Auditory Processing in Children with Specific Language Impairment: A FFR Based Study

Abstract

Specific language impairment (SLI) is diagnosed when a child has difficulty in producing or understanding spoken language for no apparent reason. The study attempted to assess the sub-cortical encoding in children with SLI using speech-evoked auditory brainstem response (ABR). The objective of the study was to compare the amplitude and latency of the frequency following response (FFR) parameters between the children with SLI and typically developing children. The frequency following response was recorded using/da/stimuli from ten ears of children diagnosed with SLI. The amplitude and the latencies of the different peaks of FFR in children with SLI were compared with those of typically developing children. The results of the study showed that the latencies of wave C and D were significantly prolonged in children with SLI compared to typically developing children. The waveforms obtained from the typically developing (TD) children were clearer and easily identifiable, with larger negativity observed in the troughs. The waveform morphology was poorer in children with SLI with shallower peaks. Thus, it can be concluded that speech evoked ABR gives an insight into the auditory processing ability of children with SLI. It indicates that signal processing in the auditory pathway of children with SLI is temporally distorted and which might affect the development of language.

Introduction

Specific language impairment (SLI) is diagnosed when a child has difficulty in producing or understanding spoken language for no apparent reason [1]. By definition, children with SLI are thought to have no obvious hearing, cognitive, or neurological deficits [2]. It is widely accepted that the causes of SLI are predominantly neurobiological rather than socio-cultural [3]. Some studies report that the difficulty faced by children with SLI could be related to subtle auditory perceptual problems or auditory agnosia [4]. It has been proposed that SLI is the consequence of low-level abnormalities in auditory perception [1]. Studies are reporting impaired auditory neural processing of both speech and non-speech stimuli in groups of children and adults with various language difficulties [5]. This can be because of the inability to perceive the acoustic cues present in the speech signal. This poor perception of acoustic cues that includes the temporal and spectral characteristics of speech leads to an unstable representation of phoneme in the brain. The speech perception ability and the development of phonology, syntax, and semantics of the language are affected by these unstable representations of phonemes leading to language impairment [6, 7].

Electrophysiological tests are essential to assess the auditory neural structures which are responsible for these subtle auditory functions. The studies on the electrophysiological tests in children with SLI highlight the subtle auditory processing deficits, which in turn affect the receptive and expressive language development in this population. The information regarding the sequence, timing, and location of neural events that are involved in auditory processing can be obtained through electrophysiological studies [8]. Subtle auditory processing deficits in children with SLI can be identified using auditory evoked potentials [1, 9]. Irrespective of a child’s developmental age or language, motivation or attention level, auditory evoked potentials can be recorded [10]. Fewer studies have been conducted on electrophysiological tests using complex signals such as speech to assess the accuracy of brainstem representation of timing events.

Several complex stimuli can be used to elicit the ABR. But speech (specifically/da/) is the most commonly used stimuli because it provides us cues regarding the temporal and spectral features preserved in the brainstem [11]. In recent years, speech evoked ABR is being used as a means to study the brainstem encoding of speech sounds [12, 13]. It is a valid and reliable tool to assess the neural transmission of speech stimuli at the brainstem level. When speech ABR is elicited with the stimulus/da/, the subcortical response emerges as a waveform of seven identifiable peaks, labeled V, A, C, D, E, F, and O. This response is known as the speech evoked ABR. The two measures in speech evoked ABR is the transient and frequency-following responses (FFR), which provides the information regarding auditory pathway encoding of consonant and vowel portion of the stimuli. The response to the onset of the speech stimulus includes a positive peak (wave V), followed by a negative trough (wave A). Following the onset response peaks ‘C’ ‘D’ ‘E’ ‘F’ are present in the FFR portion. Wave O is the offset of the response [11, 14].

Understanding neural processing at the brainstem level may assist in understanding outcomes in various populations such as individuals with hearing loss, language disorders, and learning deficits. The speech-evoked ABR provides a physiologic representation of poor speech encoding evident in children with language, literacy, reading, and learning deficits [15, 16]. Thus, the present study aimed to assess the sub-cortical encoding in children with SLI using speech-evoked ABR. The objective of the study was to compare the different parameters of speech-evoked ABR between the children with SLI and typically developing children.

Method

The study was conducted on ten children in the age range of 4–7 years with bilateral normal peripheral hearing. The children were divided into two groups: a control group consisting of five children with normal speech and language development and a study group consisting of five children with specific language impairment. The children were evaluated through a diagnostic protocol that included Clinical Evaluation of Language Fundamentals (CELF - Preschool 2) language test [17], which was assessed by an experienced speech-language pathologist.

Control group: Children in this group were selected from a normal school who had normal intelligence and intact bilateral hearing based on the information obtained by interviewing their respective teachers. The results from the language test CELF showed that the children in the control group had normal language skills.

SLI group: The children in this group had an Intelligence Quotient of > 85, which was assessed by an experienced psychologist. Children in this group were selected using the inter-national reference criteria [18], namely: persistent speech and/or language difficulty, with normal bilateral hearing; no changes in cognitive development; speech motor development impairment; comprehensive developmental disorders, syndromes, and sensorineural changes; and acquired neurological lesions [19, 20]. Children who had come for the assessment of their speech and language skills and diagnosed as having specific language disorder or impairment were also included in the study. There was a significant difference in CELF scores between the typically developing children (mean = 148.8) and children with SLI (mean = 56.3). Children with SLI scored poorly on the CELF language test. The significant difference was observed in the core language scores (|z| = 2.9; p = 0.004) and indices such as Receptive Language Index (|z| = 2.7; p = 0.006), Expressive Language Index (|z| = 2.9; p = 0.004), Language Content Index (|z| = 2.4; p = 0.016) and Language Structure Index (|z| = 2.9; p = 0.004) between the two groups.

The neuro-audiological test included recording speech-evoked auditory brainstem response (Frequency Following Response using speech stimuli), which assesses the sub-cortical area of the auditory pathway, which is responsible for speech encoding. Children were seated in an attenuated sound chamber and were instructed to ignore the auditory stimuli. Biologic Navigator Pro evoked potential system was used to elicit the auditory evoked potential. FFR was obtained using a single channel recording with an inverting electrode on the left/right ear mastoid, non-inverting (Fz) on the upper forehead, and the ground electrode on the right/left ear mastoid depending upon the ear in which the stimulus was presented. The stimulus was presented through electrically shielded insert earphones. A 40 ms speech stimuli/da/generated by Nina Kraus at Northwestern University was used to record FFR [14]. The stimulus was presented at 80 dB nHL at a repetition rate of 7.1/s. 3000 number of stimuli were used to elicit FFR. The filter setting was set between 30 and 1500 Hz, with the amplification of 75,000 times, and a time window of 100 ms was set. The latency and absolute amplitude of waves V, A, C, D, E, F, and O of FFR were obtained. Latency was calculated by the Biologic system as the time duration between the onset of the stimulus and the response obtained from the auditory pathway in the form of waveforms. The absolute amplitude was measured as the trough which occurred within a given latency range and had the greatest negative amplitude from the baseline. The accuracy of the markings of the latency and amplitude of the waveforms were rechecked by two experienced audiologists.

Results

The waveform morphology, latency, and amplitude of the peaks of speech-evoked ABR were studied. The collected data was analyzed using Statistical Package for the Social Sciences SPSS software. The mean and standard deviation of the variables were calculated in both groups. The mean and standard deviation (SD) of latency and amplitude in the right and left ears in children with SLI and typically developing children are shown in Tables 1 and 2, respectively.

Table 1.

Mean and SD of latency in children with SLI and Typically developing children

Components of FFR	Children with SLI				Typically developing children
	Mean in ‘ms’		SD		Mean in ‘ms’		SD
	Right	Left	Right	Left	Right	Left	Right	Left
V	6.32	6.21	.34	.21	6.1	6.13	.4	.26
A	7.78	7.92	1.09	.94	7.62	7.94	1.40	1.37
C	19.65	19.57	1.95	2.55	16.08	15.47	2.46	2.06
D	28.18	27.42	3.33	3.41	22.26	22.47	1.71	1.84
E	33.46	34.39	2.34	2.71	30.63	30.98	3.56	4.08
F	40.12	40.17	3.52	1.22	39.51	39.25	3.95	3.92
O	47.89	47.80	2.36	3.03	50.17	49.97	1.08	.66

Table 2.

Mean and SD of amplitude in children with SLI and Typically developing children

Components of FFR	Children with SLI				Typically developing children
	Mean in ‘μv’		SD		Mean in ‘μv’		SD
	Right	Left	Right	Left	Right	Left	Right	Left
V	.24	.3	.12	.14	.25	.26	.14	.07
A	− .30	− .37	.17	.23	− .15	− .25	.09	.12
C	− 2.54	− 001.81	2.73	2.04	− .31	− .17	.37	.33
D	.15	.03	.74	.6	− .71	− .25	.82	.16
E	.16	.14	.95	.93	− .14	− .15	.45	.25
F	− .98	− .64	.87	.66	− .29	− .23	.16	.15
O	− 1.3	− 1.13	1.4	.92	− .4	− .31	.41	.31

The mean latency obtained in the SLI group is prolonged than the mean latency obtained in typically developing children, which suggests that there is an overall delay in the transmission of information to the higher centers. However, a constant amplitude pattern could not be observed between the two groups across the peaks. Mann–Whitney U test was carried out to assess the significant difference between the groups where the latency and amplitude of FFR were the dependent variables. The latency of the ‘C’ wave of FFR was significantly delayed in the right (|z| = 2.4; p = 0.017) and left (|z| = 2.5; p = 0.014) ears in children with SLI. Similarly, the latency of the ‘D’ wave was significantly delayed in the right (|z| = 3.1; p = 0.002) and left (|z| = 2.6, p = 0.01) ears in children with SLI. There was no significant difference (p > 0.05) in right ear latency of waves V (|z| = 0.51; p = 0.6), A (|z| = 0.25; p = 0.79), E (|z| = 0.9; p = .36), F (|z| = 0.9; p = 0.6) and O (|z| = 1.87; p = 0.06). Similar results were also found in the left ear latency of waves V (|z| = 1.03; p = 0.3), A (|z| = 0.06; p = 0.94), E (|z| = 1.42; p = 0.15), F (|z| = 0.51; p = 0.6) and O (|z| = 1.03; p = 0.3). This can be visualized in Fig. 1. The amplitude of the components of FFR in right ear, V (|z| = 0; p = 1), A (|z| = 1.81; p = 0.06), C (|z| = 1.42; p = 0.15), D (|z| = 1.42; p = 0.15), E (|z| = 0.0; p = 1), F (|z| = 1.16; p = 0.24) and O (|z| = 0.9; p = 0.36) also showed no significant difference between the two groups. Similar results were also obtained for the amplitude of the waves of left ear FFR, V (|z| = 0.88; p = 0.37), A (|z| = 1.55; p = 0.12), C (|z| = 1.55; p = 0.12), D (|z| = 0.96; p = 0.33), E (|z| = 0.12; p = 0.89), F (|z| = 1.03; p = 0.3) and O (|z| = 1.67; p = 0.09). This can be seen in Fig. 2.

Fig. 1.

Comparison of latency in both the ears between the groups

Fig. 2.

Comparison of amplitude in both the ears between the groups

The waveforms obtained from the typically developing children were clearer and easily identifiable, with larger negativity observed in the troughs. Sharper and smooth VA complex was seen. The waveform morphology was relatively poorer in most of the children with SLI. Shallower peaks, particularly shallower VA slopes, were obtained. The waveform morphology of FFR in a child with SLI and a typically developing child are shown in Fig. 3.

Fig. 3.

Waveform morphology of FFR in a child with SLI and a TD child

Discussion

The occurrence of SLI might be due to the difficulty in the perception of acoustic cues present in the speech signal. This is related to the changes in the ability to process the signals and also to the abnormalities of neural coding of the information present in the speech signal [18, 21]. A significant difference in C and D wave latencies indicate abnormal processing of the transition portion of the stimulus. Russo, Nicol, Musacchia, & Kraus [12] and Johnson, Nicol & Kraus [22] have reported that the generation of wave ‘C’ is mainly due to the transition portion of the acoustic stimuli. Thus the delay in wave ‘C’ latency in children with SLI indicates that there is a delayed onset of voicing. Delayed initiation of fundamental frequency (Fo) processing is reflected in delayed latency of the ‘D’ wave [12, 22]. There was no significant difference observed in the other components of speech evoked ABR, V, A, E, F and O. This indicated that children with SLI did not exhibit difficulty in processing the onset response which is depicted by peaks V and A, which represents the burst onset of a voiced consonant. They have intact processing of the harmonic portion of the speech stimulus as depicted by the waves E and F. Intact processing of the offset of the stimulus as depicted by the wave O [11, 14]. In concordance with the present study, studies have shown delayed latencies and reduced amplitudes in few or all the components of speech evoked ABR in children with SLI [23, 24]. Previous research has also reported that children with known language-based learning problems exhibited delayed latencies for waves C, O, and A [16, 25] compared to their normal learning peers. Altogether, these studies show a trend that difficulties in language, literacy, reading, and learning affect the subcortical representation of speech and that delayed response latencies tend to be associated with these difficulties. Also, poor waveform morphology in the SLI group indicates a deficit in the synchronization of auditory neurons of the subcortical brainstem area that encodes rapidly presented speech stimulus [24]. Poor synchronous firing of neurons suggests abnormal temporal processing in this population. As a result of this study, we can consider temporal processing as an important factor for speech perception in children with SLI [26].

Children with SLI have difficulty in perceiving and processing the rapid changes of the spectral characteristics in the speech signal along the auditory pathway, which are essential for language development. There is instability in the representation of speech sounds (phonemes) as a result of changes in auditory processing skills [6, 7]. Therefore, the development of phonology, semantics, or syntax of a language might be disrupted because of weak representations of phonemes, which in turn disrupt the understanding of speech in children with SLI [6]. All these results independently or together suggest that the quality of speech presented at the level of the auditory cortex might be affected in children with SLI. This might deteriorate the processing of phonemes or distort the phoneme perception leading to SLI.

Conclusion

Thus, it can be concluded that speech evoked FFR gives an insight into the auditory processing ability of children with SLI. One can expect abnormal processing of rapidly changing acoustic perception of the stimulus, which is reflected either in wave ‘C’ or initiation of coding of F0 as reflected in wave ‘D.’ However, the absolute measure of amplitude may not be a suitable tool to predict abnormal auditory processing. Morphology of the waveform, which is a better indicator of the synchronized activity of the auditory nervous system, can be used to understand auditory temporal processing. Thus it indicates that signal processing in the auditory pathway of children with SLI is temporally distorted and might be leading to impaired language development. However, this study was conducted with a small sample size. Research needs to be carried out on a larger population for better generalization of the results.

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis of the results and writing the manuscript were performed by Animesh Barman, Prashanth Prabhu, Mekhala V G, Kavya Vijayan and Swapna N. The first draft of the manuscript was written by Mekhala V G and all authors commented on previous versions of the manuscript. All the authors have read and approved the final manuscript, drafted the work or revised it critically for important intellectual content, approved the version to be published and have agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

The present study is a part of an ongoing project funded by the Department of Science and Technology, Government of India, India.

Compliance with ethical standard

Conflicts of interest

The authors do not have any conflicts of interests to disclose. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Consent to Participate

All the procedures were explained to the parents of the participants, and informed consent was obtained from the parents of all the participants as they were below 10 years of age (age range: 4 – 7 years of age) of the study.

Consent for Publication

Additional informed consent for publication was obtained from the parents of all individual participants.

Ethics Approval

In the present study, all the testing procedures were carried out using non-invasive techniques, adhering to the guidelines of the Ethics Approval Committee of the institute (SH/ERB/PB-112).

Data Transperency

All the data in the present study comply with the field standards.