Neurobiologic responses to speech in noise in children with learning problems: deficits and strategies for improvement

Abstract

Objectives: Some children with learning problems (LP) experience speech-sound perception deficits that worsen in background noise. The first goal was to determine whether these impairments are associated with abnormal neurophysiologic representation of speech features in noise reflected at brain-stem and cortical levels. The second goal was to examine the perceptual and neurophysiological benefits provided to an impaired system by acoustic cue enhancements.

Methods: Behavioral speech perception measures (just noticeable difference scores), auditory brain-stem responses, frequency-following responses and cortical-evoked potentials (P1, N1, P1′, N1′) were studied in a group of LP children and compared to responses in normal children.

Results: We report abnormalities in the fundamental sensory representation of sound at brain-stem and cortical levels in the LP children when speech sounds were presented in noise, but not in quiet. Specifically, the neurophysiologic responses from these LP children displayed a different spectral pattern and lacked precision in the neural representation of key stimulus features. Cue enhancement benefited both behavioral and neurophysiological responses.

Conclusions: Overall, these findings contribute to our understanding of the preconscious biological processes underlying perception deficits and may assist in the design of effective intervention strategies.

Introduction

Recent studies indicate that speech-sound perception deficits may contribute to the learning problems (LP) of some children. In particular, these children have difficulty discriminating between acoustically similar sounds (Tallal and Piercy, 1974, Tallal, 1980, Elliott et al., 1989, Stark and Heinz, 1996, Kraus et al., 1996, Bradlow et al., 1999). Moreover, these deficits become worse in the presence of background noise (Nabelek and Pickett, 1974, Elliott, 1979, Bellis, 1996, Chermak and Musiek, 1997). While the underlying cause of speech-sound perception deficits remains controversial (Tallal and Piercy, 1974, Tallal, 1980, Nittrouer, 1992, Studdert-Kennedy and Mody, 1995, Denneberg, 1999), new evidence suggests that basic neurophysiologic processes related to stimulus encoding and discrimination may be involved. Three recent studies have begun to elucidate the biological bases of impaired speech perception in some individuals with LP. First, poor readers differed from good readers in neural recovery time of auditory cortical responses to rapidly presented stimuli (Nagarajan et al., 1999). Second, dyslexic individuals displayed significantly smaller far-field EEG amplitude modulated following responses than normal subjects (McAnally and Stein, 1997). Finally, a subset of children with LP showed a significant reduction in cortical responses to speech-sound contrasts differing in rapid spectro-temporal elements, consistent with their impaired behavioral discrimination of those stimuli (Kraus et al., 1996).

Despite general acknowledgement that background noise excessively taxes perception in most children with LP, little is known about the underlying neurobiologic processes. The first goal of this investigation was to determine whether speech perception deficits in some LP children are associated with abnormal neurophysiologic representation of rapidly changing speech features in noise reflected by potentials generated at brain-stem and cortical levels. To accomplish this aim, auditory brain-stem responses (ABR), frequency-following responses (FFR) and cortical-evoked potentials were studied in a group of LP children and compared to responses in normal children.

Evaluation of these electrophysiologic measures separately and in combination provides a unique opportunity to assess the integrity of central auditory stimulus-timing mechanisms at various levels of the auditory pathway. For instance, the ABR reflects neural activity synchronized to the stimulus onset. It is generated by action potentials traveling along axons in a pattern of short-duration, biphasic responses. The magnitude of the ABR depends on a high degree of synchronized firing among the neurons, such that deviations of tenths of milliseconds are considered diagnostic of brain-stem pathology (Starr and Don, 1988). If there is excessive neural ‘jitter’, which might occur in an impaired auditory system, the separation of individual neural responses by even a fraction of a millisecond could cause responses to cancel each other out. The FFR also depends on a high degree of neural synchrony. It reflects brain-stem-generated, phase-locked responses to the low frequency components of a stimulus (less than 800 Hz) (Worden and Marsh, 1968). Differences between brain-stem and cortical responses are particularly apparent in the overall spectra of the evoked responses (ABR ∼1 kHz, Boston and Moller, 1985; cortical potentials ∼10 Hz, Moller, 1994). Cortical responses reflect the summation of excitatory post-synaptic potentials originating from multiple generator sites in response to stimulus onset and other acoustic features of the stimulus. These slow dendritic events can be separated by several milliseconds and will still sum constructively. Nevertheless, cortical potentials do depend on stimulus-locked synchronous firing across neural ensembles.

The second goal of this study was to examine the perceptual and neurophysiological benefits provided to an impaired system by acoustic cue enhancements typical of ‘clear’ speech (Picheny et al., 1986) in order to gain a deeper understanding of how processing deficits can be overcome by the speech signal. Research has shown that speakers naturally alter the acoustic characteristics of their speech from a ‘conversational’ to a ‘clear’ speaking style when the listener is known to have speech perception difficulties. The acoustic characteristics of ‘conversational’ and ‘clear’ speech have been well described (Picheny et al., 1986). The perceptual benefits of ‘clear’ speech have also been established (Picheny et al., 1985, Gordon-Salant, 1986, Hazan and Simpson, 1998), and some of these features have been incorporated into commercially available auditory training programs designed for LP children (Merzenich et al., 1996, Tallal et al., 1996).