Elsevier

NeuroImage

Volume 175, 15 July 2018, Pages 70-79
NeuroImage

Atypical cortical entrainment to speech in the right hemisphere underpins phonemic deficits in dyslexia

https://doi.org/10.1016/j.neuroimage.2018.03.072Get rights and content

Highlights

  • Children with dyslexia show atypical low-frequency cortical entrainment to speech.

  • Impaired entrainment was due to dyslexia and not to reduced reading experience.

  • Dyslexia both reduced and enhanced speech entrainment in right hemisphere locations.

  • The right hemisphere effects reflected impaired phoneme-level entrainment.

  • Impaired entrainment was significantly related to impaired phonological awareness.

Abstract

Developmental dyslexia is a multifaceted disorder of learning primarily manifested by difficulties in reading, spelling, and phonological processing. Neural studies suggest that phonological difficulties may reflect impairments in fundamental cortical oscillatory mechanisms. Here we examine cortical mechanisms in children (6–12 years of age) with or without dyslexia (utilising both age- and reading-level-matched controls) using electroencephalography (EEG). EEG data were recorded as participants listened to an audio-story. Novel electrophysiological measures of phonemic processing were derived by quantifying how well the EEG responses tracked phonetic features of speech. Our results provide, for the first time, evidence for impaired low-frequency cortical tracking to phonetic features during natural speech perception in dyslexia. Atypical phonological tracking was focused on the right hemisphere, and correlated with traditional psychometric measures of phonological skills used in diagnostic dyslexia assessments. Accordingly, the novel indices developed here may provide objective metrics to investigate language development and language impairment across languages.

Introduction

Developmental dyslexia (hereafter dyslexia) is a learning disorder manifested in difficulties in the acquisition of reading and spelling that affects 5–10% of school-aged children and that can arise despite an adequate learning environment and otherwise normal intellectual and sensory functioning (Snowling, 2000). Children affected by dyslexia usually perform poorly in tests of phonological awareness, verbal short-term memory, and lexical-access (e.g., rapid naming), highlighting the relationship between this disorder and linguistic components involved in reading (Vellutino et al., 2004). Dyslexia can impact other spheres of a person's life, such as substantially higher rates of depression and anxiety, juvenile delinquency, school dropout, and a lower chance of future employment (Baker and Ireland, 2007; Brooks, 2014; Daniel et al., 2006; McNulty, 2003; Sabornie, 1994; Wiener and Schneider, 2002). The identification of the root causes of dyslexia could better inform the necessary conditions for environmental enrichment and improve the development of clinical tools for its early diagnosis. This, in turn could benefit the education and future employment of millions of children (Goswami, 2014).

A rich literature indicates a link between dyslexia and sensory dysfunctions in both the visual and auditory domains (Giraud and Ramus, 2013; Gori and Facoetti, 2015; McArthur and Bishop, 2001; Rosen, 2003; Schulte-Körne and Bruder, 2010; Valdois et al., 2004). However, it has been argued that such a link does not necessarily inform us about the root cause of dyslexia, which is currently under debate (Goswami, 2014). One reason is that learning to read moulds the brain by training its sensory and attentional neural networks, so sensory dysfunctions may be a result of diminished reading experience, rather than being directly linked with dyslexia (Bishop et al., 2012). This poses an additional challenge for the study of potential sensory causes of dyslexia, and highlights the importance of including a typically-developing control group matched in reading achievement to the children with dyslexia (the ‘reading-level-match’ research design, to control for the effects of reading experience on the brain). One area of relative agreement concerns the hypothesis that a proximal cause of dyslexia is a behavioural ‘phonological deficit’, encompassing all levels of phonology (prosody, syllables, rhymes, and the short categorical speech units reflected by the alphabet, phonemes) (Clark et al., 2014; Goswami, 2014, 2011; Goswami and Leong, 2013; Lehongre et al., 2013; Richlan, 2012). Further, it is generally agreed that the speech processing mechanisms that yield these phonological units are underpinned by a hierarchical system, whose critical mechanisms lie in the infrastructure provided by neuronal oscillations (Giraud and Poeppel, 2012; Leong and Goswami, 2014). It is possible that phonological deficits associated with dyslexia stem from an impairment in these fundamental mechanisms related to neuronal oscillations, such as the temporal sampling of the auditory input (temporal sampling framework, Goswami, 2011).

Stimulus-induced modulations in delta-, theta-, and gamma-bands (1–4 Hz, 4–8 Hz, and >25 Hz respectively) have been shown to reflect successful speech comprehension and processing related to different speech units (e.g., phrasal, syllabic, phonemic) (Ghitza, 2011; Poeppel, 2003). Furthermore, recent studies have demonstrated links between dyslexia and anomalies in specific neural oscillations. For example, deficits in the processing of slow temporal modulations (e.g., <8 Hz) have been related to difficulties in perceiving both syllable stress and the phonemic constituents of syllables, difficulties which would likely impair the development of well-specified phonological representations of words (Goswami, 2011; Molinaro et al., 2016; Poeppel, 2003; Power et al., 2016). In addition, another factor that has been linked with dyslexia is working memory, the temporary storage system necessary for a wide range of complex cognitive activities, including speech and language processing (Baddeley, 2003). Recent research has shown stronger cortical entrainment for high-frequency oscillations (>40 Hz) in adults with dyslexia (Lehongre et al., 2011) and it has been suggested that this is due to oversampling of speech information, causing an excessive demand on working memory, and consequent impairment of its related cortical functions (Giraud and Poeppel, 2012).

These and other studies with both adults and children using both speech and non-speech stimuli have demonstrated differences in both amplitude and hemispheric bias between individuals with dyslexia and control groups (Abrams et al., 2009; Cutini et al., 2016; Goswami, 2011; Heim et al., 2003; Peter et al., 2016; Vanvooren et al., 2014). However, the root causes of such processing biases remain unclear. One reason for this lack of clarity is that most studies did not control for differences in reading skills via a reading-level-match control group and, therefore, did not provide the conditions necessary to assess possible causality regarding dyslexia (Goswami, 2014). A second reason for this lack of clarity is that, although dyslexia has been shown to affect phonological skills regardless of age, different phonological skills may be affected in different age-groups, which would likely be reflected in the neural responses to speech (Miller-Shaul, 2005). A third reason is that neurophysiological studies have usually been conducted using non-naturalistic stimuli such as isolated syllables (Power et al., 2013), modulated noise (Lehongre et al., 2011), and noise-vocoded speech (Power et al., 2016).

While a tailored experimental design is sufficient to overcome the first two limitations, the use of more naturalistic stimuli in neurophysiological studies is less straightforward. Indeed, the ability to derive objective neural measures of phonological processing using natural speech could be key to clarifying the cortical underpinnings of dyslexia and, in particular, of the corresponding phonological deficit. The complexity of natural speech and associated cortical responses poses a challenge that has only been tackled quite recently. Specifically, recent research has demonstrated that cortical oscillations track the low-frequency rhythms of incoming speech stimuli. In a growing body of literature this entrainment phenomenon is being investigated by focusing on the mapping between the temporal envelope of speech and neurophysiological recordings such as electroencephalography (EEG; Aiken and Picton, 2008; Ding et al., 2017), magnetoencephalography (MEG; Ahissar et al., 2001), and electrocorticography (ECoG; Pasley et al., 2012). To date, low-frequency cortical oscillatory mechanisms have been thought primarily to aid syllable parsing and the identification of stressed syllables. However, recent research from Di Liberto and colleagues has demonstrated that low-frequency EEG signals track also phoneme categories. This provides us for the first time with a methodology to isolate quantitative measures of children's phoneme-level processing using natural continuous speech (Crosse et al., 2016; Di Liberto et al., 2015).

Here we used this novel methodology to objectively measure whether impaired cortical tracking of the temporal envelope of speech directly affects the representation of phoneme-level units in dyslexia. We investigated this in dyslexic and non-dyslexic school-aged children using EEG, controlling for the effects of age and reading level. In addition, correlational analyses were conducted between the neural measures at individual scalp electrodes and the results of a standard battery of behavioural tests of language skills, memory capacity, and attention used in dyslexia diagnosis.

Section snippets

Material and methods

Seventy children (26 female) aged between 6 and 12 years (mean = 8.6 years, SD = 1.5), who were monolingual speakers of Australian English, participated in the experiment. The ethics committee for Human Research at Western Sydney University (Approval Number H9660) approved all the experimental methods used in the study. Informed consent was obtained from the parents of all the participants. Children also gave verbal assent for the study.

Results

70 children (6–12 years of age) undertook a standard battery of behavioural tests of phonological and language skills, memory capacity, IQ, and attention. These behavioural tests identified 25 participants with the typical symptoms of dyslexia (DX group). In a separate session, 129-channel EEG was recorded as participants listened to an audio-story for 9 min, while watching the corresponding cartoon. Scalp electrical signals were analysed to test the hypothesis that dyslexia is linked to an

Discussion

This study investigated the cortical underpinnings of developmental dyslexia, a developmental disorder of learning whose root causes remain debated, but with a reasonable consensus around impairments in phonological processing (Snowling, 2000). Recently, it has been suggested that phonological impairments in individuals with dyslexia may relate to atypical function of the cortical oscillatory mechanisms of temporal sampling and phase-locking at <10 Hz that underpin continuous speech

Conflicts of interest

None declared.

Funding sources

This study was supported by an Irish Research Council Government of Ireland Postgraduate Scholarship (GOIPG/2013/1249) awarded to the first and sixth authors and an Australian Research Council Discovery Project grant (DP110105123) awarded to the fourth and fifth authors.

Author contributions

The study was conceived by V.P., M.K., D.B., and G.D.L.; V.P. programmed the tasks; V.P. and M.K. collected the data; G.D.L. analysed the data; M.K. and V.P. analysed the behavioural data; G.D.L. wrote the first draft of the manuscript; V.P., M.K., D.B., U.G. and E.C.L. edited the manuscript.

Acknowledgements

This study was supported by an Irish Research Council Government of Ireland Postgraduate Scholarship to the first author (GOIPG, 2013–2017), and an Australian Research Council Discovery Project grant (DP110105123) to the fourth and fifth authors. The authors would like to thank all the children and their families who invested so many long hours during the testing sessions for this study. The authors also thank Maria Christou-Ergos, Scott O'Loughlin, Samra Alispahic and Elizabeth Byron for their

References (58)

  • E. Maris et al.

    Nonparametric statistical testing of EEG-and MEG-data

    J. Neurosci. Methods

    (2007)
  • V. Peter et al.

    Neural processing of amplitude and formant rise time in dyslexia

    Dev. Cogn. Neurosci.

    (2016)
  • D. Poeppel

    The analysis of speech in different temporal integration windows: cerebral lateralization as “asymmetric sampling in time.”

    Speech Commun.

    (2003)
  • A.J. Power et al.

    Neural encoding of the speech envelope by children with developmental dyslexia

    Brain Lang.

    (2016)
  • S. Rosen

    Auditory processing in dyslexia and specific language impairment: is there a deficit? What is its nature? Does it explain anything?

    J. Phon.

    (2003)
  • G. Schulte-Körne et al.

    Clinical neurophysiology of visual and auditory processing in dyslexia: a review

    Clin. Neurophysiol.

    (2010)
  • D.A. Abrams et al.

    Abnormal cortical processing of the syllable rate of speech in poor readers

    J. Neurosci.

    (2009)
  • E. Ahissar et al.

    Speech comprehension is correlated with temporal response patterns recorded from auditory cortex

    Proc. Natl. Acad. Sci. U. S. A.

    (2001)
  • S.J. Aiken et al.

    Human cortical responses to the speech envelope

    Ear Hear

    (2008)
  • Y. Benjamini et al.

    The control of the false discovery rate in multiple testing under dependency

    Ann. Stat.

    (2001)
  • D.V.M. Bishop

    The Children's communication checklist: CCC-2

    ASHA

    (2003)
  • D.V.M. Bishop et al.

    Auditory deficit as a consequence rather than endophenotype of specific language impairment: electrophysiological evidence

    PLoS One

    (2012)
  • M.Y. Brooks

    School, Disability Status, and Delinquency: an Examination of Delinquency Among Rural Adolescents

    (2014)
  • K.A. Clark et al.

    Neuroanatomical precursors of dyslexia identified from pre-reading through to age 11

    Brain

    (2014)
  • M.J. Crosse et al.

    The multivariate temporal response function (mTRF) toolbox: a MATLAB toolbox for relating neural signals to continuous stimuli

    Front. Hum. Neurosci.

    (2016)
  • S.S. Daniel et al.

    Suicidality, school dropout, and reading problems among adolescents

    J. Learn Disabil.

    (2006)
  • G.M. Di Liberto et al.

    Low-frequency cortical entrainment to speech reflects phoneme-level processing

    Curr. Biol.

    (2015)
  • N. Ding et al.

    Characterizing neural entrainment to hierarchical linguistic units using electroencephalography (EEG)

    Front. Hum. Neurosci.

    (2017)
  • O. Ghitza

    Linking speech perception and neurophysiology: speech decoding guided by cascaded oscillators locked to the input rhythm

    Front. Psychol.

    (2011)
  • Cited by (102)

    View all citing articles on Scopus
    1

    These authors contributed equally.

    2

    Co-senior authors.

    View full text