare was the only variable to show a consistent and marked improvement with tinted lenses. According to Wilkins,49 the benefit from the prescribed coloured filters in migraine is the alleviation of cortical excitability and in another randomised control study,55 the frequency of migraine headaches was much lower with the optimal tint than a control tint. A recent study on VER in children with visual stress found a larger amplitude VER response in children with symptoms of severe headaches or migraine, when wearing the prescribed colour tint compared to children without symptoms of headaches and without the tint.73

There is also a suggestion of a link between the magnocellular deficit theory and MISVS. It is now known that, although the magnocellular system does not contribute to the perception of colour, nevertheless in daylight, M-ganglion cells probably receive summed input from all three classes of cone roughly in proportion to their density in the retina, that is, 45 per cent from L- (red), 45 per cent from M- (green) and 10 per cent from S- (blue) cones.74 Coloured filters are described as blocking some of the light, which would normally activate the parvocellular pathway. About 10 per cent of children with visual reading problems, particularly those complaining of glare, were not helped by yellow filters but were helped by deep blue (‘negative yellow’).

According to Stein,75 this small group of children seems to have weaker S-cone input to M-cells than normal, so that blue filters probably help them by cutting down the relative stimulation of their L- and M-cones and this may help the system to return to an appropriate balance of L-, M- and S-inputs. Results from experiments involving colour and reading show that red light impairs reading performance under normal luminance and contrast conditions. Alternatively, an isoluminant coloured text, designed to selectively activate the parvocellular pathway, was found to be easier to read under red light. Chase and associates29 used these apparently discrepant results to suggest that the magnocellular pathway is the dominant visual pathway for text perception and also argue that the magnocellular pathway deficits are responsible for beneficial effects of coloured lenses. Yellow filters have been suggested to increase motion sensitivity, convergence and accommodation in children with reading deficits,76 although Ciuffreda and co-workers77 did not find any improvement in accommodation accuracy with coloured spectacle lenses. Thus, the link between reading disability and M-pathway deficits is still uncertain.

Conclusions regarding the two theories

In appropriately selected patients, individually prescribed coloured filters may have a beneficial effect on symptoms as well as reading performance, though many questions about the use of coloured filters for people with reading problems remain unanswered. Among these are questions dealing with whether the M-pathway anomalies found in some dyslexics are also present to a lesser degree in other cases of reading disability. Meanwhile this method of prescribing coloured overlays or filters is practised in many parts of the world. Hopefully, ongoing research will give clearer evidence in support (or against) these theories and a clearer explanation of the mechanism by which the filters are effective.


Returning to the broader picture regarding dyslexia, it is generally agreed that there is no single underlying cause for reading disability. In most individuals, there are several causative or contributory factors. The consensus is that visual factors are contributory, rather than causative.3 Thus, there is a need for a psycho-educational (or educational) assessment, which includes a full investigation into the individual's social, medical, sensory and academic standing. Tests of IQ, perceptual skills, attention and memory are included as well as assessments of behavioural dimensions. The purpose is to identify the various factors that may be causative or contributory to the reading difficulty in an individual.

As optometrists, our first responsibility is to undertake a full and comprehensive eye examination. This should include assessments of accommodation, binocular vision and possibly a cycloplegic refraction. The optometrist should prescribe the usual therapy or correction for any eye disorder that may be present,3 including refractive errors, accommodative dysfunction, fusional vergences, heterophorias et cetera that have been shown to be non-causal correlates in dyslexia. A comprehensive educational or psycho-educational assessment should be recommended, if this is not already in place. Optometrists should remember that most cases of dyslexia are linguistic in their foundations and the existence of contributory or causative factors of a reading difficulty such as attention disorders, emotional or behavioural problems or social factors should be ruled out. The optometrist's input into the psycho-educational assessment is a full report of visual function. The primary purpose of all this assessment is to develop a ‘long-term’ remedial plan to help an individual with dyslexia, regardless of the origin of his or her reading difficulties. Special instruction modules could be used on individuals with specified reasons for their reading disabilities. Still most experts agree that the primary treatment for dyslexia (once contributory factors such as visual disorders and other causes have been eliminated or treated) is more instruction in reading in the early grades and educational ‘accommodations’ in the later grades. Accommodations are the application of alternative methods of learning, so that the child is not held back in other academic areas because of a reading disorder.

In cases where no other visual disorder is present or when other visual disorders have already been corrected, the use of coloured filters may be considered. Coloured overlays are available from The Institute of Optometry, London.78 As there is a strong placebo effect (which may be useful in itself), we recommend that part of the counselling to the parents should be undertaken without the child present, so that this valuable placebo effect may not be lost. Inexpensive coloured overlays should be tried in the first instance to determine whether individually prescribed coloured lenses may be beneficial. It must be noted that the colour of the overlays does not give a reliable guide to the optimal lens colour.58