Multisensory integration deficits in developmental dyslexia

P-annotatePDF-v11

INSTRUCTIONS ON THE ANNOTATION OF PDF FILES

To view, print and annotate your article you will need Adobe Reader version 9 (or higher). This program is freely available for a whole series of platforms that include PC, Mac, and UNIX and can be downloaded from http://get.adobe.com/reader/. The exact system requirements are given at the Adobe site: http://www.adobe.com/products/reader/tech-specs.html.

Note: if you opt to annotate the file with software other than Adobe Reader then please also highlight the appropriate place in the PDF file.

PDF ANNOTATIONS

Adobe Reader version 9 Adobe Reader version X and XI

When you open the PDF file using Adobe Reader, the Commenting tool bar should be displayed automatically; if not, click on ‘Tools’, select ‘Comment & Markup’, then click on ‘Show Comment & Markup tool bar’ (or ‘Show Commenting bar’ on the Mac). If these options are not available in your Adobe Reader menus then it is possible that your Adobe Acrobat version is lower than 9 or the PDF has not been prepared properly.

(Mac) PDF ANNOTATIONS (Adobe Reader version 9)

The default for the Commenting tool bar is set to ‘off’ in version 9. To change this setting select ‘Edit | Preferences’, then ‘Documents’ (at left under ‘Categories’), then select the option ‘Never’ for ‘PDF/A View Mode’.

(Changing the default setting, Adobe version 9)

To make annotations in the PDF file, open the PDF file using Adobe Reader XI, click on ‘Comment’.

If this option is not available in your Adobe Reader menus then it is possible that your Adobe Acrobat version is lower than XI or the PDF has not been prepared properly.

This opens a task pane and, below that, a list of all Comments in the text. These comments initially show all the changes made by our copyeditor to your file.

Action

HOW TO...

Adobe Reader version 9 Adobe Reader version X and XI Insert text Click the ‘Text Edits’ button on the

Commenting tool bar. Click to set the cursor location in the text and simply start typing. The text will appear in a commenting box. You may also cut-and-paste text from another file into the commenting box. Close the box by clicking on ‘x’ in the top right-hand corner.

Click the ‘Insert Text’ icon on the Comment tool bar. Click to set the cursor location in the text and simply start typing. The text will appear in a commenting box. You may also cut-and-paste text from another file into the commenting box. Close

the box by clicking on ‘_’ in the top right-hand corner.

Replace text Click the ‘Text Edits’ button on the Commenting tool bar. To highlight the text to be replaced, click and drag the cursor over the text. Then simply type in the replacement text. The replacement text will appear in a commenting box. You may also cut-and-paste text from another file into this box. To replace formatted text (an equation for example) please Attach a file (see below).

Click the ‘Replace (Ins)’ icon on the Comment tool bar. To highlight the text to be replaced, click and drag the cursor over the text. Then simply type in the replacement text. The replacement text will appear in a commenting box. You may also cut-and-paste text from another file into this box. To replace formatted text (an equation for example) please Attach a file (see below).

Remove text Click the ‘Text Edits’ button on the Commenting tool bar. Click and drag over the text to be deleted. Then press the delete button on your keyboard. The text to be deleted will then be struck through.

Click the ‘Strikethrough (Del)’ icon on the Comment tool bar. Click and drag over the text to be deleted. Then press the delete button on your keyboard. The text to be deleted will then be struck through.

Highlight text/ make a comment

Click on the ‘Highlight’ button on the Commenting tool bar. Click and drag over the text. To make a comment, double click on the highlighted text and simply start typing.

Click on the ‘Highlight Text’ icon on the Comment tool bar. Click and drag over the text. To make a comment, double click on the highlighted text and simply start typing.

Attach a file

Click on the ‘Attach a File’ button on the Commenting tool bar. Click on the figure, table or formatted text to be replaced. A window will automatically open allowing you to attach the file. To make a comment, go to ‘General’ in the ‘Properties’ window, and then ‘Description’. A graphic will appear in the PDF file indicating the insertion of a file.

Click on the ‘Attach File’ icon on the Comment tool bar. Click on the figure, table or formatted text to be replaced. A window will automatically open allowing you to attach the file. A graphic will appear indicating the insertion of a file.

Leave a note/ comment Click on the ‘Note Tool’ button on

the Commenting tool bar. Click to set the location of the note on the document and simply start typing. Do not use this feature to make text edits.

Click on the ‘Add Sticky Note’ icon on the Comment tool bar. Click to set the location of the note on the document and simply start typing. Do not use this feature to make text edits.

Action

HOW TO...

Adobe Reader version 9 Adobe Reader version X and XI Review To review your changes, click on the ‘Show’

button on the Commenting tool bar. Choose ‘Show Comments List’. Navigate by clicking on a correction in the list. Alternatively, double click on any mark-up to open the commenting box.

Your changes will appear automatically in a list below the Comment tool bar. Navigate by clicking on a correction in the list. Alternatively, double click on any mark-up to open the commenting box.

Undo/delete change

To undo any changes made, use the right click button on your mouse (for PCs, Ctrl-Click for the Mac). Alternatively click on ‘Edit’ in the main Adobe menu and then ‘Undo’. You can also delete edits using the right click (Ctrl-click on the Mac) and selecting ‘Delete’.

To undo any changes made, use the right click button on your mouse (for PCs, Ctrl-Click for the Mac). Alternatively click on ‘Edit’ in the main Adobe menu and then ‘Undo’. You can also delete edits using the right click (Ctrl-click on the Mac) and selecting ‘Delete’.

SEND YOUR ANNOTATED PDF FILE BACK TO ELSEVIER

Save the annotations to your file and return as instructed by Elsevier. Before returning, please ensure you have answered any questions raised on the Query Form and that you have inserted all corrections: later inclusion of any subsequent corrections cannot be guaranteed.

FURTHER POINTS

Any (grey) halftones (photographs, micrographs, etc.) are best viewed on screen, for which they are optimized, and your local printer may not be able to output the greys correctly.

If the PDF files contain colour images, and if you do have a local colour printer available, then it will be likely that you will not be able to correctly reproduce the colours on it, as local variations can occur.

If you print the PDF file attached, and notice some ‘non-standard’ output, please check if the problem is also present on screen. If the correct printer driver for your printer is not installed on your PC, the printed output will be distorted.

Current Biology 24, 1–5, March 3, 2014 ª2014 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2014.01.029

ReportMultisensory Integration and Attentionin Developmental Dyslexia

Vanessa Harrar,1,* Jonathan Tammam,2

Alexis Perez-Bellido,3 Anna Pitt,2 John Stein,2

and Charles Spence11Department of Experimental Psychology, University ofOxford, Oxford OX1 3UD, UK2Department of Physiology, Anatomy, and Genetics,University of Oxford, Oxford OX1 3QX, UK3Departament de Psicologia Basica, Universitat de Barcelona,08035 Barcelona, Spain

Summary

Developmental dyslexia affects 5%–10% of the population

[1], resulting in poor spelling and reading skills. While thereare well-documented differences in the way dyslexics pro-

cess low-level visual [2, 3] and auditory [4, 5] stimuli, it ismostly unknown whether there are similar differences in

audiovisual multisensory processes. Here, we investigatedaudiovisual integration using the redundant target effect

(RTE) paradigm. Some conditions demonstrating audiovi-sual integration appear to depend upon magnocellular

pathways [6], and dyslexia has been associated with defi-cits in this pathway [7]; so, we postulated that develop-

mental dyslexics (‘‘dyslexics’’ hereafter) would showdifferences in audiovisual integration compared with con-

trols. Reaction times (RTs) to multisensory stimuli werecompared with predictions from Miller’s race model [8, 9].

Dyslexics showed difficulty shifting their attention between

modalities; but such ‘‘sluggish attention shifting’’ (SAS)[10] appeared only when dyslexics shifted their attention

from the visual to the auditory modality. These resultssuggest that dyslexics distribute their crossmodal attention

resources differently from controls, causing differentpatterns in multisensory responses compared to controls.

From this, we propose that dyslexia training programsshould take into account the asymmetric shifts of cross-

modal attention.

Results and Discussion

When asked to make a motor response to a visual, auditory, oraudiovisual stimulus, 17 dyslexic adults responded signifi-cantly more slowly than matched controls (F1,33 = 11.55, p =0.002, h2

p = 0.259). Figure 1A shows the median unisensoryRTs that were used in a mixed-model ANOVA. This revealedsignificant differences between the modalities (F1,33 = 16.69,p < 0.001, h2

p = 0.336) with RTs to visual stimuli being signifi-cantly slower than to auditory stimuli. Importantly, there wasno significant interaction between group (dyslexic or control)and unisensory stimulus type. While it has been known forsome time that dyslexics respond considerably more slowlythan matched controls [12], here we report that the delay inresponding is similar across modalities. To properly comparemultisensory RTs between groups, it was necessary to factorout this constant unisensory delay.

RTs to combined sensory stimuli were faster than to eitherthe visual or auditory stimulus (df = 34, tAudio = 7.77, tvisual =18.15, p < 0.001). This is known as the redundant target effect(RTE) because people respond more rapidly when there aremultiple (redundant) signals. This redundancy effect hasbeen modeled by Miller [8, 9], who suggested comparing theobserved RT with the minimum sum of two unisensoryresponse distributions (the ‘‘race’’ model). RTs are fasterwhen the simultaneously presented unisensory signals raceto the finish line (detection threshold) as compared to wheneach stimulus is presented individually. As can be seen fromFigure 1B,medianmultisensory RTswere even faster than pre-dicted from the race model (F1,33 = 31.35, p < 0.001, h2

p =0.487). Violation of the race model is a classic multisensoryeffect [13], which initially led researchers to concludethat the unisensory signals are combined early in sensoryprocessing—i.e., prior to detection.More recent, parsimonious interpretations of violations of

the race model assume that the signals interact at a neuronallevel while still being processed independently and in parallel[14, 15]. Here, we assume that RTs that are faster than therace model (the combination of visual or auditory RTs alone)are due to signals being processed in parallel, but interactingat the neuronal level. Therefore, our definition of multisensoryintegration does not require the stimuli to be combined;integration is anything that is different from (or cannot beexplained only by) unisensory processing alone (as in [16]).While the violation of the race model is similar at the median

for dyslexics and controls (w10 ms; Figure 1B), differencesbecame apparent when the entire distribution of multisensoryRTs is compared to the race model. Dyslexics’ multisensoryRTs were faster than expected from the race model less oftenthan controls (Figure 1C; one-tailed t test, t33 = 2.379, p =0.012). Furthermore, the discrepancy between our partici-pants’ literacy and nonverbal reasoning correlated with theproportion of RTs that were faster than the race model predic-tions (r = 20.501, p = 0.003). Our results demonstrate, for thefirst time, that dyslexics’ multisensory RTs violate the racemodel prediction less often than controls, and this behavioralmeasure of integration is correlated with literacy. (Furtherdetails on calculating the proportion of RTs faster than therace model and further descriptions of correlations withliteracy discrepancy scores can be found in Figures S1 andS2, available online.)One of the effects driving the RTE is the cost of switching

attention between modalities [14, 17]. In the current experi-ment, auditory, visual, and audiovisual RT trials were randomlyordered within a block. As can be seen in Figure 2, when thesame stimulus was presented on successive trials, responseswere fastest. In contrast, when the stimuli in successive trialswere different (e.g., when the participant was presented witha visual stimulus directly after a unisensory-auditory trial),attention would need to shift from one modality to another.The RT cost associated with shifting attention betweenmodal-ities on successive trials is known as the modality shift effect(MSE) [18]. Since auditory and visual stimuli were colocalizedat the center of the screen, with loudspeakers on either sideof the screen, the MSE calculated here is not confoundedwith costs associated with having to shift attention spatially.*Correspondence: [email protected]

CURBIO 10904

Please cite this article in press as: Harrar et al., Multisensory Integration and Attention in Developmental Dyslexia, Current Biology(2014), http://dx.doi.org/10.1016/j.cub.2014.01.029

While both groups demonstrated a classic MSE, there wasalso a significant three-way interaction between the MSE,modality, and group (F1,31 = 4.145, p = 0.050, h2

p = 0.118).This interaction suggested that we should analyze MSEsseparately in each group.

Controls demonstrated typical results: RTs were faster toauditory than visual targets (main effect of modality: F1,16 =20.37, p < 0.001, h2

p = 0.560); also, RTs were faster when thesame stimulus was repeated, and slower when the previousstimulus was different (MSE: F1,16 = 52.51, p < 0.001, h2

p =0.776), but did not demonstrate an interaction between thetwo (F < 1; see Figure 2B; similar results in [17]).

Dyslexics, by contrast, demonstrated an interactionbetween the two; dyslexics showed a smaller MSE for visualtargets as compared with auditory targets (F1,15 = 6.096, p =0.026, h2

p = 0.289). That is, RTs to visual stimuli were 18 msfaster when preceded by another visual target as opposed towhen preceded by an auditory or multisensory target, whileRTs to auditory targets were 35 ms faster when preceded bythe same stimulus, as opposed to a different stimulus.Dyslexics appear to find it harder to shift their attention awayfrom visual stimuli or toward auditory stimuli. This pattern isalso present in previous reports of dyslexics’ performance

on crossmodal association tasks [19, 20], though discussedin the framework of association rather than attention shifting.The most robust difference between dyslexics and controlsoccurred when auditory stimuli were presented after visualstimuli (dyslexics have the largest cost in this order).A larger MSE would result in dyslexics having more

variability in their unisensory RT distributions. We found thatthe MSE accounted for a larger proportion of the variabilityin the RTs in the control group than it did in the dyslexicgroup (the MSE divided by the interquartile range of the RTdistribution was larger for controls than for the dyslexics:F1,33 = 4.256, p = 0.047, h2

p = 0.114). Thus, for the controls, afast response to an auditory stimulus was more likely to befollowed by a slower response (if the next trial was a differentstimulus) than for the dyslexics. For dyslexics, on the otherhand, the MSE accounted for a smaller proportion of thevariability in the distribution of RTs. The large variability indyslexics’ RTs is more likely related to the wider neural tuningof their sensory processes [21].Other patient groups with temporal processing deficits

show a similar MSE pattern as dyslexics [22]), but thepattern is in the opposite direction to controls respondingto ‘‘equated’’ auditory and visual stimuli [23] and opposite to

A B C

Figure 1. RTs to Unisensory and Multisensory Stimuli, and Comparisons with the Race Model

For each individual participant, unisensory outlier RTs were removed if they were greater than three SDs from the mean (calculated separately for each

modality). Error bars indicate the SEM. (A) and (B) demonstrate that controls were faster than dyslexics and that median RTs to audiovisual stimuli were

faster than predicted by the median race model for both groups.

(A) Median (50th percentile) RTs for each modality as a function of group.

(B)Median (50th percentile) RTs for audiovisual stimuli are compared to themedian RTs predicted from the racemodel (calculated using theMATLAB routine

provided in [11]).

(C) The proportion of percentiles (out of ten) where RTswere faster than predicted by the racemodel plotted as a function of group (see further description in

Figure S1). Larger values indicate that a greater proportion of RTs were faster than predicted. This suggests that dyslexics have less multisensory RT

facilitation than controls.

See also Figures S1 and S2.

A B Figure 2. Modality Shift Effects

For each participant, unisensory RTs were calcu-

lated when the previous stimulus was the same

(e.g., two successive visual trials) or different

(e.g., an auditory or multisensory trial followed

by a visual trial). In (A), median RTs are plotted

demonstrating the main trends: dyslexics exhibit

slower RTs overall, and RTs are slower when the

previous stimulus is different. In (B), the inter-

action between these factors for dyslexics is

more easily seen. Unisensory switch costs were

calculated by taking the difference between RTs

to consecutive trials with the same target, and

RTs when the previous trial was different. Visual

and auditory MSEs were the same magnitude

for controls, but dyslexics have a larger MSE for

auditory responses compared to visual ones.

Error bars indicate the SEM.

Current Biology Vol 24 No 52

CURBIO 10904

Please cite this article in press as: Harrar et al., Multisensory Integration and Attention in Developmental Dyslexia, Current Biology(2014), http://dx.doi.org/10.1016/j.cub.2014.01.029

the pattern reported for people with other developmental dis-orders [24]. SinceMSEs are known to varywith attention [22], itis possible that comorbid attention deficit hyperactivity disor-der (ADHD) may have mediated MSE in the current sample ofdyslexics.

We expect that the majority of dyslexics in the presenttypical sample also had ADHD since the rates of comorbidityin dyslexia are at least 60% [25]. People with ADHD havebeen shown to preferentially divert their attention towardvisual stimuli, resulting in auditory stimuli being processedmore slowly (see shifted simultaneity values in [26]). However,there do not appear to be previous reports of MSE in ADHD, sowe cannot comment on the relative contributions of ADHD, orother comorbid deficits, to the MSE reported here for dys-lexics. Whatever the cause, these results demonstrate cleardifferences in how attention is distributed between modalitiesfor a typical sample of dyslexics compared with controls.

TheMSE demonstrates that RTs on successive trials are notindependent. We applied Otto’s model [14, 15] to calculate thecorrelation between auditory and visual RTs. While Miller’srace model assumes a perfect negative correlation betweenvisual and auditory RTs ([9], p. 337), Otto’s analysis of multi-sensory RTs allows the correlation to vary. Previous worksuggests that the correlations are highly dependent onexperimental methods that affect the MSE ([15], p. 7472), butcorrelations have not yet been compared in populations thathave different MSEs. Figure 3A demonstrates that the correla-tions were indeed significantly stronger (i.e., closer to 21) forcontrol participants than for the dyslexics, probably becausethe MSE contributes less to the variability in dyslexics’ RTs(one-tailed t test: t33 = 2.715, p = 0.010).

We also compared the additional noise parameter estimatedby the model [14] in the two groups. Unisensory RT distribu-tions provide an estimate of the variability in the firing rate ofa neuron when presented with a single signal. With redundantsignals, two pools of neurons would accumulate evidenceseparately. When multiple pools are simultaneously active,they might interact and cause a further increase in activity(plus variability) in each pool of neurons, which would not bepresent when evidence was accumulating for each signalseparately. This increased noise would suggest a capacityfor parallel processing, because increased unisensory vari-ability would predict a faster mean multisensory RT comparedto decreased unisensory variability (see [14] for further dis-cussion). However, we did not find a reliable difference in theadditional noise estimated (h) for the two groups (F1,33 =1.71, p = 0.200, Figures 3B and S3). We conclude that the

difference between groups is, mostly, attributable to the wayin which dyslexics shift their attention between trials contain-ing different stimuli (MSE), rather than differences in neuronalnoise on a given multisensory trial.There are very few previous reports of how multisensory

stimuli are processed by dyslexics, and many of them haveconfounded spatial with crossmodal effects by using head-phones to present the auditory stimuli [28, 29]. Whenaudiovisual stimuli are not colocalized, attention is distributedindependently across modalities and across space [18]. Giventhat dyslexics have difficulty allocating their unisensory visualattention resources across space [30, 31], it is all the moreimportant to control for spatial location when investigatingcrossmodal shifts of attention. Previous multisensory integra-tion studies in dyslexics have generated conflicting results,most likely, in large part, due to such confounding of spatialand modality attention effects. For example, meta-analysisof [32, 33] shows that integration differences between groupsare smallest for colocalized audiovisual stimuli, as comparedto spatially separate audiotactile and visualtactile combina-tions. Location coding, rather than modality coding per se,appears to be driving many of the group differences inprevious crossmodal studies in dyslexia (see a similar pointmade in [34]). Indeed, we report the largest differencesbetween dyslexics and controls not when audiovisual stimuliwere presented together (NS effect of noise, Figure 3B; seealso NS difference in [35]), but instead when auditory stimulifollowed visual ones (resulting in smaller correlations; seeFigure 3A).These results fit with the idea that dyslexics have a visual

attention disorder [36]. Indeed, dyslexics have impaired con-nections between prefrontal attention areas and visual areas[37], resulting in, for example, an asymmetric distribution ofvisual attention across the visual field [38].Hari and Renval [10] suggested that dyslexics suffer from

‘‘sluggish attention shifting’’ (SAS), which impairs processingof rapid stimuli in all modalities [29]. The present crossmodalresults demonstrate that dyslexics distribute attention asym-metrically between auditory and visual modalities, more sothan controls; it is difficult for dyslexics to disengage theirattention from visual stimuli and shift it to auditory stimuli. Toour knowledge, ours is the first nonspatially confounded inves-tigation into the distribution of attention between modalities indyslexics. Our results suggest an important feature of SAS: forcrossmodal shifts of attention, it is probable that many dys-lexics dwell more on the visual stimuli in conditions that requireshifts of attention from visual to auditory stimuli. For them, SAS

A B Figure 3. Correlation and Increased Noise

Estimates Explaining Why the RTs Violate the

Race Model

(A) Mean correlation for each group obtained

using Otto et al.’s analysis of RTs [14]. Controls

demonstrate significantly stronger negative

correlations than dyslexics.

(B) There was no significant difference between

controls and dyslexics in terms of the estimated

increases in additional noise (h) (not significant

[NS], df = 33, t = 1.3).

Nota bene: the data from twomethods of analysis

(those presented in Figure 1C and those pre-

sented here using Otto’s analysis) demonstrate

the same pattern of results (and statistical signif-

icance), which was also obtained by following

the method laid out in [27]. Errors bars indicate

the SEM. See also Figure S3.

Multisensory Integration in Dyslexia3

CURBIO 10904

Please cite this article in press as: Harrar et al., Multisensory Integration and Attention in Developmental Dyslexia, Current Biology(2014), http://dx.doi.org/10.1016/j.cub.2014.01.029

only appears to be problematic for attention shifts from visionto audition—and, importantly, not vice versa—causing espe-cially delayed responses to auditory stimuli that directly followvisual stimuli. We speculate that dyslexics presenting withmore apparent auditory processing deficits may haveopposite crossmodal SAS asymmetry effects compared withdyslexics who are more affected by difficulties processingvisual stimuli.

The neural circuitry responsible for sluggish shifting is likelysimilar for sluggish spatial and crossmodal shifts, though inthe latter condition, modality-specific areas would likely alsobe active. Violations of the race model have been associatedwith auditory inputs modifying early visual sensory processingin the parieto-occipital region [13]. Visual-spatial attentionshifts rely on the dorsal stream, more specifically on theoccipitoparietal areas [39, 40]. Others have suggested thatthe dorsal pathway might be fed, to a large extent, by themagnocellular pathway [30, 41]. This is compatible with themagnocellular deficit theory of dyslexia, which can accountfor many of the transient, temporal, spatial, unisensory, andcrossmodal processing differences in dyslexics [7].

Conclusions

The results of the present study highlight differences in theway dyslexic participants respond to multisensory stimuli ascompared to controls. Dyslexics are faster than the racemodelpredictions less often than are controls, thus demonstratingthat they experience less interaction between the senses.This difference is probably caused by a combination of factors,including dyslexics having larger variability in their responses,a larger MSE for auditory stimuli, and a weaker correlationbetween RTs. Together, these results clarify the SAS charac-terization of dyslexics’ responses, which is not symmetric forshifts of attention between vision and audition. Since limitedattention resources appear to be unevenly distributed in favorof vision, future investigations into conditions where visualattention dominates might shed light on the causes of dyslexiaand indicate training programs for it.

Dyslexia training programs fall into two broad categories:reading based and low level. Reading-based training improvesphonological awareness—for example, ‘‘seeing stars’’ (seediscussion in [42]) or slowed auditory speech stimuli thatmay compensate for dyslexic’s temporal deficits (see [43])—but this training has had limited success [44]. Our hypothesis,which needs to be tested, is that dyslexics might learn audio-visual phonological associations faster if they first hear thesound, and then see the corresponding letter/word—sincecrossmodal shifts from audition to vision were not sluggish.Phonological training may also improve low-level magnocellu-lar functions (e.g., [42]), but the reversemight also be possible,and perhaps would be more straightforward. That is, percep-tual learning to train low-level, basic neurological processescould lead to benefits in upstream functions such as atten-tional networks (e.g., crossmodal attention and cross-spatialattention) and, eventually, reading [45]. Training phonologicalawareness is analogous to treating the symptoms, whereastraining the low-level basic processesmay treat the underlyingcauses of dyslexia. Video games designed to speed allocationof spatial attention have proved successful in improvingreading in dyslexics [46]. The colocalized multisensory stimuliin these video games most likely also improved the crossmo-dal attention deficits. The contribution of crossmodal atten-tion, cross-spatial attention, and motion detection deficits onreading remains unclear. However, these video games are

potentially an excellent method for providing perceptuallearning to dyslexics in order to empirically test the involve-ment of (and improve) the neurological processes critical foradvanced reading.

Supplemental Information

Supplemental Information includes Supplemental Experimental Procedures

and three figures and can be found with this article online at http://dx.doi.

org/10.1016/j.cub.2014.01.029.

Acknowledgments

Our thanks go to Thomas Otto for extensive analysis support. In addition,

wewould like to thankHans Colonius for analysis recommendations, Rachel

Hulatt for help with data collection and literature review, and the Dyslexia

Research Trust and its clients for participating. V.H. held the Mary

Somerville Junior Research Fellowship from Somerville College, Oxford

University, UK.

Received: September 16, 2013

Revised: December 9, 2013

Accepted: January 13, 2014

Published: February 13, 2014

References

1. Shaywitz, S.E., Shaywitz, B.A., Fletcher, J.M., and Escobar, M.D. (1990).

Prevalence of reading disability in boys and girls. Results of the

Connecticut Longitudinal Study. JAMA 264, 998–1002.

2. Cornelissen, P.L., Hansen, P.C., Hutton, J.L., Evangelinou, V., and Stein,

J.F. (1998). Magnocellular visual function and children’s single word

reading. Vision Res. 38, 471–482.

3. Livingstone, M.S., Rosen, G.D., Drislane, F.W., and Galaburda, A.M.

(1991). Physiological and anatomical evidence for a magnocellular

defect in developmental dyslexia. Proc. Natl. Acad. Sci. USA 88,

7943–7947.

4. Helenius, P., Uutela, K., and Hari, R. (1999). Auditory stream segregation

in dyslexic adults. Brain 122, 907–913.

5. McAnally, K.I., and Stein, J.F. (1996). Auditory temporal coding in

dyslexia. Proc. Biol. Sci. 263, 961–965.

6. Perez-Bellido, A., Soto-Faraco, S., and Lopez-Moliner, J. (2013).

Sound-driven enhancement of vision: disentangling detection-level

from decision-level contributions. J. Neurophysiol. 109, 1065–1077.

7. Stein, J., andWalsh, V. (1997). To see but not to read; the magnocellular

theory of dyslexia. Trends Neurosci. 20, 147–152.

8. Miller, J. (1982). Divided attention: evidence for coactivation with

redundant signals. Cognit. Psychol. 14, 247–279.

9. Miller, J. (1986). Timecourse of coactivation in bimodal divided

attention. Percept. Psychophys. 40, 331–343.

10. Hari, R., and Renvall, H. (2001). Impaired processing of rapid stimulus

sequences in dyslexia. Trends Cogn. Sci. 5, 525–532.

11. Ulrich, R., Miller, J., and Schroter, H. (2007). Testing the race

model inequality: an algorithm and computer programs. Behav. Res.

Methods 39, 291–302.

12. Ramus, F., Pidgeon, E., and Frith, U. (2003). The relationship between

motor control and phonology in dyslexic children. J. Child Psychol.

Psychiatry 44, 712–722.

13. Molholm, S., Ritter, W., Murray, M.M., Javitt, D.C., Schroeder, C.E., and

Foxe, J.J. (2002). Multisensory auditory-visual interactions during early

sensory processing in humans: a high-density electrical mapping study.

Brain Res. Cogn. Brain Res. 14, 115–128.

14. Otto, T.U., and Mamassian, P. (2012). Noise and correlations in parallel

perceptual decision making. Curr. Biol. 22, 1391–1396.

15. Otto, T.U., Dassy, B., and Mamassian, P. (2013). Principles of multi-

sensory behavior. J. Neurosci. 33, 7463–7474.

16. Stein, B.E., Burr, D., Constantinidis, C., Laurienti, P.J., Alex Meredith,

M., Perrault, T.J., Jr., Ramachandran, R., Roder, B., Rowland, B.A.,

Sathian, K., et al. (2010). Semantic confusion regarding the development

of multisensory integration: a practical solution. Eur. J. Neurosci. 31,

1713–1720.

Current Biology Vol 24 No 54

CURBIO 10904

Please cite this article in press as: Harrar et al., Multisensory Integration and Attention in Developmental Dyslexia, Current Biology(2014), http://dx.doi.org/10.1016/j.cub.2014.01.029

17. Gondan, M., Lange, K., Rosler, F., and Roder, B. (2004). The redundant

target effect is affected by modality switch costs. Psychon. Bull. Rev.

11, 307–313.

18. Spence, C., Nicholls, M.E., and Driver, J. (2001). The cost of expecting

events in the wrong sensory modality. Percept. Psychophys. 63,

330–336.

19. Fox, E. (1994). Grapheme-phoneme correspondence in dyslexic and

matched control readers. Br. J. Psychol. 85, 41–53.

20. Snowling, M.J. (1980). The development of grapheme-phoneme

correspondence in normal and dyslexic readers. J. Exp. Child

Psychol. 29, 294–305.

21. Geiger, G., Cattaneo, C., Galli, R., Pozzoli, U., Lorusso, M.L., Facoetti,

A., and Molteni, M. (2008). Wide and diffuse perceptual modes charac-

terize dyslexics in vision and audition. Perception 37, 1745–1764.

22. Sutton, S., Hakerem, G., Zubin, J., and Portnoy, M. (1961). Effect of shift

of sensory modality on serial reaction-time - a comparison of schizo-

phrenics and normals. Am. J. Psychol. 74, 224.

23. Sandhu, R., and Dyson, B.J. (2012). Re-evaluating visual and auditory

dominance throughmodality switching costs and congruency analyses.

Acta Psychol. (Amst.) 140, 111–118.

24. Williams, D.L., Goldstein, G., and Minshew, N.J. (2013). The modality

shift experiment in adults and children with high functioning autism.

J. Autism Dev. Disord. 43, 794–806.

25. Willcutt, E.G., and Pennington, B.F. (2000). Comorbidity of reading

disability and attention-deficit/hyperactivity disorder: differences by

gender and subtype. J. Learn. Disabil. 33, 179–191.

26. Donohue, S.E., Darling, E.F., and Mitroff, S.R. (2012). Links between

multisensory processing and autism. Exp. Brain Res. 222, 377–387.

27. Colonius, H., and Diederich, A. (2006). The race model inequality:

interpreting a geometric measure of the amount of violation. Psychol.

Rev. 113, 148–154.

28. Hairston, W.D., Burdette, J.H., Flowers, D.L., Wood, F.B., and Wallace,

M.T. (2005). Altered temporal profile of visual-auditory multisensory

interactions in dyslexia. Exp. Brain Res. 166, 474–480.

29. Facoetti, A., Lorusso, M.L., Cattaneo, C., Galli, R., and Molteni, M.

(2005). Visual and auditory attentional capture are both sluggish in

children with developmental dyslexia. Acta Neurobiol. Exp. (Warsz.)

65, 61–72.

30. Vidyasagar, T.R. (1999). A neuronal model of attentional spotlight:

parietal guiding the temporal. Brain Res. Brain Res. Rev. 30, 66–76.

31. Franceschini, S., Gori, S., Ruffino, M., Pedrolli, K., and Facoetti, A.

(2012). A causal link between visual spatial attention and reading

acquisition. Curr. Biol. 22, 814–819.

32. Laasonen, M., Service, E., and Virsu, V. (2002). Crossmodal temporal

order and processing acuity in developmentally dyslexic young adults.

Brain Lang. 80, 340–354.

33. Virsu, V., Lahti-Nuuttila, P., and Laasonen, M. (2003). Crossmodal

temporal processing acuity impairment aggravates with age in develop-

mental dyslexia. Neurosci. Lett. 336, 151–154.

34. Jones, M.W., Branigan, H.P., Parra, M.A., and Logie, R.H. (2013).

Cross-modal binding in developmental dyslexia. J. Exp. Psychol.

Learn. Mem. Cogn. 39, 1807–1822.

35. de Boer-Schellekens, L., and Vroomen, J. (2012). Sound can improve

visual search in developmental dyslexia. Exp. Brain Res. 216, 243–248.

36. Valdois, S., Bosse, M.L., and Tainturier, M.J. (2004). The cognitive

deficits responsible for developmental dyslexia: review of evidence

for a selective visual attentional disorder. Dyslexia 10, 339–363.

37. Finn, E.S., Shen, X., Holahan, J.M., Scheinost, D., Lacadie, C.,

Papademetris, X., Shaywitz, S.E., Shaywitz, B.A., and Constable, R.T.

(2013). Disruption of Functional Networks in Dyslexia: A Whole-Brain,

Data-Driven Analysis of Connectivity. Biol. Psychiatry. Published online

October 11, 2013. http://dx.doi.org/10.1016/j.biopsych.2013.08.031.

38. Facoetti, A., and Molteni, M. (2001). The gradient of visual attention in

developmental dyslexia. Neuropsychologia 39, 352–357.

39. Behrmann, M., Geng, J.J., and Shomstein, S. (2004). Parietal cortex and

attention. Curr. Opin. Neurobiol. 14, 212–217.

40. Yantis, S., Schwarzbach, J., Serences, J.T., Carlson, R.L., Steinmetz,

M.A., Pekar, J.J., and Courtney, S.M. (2002). Transient neural activity

in human parietal cortex during spatial attention shifts. Nat. Neurosci.

5, 995–1002.

41. Livingstone, M., and Hubel, D. (1988). Segregation of form, color,

movement, and depth: anatomy, physiology, and perception. Science

240, 740–749.

42. Olulade, O.A., Napoliello, E.M., and Eden, G.F. (2013). Abnormal visual

motion processing is not a cause of dyslexia. Neuron 79, 180–190.

43. Tallal, P., Miller, S.L., Bedi, G., Byma, G., Wang, X., Nagarajan, S.S.,

Schreiner, C., Jenkins, W.M., and Merzenich, M.M. (1996). Language

comprehension in language-learning impaired children improved with

acoustically modified speech. Science 271, 81–84.

44. Strong, G.K., Torgerson, C.J., Torgerson, D., and Hulme, C. (2011). A

systematic meta-analytic review of evidence for the effectiveness of

the ‘Fast ForWord’ language intervention program. J. Child Psychol.

Psychiatry 52, 224–235.

45. Gori, S., and Facoetti, A. (2013). Perceptual learning as a possible

new approach for remediation and prevention of developmental

dyslexia. Vision Res. Published online December 8, 2013. http://dx.

doi.org/10.1016/j.visres.2013.11.011.

46. Franceschini, S., Gori, S., Ruffino, M., Viola, S., Molteni, M., and

Facoetti, A. (2013). Action video games make dyslexic children read

better. Curr. Biol. 23, 462–466.

Multisensory Integration in Dyslexia5

CURBIO 10904

Please cite this article in press as: Harrar et al., Multisensory Integration and Attention in Developmental Dyslexia, Current Biology(2014), http://dx.doi.org/10.1016/j.cub.2014.01.029