Elsevier

Brain Research

Volume 1205, 18 April 2008, Pages 55-69
Brain Research

Research Report
Neural mechanisms of selective auditory attention are enhanced by computerized training: Electrophysiological evidence from language-impaired and typically developing children

https://doi.org/10.1016/j.brainres.2007.10.108Get rights and content

Abstract

Recent proposals suggest that some interventions designed to improve language skills might also target or train selective attention. The present study examined whether six weeks of high-intensity (100 min/day) training with a computerized intervention program designed to improve language skills would also influence neural mechanisms of selective auditory attention previously shown to be deficient in children with specific language impairment (SLI). Twenty children received computerized training, including 8 children diagnosed with SLI and 12 children with typically developing language. An additional 13 children with typically developing language received no specialized training (NoTx control group) but were tested and retested after a comparable time period to control for maturational and test–retest effects. Before and after training (or a comparable delay period for the NoTx control group), children completed standardized language assessments and an event-related brain potential (ERP) measure of selective auditory attention. Relative to the NoTx control group, children receiving training showed increases in standardized measures of receptive language. In addition, children receiving training showed larger increases in the effects of attention on neural processing following training relative to the NoTx control group. The enhanced effect of attention on neural processing represented a large effect size (Cohen's d = 0.8), and was specific to changes in signal enhancement of attended stimuli. These findings indicate that the neural mechanisms of selective auditory attention, previously shown to be deficient in children with SLI, can be remediated through training and can accompany improvements on standardized measures of language.

Introduction

For some children, the development of oral language skills does not proceed as expected. Despite normal intelligence, this small but significant minority of otherwise typically developing children – roughly 7% of kindergarteners – experiences considerable difficulty with language acquisition (Leonard, 1998, Tomblin et al., 1997). The particular areas of linguistic weakness vary from one child to the next, and many of these children also experience subtle deficits in nonlinguistic aspects of sensory processing (for a review, see Leonard, 1998). Given the heterogeneity of deficits in children with specific language impairment (SLI), a major challenge for intervention research is to identify what skills should be targeted by language training programs and what benefits can be expected to result from such training.

Recent proposals suggest that some interventions that have been developed to improve language skills might also train, or improve language in part by training, selective attention (Gillam, 1999, Gillam et al., 2001a, Gillam et al., 2001b, Hari and Renvall, 2001, Sundberg and Lacerda, 2003). However, to date no studies have examined the effects of putative language training programs on measures of selective attention. The overarching goal of the present study was to examine whether intensive training with a computerized intervention program designed to improve language skills also leads to changes in the neural mechanisms of selective auditory attention previously shown to be deficient in children with SLI (Stevens et al., 2006).

Current hypotheses about the underlying causes of SLI differ in the relative emphasis placed on linguistic versus sensory/cognitive mechanisms. For example, several researchers have proposed that SLI is characterized by deficits in aspects of syntax and morphology (Chiat and Hirson, 1987, Rice, 1997, Rice et al., 1998, van der Lely et al., 2004). Although different researchers emphasize different linguistic processes, these theories hold in common the claim that SLI, at least in some children, is fundamentally a language-specific disorder; that is, a disorder specific to the linguistic system. In contrast, other researchers have proposed that SLI emerges from more domain-general deficits in sensory/cognitive mechanisms, such as reduced speed of processing or difficulty integrating rapidly presented auditory stimuli (Bishop and McArthur, 2005, Leonard, 1998, McArthur and Bishop, 2005, Tallal et al., 1996). Such domain-general deficits are proposed to have particularly profound consequences for those aspects of language that place heavy processing demands on less perceptually salient phonemes and morphemes (Leonard, 1998, Leonard, 1989).

Consistent with the hypothesis that domain-general deficits are characteristic of SLI, several recent behavioral studies have reported deficits in selective attention in children with disorders of language, including SLI and dyslexia (Asbjørnsen and Bryden, 1998, Asbjørnsen et al., 2003, Atkinson, 1991, Cherry and Kruger, 1983, Klein and D'Entremont, 1999, Sperling et al., 2005). These studies indicate that children with disorders of language have particular difficulty selectively attending to task-relevant stimuli when co-present task-irrelevant stimuli must be actively filtered. Importantly, the observed attention deficits generalize across both linguistic and nonlinguistic contexts in both the auditory and visual modalities, suggesting that the attention deficit implicated in children with disorders of language is both domain-general and multi-sensory.

The range of deficits proposed to underlie SLI has given rise to interventions focusing on different aspects of linguistic or sensory/cognitive systems including grammar, phonological awareness, and rapid auditory processing. Although purely attention-based interventions have not been used with children with SLI, it has been proposed that some interventions for children with SLI may also influence selective attention (Gillam, 1999, Gillam et al., 2001a, Gillam et al., 2001b, Hari and Renvall, 2001, Sundberg and Lacerda, 2003). This proposal has been most prominent in discussions of the Fast ForWord intervention products. The Fast ForWord – Language program, hereafter referred to simply as FFW, targets oral language skills through intensive, computer-based activities with acoustically modified speech and nonspeech sounds (Tallal, 2004). The program is based on the theory that language deficits arise from more basic perceptual deficits in processing, and specifically in processing sounds that are separated by brief durations or are themselves short in duration (Tallal, 2004, Tallal et al., 1998). Children receiving the FFW intervention typically train with the software for 100 min per day, five days per week, for four to six weeks. To date, over half a million children in thousands of public schools across the United States have used the Fast ForWord programs (Scientific Learning Corporation, 2007).

Although FFW is typically considered an intervention targeting rapid auditory processing, the creators of FFW contend that the program also influences the domain-general systems of memory and attention.3 Other researchers have also suggested that when FFW is effective, it may work in part by training attention or other domain-general mechanisms (Gillam, 1999, Hari and Renvall, 2001, Sundberg and Lacerda, 2003). When FFW is implemented as intended, children spend sustained amounts of time (100 min/day) engaging with and attending to auditory stimuli. Thus, it is reasonable to hypothesize as some clinicians have (Gillam, 1999), that FFW training improves attention skills in children.

Recent reviews indicate that the evidence for the efficacy of FFW remains controversial (e.g., see Gillam, 1999, Gillam et al., 2001b, Troia and Whitney, 2003). Studies of individual children receiving FFW training with one-on-one coaching indicate that gains in standardized assessments of language can approach or exceed one standard deviation in magnitude (Friel-Patti et al., 2001, Loeb et al., 2001). However, the particular areas of language improvement are inconsistent across children, and there is little evidence that language gains are related to performance measures on FFW tasks (Borman and Benson, 2006, Cohen et al., 2005, Friel-Patti et al., 2001, Loeb et al., 2001). When FFW has been directly compared to other intervention programs (e.g., Lindamood Phoneme Sequencing Program, Laureate Learning Systems, Earobics), the gains from different interventions are generally indistinguishable (Cohen et al., 2005, Gillam et al., 2001a, Pokorni et al., 2004, Sundberg and Lacerda, 2003). Although there is great variability in the presence and magnitude of standardized test gains depending upon the study and particular implementation procedure, it is remarkable that training effects are seldom treatment-specific. It has been suggested (Gillam et al., 2001a, Gillam et al., 2001b, Sundberg and Lacerda, 2003) that the similar gains observed across FFW and other computer-based (Cohen et al., 2005, Pokorni et al., 2004, Sundberg and Lacerda, 2003) and interpersonally-delivered (Gillam et al., 2001a, Pokorni et al., 2004) interventions might reflect the common influence of these interventions on selective attention.

A recent neuroimaging study provides preliminary support for the hypothesis that FFW may influence attentional systems (Temple et al., 2003). In this study, twenty adolescents with dyslexia received Fast ForWord training. Before and after training, children completed a phonological processing task during functional magnetic resonance imaging (fMRI) scanning. Following training, the children with dyslexia showed increased recruitment of brain regions associated with phonological processing, as well as increased activation in the anterior cingulate, an area associated with attentional control (Bush et al., 2000). The same changes were not observed in a typically developing, no-treatment control group retested after a comparable period of time. As the authors noted, the observed increased activation in the anterior cingulate could be an indication of changes in attentional skills related to FFW training. However, to our knowledge no studies have directly examined whether training with FFW influences the neural mechanisms of selective attention.

Event-related brain potentials (ERPs) are changes in the electrical activity of the brain in response to specific events. ERPs can be recorded with surface electrodes placed on the scalp and, for over 30 years, have been used as a noninvasive method to examine the mechanisms of selective auditory attention in adult humans. In a classic selective auditory attention ERP paradigm, separate streams of auditory stimuli (e.g., tone pips) are presented to each ear (Hillyard et al., 1973). Participants attend to one of the two streams to detect rare target events, and ERPs are recorded to standard (non-target) stimuli in the attended and ignored stream. A comparison of the ERPs elicited by stimuli in the attended and ignored streams reveals the effects of selective attention on sensorineural processing: A number of studies have reported that approximately 100 ms after sounds are presented, the first negative wave (N1) of the ERP is amplified (i.e., more negative) in response to stimuli in the attended relative to the ignored stream (Hillyard et al., 1973, Hillyard et al., 1987, Woldoff and Hillyard, 1991). This early amplification likely results from the joint processes of signal enhancement of the attended stimuli and suppression of the competing stimuli presented in the ignored channel. Several studies have indicated that these early ERP attention effects have a spatial gradient (Teder-Salejarvi and Hillyard, 1998), can select for non-spatial auditory dimensions (Woods et al., 1991), and are apparent in the visual modality, as well (Hillyard and Anllo-Vento, 1998). This ERP attention effect is associated with improved behavioral performance on detection tasks, as measured by response accuracy (d-prime) and reaction time (Neville and Lawson, 1987, Roder et al., 1999, Squires et al., 1973, Teder-Salejarvi and Hillyard, 1998, Teder-Salejarvi et al., 2005).

We have recently conducted studies using similar ERP paradigms with children (Coch et al., 2005, Sanders et al., 2006). In these studies, separate narrative stories are presented to each ear, and ERPs are recorded to probe stimuli superimposed on the attended and ignored story. Although children show a broad positivity (rather than an N1) in response to probe stimuli approximately 100 ms after stimuli are presented, the broad positivity is amplified (i.e., more positive) with attention in children as young as three years of age (Sanders et al., 2006). Interestingly, unlike typically developing children, children with SLI do not show evidence of early attentional modulation in this paradigm, even when performing the task as directed (Stevens et al., 2006). Further, the deficits in SLI are linked specifically to reduced amplification of the neural response to probes in the attended channel (i.e., signal enhancement) rather than difficulties in suppression of responses to probe stimuli in the ignored channel (i.e., distracter suppression) (Stevens et al., 2006).

Given the evidence that attention skills are compromised in children with language or reading impairments (Asbjørnsen and Bryden, 1998, Asbjørnsen et al., 2003, Atkinson, 1991, Cherry and Kruger, 1983, Klein and D'Entremont, 1999, Sperling et al., 2005, Stevens et al., 2006), any effect of FFW training on attention skills may be particularly beneficial to these children. The electrophysiological data reviewed above suggest that SLI is associated with specific deficits in the neural mechanism of early auditory signal enhancement. Thus, if FFW does influence the mechanisms of selective attention, these changes might be evident in electrophysiological indices of selective auditory attention.

The present study examined whether training with Fast ForWord, when delivered in a controlled laboratory setting by trained FFW coaches, would influence the neural mechanisms of selective auditory attention previously shown to be deficient in children with SLI. Before and after receiving FFW training, children completed an electrophysiological assessment of selective auditory attention. Children were cued to attend to one of two simultaneously presented children's stories differing in content, location (a speaker located to their right or to their left), and narrator voice (male or female). ERPs were recorded to probe stimuli embedded in the attended and unattended stories. The difference in mean amplitude response to probe stimuli when attended versus unattended was compared from pre- to post-intervention. Children also completed the receptive and expressive subscales of the standardized Clinical Evaluation of Language Fundamentals-3 (CELF-3, Semel et al., 1995) assessment before and after training.

We predicted that training with FFW would increase the effects of attention on early (100-200 ms post-stimulus onset) neural processing, and that these increases would exceed any changes observed in a no-contact control group retested after a comparable period of time. Further, changes were expected to result from increases in the amplitude of response to attended stimuli (i.e., improvements in signal enhancement), as this has been identified as the specific locus of deficits in attentional modulation in children with SLI (Stevens et al., 2006). We also predicted that FFW training would lead to gains in standardized measures of receptive and expressive language.

Two groups of children received FFW training: a group diagnosed with SLI and presenting with poor receptive language skills (n = 8) and a group of typically developing children (n = 12). The inclusion of the latter group provided a strong test of the hypothesis that training with FFW would also confer advantages to children without poor language or selective attention skills. A no-treatment control group of typically developing children (n = 13) was also tested twice, at comparable time points, to control for any maturational or test–retest effects.

Section snippets

Results

Three dependent measures were examined: CELF-3 receptive language composite scores, CELF-3 expressive language composite scores, and the ERP index of selective auditory attention. Standard scores, which control for participants' age, were used in all analyses of standardized test data. A schematic of the ERP experimental paradigm is provided in Fig. 1. Table 1 presents demographic characteristics of the three groups of children (FFW-LI, FFW-TD, NoTx), separately for children with data available

Discussion

The present study provides the first direct assessment of changes in the neural mechanisms of selective auditory attention following training with the Fast Forward – Language program. Following FFW training, children with SLI showed evidence for increased effects of selective auditory attention on neural processing during a language listening task, and typically developing children showed a similar trend. These increases were nearly one standard deviation in magnitude and were not observed in a

Participants

A total of 33 children participated in this study, either as participants in FFW training or as members of a no-treatment control group. Twenty children received FFW training, including 8 children diagnosed with specific language impairment (FFW-LI) and 12 children with typically developing language (FFW-TD). An additional 13 children with typically developing language received no specialized training (NoTx control group) but were tested and retested at time points comparable to the FFW

Acknowledgments

We are grateful to the school districts, parents, and children who so generously gave of their time to participate in the study. This research was supported by NIH/NIDCD grant DC00481 to HJN. CS was supported by an NSF Graduate Research Fellowship and DC by NIH/NICHD NRSA Postdoctoral Fellowship (HD08598). The Fast ForWord licenses were provided free of charge by the Scientific Learning Corporation for the purposes of this independent research study. Paul Compton and Ray Vukcevich provided

References (66)

  • SandersL. et al.

    Selective auditory attention in 3- to 5-year-old children: an event-related potential study

    Neuropsychologia

    (2006)
  • StevensC. et al.

    Neurophysiological evidence for selective auditory attention deficits in children with specific language impairment

    Brain Res.

    (2006)
  • Teder-SalejarviW. et al.

    Auditory spatial localization and attention deficits in autistic adults

    Cogn. Brain Res.

    (2005)
  • TroiaG. et al.

    A close look at the efficacy of Fast ForWord Language for children with academic weaknesses

    Contemp. Educ. Psychol.

    (2003)
  • van der LelyH. et al.

    Grammatical language impairment and the specificity of cognitive domains: relations between auditory and language abilities

    Cognition

    (2004)
  • AsbjørnsenA.E. et al.

    The role of dichotic listening performance and tasks of executive functions in reading impairment: a discriminant function analysis

    Child Neuropsychol.

    (2003)
  • Astheimer, L., Sanders, L., in press. Listeners modulate temporally selective attention during natural speech...
  • AtkinsonJ.

    Review of human visual development: crowding and dyslexia

  • BavelierD. et al.

    Visual attention to the periphery is enhanced in congenitally deaf individuals

    J. Neurosci.

    (2000)
  • BormanG. et al.

    Can brain research and computers improve literacy? A randomized field trial of the Fast ForWord Language computer-based training program. (WCER Working Paper No. 2006-05)

    (2006)
  • BurgemeisterB. et al.

    Columbia Mental Maturity Scale (CMMS)

    (1972)
  • ChenaultB. et al.

    Effects of prior attention training on child dyslexics' response to composition instruction

    Dev. Neuropsychol.

    (2006)
  • CherryR. et al.

    Selective auditory attention abilities of learning disabled and normal achieving children

    J. Learn. Disabil.

    (1983)
  • ChiatS. et al.

    From conceptual intention to utterance: a study of impaired output in a child with developmental dysphasia

    Br. J. Disord. Commun.

    (1987)
  • CochD. et al.

    An event-related potential study of selective auditory attention in children and adults

    J. Cogn. Neurosci.

    (2005)
  • CohenD. et al.
  • CohenW. et al.

    Effects of computer-based intervention through acoustically modified speech (Fast ForWord) in severe mixed receptive–expressive language impairment: outcomes from a randomized field trial

    J. Speech Lang. Hear. Res.

    (2005)
  • FiegerA. et al.

    Auditory spatial tuning in late-onset blindness in humans

    J. Cogn. Neurosci.

    (2006)
  • Friel-PattiS. et al.

    Case studies of children using Fast ForWord

    Am. J. Speech Lang. Pathol.

    (2001)
  • GillamR.

    Computer assisted language intervention using Fast ForWord: theoretical and empirical considerations for clinical decision-making

    Lang. Speech Hear. Serv. Sch.

    (1999)
  • GillamR. et al.

    Language change following computer-assisted language instruction with Fast ForWord or Laureate Learning Systems software

    Am. J. Speech Lang. Pathol.

    (2001)
  • GillamR. et al.

    Looking back: a summary of five exploratory studies of Fast ForWord

    Am. J. Speech Lang. Pathol.

    (2001)
  • GreenC. et al.

    Action video game modifies visual attention

    Nature

    (2003)
  • Cited by (166)

    • Development of auditory perception in preschool children

      2020, International Journal of Pediatric Otorhinolaryngology
    View all citing articles on Scopus
    1

    Now at Dartmouth College, USA.

    2

    Now at University of Massachusetts, Amherst, USA.

    View full text