Large-scale neural correlates of developmental dyslexia

European Child & Adolescent Psychiatry (2004) 13:125–140 DOI 10.1007/s00787-004-0361-7

Sabine Heim Andreas Keil

■ Abstract Recent work in the field of developmental dyslexia has emphasized the large-scale neural

Accepted: 3 July 2003 Dr. S. Heim (
) · Dr. A. Keil Department of Psychology University of Konstanz PO Box D23 78457 Konstanz, Germany E-Mail: [email protected] [email protected] Dr. S. Heim Center for Molecular and Behavioral Neuroscience Rutgers, The State University of New Jersey 197 University Avenue Newark, New Jersey 07102, USA

REVIEW

Large-scale neural correlates of developmental dyslexia

aspects of the disorder as measured by means of contemporary imaging techniques, electrophysiology, and post-mortem analyses. This article presents findings from these research domains and comprehensively reviews their relevance with respect to the behavioral and cognitive profiles of dyslexia. Large-scale alterations were observed in the perisylvian region across paradigms. Convergent evidence has also been reported in terms of hemispheric balance. Specifically, deviances in cerebral asymmetry associated with atypical organization of the left hemisphere were found in both children and adults with dyslexia.

Introduction

■ Key words dyslexia – magnetoencephalography – electroencephalography – neuroimaging – postmortem studies – temporal dynamics

with academic achievement or activities of daily living requiring reading or spelling skills. In addition to difficulties in the literacy domain, dyslexia may be associated with psychosocial problems, abnormalities in cognitive processing, and clinically relevant conditions (cf. DSM-IV, ICD-10). Deficits in cognitive processing that often precede or are associated with dyslexia include inter alia poor visual discrimination, weakness in auditory segmenting, limitations in working memory, linguistic disturbances (e. g., misarticulation of sounds, impairment in receptive and/or expressive language abilities), or a combination of these. The disorder is often associated with a higher rate of attention-deficit/hyperactivity disorder, emotional disorders, or developmental coordination disorder (cf. DSM-IV, ICD-10). Thus, the heterogeneity of dyslexia at the phenotype level represents a challenge for subject recruitment and interpretation of experimental results.

ECAP 361

Developmental dyslexia is a language-based learning disorder that affects an individual’s written language skills. Its prevalence rates have been estimated to vary from 3 to 10 % (e. g., [39, 101, 111]) exemplifying the epidemiological validity of the condition. Dyslexia has often been defined on the basis of a specific reading disorder [1]; (DSM-IV: 315.00) or as a combined specific reading and spelling disorder [125]; (ICD-10: F81.0).According to standard definitions, dyslexia is a disability in learning to read, spell, and write despite normal intellectual capacity and educational resources, as well as adequate sociocultural opportunities. At the same time sensory deficits, neurological pathology, and other impediments to attaining literacy skills are absent. Disturbances in reading and spelling significantly interfere

Emerging research encompassing high-temporal resolution methods such as magnetoencephalography (MEG) suggests right-hemisphere involvement and points to the complexity of the developmental disorder. A combined approach of structural imaging and MEG, and most importantly theory driven behavioral tasks may shed light on dynamics and trajectories of the neurobiology of dyslexia.

126

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

Although its biological basis is still under debate, dyslexia is believed to reflect neurological difficulties and tends to run down generations (for reviews see [86, 87, 105]). During the past 30 years various neurobiological correlates of dyslexia have been suggested by numerous research teams. It is conceivable that the various deficits making up the complex disorder might be associated with different neural bases. There has been a spurt in literature monitoring neurophysiological correlates of intervention techniques. These studies might have important implications in understanding the neural underpinnings of dyslexia and are discussed in detail elsewhere [45]. In what follows, we review findings from post-mortem, neuroimaging, and electrophysiological studies contrasting individuals with dyslexia and normal controls. For the purpose of this article, we focus on data allowing for the integration of neuroanatomical structure and temporal dynamics.

Post-mortem studies Galaburda and colleagues have performed the only comprehensive post-mortem studies to date of diagnosed cases of developmental dyslexia [28, 31, 32, 54, 58, 75]. All five cases (one female and four males) showed evidence of small areas of cortical dysgenesis including ectopias (small nests of abnormally placed neurons) and dysplasia (focally distorted cortical lamination). The dysgenesis varied in number and location from brain to brain and tended to involve the language-relevant perisylvian cortex. Furthermore, the structural deviances tended to be lateralized to the left hemisphere in male specimens [28],whereas in the female brain a fairly symmetric distribution was observed [54]; (NB: specimen ORT-20–87). Because ectopias or dysplasias are found only rarely in routine autopsy analyses (or in other developmental disorders), usually omit perisylvian regions, and are located more frequently in the right side of the brain than in the left, Galaburda [24–26, 114] considered the malformations to be specifically associated with dyslexia. The observed dysgenesis might reflect neuronal migration errors that may have occurred during fetal development. Changes in the pattern of hemispheric asymmetry of the planum temporale also were seen in dyslexic brains. The planum temporale is a triangular landmark situated on the supratemporal surface just posterior to the first Heschl’s gyrus, inside the Sylvian fissure. The left planum coincides with part of Wernicke’s speech comprehension area (e. g., [26, 110]). Large post-mortem studies [36, 124] including 100 normal adult brains found that it was symmetrically sized between the hemispheres in 16 %, whereas 10.5 % showed a rightward asymmetry and 73.5 % a leftward. Corresponding figures reported on 307 ordinary fetal or neonatal speci-

mens were 29 %, 16 %, and 54 % [12, 124]. Consequently, the planum temporale is thought to be an important substrate of left-hemispheric language lateralization [36, 124]. Going back to the five dyslexic brains [28, 54], none was reported to show the typical planar asymmetry favoring the left side; instead these autopsy specimens exhibit the symmetrical type due to an enlarged right-hemisphere planum. Galaburda et al. [29] have proposed that symmetry reflects reduced cell death in the right planum temporale during late fetal development, which leads to enhanced survival of neurons, forming improper connections and resulting in a redefinition of the cortical architecture. Another set of post-mortem examinations was performed on thalamic structures, i. e., the lateral geniculate nucleus (LGN) of the visual pathway and the medial geniculate nucleus (MGN) of the auditory pathway. The magnocellular layers of the LGN were found to be more disorganized in dyslexic than in non-dyslexic brains [75]. Furthermore, magnocell bodies were on average 27 % smaller and appeared more variable in size and shape in the brain of dyslexic individuals relative to those of controls. Neither the parvocellular lamination nor the parvocell sizes of the LGN differed between the population specimens. In the auditory system, Galaburda et al. [32] reported significantly smaller MGN neurons on the left side compared with the right in the same dyslexic autopsy specimens. No hemispheric asymmetry in MGN neuronal size was observed in ordinary brains. In addition, brains of dyslexic individuals were said to exhibit a relative excess of small neurons and a relative paucity of large neurons on the left side as compared to control brains. Galaburda et al. [33] claim that the structural deviances found in the LGN of dyslexic brains may be at the core of slowness in early segments of the magnocellular channels, whereas the MGN differences may relate to the auditory temporal processing abnormalities described in language-impaired children. Autopsy data on neuronal tissue in the primary visual cortex (area 17) were presented in a recent work by Jenner et al. [58]. In contrast to the atypical organization in the magnocellular layers of the LGN, the (five) dyslexic brains did not show consistent changes in the size of cortical neurons receiving thalamic magno input. The researchers suggest that this inconsistency may in part be due to blending of magnocellular and parvocellular pathways or functional effects of cortico-cortical topdown projections.On the other hand,another example of changes in hemispheric asymmetry similar to that of the planum temporale was observed. That is, brains of nondyslexic individuals comprised larger neurons in the left hemisphere than in the right, whereas dyslexic brains showed no lateralization. According to Jenner et al. [58], the neuronal symmetry in primary visual cortex is associated with abnormality in circuits involved in reading.

S. Heim et al. Neural correlates of dyslexia

To date, Galaburda’s group has presented autopsy data on nine brains of individuals (six males and three females) with a history of developmental dyslexia1. Three of the male and one of the female patients were reported to have histories of delayed language acquisition [28, 54]. All dyslexic brains have displayed evidence of symmetric plana temporali [24, 25, 54]. Neuronal ectopias and architectonic dysplasias were observed in all male cases and two of the females [26]. Other cerebrocortical deviances in dyslexic autopsy specimens such as microgyria and cortical scars were less uniform than the pattern of dysgenesis [26]. Overall, dyslexic female brains showed fewer and differently located microcortical malformations when compared to male brains [54]. Histological differences in thalamic structures and the primary visual cortex are hitherto limited to reports on five dyslexic brains versus five [58, 75] or seven control brains [32]. In interpreting the study results, Galaburda [24–26, 33] has hypothesized that dyslexia is an outcome of anomalous neural development, which might derive from brain injury during the prenatal stage. Here, the chemical environment and maturation rate of relevant brain areas are assumed to interact. Despite the robustness of most of the findings provided by Galaburda and colleagues, there are methodological issues complicating the interpretation of the results. Many subjects with dyslexia had a history of comorbid disorders or prior head injuries which would have prevented their participation in neuroimaging studies [28, 54]. Another concern may be seen in storage duration of post-mortem brains. It is conceivable that the brains of dyslexic subjects have been stored for a longer period of time than those of the control subjects putting them at higher risk of cell shrinkage2. Furthermore, the number of autopsy specimens examined so far is small. In post-mortem studies, reliable identification of microanatomical deviances in general and the boundaries of the planum temporale in particular has often proved to be difficult (e. g., [60, 110]).

Neuroimaging studies Based on neurobiological and cognitive theories, structural as well as functional brain-imaging studies in people with dyslexia have focused on areas subserving language functions. Because findings of atypical cortical asymmetry in known language regions may be related to deviances in interhemispheric transfer of information, the morphology of the corpus callosum has been another point of interest.

1 Figures were taken from various journals and books indexed by the bibliography databases Medline and Psychinfo. 2 We would like to thank an anonymous referee for this comment.

127

■ Structural neuroimaging Magnetic resonance imaging (MRI) studies have shown that individuals with dyslexia have a higher incidence of reduced or reversed asymmetry of temporo-parietal language regions than the normal population [13, 17, 56, 68, 70, 92, 95]. Similar to the post-mortem findings by Galaburda and colleagues, in vivo MRI studies demonstrated unusual asymmetry (i. e., right = left or right > left) of the planum temporale in people with dyslexia [23, 56, 70]. Larsen et al. [70] found that 13 out of 19 dyslexic adolescents displayed symmetric plana as compared to only 5 out of 17 normal readers.Among the dyslexic readers exhibiting “pure phonological dysfunction”([70], p. 297), all showed absence of typical leftward asymmetry of the planum. This led the authors to hypothesize that symmetrical plana temporali might be a possible neurobiological substrate for phonological processing impairment in dyslexia. While in the studies of Larsen et al. [70] and Flowers [23] atypical asymmetry of the planum temporale was due to an increase in size on the right side (which is consistent with Galaburda’s postmortem results), Hynd et al. [56] have shown differences due to a shorter left planum length. More recent MRI research has challenged the view of altered planar asymmetry in dyslexia [5, 40, 72, 92, 100, 109]. For instance, Leonard et al. [72] reported an exaggerated leftward asymmetry in a small group of compensated dyslexics compared with unaffected relatives and controls. Best and Demb [5] observed that dyslexic adults with a magnocellular pathway deficit did not depart from the left-lateralized planum temporale type. According to Best and Demb, planar asymmetry may be associated with a subgroup of dyslexia. Structures of the perisylvian area other than the planum temporale have also been found to be different in dyslexia [40, 92]. Robichon et al. [92] demonstrated stronger right-hemisphere preponderance for Broca’s region in 16 adult male dyslexics compared to 14 controls. Heiervang et al. [40] in their study targeting primarily posterior language regions found no changes in the planum+ (which includes both planum temporale and planum parietale) lateralization. Analyzing the vertical part of the planum+, the posterior ascending ramus or so-called planum parietale, they found that dyslexic boys were less likely to show the expected rightward asymmetry than normally reading controls. Several studies in dyslexia reported no alterations in planum parietale asymmetry, however [72, 100]. Work on sulcal pattern morphology at perisylvian sites revealed no systematic relationship with diagnosis of developmental dyslexia [50, 92]. Examining the microstructure of temporo-parietal areas using diffusion tensor imaging, Klingberg et al. [61] observed reduced integrity of white matter tied to the left hemisphere as a function of reading impairment.

128

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

In summary, changes in perisylvian-language regions have been reported for dyslexia. The planum temporale has been the most prominent landmark investigated in dyslexia. While some studies indicated reduced or absent left-right planar asymmetry, others did not. The inconsistency of the studies examining the planum temporale may be attributed to several factors: 1) Research groups disagree on how to define the boundaries of the planum temporale; its structural ambiguity has led to imaging measurements of unidimensional lengths rather than surface area (for a review see [110]). 2) Different measurement techniques used to acquire images and to measure anatomical regions are associated with considerable variability in planar surface areas across studies. For example, Best and Demb [5] compared three measurement methods on the planum temporale in five dyslexic and five normally literate adults. The first two methods adopted from relevant MRI studies included the tissue between Heschl’s sulcus and the terminal upswing of the posterior ascending Sylvian ramus, though the second one took into account neither the shape of the planum nor the small sulci on its surface. The third method approximated those used in Galaburda’s post-mortem work revealing solely the bidimensional area on the superior surface of the temporal lobe. The results showed that both participant groups became less left-lateralized using the second and third procedure that exclude sulcal tissue to an increasing degree. 3) Variation in certain characteristics of the participants (e. g., handedness, gender, intellectual capacity, oral language skills, or socioeconomic background) across studies might obscure the relation between planar asymmetry and dyslexia [19], (for reviews see [3, 18, 69, 110]). For instance, given that non-right-handedness is related to reduced or reversed asymmetry of the planum temporale and given that most studies reported normal distributions of handedness within the dyslexic group, careful control for handedness is essential in imaging studies of dyslexia. 4) Finally, certain methodological flaws, such as small sample sizes, criteria used to define dyslexia, the heterogeneity of the disorder, and codiagnoses (e. g., attention-deficit/hyperactivity disorder) might also contribute to conflicting information regarding morphometric changes in this landmark (e. g., [18, 110]). One suggestion proposed to explain the observed reduction in cerebral asymmetry in dyslexia is that it might be the result of anomalous interhemispheric pathways coursing through the corpus callosum to the perisylvian-language regions [22]. The corpus callosum subserves communication and integration between the hemispheres and has been shown to be topographically organized with projections from specific cortical areas to specific callosal regions [14, 84]. Based on animal models, Galaburda’s group [30, 94] has hypothesized that more symmetric brains have a stronger interhemi-

spheric connectivity, which may be reflected by a larger size of the corpus callosum and vice versa. To date, there are only a few studies using MRI that compare the size of corpus callosum between individuals with dyslexia and non-dyslexic controls [17, 57, 71, 91, 98, 123]. Duara et al. [17] observed that the most posterior segment of the corpus callosum termed splenium was larger in a group of 21 dyslexic adults than in 29 controls. However, this effect was primarily accounted for by dyslexic female participants. In addition, both the (most anterior) genu area and the corpus callosum in general were larger in female than in male dyslexic adults. Larsen et al. [71] failed to find differences of the total callosal area or the splenium in a predominantly male sample of 19 dyslexic adolescents and 17 normal readers.They also reported no deviances in size of the corpus callosum in subgroups of dyslexia related to reading profile or symmetry/asymmetry of the planum temporale. Studying children, Hynd et al. [57] noted completely different results with the dyslexic group (n = 16) showing a smaller genu region than the equally-sized control group. Furthermore, moderate positive correlations were found between overall reading achievement and the (region of interest) measurements for the genu (r = 0.40) and splenium (r = 0.35) in these children. Rumsey et al. [98] reported an increase in the area of the posterior third of the corpus callosum – roughly corresponding to the splenium and its rostrally adjacent segment, the isthmus – in dyslexic men (n = 21). Likewise in a group of 16 adult male dyslexics, Robichon and Habib [91] showed a larger total callosal area, in particular in the isthmus but found that this result was accounted for by right-handed participants. Taken together, of the six studies quoted above three found an increase in size of the corpus callosum in adults with dyslexia, especially in the splenium [17, 98] and the isthmus [91, 98]. The isthmus contains fibers from the superior temporal and posterior parietal regions; the splenium involves all of the fibers connecting occipital cortex, but also links the superior parietal lobules and the temporo-parieto-occipital junctional area, the region including the planum temporale [14, 84]. Thus, these callosal segments are associated with posterior language regions, in which atypical cerebral asymmetries and other cytoarchitectonic deviances have been reported in dyslexia (e. g., [13, 23, 56, 68, 70, 72]). The studies of Larsen et al. [71] and Hynd et al. [57] on corpus callosum morphometry in dyslexia provide conflicting results, however. In explaining this, differences in subject characteristics (e. g., age, gender, handedness, comorbidity, or intellectual ability) as well as procedural variations in the methods used to acquire the scans and to define and measure the callosal subregions (of interest) may play an important role [3, 22, 69]. As repeatedly stressed, it is apparent that no consistent structural correlates have been associated with de-

S. Heim et al. Neural correlates of dyslexia

velopmental dyslexia. Several factors possibly accounting for the inconsistent findings have been outlined earlier. Because the relationship between neurostructural deviances and behavioral measures is under discussion, more insight has been expected from functional brainimaging methods.

■ Functional neuroimaging Various functional studies using positron emission tomography (PET) or functional magnetic resonance imaging (fMRI) have primarily targeted processes hypothesized to be compromised in dyslexia: phonological processing, auditory temporal processing, and visual motion perception (magnocellular system).

Brain activation during phonological tasks One of the first functional imaging studies examining phonological processing in dyslexia was that of Rumsey et al. [96]. PET scans were obtained from 14 adult male dyslexics and 14 controls while performing two tasks: a phonemic awareness task in which participants were asked to press a button if two auditorily presented words rhymed with each other, and a non-phonologic attentional task in which they were required to push a key whenever a target tone in a series of (simple) tones was detected. In controls, the left temporo-parietal cortex (angular/supramarginal gyrus) was activated during rhyme judgment but not during tone detection. Dyslexic subjects showed reduced blood flow in the left temporoparietal regions activated in controls while performing the phonological task but did not differ from controls in these regions during rest or attentional testing. Thus, dyslexic individuals demonstrated a left temporo-parietal dysfunction associated with phonological demands of the rhyming task. A subsequent PET study [85] employed two visually presented phonological tasks: a rhyming task (Does the letter rhyme with B?) and a short-term memory task (Was K among the last 6 letters you saw?). In normally literate men (n = 5; all right-handed), both tasks activated a number of perisylvian structures of the left hemisphere including Broca’s area, Wernicke’s area, and the insula, whereas parietal operculum activation was specific to the phonological memory task. In dyslexic men (n = 5; all right-handed), only a subset of brain regions normally involved in phonological processing was activated: Broca’s area during rhyme judgment, left temporo-parietal cortex during short-term memory demands, but the insula of the left hemisphere never. Paulesu and co-workers thought the left insular cortex to be crucial to convert whole-word phonology (temporo-parietal regions) to segmented phonology (inferior-frontal regions). They speculated that phonological

129

deficits in dyslexia may result from a weak connectivity between anterior and posterior language areas. The study of Paulesu et al. [85] supports the findings of Rumsey et al.[96] to the effect that they found reduced activity in the left temporo-parietal regions in dyslexic adults while performing simple rhyming tasks, but extended it in showing task-dependent activations of only a subset of left-hemispheric perisylvian-language areas. It deserves mention, however, that the two PET studies used different methodological approaches. While Rumsey et al. [96] employed a region of interest method, which is governed by preconceived anatomical considerations, Paulesu et al. [85] exploited whole-brain scanning and voxel-based image analysis permitting more detailed investigation of brain areas. Using whole-head PET scanning, Rumsey et al. [99] compared 17 right-handed dyslexic men and 14 non-impaired controls who performed two kinds of print tasks with stress on phonological or orthographic features. The first type of task, referred to as ‘pronunciation’ included (phonological) decoding of pseudowords (e. g., phalbap, chirl) and (orthographic) reading of low-frequency irregularly spelled words (e. g., pharaoh, choir). The second type of task involved decision making with a phonological instruction (Which one sounds like a real word?; e. g., jope-joak) or an orthographic instruction (Which one is a real word?; e. g., thurd-third). In comparison to controls, dyslexic subjects displayed reduced blood flow in temporal regions bilaterally and in inferior parietal cortex, mainly on the left, during both pronunciation and decision making. Their activation of left inferior frontal cortex (Broca’s area) during both phonological- and orthographic-decision making did not differ from the control group. Thus, the Rumsey et al. [99] results contrast with the data of their earlier study [96] as well as with those of Paulesu et al. [85]. Rumsey et al. [99] have discussed the absence of different activation loci in phonological versus orthographic tasks which might result, they suggested, from more basic deficits in phonemic awareness. They further hypothesized that the dyslexic group may have approached unknown irregular words in a manner closely resembling the letter-by-letter reading of unfamiliar pseudowords. In line with the latter interpretation by Rumsey and associates, Shaywitz et al. [112] designed a set of hierarchically organized print tasks, intended to progressively increase demands on phonological analysis. The tasks required same-different judgments concerning: (i) line orientation (e. g., [\\V]-[\\V]), presumed to reflect visual-spatial processing; (ii) letter case (e. g., [bbBb][bbBb]), thought to predominantly explore orthographic processing; (iii) single-letter rhyme (e. g., [T][V]); as well as (iv) pseudoword rhyme (e. g., [leat][jete]), assumed to add increasingly more phonological processing demands; and (v) semantic category (e. g., [corn]-[rice]), believed to make demands on transcod-

130

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

ing from print to phonology, but requires also activation of the mental lexicon to determine the meaning of the word. Brain activation patterns of 17 regions of interest per hemisphere were measured by means of fMRI in 29 right-handed dyslexic adults and 32 controls. On tasks making explicit demands on phonological processing (e. g., pseudoword rhyming), dyslexic individuals showed a relative underengagement of left posterior perisylvian and occipital sites (Wernicke’s area, the angular gyrus, and striate cortex), coupled with a disproportionately elevated response in a left anterior region (inferior frontal gyrus) than non-impaired readers. This has been suggested to be a reflection of functional disruption in the posterior cortical systems engaged in phonological decoding and a compensatory reliance on Broca’s area, respectively. In a further analysis of the Shaywitz et al. [112] data, Pugh et al. [89] turned their attention to the functional connectivity of the angular gyrus. Their reasons were twofold: First, the (left) angular gyrus is considered pivotal in mapping visually presented inputs onto phonologic representations. Second, dyslexic males have been reported to show a functional disconnection between the left angular gyrus and related posterior regions during reading [53]3. However, it could not be determined whether this disruption is specific to phonological decoding engaged by reading tasks. The re-analyzed data by Pugh et al. [89] revealed significant correlations between angular gyrus and occipital and temporal-lobe sites on pseudoword rhyme and semantic category judgments in controls, but not in the dyslexic group. In the right hemisphere, corresponding correlations were significant for both reading groups. Thus, this pattern of results suggests a breakdown in left-hemisphere connectivity in reading, when substantial phonological decoding (or “phonological assembly”; [89], p. 51) was required, whereas right-hemisphere homologues seem to work in a compensatory manner for dyslexic readers. Comparable evidence for an atypical left-hemispheric brain activation pattern thought to reflect a fundamental disruption of phonological processing for poor reading was offered from the PET study by Brunswick et al. [8]. The researchers compared six nonimpaired adult readers with six readers having a childhood history of developmental dyslexia (all righthanded males) on word and pseudoword naming. Dyslexic individuals showed less activation in ventral occipito-temporal sites and greater engagement of left inferior frontal gyrus than non-impaired controls.Using identical sets of stimuli, the same laboratory tested verbal repetition in eight dyslexic men and six controls [77]. Here, the dyslexic group demonstrated less hemodynamic response compared with the control group in 3 Horwitz et al. (53) used data from the Rumsey et al. (99) PET study.

the right superior temporal and right post-central gyri. Since studies in healthy individuals indicate that attending to the phonetic structure of speech is associated with a decrease in right-hemisphere processing, McCrory and colleagues concluded that reduced right-hemisphere activation in the dyslexic group indicate an attentional bias towards phonetic elements of the auditory input. That is, less processing of non-phonetic aspects of speech may favor greater salience of the phonological structure of attended speech for dyslexic readers4. Regarding the findings of an atypical brain activation profile observed either in the left hemisphere [8] or the right [77], the authors proposed that the neural manifestation of phonological disruption in dyslexia is taskspecific, i. e., functional rather than structural in nature. Taken together, during print tasks tapping phonological processing, dyslexic adults have shown usual or enhanced activity in left-hemisphere frontal-lobe language regions, but reduced or absent activity in left temporo-parietal language areas [8, 85, 99, 112]. Furthermore, the left angular gyrus has been found to be functionally disconnected from related temporal and occipital regions [53, 89]. More recent fMRI studies aimed at examining whether digressions from typical cerebral response patterns in dyslexia reflect a fundamental deficit of phonological processing or rather a compensation for poor reading in adulthood [34, 35, 113, 122]. Temple et al. [122] recorded whole-brain imaging data in 24 dyslexic and 15 normally reading children (8–12 years old) during phonological and orthographic tasks of rhyming and matching visually presented consonant letter pairs (e. g., Do T and D rhyme? and Are P and P the same?, respectively). During letter rhyming, activity in left frontal-lobe regions was evident in both groups, whereas activity in left temporo-parietal cortex was only observed in control children. During letter matching, the control group demonstrated activity throughout extrastriate visual cortex, whereas the dyslexic group showed reduced extrastriate responses. Thus, altered temporo-parietal activation probed by rhyme letters in dyslexic children parallel prior findings in dyslexic adults indicating a core phonological deficit. Moreover, childhood dyslexia may be characterized by impaired extrastriate activity thought to be important for orthographic processing. In a study by Georgiewa et al. [34], 17 (German) dyslexic adolescents and 17 non-impaired control subjects with an average age of 14 years (all right-handed) 4 In the Brunswick et al. [8] as well as McCrory et al. [77] study, no group differences were observed between the word and pseudoword versions of the tasks. The authors pointed out that the pseudowords used in the studies were highly word-like (e. g., carrot vs. cappot) and that differences might emerge when the lexical credibility of the pseudowords is reduced.

S. Heim et al. Neural correlates of dyslexia

were scanned while silently performing several tasks: viewing of letter strings, reading of nonwords (e. g., bnams) and frequent words (e. g., Blume, Engl.: flower), and phonological transformation (Move the first letter to the end of the word and add the common German suffix ‘-ein’; e. g., Blume → lume-bein). Dyslexic participants were found to show reduced activation in inferior frontal regions (in particular Broca’s area) and in lefthemisphere inferior temporal-lobe sites during tasks that invoke substantial grapheme-phoneme conversions and phonological awareness (i. e., nonword reading and phonological transformation). Neither group displayed temporo-parietal activity, however. This finding is in contrast to what has been observed in younger dyslexic children [122] as well as in adult samples (see above). The differences could be task-related (covert behavioral response in the Georgiewa study versus overt response in the other studies), which has also been suggested by a more recent fMRI experiment of Georgiewa et al. [35]. In a study involving a large sample of 144 righthanded children aged 7–18 years, Shaywitz et al. [113] showed that dyslexic subjects demonstrated lesser activation than normal controls both in posterior brain regions (including parieto-temporal sites and sites in the occipito-temporal area) and in the inferior frontal gyri during tasks (see [112]) relying on phonology. The reduced activity in anterior regions contrasted their earlier findings in adults [112]. However, a positive correlation (r ≈ 0.32) between chronological age and bilateral activation in the inferior frontal gyri of dyslexic children led them to suggest that frontal sites become increasingly incorporated with age in compensating for the disrupted posterior regions. Although the notion of compensation is promising, the influence of chronological age as an important factor needs to be examined in greater detail.

Brain activation during auditory temporal processing tasks To date, there are only a few functional neuroimaging studies reporting on auditory temporal processing in dyslexia [97, 121]. Rumsey et al. [97]5, employing PET contrasted brain activation in 15 right-handed dyslexic men and 18 normal readers during performance of a tonal matching task. The task demanded the participants to press a button if tonal sequences (3–4 tones) in a pair were identical. During tonal matching, dyslexic and normally reading adults displayed similar lefthemisphere temporal activation, but the dyslexic group exhibited reduced blood flow in right fronto-temporal regions. Along with this physiological difference, the

5 It should be noted that the PET study by Rumsey et al. (97) was not explicitly designed to test auditory temporal processing.

131

dyslexic group was significantly impaired in performing the task. Since the task involved fast-paced stimulus presentation (16 tonal pairs/min), the authors considered the finding of impaired right-hemisphere activation as being consonant with hypothesized deficits in rapid temporal processing in dyslexia. Concerning the fact that many subjects had participated in the Rumsey et al. [96] PET study of phonological processing (see above), Rumsey and colleagues proposed that dyslexic individuals may have more widespread deficits encompassing left- as well as right-hemisphere temporal cortex. In a recent study employing fMRI, Temple et al. [121] have examined whether adults with dyslexia exhibit deviances in the neural response to rapidly changing acoustic information. Stimuli employed were nonspeech analogues of consonant-vowel-consonant syllables with either brief (= rapid) or temporally extended (= slow) acoustic transitions. In each stimulus condition, subjects were asked to press a key for high-pitched but not for low-pitched sounds. While normal readers displayed increased activity in the left prefrontal cortex in response to rapid relative to slow transitions, dyslexic individuals showed no differential activity. Further, magnitude of the differential response was inversely correlated with performance in rapid auditory processing as measured by the threshold needed for sequencing three 20-ms tones presented at different rates. Following these results, the authors point to the possible role of left prefrontal regions as mediating rapid auditory processing. The two studies of Rumsey et al. [97] and Temple et al. [121] provide contrasting results. However, differences in imaging technologies, stimulus materials, tasks, and brain regions of interests between the studies as well as the lack of further research aggravate concluding remarks on neuronal responses to auditory temporal processing in dyslexia.

Brain activation during visual motion perception Several published fMRI articles presented evidence for a selective deficit in the magnocellular system in adults with dyslexia (e. g., [15, 16, 20]). To compare cerebral activation in six right-handed dyslexic males and eight normal controls, Eden et al. [20] measured local bloodoxygenation level-dependent (BOLD) contrast signals, while the participants passively viewed either a coherently moving random-dot stimulus (magnocellular stimulus) or a stationary pattern (parvocellular stimulus). Moving stimuli were expected to elicit strong responses in area V5 (MT) that is located in an extrastriate region at the junction of occipital and temporal lobes [127]. In normal readers the magnocellular stimulus activated V5/MT bilaterally but it failed to activate this area in dyslexic readers. Parvocellular stimulus did not elicit any differences between the two groups.

132

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

Demb et al. [15, 16] measured BOLD signals in response to low-luminance moving gratings (magnocellular stimuli) as opposed to control stimuli “designed to stimulate multiple pathways” [15], (p. 13363). Dyslexic individuals (n = 5; all right-handed) showed reduced activity relative to normal readers both in primary visual cortex (V1) and several extrastriate regions (inter alia MT+) in response to moving gratings of various contrasts. Moreover, participants exhibiting stronger V1 and MT+ activity demonstrated better moving discrimination performance and tended to be faster readers. In summary, the fMRI results obtained in the visual system in dyslexia point to a functional deficit in the magnocellular pathway. However, as for the previous imaging findings on auditory and phonological processing, brain areas showing significantly atypical responses vary between the studies. Thus, discrepancies might be attributable to differences in the subject populations, the stimuli, or the procedures used for localizing visual brain areas. While the phonological hypothesis of dyslexia has received valuable support from recent PET and fMRI research, hemodynamic neuroimaging studies on both magnocellular and auditory temporal processing are limited.

Electrophysiological studies Electrophysiological recording techniques such as electroencephalography (EEG) and magnetoencephalography (MEG) excel in examining brain processes with high temporal resolution. Event-related potentials (ERPs) of the EEG elicited by various verbal and non-verbal stimuli have been analyzed in numerous studies of languagebased learning impairments. For a review of auditory ERPs in dyslexia the reader is referred to Leppänen and Lyytinen [73]. A brief survey of ERPs to visually and auditorily presented stimuli in these populations is provided by Habib [38]. For more recent studies on visual ERPs in the framework of the magnocellular deficit theory of dyslexia, the reader is referred to Johannes et al. [59], Schulte-Körne et al. [107], Kuba et al. [64], and Romani et al. [93]. There is now growing literature in the field of dyslexia research that uses the MEG technology. Most studies have been concerned with magnetic source imaging during performance of various reading tasks [46, 47, 102, 115–117]; (for a review see [103]). Here, neuronal source activity within a predefined time window of visual event-related fields (ERFs) is determined and projected onto structural brain images. In general, these studies support the findings of hemodynamic deviances in language-related brain sites in dyslexia. For instance, Salmelin et al. [102] found that print processing was associated with enhanced source activity in the left-hemisphere inferior temporo-occipital border at about

180 ms following stimulus onset in control subjects but not in dyslexic adults. Also, between 200 and 400 ms, normal readers showed activation in the left temporal region while dyslexic participants demonstrated activation of the left inferior frontal cortex (approximately in Broca’s area). As with the fMRI findings in children and adults with dyslexia [8, 112, 122], this pattern was suggested to reflect posterior cortical anomaly and compensatory reliance on frontal-lobe systems. Atypical source activity in posterior brain sites was also detected in children with dyslexia during engagement in printed pseudoword rhyme-matching and word-recognition tasks [115, 116]. Dyslexic children displayed reduced activity in left temporo-parietal cortex between 300 and 1200 ms post-stimulus onset, coupled with a high density of source clusters in homologous right-hemisphere regions as compared to normally reading children.

■ Auditory event-related potentials In the following sections, auditory ERP studies of dyslexia are reviewed. Because of the wealth of ERP components examined in this population, the survey is limited to the N100 and mismatch negativity (MMN).

N100 The N100 (or N1) is the most prominent peak of auditory ERPs elicited by simple repetitive stimuli such as tones or syllables. Differences in latency or amplitude of the auditory N100 have been reported in children with reading difficulties [7, 82, 88]. Amplitude reduction of the N100 was found in a group of 14 boys with difficulties in reading, writing, and spelling (designated ‘poor readers’) as compared to 18 ‘good readers’ (all 8–9 years old) in a study by Pinkerton et al. [88]. Cortical auditory ERPs were recorded in response to 2000-Hz tone bursts while participants watched silent films. Reduced N100 amplitudes (around 160 ms) in poor readers were observed at three of four scalp locations. For the whole sample, N100 amplitude was correlated positively with performance IQ, spelling scores, reading accuracy and comprehension, as well as arithmetics. In interpreting the data, Pinkerton and colleagues suggested that the decreased N100 magnitude could be associated with impaired processes mediating selective attention. Brunswick and Rippon [7] contrasted 15 dyslexic boys (7–11 years old) and 15 normally reading controls (8–10 years old) on ERPs to stop consonant-vowel syllables presented in a dichotic listening paradigm. The participants were asked to report simultaneously presented syllables as accurately as possible. No significant group differences were observed either in the right or in the left ear responses. However, normally reading children exhibited larger N100 amplitudes at left temporal-elec-

S. Heim et al. Neural correlates of dyslexia

trode sites than the dyslexic children who showed less lateralized temporal N100 magnitude. N100 lateralization was also found to be positively related to performance on a phonological awareness task, viz., rhyme oddity detection among words differed in their last sounds (e. g., pin, win, sit, fin). According to Brunswick and Rippon, the deviances in N100 laterality are associated with abnormal cerebral lateralization of language functions in dyslexia. The failure of the dichotic listening task to discriminate between dyslexic and normal readers in spite of the N100 laterality differences was suggested to indicate that laterality does not affect processing of the stimuli per se but appears to be associated with later aspects of phoneme analysis. However, in view of the fact that the N100 has been considered a basic index of adequate sensory registration, Leppänen and Lyytinen [73] proposed that an altered N100 response might reflect inaccurate tuning of sensory information resulting in less reliable auditory representations that are, in turn, manifested in poor performance on language tests. Yingling et al. [126], on the other hand, did not find any differences between 38 severely dyslexic boys (mean age 13.3 years) and 38 non-impaired peers in ERPs following stimulation with auditory clicks. Bernal et al. [4] observed no deviances of the N100 to pure tones in a group of 20 poor readers (10–12 years old), but reported larger amplitudes in two later components, the N200 and the P200 as compared to 20 normally reading children. In a recent study, Molfese [79] presented evidence that auditory ERPs recorded within 36 h of birth discriminated between newborns who 8 years later would be classified as dyslexic,poor,or normal readers.The auditory ERPs analyzed by Molfese included the N1-P2-N2 waves elicited by speech and non-speech syllables with mean peak latencies of 174,309,and 458 ms,respectively. The left-hemisphere N1 latency at birth was found to be shortest for the normally reading children and longest for the poor readers. Neither the dyslexic nor the poor readers displayed a well-defined N1 component. Righthemisphere N2 peak amplitudes were largest for the dyslexic children and smallest for the poor readers. In particular the group differences in the N1 latency might point, as suggested by Molfese, to an underlying perceptual mechanism upon which some aspects of later developing verbal and cognitive processes are based. Neville et al. [82] reported N100 deviances in a subset of language-impaired children who exhibit deficits in auditory temporal processing. Twenty-two languageimpaired children with concomitant reading disability and 12 controls who evidenced normal language development and academic achievement (all 8–10 years old) were compared on auditory and visual ERPs. The auditory paradigm involved an active oddball task in which a 1000-Hz tone was presented as the target stimulus (10 % probability) among 2000-Hz standard stimuli at

133

one of three ISIs (200, 1000, and 2000 ms) and at one of three different stimulus positions (left ear,both ears,and right ear). Since no group differences were obtained for the auditory ERPs to either stimulus, the reading-disabled children were subclassified into two subgroups according to their performance on an auditory rapid sequencing test. Reading-disabled children performing below the median level were classified as ‘low repetition’ (i. e., displaying auditory temporal processing problems), while those scoring above were classified as ‘high repetition’. Following that, the N140 component to standard tones was found to be significantly diminished over the right hemisphere at the shortest interstimulus interval in the low-repetition group compared to both the language-normal controls and the high-repetition reading-disabled group. In addition, the latency of the standard N140 was significantly delayed in the low-repetition reading-disabled group, especially over temporal and parietal sites of the left hemisphere. Neville and coworkers considered the N140 component equivalent to the adult N100. A contralateral (to the stimulated ear) and anterior distribution of the N140 response suggests to them that it reflects activity generated in the superior temporal gyrus encompassing primary and secondary auditory areas. Hence, these findings were assumed to indicate that in reading-disabled children with auditory temporal processing problems, the reduced and slowed activity within these cortical sites contributed to their language symptoms. The authors’ interpretation is not to be taken as a single-factor account of the deficits of language- and reading-impaired children, however. Thus, various deviances on visual ERPs to both language and non-language stimuli were also reported for either the whole reading-disabled group or only a subset of it. Taken together, the auditory ERP studies cited above indicate differences in N100 features between groups of children designated dyslexia or poor readers and healthy controls. While latency deviances in languagebased learning impairments may be associated with a common timing deficit, N100 amplitude differences have been related to attentional factors or inadequate sensory processing. Great individual subject variability coupled with recording techniques using only a limited number of electrodes have not seldom led to negative results or only non-significant trends, however. Furthermore, variations in stimulus materials (e. g., clicks, tones, or syllables), task paradigms (e. g., response/active vs. no-response/passive task, oddball paradigm vs. repetitive unchanging stimuli), and interstimulus intervals [e. g., 10] across the ERP studies aggravate a comparison of the findings.

MMN The MMN is a fronto-centrally negative component of the auditory ERP, usually peaking between 100 and

134

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

250 ms post-stimulus onset. It is thought to reflect a preattentive neuronal change-detection mechanism, occurring when an infrequent physically ‘deviant’ sound encounters a well-established sensory memory trace of a frequently presented ‘standard’ sound (e. g., [80]). The MMN has proven to be a suitable tool for studying auditory discrimination in both adults and children (for a review see e. g., [11, 63, 80]). It also has been demonstrated as a sensitive measure for distinguishing individuals with language-based learning impairments from healthy peers (e. g., [2, 6, 51, 52, 62, 65, 67, 104, 106, 108], for a review see [66]). Kraus and colleagues were among the first to investigate auditory phoneme processing in children with language-based learning disorders using MMN. Kraus et al. [62] sought to determine whether deficits in perception of rapid spectro-temporal changes experienced by children with learning problems derive from aberrant neuronal representations of acoustic events prior to conscious reception or from higher-level dysfunctions. Behavioral discrimination abilities and MMN responses to synthetic consonant syllables were evaluated in 6- to 15-year-old controls with normal academic performance and children exhibiting learning problems. The learning-impaired children displayed a discrepancy between intellectual capacity and psychoeducational achievement. As a group, these children showed inter alia problems on measures of listening comprehension, sound blending, reading, and spelling. The MMN was evaluated in two subgroups of the children – one comprising 21 ‘good /da/-/ga/ perceivers’ and the other 21 ‘poor /da/-/ga/ perceivers’ – based on behavioral data. A prominent MMN in response to just-perceptibly different variants of /da/ and /ga/ was evident in good /da/-/ga/ perceivers, but not in poor perceivers. Correlational analyses performed for all the 42 children revealed moderate but significant relationships between behavioral /da/-/ga/ discrimination scores and both MMN duration and mean amplitude (r = –0.40 and r = –0.42, respectively). That is, accurate discrimination on /da/ versus /ga/ was associated with robust MMNs; limited performance, on the other hand, was related to diminished mismatch responses. In addition, both good and poor /da/-/ga/ perceivers were easily able to discriminate a /ba/-/wa/ contrast and, as was evaluated in 14 children of each subgroup, displayed a robust MMN to just-perceptibly different variants of /ba/ and /wa/. According to Kraus and colleagues, the findings indicate that the speech-sound discrimination deficits exhibited by children with learning problems probably have their origins in the auditory pathways and may be pre-attentive in nature. Moreover, the selective impairment in neuronal representation and behavioral discrimination of the /da/-/ga/ syllables compared to the /ba/-/wa/ stimuli, suggested to them that the two rapid spectro-temporal contrasts tap separate and distinct neuronal

mechanisms which may be differentially vulnerable to disruption. Identification of disturbed mismatch responses may thus have implications for differential diagnosis and targeted intervention strategies for children with learning impairments and attentional disorders [62]. Given their correlational nature, caution is warranted however regarding interpretation of electrophysiological parameters as indicators of causal relationships. In a subsequent study of the same research team, Bradlow et al. [6] aimed at investigating the precise acoustic-phonetic features that pose perceptual difficulties for children displaying similar learning problems as detailed above. Seventy-two controls with normal academic achievement and 32 learning-impaired children ranging in age from 6 to 16 years participated here. Consistent with previous findings [62], children with learning problems displayed smaller MMNs relative to their non-impaired age-mates to the /da/-/ga/ pair when the formant transition duration was short (40 ms). In the learning-impaired group, larger mismatch activity to temporally extended (80 ms) compared to short transitional syllables was found, although behavioral discrimination performance remained significantly impaired irrespective of formant transition length. In accord with their performance levels, normally learning children showed similar MMN responses to both short- and lengthened-transition /da/-/ga/ pairs. While extending the formant transition duration did not improve behavioral discrimination, the MMN data were thought to indicate that, at a pre-attentive neural level, the long-transition syllables were represented more accurately than the short-transition stimuli in children with learning problems. In further interpreting the results, Bradlow et al. [6] refer to the speech training program by Merzenich and Tallal (e. g., [78, 120]) in which formant transitions of consonant stimuli had been lengthened in time so as to make them more discernible during training. The authors suggested that the better neural representation of the longer duration syllables may underlie the success of acoustically modified speech training, whereas short-transition stimuli – which are poorly encoded – may be difficult for children to access for learning purposes. Risk for language-based learning impairments also has been studied exploiting the MMN paradigm in infants. Leppänen and Lyytinen [73] compared 6-monthold infants born into families with a history of dyslexia (n = 18) and control babies with no such background (n = 17). ERP differences were found in response to the (Finnish) nonsense word /atta/ while the standard was the shorter-duration nonsense word /ata/: infants with a positive family history showed a smaller MMN-like response over the left, but not right, hemisphere than the control group. The Jyväskylä Longitudinal study of Dyslexia follows children from birth to reading age

S. Heim et al. Neural correlates of dyslexia

(http://psykonet.jyu.fi/humander/JLD.htm). It will be interesting to see whether the mitigated MMN-like component signals an elevated risk of developmental dyslexia. Schulte-Körne and co-workers [104, 106, 108] have used the MMN to address the question of whether the perceptual deficits in people with dyslexia are of a general auditory or speech specific nature. In the SchulteKörne et al. [104] study, 19 dyslexic boys and 15 normal spellers (mean age 12.6 years) were presented with either synthetic stop-consonant syllables (standard /da/ vs. deviant /ba/) or pure tones (standard 1000 Hz vs. deviant 1050 Hz).While there were no group differences to the frequency change in tones, the dyslexic children showed a significantly reduced MMN amplitude to the change in syllables. Consequently, the results were assumed to point to a specific deficit at a pre-attentive sensory level. Comparable results were obtained when contrasting 12 dyslexic adults and 13 controls with normal spelling skills on stimulus series encompassing either the synthetic stop-consonant syllables /da/ and /ga/ or the 2200-Hz and 2640-Hz sinusoidal tones [108]. MMN responses differed between the two adult groups only in the syllable condition. The researchers conclude that speech perception at a pre-conscious stage constitutes one major player in dyslexia not only in children but also in adults. However, as conceded by Schulte-Körne et al. [104, 108] the findings leave open the question whether the group differences in the speech condition were due to the temporal information embedded in consonant stimuli. Pursuing this issue further, Schulte-Körne et al. [106] evaluated the mismatch activity to rapidly changing tone-burst patterns in 15 dyslexic adults and 20 normal spellers. The tone-burst patterns consisted of four frequency segments, 720–815–1040–815 Hz. In the standard pattern, the duration of the single frequencies was 50–90–25–50 ms, respectively. In the deviant pattern, the two segments of identical frequency (viz., 815 Hz) had been exchanged resulting in the duration sequence 50–50–25–90 ms. Dyslexic adults were observed to show an attenuated MMN relative to non-impaired controls. Schulte-Körne et al. [106] concluded that impaired neural discrimination of temporal, rather than phonetic, information may be pivotal for the findings of reduced MMN to stop-consonant syllables in dyslexia. One of the most informative studies in testing the major competing etiology hypotheses in dyslexia – a linguistic versus a more general processing deficit – is provided by Kujala et al. [65]. Two sets of stimuli were presented to eight dyslexic adults and eight normal readers: the tone-pattern stimuli consisted of four 500-Hz tones with silent intertone intervals of either 200, 150, and 50 ms (standard pattern) or 200, 50, and 150 ms (deviant pattern), the tone-pair stimuli comprised 500-Hz tone pairs separated by either 150 ms (standard pair) or

135

50 ms (deviant pair). No group differences were found in the MMN amplitude to the temporal change in the tonepair condition.In the tone-pattern condition,for normal subjects biphasic MMNs were elicited but dyslexic readers showed only the second MMN. The biphasic MMN reflected a response to the two deviations in the pattern stimuli: one, a quick tone following the second sound, and two, a delayed tone following the third sound. The auditory system of the dyslexic readers seems to discriminate only the second deviation. In agreement with the MMN data, behavioral discrimination performance was found to be normal in the tone-pair task but impaired in the tone-pattern task. Kujala et al. concluded that dyslexic subjects have problems in processing auditory temporal information only when presented in a complex context as in the case of the linguistic domain (cf., phonemes in words), not otherwise. Another MMN study has found a selective deficit in frequency discrimination in adults with dyslexia [2]. Baldeweg et al. contrasted 10 dyslexic adults and 10 normal readers on their MMN responses to either frequency changes or duration changes in pure tones. Mismatch activity to duration decrement did not differ between the two participant groups. In the frequency condition, however, dyslexic adults showed delayed and reduced MMN potentials relative to normal readers. This neuronal dysfunction was mirrored in a similarly specific behavioral impairment in discriminating tone frequency, but not tone duration. Furthermore, the frequency-discrimination deficit and MMN delay correlated with the degree of impairment in phonological skills, as reflected in reading errors of regular words and pseudowords. Although the authors pointed out that the study was not designed to investigate the ability to process rapidly presented auditory stimuli, some physical features of auditory events (viz., frequency) were assumed to add more than others (viz., duration) to the temporal processing deficit observed in some dyslexic individuals. The survey of MMN studies indicates that the mismatch response provides a powerful method for studying auditory discrimination and memory in children and adults with language-based learning disorders. Some of the experiments were designed to test whether the neuronal mismatch pattern favors a general auditory dysfunction hypothesis or one of a linguistic processing deficit. Other MMN studies sought to determine those precise acoustic features which provoke the perceptual difficulties experienced by some individuals with language-based learning disabilities. Taken together, the findings suggest deviances in the accuracy of the neuronal representation of speech and non-speech sounds as well as of certain auditory features in the language learning-impaired population.

136

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

■ Auditory event-related fields Some recent MEG experiments in the auditory domain have examined the magnetic counterpart of the N100, the N100m, elicited by repetitive stimulation with tones or speech sounds [41, 43, 48, 49, 81, 90]. Nagarajan et al. [81] compared seven adult poor readers and seven normally reading controls on tone-sequence perception. Sequences of two brief (20-ms duration) sinusoidal tones (high-high, high-low, low-high, or low-low) were presented at each of three different interstimulus intervals, or ISIs (100, 200, and 500 ms), and participants were asked to press the appropriate buttons to indicate the correct order. N100m amplitudes for the second stimulus of a sequence were weaker in poor readers than in control participants, for ISIs of 100 or 200 ms, but not for the ISI of 500 ms. This neuronal deviance was corroborated by a similar performance profile on tasks measuring perceptual interference between rapidly successive stimuli. The findings were put forth as further evidence in support of the general auditory temporal processing deficit. A study by Heim et al. [41] included 10 children with dyslexia and 9 normally literate controls (aged 8–14 years). Using a passive oddball paradigm, subjects were stimulated with frequency changes (1000-Hz standard vs. 1200-Hz deviant) or naturally produced stop consonant-syllables (/da/ standard vs. /ga/ deviant and vice versa). The sources of the magnetic waves in response to both speech and non-speech (standard) events 210 ms after stimulus onset were found to be located about 1.5 cm more anterior in children with dyslexia than in controls. The magnetic response at 210 ms was assumed to be equivalent to the adult N100m, which has recently been located to the planum temporale [76, 83]. Heim and colleagues speculated that the source configuration of the 210-ms responses might index activity within the planum temporale in the control children, whereas in the dyslexic children sources might be tied to sites anterior to the planum. Since this study focused on recordings from the left supratemporal cortex, it is not clear if dyslexic individuals also differ in source configurations of the right hemisphere. A following study was carried out using a whole-head magnetometer in order to address this very question [44]. Twenty-six subjects aged 8–15 years passively listened to synthetic stop consonants. In children with dyslexia, the right-hemispheric sources of the magnetic waves in response to the syllable /ba/ 100 ms after onset were found to be located more posterior than in controls. No such group difference was observed in the left hemisphere. Thus, the dyslexic group displayed a rather symmetrical source configuration between the hemispheres, whereas the control group showed the right-left asymmetry typical for the adult N100m. Indeed, data of normally literate adults (n = 10) indicate that the N100m

sources to the syllable /ba/ are distributed asymmetrically with a more anterior localization in the right than in the left hemisphere. In contrast, the N100m dipoles in dyslexic adults (n = 10) did not exhibit the same interhemispheric asymmetry. While there was no significant between-group difference in the center of activity over the left hemisphere, the dyslexic subjects’ N100m source of the right hemisphere was positioned ≈0.70 cm posterior to the source in the control participants [43]. Concerning findings offered from intracerebral recordings and magnetic source imaging techniques, the N100m deviances might be tied to Heschl’s gyrus and adjacent regions, in particular the planum temporale [37, 74, 76, 83]. The hemispheric balance in the source locations of the N100m found in our dyslexic group might agree with structural brain studies suggesting atypical asymmetry of the planum temporale in reading disabled individuals. Less N100m lateralization might also reflect altered neuronal morphology in temporal-plane sites of dyslexic individuals [24–26]. It is conceivable that based on these alterations other (right posterior) perisylvian regions become involved in auditory processing. These substituted regions may not perform the task as efficiently as a normally developed planum temporale would. Traditional views such as this one suggest that a structural deficit is the cause and functional deviance the consequence. In the course of neural plasticity studies of the human brain, the possibility has been acknowledged that functional alterations arising presumably from behavioral or environmental demands trigger morphological changes [21]. In a recent study, Helenius et al. [48] investigated N100m response to naturally spoken words presented in the context of sentences in 9 normal readers and 10 adults with a history of dyslexia. The activation peaked at 100 ms post-stimulus was specifically enhanced in the left supratemporal plane of dyslexic individuals. This effect was not associated with semantic appropriateness of the word, and thus was considered reflecting pre-semantic processing. A subsequent study [49] with the same sample evidenced a similar activation pattern of the N100m following stimulation with natural bisyllabic pseudowords (/ata/, /atta/, and /a a/). Dipole moment strength again displayed a left-hemisphere preponderance in the dyslexic group. In addition, the authors reported delayed N100m response in adults with dyslexia, when the duration of the silent gap between an attended initial vowel /a/ and subsequent syllable /ta/ was 175 ms. This group difference was absent for short gaps (95 ms) and unattended syllables. The two studies by Helenius and colleagues thus point to an early atypical left-hemisphere activation associated with natural speech perception.The observed N100m latency difference for long gaps appears surprising in the framework of Tallal’s notion of a deficit in perceiving rapidly occurring events – within tens of milliseconds. For instance, Tallal and

S. Heim et al. Neural correlates of dyslexia

Piercy [118, 119] observed that school-age children with specific language impairment may succeed in discriminating stop-consonant syllables when the fast transitional elements were artificially lengthened. The magneto-cortical correlates of this phenomenon were examined using the mismatch field (MMF) in two subgroups of children and adolescents with dyslexia [42]: the benefiters (n = 8) who displayed improved discrimination on extended formant transition syllables (/ba/ and /da/) compared to rapid formant transitions and the non-benefiters (n = 14) who did not show any difference. The benefiters showed an increase in MMF amplitude to extended- versus rapid-transition syllables in the right hemisphere while the non-benefiter type displayed no formant transition-related MMF enhancement in either hemisphere. Normal controls (n = 21) demonstrated an increase in MMF to extended syllables only over the left hemisphere. It seems likely then that there might be multiple subtypes of dyslexia with different underlying neuronal profiles. This subtype difference may account for negative MMF findings reported in a previous MEG study in children with dyslexia that focused on left auditory cortex [41]. The auditory ERF studies reviewed above suggest deviant functional neuroanatomy both in the left and right perisylvian language regions. These findings complement structural data of the dyslexic brain by a spatiotemporal aspect. This is not surprising in the context of the notion of the brain as a highly dynamic system (e. g.,

137

[21]). For instance, formal learning to read and write may be related to dramatic changes in brain organization and development as measured in literate and nonliterate adults [9].

Conclusion Structural and functional studies of the human brain have shown altered neuro-architecture within as well as between cerebral hemispheres in people with developmental dyslexia. Paralleling behavioral data, large-scale neural measures have not yielded evidence in favor of a single etiological theory of dyslexia. Sources for divergent findings are the heterogeneity of the study samples and differences in measurement procedures within a given research domain described in this article. Consistent results across studies were obtained however, when authors used comparable paradigms, encompassing similar task materials, stimulus modality, and theoretical background. These characteristics point to the complexity of speech and language functions. A complete picture therefore should include both high-temporal and spatial resolution data as provided by a combined approach of structural imaging and MEG technique, and most importantly theory driven behavioral tasks. New data emerging from psychophysiological work may shed light on dynamics and trajectories of the neurobiology of dyslexia.

References 1. American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders, DSM-IV. 4th edn. Washington DC: American Psychiatric Association 2. Baldeweg T, Richardson A, Watkins S, Foale C, Gruzelier J (1999) Impaired auditory frequency discrimination in dyslexia detected with mismatch evoked potentials. Ann Neurol 45: 495–503 3. Beaton AA (1997) The relation of planum temporale asymmetry and morphology of the corpus callosum to handedness, gender, and dyslexia: A review of the evidence. Brain and Language 60:255–322 4. Bernal J, Harmony T, Rodriguez M, Reyes A, Yanez G, Fernandez T, Galan L, Silva J, Fernandez-Bouzas A, Rodriguez H,Guerrero V,Marosi E (2000) Auditory event-related potentials in poor readers. Intern J Psychophysiol 36:11–23 5. Best M, Demb JB (1999) Normal planum temporale asymmetry in dyslexics with a magnocellular pathway deficit. Neurorep 10:607–612

6. Bradlow AR, Kraus N, Nicol TG, McGee TJ, Cunningham J, Zecker SG, Carrell TD (1999) Effects of lengthened formant transition duration on discrimination and neural representation of synthetic CV syllables by normal and learning-disabled children. J Acoustical Soc Am 106:2086–2096 7. Brunswick N, Rippon G (1994) Auditory event-related potentials, dichotic listening performance and handedness as indices of lateralisation in dyslexic and normal readers. Intern J Psychophysiol 18:265–275 8. Brunswick N, McCrory E, Price CJ, Frith CD, Frith U (1999) Explicit and implicit processing of words and pseudowords by adult developmental dyslexics. A search for Wernicke’s Wortschatz? Brain 122:1901–1917 9. Castro-Caldas A, Miranda PC, Carmo I, Reis A, Leote F, Ribeiro C, Ducla-Soares E (1999) Influence of learning to read and write on the morphology of the corpus callosum. Europ J Neurol 6: 23–28

10. Ceponiene R, Cheour M, Näätänen R (1998) Interstimulus interval and auditory event-related potentials in children: Evidence for multiple generators. Electroencephalography and Clin Neurophysiol 108:345–354 11. Cheour M, Leppänen PH, Kraus N (2000) Mismatch negativity (MMN) as a tool for investigating auditory discrimination and sensory memory in infants and children. Clin Neurophysiol 111:4–16 12. Chi JG, Dooling EC, Gilles FH (1977) Left-right asymmetries of the temporal speech areas of the human fetus. Arch Neurol 34:346–348 13. Dalby MA, Elbro C, Stødkilde-Jørgensen H (1998) Temporal lobe asymmetry and dyslexia: An in vivo study using MRI. Brain and Language 62:51–69 14. De Lacoste MC, Kirkpatrick JB, Ross ED (1985) Topography of the human corpus callosum. J Neuropathol Experiment Neurol 44:578–591

138

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

15. Demb JB, Boynton GM, Heeger DJ (1997) Brain activity in visual cortex predicts individual differences in reading performance. Proceedings of the National Academy of Sciences USA. 94: 13363–13366 16. Demb JB, Boynton GM, Heeger DJ (1998) Functional magnetic resonance imaging of early visual pathways in dyslexia. J Neurosci 18:6939–6951 17. Duara R, Kushch A, Gross-Glenn K, Barker WW, Jallad B, Pascal S, Loewenstein DA, Sheldon J, Rabin M, Levin B, Lubs H (1991) Neuroanatomic differences between dyslexic and normal readers on magnetic resonance imaging scans. Arch Neurol 48:410–416 18. Eckert MA, Leonard CM (2000) Structural imaging in dyslexia: The planum temporale. Ment Retard Devel Disab Res Rev 6:198–206 19. Eckert MA, Lombardino LJ, Leonard CM (2001) Planar asymmetry tips the phonological playground and environment raises the bar. Child Develop 72: 988–1002 20. Eden GF, VanMeter JW, Rumsey JM, Maisog JM,Woods RP, Zeffiro TA (1996) Abnormal processing of visual motion in dyslexia revealed by functional brain imaging. Nature 382:66–69 21. Elbert T, Heim S, Rockstroh B (2001) Neural plasticity and development. In: Nelson C, Luciana M (eds) Handbook of developmental cognitive neuroscience. Cambridge, MA: MIT Press, pp 191–202 22. Filipek PA (1995) Neurobiologic correlates of developmental dyslexia: How do dyslexics’ brains differ from those of normal readers? J Child Neurol 10 (suppl. 1):62–69 23. Flowers DL (1993) Brain basis for dyslexia: a summary of work in progress. J Learning Disab 26:575–582 24. Galaburda AM (1988) The pathogenesis of childhood dyslexia. In: Plum F (ed) Language, communication, and the brain. New York: Raven Press, pp 127–138 25. Galaburda AM (1989) Ordinary and extraordinary brain development: Anatomical variation in developmental dyslexia. Ann Dyslexia 39:67–80 26. Galaburda AM (1993) Neuroanatomical basis of developmental dyslexia. Neurologic Clin 11:161–173 27. Galaburda AM (1994) Developmental dyslexia and animal studies: At the interface between cognition and neurology. Cognition 50:133–149 28. Galaburda AM, Sherman GF, Rosen GD, Aboitiz F, Geschwind N (1985) Developmental dyslexia: Four consecutive patients with cortical anomalies. Ann Neurol 18:222–233 29. Galaburda AM, Corsiglia J, Rosen GD, Sherman GF (1987) Planum temporale asymmetry, reappraisal since Geschwind and Levitsky. Neuropsychol 25:853–868

30. Galaburda AM, Rosen GD, Sherman GF (1990) Individual variability in cortical organization: Its relationship to brain laterality and implications to function. Neuropsychol 28:529–546 31. Galaburda AM, Livingstone M (1993) Evidence for a magnocellular defect in developmental dyslexia. Annals of the New York Academy of Sciences 682: 70–82 32. Galaburda AM, Menard MT, Rosen GD (1994) Evidence for aberrant auditory anatomy in developmental dyslexia. Proceedings of the National Academy of Sciences USA. 91:8010–8013 33. Galaburda AM, Schrott LM, Sherman GF, Rosen GD, Denenberg VH (1996) Animal models of developmental dyslexia. In: Chase CH, Rosen GD, Sherman GF (eds) Developmental dyslexia. Baltimore, MD: York Press, pp 3–14 34. Georgiewa P, Rzanny R, Hopf JM, Knab R, Glauche V, Kaiser WA, Blanz B (1999) fMRI during word processing in dyslexic and normal reading children. Neurorep 10:3459–3465 35. Georgiewa P, Rzanny R, Gaser C, Gerhard UJ, Vieweg U, Freesmeyer D, Mentzel HJ, Kaiser WA, Blanz B (2002) Phonological processing in dyslexic children: A study combining functional imaging and event-related potentials. Neurosci Letters 318:5–8 36. Geschwind N, Levitsky W (1968) Human brain: Left-right asymmetries in temporal speech region. Science 161: 186–187 37. Godey B, Schwartz D, de Graaf JB, Chauvel P, Liégeois-Chauvel C (2001) Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials: A comparison of data in the same patients. Clin Neurophysiol 112:1850–1859 38. Habib M (2000) The neurological basis of developmental dyslexia: An overview and working hypothesis. Brain 123: 2373–2399 39. Haffner J, Zerahn-Hartung C, Pfüller U, Parzer P, Strehlow U, Resch F (1998) Auswirkungen und Bedeutung spezifischer Rechtschreibprobleme bei jungen Erwachsenen – empirische Befunde in einer epidemiologischen Stichprobe. Zeitschrift für Kinder und Jugendpsychiatrie 26:124–135 40. Heiervang E, Hugdahl K, Steinmetz H, Smievoll AI, Stevenson J, Lund A, Ersland L, Lundervold A (2000) Planum temporale, planum parietale and dichotic listening in dyslexia. Neuropsychol 38:1704–1713 41. Heim S, Eulitz C, Kaufmann J, Füchter I, Pantev C, Lamprecht-Dinnesen A, Matulat P, Scheer P, Borstel M, Elbert T (2000) Atypical organisation of the auditory cortex in dyslexia as revealed by MEG. Neuropsychol 38:1749–1759

42. Heim S, Eulitz C, Elbert T (2002) Psychophysiological differentiation of language dysfunction in dyslexia. Psychophysiol 39 (suppl. 1):41 43. Heim S, Eulitz, C, Elbert T (2003) Altered hemispheric asymmetry of auditory N100m in adults with developmental dyslexia. Neurorep 14:501–504 44. Heim S, Eulitz C, Elbert T (2003) Altered hemispheric asymmetry of auditory P100m in dyslexia. Europ J Neurosc 17:1715–1722 45. Heim S, Lakshminarayanan K (2003) Hemispheric involvement in dyslexia: Neuromagnetic responses to auditory stimuli. In: Columbus F (ed) Progress in dyslexia research. Hauppauge, NY: Nova Science Publishers (in press) 46. Helenius P, Tarkiainen A, Cornelissen P, Hansen PC, Salmelin R (1999) Dissociation of normal feature analysis and deficient processing of letter-strings in dyslexic adults. Cerebral Cortex 9: 476–483 47. Helenius P, Salmelin R, Service E, Connolly JF (1999) Semantic cortical activation in dyslexic readers. J Cogn Neurosc 11:535–550 48. Helenius P, Salmelin R, Service E, Connolly JF, Leinonen S, Lyytinen H (2002) Cortical activation during spokenword segmentation in nonreading-impaired and dyslexic adults. J Neurosc 22:2936–2944 49. Helenius P, Salmelin R, Richardson U, Leinonen S, Lyytinen H (2002) Abnormal auditory cortical activation in dyslexia 100 msec after speech onset. J Cogn Neurosc 14:603–617 50. Hiemenz JR, Hynd GW (2000) Sulcal/ gyral pattern morphology of the perisylvian language region in developmental dyslexia. Brain and Language 74:113–133 51. Holopainen IE, Korpilahti P, Juottonen K, Lang H, Sillanpää M (1997) Attenuated auditory event-related potential (mismatch negativity) in children with developmental dysphasia. Neuropediatr 28:253–256 52. Holopainen IE, Korpilahti P, Juottonen K, Lang H, Sillanpää M (1998) Abnormal frequency mismatch negativity in mentally retarded children and in children with developmental dysphasia. J Child Neurol 13:178–183 53. Horwitz B, Rumsey, JM, Donohue BC (1998) Functional connectivity of the angular gyrus in normal reading and dyslexia. Proceedings of the National Academy of Sciences USA. 95: 8939–8944 54. Humphreys P, Kaufmann WE, Galaburda AM (1990) Developmental dyslexia in women: Neuropathological findings in three patients. Ann Neurol 28:727–738 55. Hynd GW, Semrud-Clikeman M (1989) Dyslexia and brain morphology. Psychol Bull 106:447–482

S. Heim et al. Neural correlates of dyslexia 56. Hynd GW, Semrud-Clikeman M, Lorys AR, Novey ES, Eliopulos D (1990) Brain morphology in developmental dyslexia and attention deficit disorder/hyperactivity. Arch Neurol 47:919–926 57. Hynd GW, Hall J, Novey ES, Eliopulos D, Black K, Gonzalez JJ, Edmonds JE, Riccio C, Cohen M (1995) Dyslexia and corpus callosum morphology. Arch Neurol 52:32–38 58. Jenner AR, Rosen GD, Galaburda AM (1999) Neuronal asymmetries in primary visual cortex of dyslexic and nondyslexic brains. Ann Neurol 46: 189–196 59. Johannes S, Kussmaul CL, Münte TF, Mangun GR (1996) Developmental dyslexia: Passive visual stimulation provides no evidence for a magnocellular processing defect. Neuropsychol 34: 1123–1127 60. Kirch DG, Weinberger DR (1986) Anatomical neuropathology in schizophrenia: Post-mortem findings. In: Nasrallah HA (ed) Handbook of schizophrenia. Amsterdam: Elsevier, pp 325–348 61. Klingberg T, Hedehus M, Temple E, Salz T, Gabrieli JD, Moseley ME, Poldrack RA (2000) Microstructure of temporoparietal white matter as a basis for reading ability: Evidence from diffusion tensor magnetic resonance imaging. Neuron 25:493–500 62. Kraus N, McGee TJ, Carrell TD, Zecker SG, Nicol TG, Koch DB (1996) Auditory neurophysiologic responses and discrimination deficits in children with learning problems.Science 273:971–973 63. Kraus N, Cheour M (2000) Speech sound representation in the brain. Audiology & Neuro-Otology 5:140–150 64. Kuba M, Szanyi J, Gayer D, Kremlacek J, Kubova Z (2001) Electrophysiological testing of dyslexia. Acta Medica 44: 131–134 65. Kujala T, Myllyviita K, Tervaniemi M, Alho K, Kallio J, Näätänen R (2000) Basic auditory dysfunction in dyslexia as demonstrated by brain activity measurements. Psychophysiol 37:262–266 66. Kujala T, Näätänen R (2001) The mismatch negativity in evaluating central auditory dysfunction in dyslexia. Neurosci Biobehav Rev 25:535–543 67. Kujala T, Karma K, Ceponiene R, Belitz S, Turkkila P, Tervaniemi M, Näätänen R (2001) Plastic neural changes and reading improvement caused by audiovisual training in reading-impaired children. Proceedings of the National Academy of Sciences USA. 98:10509–10514 68. Kushch A, Gross-Glenn K, Jallad B, Lubs H, Rabin M, Feldman E, Duara R (1993) Temporal lobe surface area measurements on MRI in normal and dyslexic readers. Neuropsychol 31:811–821

69. Lambe EK (1999) Dyslexia, gender, and brain imaging. Neuropsychol 37: 521–536 70. Larsen JP, Høien T, Lundberg I, Ødegaard H (1990) MRI evaluation of the size and symmetry of the planum temporale in adolescents with developmental dyslexia. Brain and Language 39:289–301 71. Larsen JP, Høien T, Ødegaard H (1992) Magnetic resonance imaging of the corpus callosum in developmental dyslexia. Cognitive Neuropsychol 9: 123–134 72. Leonard CM, Voeller KKS, Lombardino LJ, Morris MK, Hynd GW, Alexander AW, Andersen HG, Garofalakis M, Honeyman JC, Mao J, Agee OF, Staab EV (1993) Anomalous cerebral structure in dyslexia revealed with magnetic resonance imaging.Arch Neurol 50:461–469 73. Leppänen PHT, Lyytinen H (1997) Auditory event-related potentials in the study of developmental language-related disorders. Audiology & NeuroOtology 2:308–340 74. Liégeois-Chauvel C, Musolino A, Badier J, Marquis P, Chauvel P (1994) Evoked potentials recorded from the auditory cortex in man: Evaluation and topography of the middle-latency components. Electroencephalogr Clin Neurophysiol 92:204–214 75. Livingstone MS, Rosen GD, Drislane FW, Galaburda AM (1991) Physiological and anatomical evidence for a magnocellular defect in developmental dyslexia. Proceedings of the National Academy of Sciences USA. 88: 7943–7947 (Published erratum appears in Proceedings of the National Academy of Sciences USA (1993) 90:2556) 76. Lütkenhöner B, Steinsträter O (1998) High-precision neuromagnetic study of the functional organization of the human auditory cortex. Audiology & Neuro-Otology 3:191–213 77. McCrory E, Frith U, Brunswick N, Price C (2000) Abnormal functional activation during a simple word repetition task: A PET study of adult dyslexics. J Cogn Neurosci 12:753–762 78. Merzenich MM, Jenkins WM, Johnston P, Schreiner C, Miller SL, Tallal P (1996) Temporal processing deficits of language-learning impaired children ameliorated by training. Science 271:77–81 79. Molfese DL (2000) Predicting dyslexia at 8 years of age using neonatal brain responses. Brain and Language 72: 238–245 80. Näätänen R (2001) The perception of speech sounds by the human brain as reflected by the mismatch negativity (MMN) and its magnetic equivalent (MMNm). Psychophysiol 38:1–21

139

81. Nagarajan S, Mahncke H, Salz T, Tallal P, Roberts T, Merzenich MM (1999) Cortical auditory signal processing in poor readers. Proceedings of the National Academy of Sciences USA. 96: 6483–6488 82. Neville HJ, Coffey SA, Holcomb PJ, Tallal P (1993) The neurobiology of sensory and language processing in language-impaired children. J Cogn Neurosci 5:235–253 83. Ohtomo S, Nakasato N, Kanno A, Hatanaka K, Shirane R, Mizoi K, Yoshimoto T (1998) Hemispheric asymmetry of the auditory evoked N100m response in relation to the crossing point between the central sulcus and Sylvian fissure. Electroencephalogr Clin Neurophysiol 108:219–225 84. Pandya DN, Seltzer B (1986) The topography of commissural fibers. In: Lepore F, Ptito M, Jasper HH (eds) Two hemispheres – one brain: Functions of the corpus callosum. New York: Liss, pp 47–73 85. Paulesu E, Frith U, Snowling M, Gallagher A, Morton J, Frackowiak RS, Frith CD (1996) Is developmental dyslexia a disconnection syndrome? Evidence from PET scanning. Brain 119:143–157 86. Pennington BF (1995) Genetics of learning disabilities. J Child Neurol 10 (suppl. 1):69–77 87. Pennington BF (1999) Toward an integrated understanding of dyslexia: Genetic, neurological, and cognitive mechanisms. Develop Psychopathol 11: 629–654 88. Pinkerton F, Watson DR, McClelland RJ (1989) A neurophysiological study of children with reading, writing and spelling difficulties. Develop Med Child Neurol 31:569–581 89. Pugh KR, Mencl WE, Shaywitz BA, Shaywitz SE, Fulbright RK, Constable RT, Skudlarski P, Marchione KE, Jenner AR, Fletcher JM, Liberman AM, Shankweiler DP, Katz L, Lacadie C, Gore JC (2000) The angular gyrus in developmental dyslexia: Task-specific differences in functional connectivity within posterior cortex. Psychol Sci 11:51–56 90. Renvall H, Hari R (2002) Auditory cortical responses to speech-like stimuli in dyslexic adults. J Cogn Neurosci 14: 757–768 91. Robichon F, Habib M (1998) Abnormal callosal morphology in male adult dyslexics: Relationships to handedness and phonological abilities. Brain and Language 62:127–146 92. Robichon F, Levrier O, Farnarier P, Habib M (2000) Developmental dyslexia: Atypical cortical asymmetries and functional significance. Europ J Neurol 7:35–46

140

European Child & Adolescent Psychiatry (2004) Vol. 13, No. 3 © Steinkopff Verlag 2004

93. Romani A, Conte S, Callieco R, Bergamaschi R, Versino M, Lanzi G, Zambrino CA, Cosi V (2001) Visual evoked potential abnormalities in dyslexic children. Functional Neurol 16: 219–229 94. Rosen GD, Sherman GF, Galaburda AM (1989) Interhemispheric connections differ between symmetrical and asymmetrical brain regions. Neurosci 33:525–533 95. Rumsey JM, Dorwart R, Vermess M, Denckla MB, Kruesi MJ, Rapoport JL (1986) Magnetic resonance imaging of brain anatomy in severe developmental dyslexia.Arch Neurol 43:1045–1046 96. Rumsey JM, Andreason P, Zametkin AJ, Aquino T, King C, Hamburger SD, Pikus A, Rapoport JL, Cohen RM (1992) Failure to activate the left temporoparietal cortex in dyslexia: An oxygen 15 positron emission tomographic study.Arch Neurol 49:527–534 (Published erratum appears in Arch Neurol (1994) 51:243) 97. Rumsey JM, Andreason P, Zametkin AJ, King AC, Hamburger SD, Aquino T, Hanahan AP, Pikus A, Cohen RM (1994) Right frontotemporal activation by tonal memory in dyslexia, an O15 PET study. Biol Psychiatry 36: 171–180 98. Rumsey JM, Casanova M, Mannheim GB, Patronas N, DeVaughn N, Hamburger SD, Aquino T (1996) Corpus callosum morphology, as measured with MRI, in dyslexic men. Biol Psychiatry 39:769–775 99. Rumsey JM, Nace K, Donohue B, Wise D, Maisog JM, Andreason P (1997) A positron emission tomographic study of impaired word recognition and phonological processing in dyslexic men. Arch Neurol 54:562–573 100. Rumsey JM, Donohue BC, Brady DR, Nace K, Giedd JN, Andreason P (1997) A magnetic resonance imaging study of planum temporale asymmetry in men with developmental dyslexia. Arch Neurol 54:1481–1489 101. Rutter M, Tizard J, Yule W, Graham P, Whitmore K (1976) Isle of Wight Studies 1964–1974. Psychol Med 6:313–332 102. Salmelin R, Service E, Kiesilä P, Uutela K, Salonen O (1996) Impaired visual word processing in dyslexia revealed with magnetoencephalography. Ann Neurol 40:157–162 103. Salmelin R, Helenius P, Service E (2000) Neurophysiology of fluent and impaired reading: a magnetoencephalographic approach. J Clin Neurophysiol 17:163–174 104. Schulte-Körne G, Deimel W, Bartling J, Remschmidt H (1998) Auditory processing and dyslexia: evidence for a specific speech processing deficit. Neurorep 9:337–340

105. Schulte-Körne G, Nöthen MM, Remschmidt H (1998) Zur Genetik der Lese-Rechtschreibstörung (Legasthenie). Medizinische Genetik 10: 402–405 106. Schulte-Körne G, Deimel W, Bartling J, Remschmidt H (1999). Pre-attentive processing of auditory patterns in dyslexic human subjects. Neurosci Letters 276:41–44 107. Schulte-Körne G, Bartling J, Deimel W, Remschmidt H (1999) Attenuated hemispheric lateralization in dyslexia: evidence of a visual processing deficit. Neurorep 10:3697–3701 108. Schulte-Körne G, Deimel W, Bartling J, Remschmidt H (2001) Speech perception deficit in dyslexic adults as measured by mismatch negativity (MMN). Intern J Psychophysiol 40:77–87 109. Schultz RT, Cho NK, Staib LH, Kier LE, Fletcher JM, Shaywitz SE, Shankweiler DP, Katz L, Gore JC, Duncan JS, Shaywitz BA (1994) Brain morphology in normal and dyslexic children: The influence of sex and age. Ann Neurol 35:732–742 110. Shapleske J, Rossell SL, Woodruff PW, David AS (1999) The planum temporale: A systematic, quantitative review of its structural, functional and clinical significance. Brain Res Rev 29: 26–49 111. Shaywitz SE, Shaywitz BA, Fletcher JM, Escobar MD (1990) Prevalence of reading disability in boys and girls: Results of the Connecticut Longitudinal Study. J Am Med Assoc 264: 998–1002 112. Shaywitz SE, Shaywitz BA, Pugh KR, Fulbright RK, Constable RT, Mencl WE, Shankweiler DP, Liberman AM, Skudlarski P, Fletcher JM, Katz L, Marchione KE, Lacadie C, Gatenby C, Gore JC (1998) Functional disruption in the organization of the brain for reading in dyslexia. Proceedings of the National Academy of Sciences USA. 95: 2636–2641 113. Shaywitz BA, Shaywitz SE, Pugh KR, Mencl WE, Fulbright RK, Skudlarski P, Constable RT, Marchione KE, Fletcher JM, Lyon GR, Gore JC (2002) Disruption of posterior brain systems for reading in children with developmental dyslexia. Biol Psychiatry 52: 101–110 114. Sherman GF, Galaburda AM (1999) Neuronal migration abnormalities in autoimmune mice: Implications for developmental dyslexia. Develop Neuropsychol 16:355–357 115. Simos PG, Breier JI, Fletcher JM, Foorman BR, Bergman E, Fishbeck K, Papanicolaou AC (2000) Brain activation profiles in dyslexic children during non-word reading: A magnetic source imaging study. Neurosci Letters 290: 61–65

116. Simos PG, Breier JI, Fletcher JM, Bergman E, Papanicolaou AC (2000) Cerebral mechanisms involved in word reading in dyslexic children: A magnetic source imaging approach. Cerebral Cortex 10:809–816 117. Simos PG, Fletcher JM, Bergman E, Breier JI, Foorman BR, Castillo EM, Davis RN, Fitzgerald M, Papanicolaou AC (2002) Dyslexia-specific brain activation profile becomes normal following successful remedial training. Neurol 58:1203–1213 118. Tallal P, Piercy M (1974) Developmental aphasia: Rate of auditory processing and selective impairment of consonant perception. Neuropsychol 12: 83–93 119. Tallal P, Piercy M (1975) Developmental aphasia: The perception of brief vowels and extended stop consonants. Neuropsychol 13:69–74 120. Tallal P, Miller SL, Bedi G, Byma G, Wang X, Nagarajan SS, Schreiner C, Jenkins WM, Merzenich MM (1996) Language comprehension in language-learning impaired children improved with acoustically modified speech. Science 271:81–84 121. Temple E, Poldrack RA, Protopapas A, Nagarajan S, Salz T, Tallal P, Merzenich MM, Gabrieli JDE (2000) Disruption of the neural response to rapid acoustic stimuli in dyslexia: Evidence from functional MRI. Proceedings of the National Academy of Sciences USA. 97:13907–13912 122. Temple E, Poldrack RA, Salidis J, Deutsch GK, Tallal P, Merzenich MM, Gabrieli JD (2001) Disrupted neural responses to phonological and orthographic processing in dyslexic children: An fMRI study. Neurorep 12:299–307 123. von Plessen K, Lundervold A, Duta N, Heiervang E, Klauschen F, Smievoll AI, Ersland L, Hugdahl K (2002) Less developed corpus callosum in dyslexic subjects – a structural MRI study. Neuropsychol 40:1035–1044 124. Wada JA, Clarke R, Hamm A (1975) Cerebral hemispheric asymmetry in humans. Cortical speech zones in 100 adults and 100 infant brains. Arch Neurol 32:239–246 125. World Health Organization (1992) Tenth Revision of the International Classification of Diseases, ICD-10, Chapter V (F): Mental and Behavioral Disorders. Bern: Huber 126. Yingling CD, Galin D, Fein G, Peltzman D, Davenport L (1986) Neurometrics does not detect ‘pure’ dyslexics. Electroencephalogr Clin Neurophysiol 63:426–430 127. Zeki S (1993) A vision of the brain. Cambridge, MA: Blackwell Scientific