1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
HEARES6804_proof ■ 10 May 2014 ■ 1/8
Hearing Research xxx (2014) 1e8
Contents lists available at ScienceDirect
Hearing Research journal homepage: www.elsevier.com/locate/heares
Research paper
Brainstem response to speech and non-speech stimuli in children with learning problems Q5
Saeed Malayeri a, *, Yones Lotfi a, Seyed Abdollah Moossavi a, Reza Rostami b, Soghrat Faghihzadeh c a b c
Department of Audiology, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran Department of Psychology, Tehran University, Tehran, Iran Department of Social Medicine, Zanjan University of Medical Sciences, Tehran, Iran
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 September 2013 Received in revised form 30 March 2014 Accepted 23 April 2014 Available online xxx
Neuronal firing synchronization is critical for recording auditory responses from the brainstem. Recent studies have shown that both click and/da/synthetic syllable (speech) stimuli perform well in evoking neuronal synchronization at the brainstem level. In the present study, brainstem responses to click and speech stimuli were compared between children with learning problems (LP) and those with normal learning (NL) abilities. The study included 49 children with LP and 34 children with NL. Auditory brainstem response (ABR) to 100-ms click stimulus and speech ABR (sABR) to/da/40-ms stimulus were tested in these children. Wave latencies III, V, and Vn and inter-peak latency (IPL) VeVn in click ABR and wave latencies I, V, and A and IPL VeA in sABR were significantly longer in children with LP than children with NL. Except IPL of IeIII, a significant positive correlation was observed between click ABR and sABR wave latencies and IPLs in children with NL; this correlation was weaker or not observed in children with LP. In this regard, the difference between correlation coefficients of wave latencies I, III, and V and IPLs I eV and VeVn/VeA was significant in the two groups. Deficits in auditory processing timing in children with LP may have probably affected ABR for both click and speech stimuli. This finding emphasizes the possibility of shared connections between processing timing for speech and non-speech stimuli in auditory brainstem pathways. Weak or no correlation between click and speech ABR parameters in children with LP may have a clinical relevance and may be effectively used for objective diagnoses after confirming its sufficient sensitivity and specificity and demonstrating its acceptable validity with more scientific evidence. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Click-evoked auditory brainstem response (ABR) is the most commonly used auditory evoked potential to assess data
List of the abbreviations: ABR, Auditory brainstem response; ADHD, Attention deficit hyperactivity disorder; AEP, Auditory evoked potential; ANSD, Auditory neuropathy spectrum disorder; ANSI, American national standard institute; CI, Confidence interval; FFR, Frequency following response; IPL, Inter peak latency; LP, Learning problem; MANOVA, Multivariate analysis of the variance; NL, Normal learning; sABR, Speech ABR; SD, Standard deviation; TD, Typically-developed; WISC-R, Wechsler intelligence scale for children - revised; WIAT-II, Wechsler individual achievement test - second edition * Corresponding author. Evin St, Daneshjo St, Koudakyar St, Tehran, Iran. Tel./ fax: þ98 21 22180066. E-mail addresses: [email protected], [email protected] (S. Malayeri), [email protected] (Y. Lotfi), [email protected] (S.A. Moossavi), rezaros@ gmail.com (R. Rostami), [email protected] (S. Faghihzadeh).
processing in auditory neural pathways in the first 10 ms after stimulus presentation. This response comprises 7 waves with positive and negative peaks that can be used to estimate auditory threshold. This response can also be used for otoneurologic assessment, particularly to diagnose possible lesions along the auditory nerve and/or in auditory brainstem pathways (Burkard and Don, 2007; Hall, 2007; Roeser et al., 2007). The brainstem is a structure with high neural synchronization, and the effects of its different pathologies usually manifest as changes in the absolute latency, inter-peak latency (IPL), and/or amplitude of waves in the click ABR. The reliability of results for auditory threshold estimation and their applications in differential diagnosis of sensory-neural hearing impairments are the most important reasons for the widespread use of click ABR along with other diagnostic tests in clinical practice (Hall, 2007). Over the past two decades, use of more complex stimuli such as speech sounds or music that challenges the brainstem has been the
http://dx.doi.org/10.1016/j.heares.2014.04.009 0378-5955/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
1 2 3 4 5 6 7 8 9 10 11 12 Q1 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 2/8
2
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
focus of study (Banai et al., 2007). Researchers have studied brainstem responses to complex stimuli and the impact of training, experience, and intervention on such responses. Effects of these factors on brainstem responses to complex stimuli have been investigated in individuals with varied disorders. The findings of most of these studies are interesting in terms of displaying capabilities and functions of the auditory brainstem. In addition, most authors are of the opinion that ABR to 40 ms/da/synthetic syllable or speech ABR (sABR) can be used an objective and non-invasive tool to study the role of the brainstem in processing complex auditory stimuli such as speech in individuals with developmental and learning difficulties (Skoe and Kraus, 2010). Furthermore, many studies have reported that most children with learning problems (LPs) and attention deficit disorders are likely to have a potential impairment in the neural encoding of complex acoustical signals compared with typically developed (TD) children with normal learning (NL) abilities (Banai et al., 2009; Cunningham et al., 2001; Hornickel et al., 2011; Kraus et al., 1996; King et al., 2002; Song et al., 2006, 2008; Wible et al., 2004). By definition, children with LPs are those who cannot be effectively taught and expected to progress when regular education programs, even in the presence of support, are implemented. They need to be “pulled out” into special settings where they can receive remedial services (Will, 1986). By the time these children reach school-going age, much of the critical time for language development has already passed. Therefore, it is not surprising that these children lag behind their peers not only in language learning and speech but also in various other skills. A substantial percentage of these children exhibit difficulties in perceiving and discriminating between auditory stimuli, including speech sounds (King et al., 2002; Mody et al., 1997; Tallal and Stark, 1981; Wible et al., 2004). Studies have shown a relationship between impairment of speech recognition tasks and deficits in cortical and brainstem auditory processing times (King et al., 2002). The present study examines neural brainstem responses to click and speech stimuli in children with LPs and compares them with the results in children with NL abilities. Approximately 9% children in the United States have reading and learning problems (King et al., 2002). In recent decades, several studies have tried to investigate the different levels of auditory neural pathways, which fail to process auditory stimuli, in such children. For instance, some studies have reported that abnormal cortical processing may be responsible for the inability to process speech (Kraus et al., 1996; Lachmann et al., 2005; Paul et al., 2006) and non-speech stimuli (Baldeweg et al., 1999; Kujala et al., 2006; Nagarajan et al., 1999). Despite the heterogeneous nature of LPs and the use of different research methods, the findings of most previously mentioned studies corroborate the theory that deficits in cortical processing of auditory stimuli are responsible for most LPs in children (Song et al., 2008). Electrophysiological studies measuring auditory processing times at the brainstem level have reported no differences in click ABR between children with LP and those with NL abilities (Grontved et al., 1988; Jerger et al., 1987; Lauter and Wood, 1993; Mason and Mellor, 1984; McAnally and Stein, 1997; Purdy et al., 2002; Tait et al., 1983). In contrast, some defects in speech stimuli processing timing, such as abnormalities in wave latencies or in the spectral content of the frequency following response (FFR) have been observed in sABR recordings (Cunningham et al., 2001; King et al., 2002; Wible et al., 2004). The sABR to the/da/synthetic syllable gives information about time-locked responses that are either transient or sustained. Transient responses are the result of transient, non-periodic characteristics of the stimulus, whereas sustained time-locked responses are due to sustained, periodic characteristics of the stimulus (Kraus and Nicol, 2005; Skoe and Kraus, 2010). While most
studies investigating sABR in children with LPs have focused on the abovementioned responses, a few studies have tried to study waves with latencies lower than wave V. Song et al. (2006) examined children with language-based LP and those with NL abilities and observed a relationship between wave latencies in click ABR and sABR. The study found no significant differences between the two groups of children for click ABR in the presence or absence of noise. However, longer latencies for waves V and A increased V-A duration, resulted in shallower V-A slope, and delayed wave III latency for sABR in the absence of noise. The result for sABR in the presence of noise was not reported in this study because earlier work (Banai et al., 2005; Cunningham et al., 2001; Johnson et al., 2005; King et al., 2002; Russo et al., 2004; Wible et al., 2004) demonstrated that the/da/-evoked response sufficiently provided means to objectively identify children at a risk of LP. Moreover, Russo et al. (2004) reported that the onset waves of the/da/-evoked response in the presence of noise are not reliably identifiable. The children participating in the study were divided into two groups- children with normal sABR and children with abnormal sABR (Banai et al., 2005), and the correlation between wave V latency for click ABR and the four measures of VeA onset response for sABR were examined (latencies of waves V and A, VeA duration, VeA slope). In the normal sABR group, a significant correlation was observed between wave V latency in click ABR and latency of wave V and A in sABR. However, these findings were not consistent with those seen in the abnormal sABR group (Song et al., 2006). Song et al. (2008) investigated the hypothesis that defects in auditory processing timing in children with LP originate either from the rostral level of the brainstem or from deficits in sensory encoding at the lower levels of neural auditory pathways. The study concluded that a significant difference existed in the rostral components of sABR (waves V and A) between children with LP and those with NL abilities. The authors of the study reported that wave I of sABR did not emerge reliably in either group. Moreover, no significant differences were observed in wave III between the two groups. They concluded that the defects in the auditory processing timing in children with LP with abnormal sABR resulted from the corticofugal modulation of subcortical activity. Based on the potential shared connections in timing neural encodings between speech and non-speech stimuli at the brainstem level and the probable effects of corticofugal modulation or top-down processing of the subcortical activity, the present study compared the timing and amplitude components of click ABR and sABR between children with LPs and those with NL abilities. The relationship between matched components of click ABR and sABR was also investigated. Findings of this study corroborate the results of previous studies regarding auditory processing timing at the brainstem level and the possible relationship between processing timing of speech and non-speech stimuli. 2. Methods 2.1. Participants The study was conducted at Rofeydeh Hospital. Eighty-three children aged 8e12 years were shortlisted to participate in the study. Among these, 49 children (33 boys, mean age: 9.88 1.42 yr) had a history of LPs and the remaining 34 children (20 boys, mean age: 9.92 1.51 yr) were typically developed with NL abilities and no history of educational problems or learning difficulties at school. All the study participants were patients at the Atiyeh Comprehensive Psychiatric Center. Children with a history of neurological problems, medical diseases, major affective disorders, or schizophrenia and those under systemic treatment were excluded from the study. All the
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 3/8
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
children had an IQ between 90 and 115 on Wechsler Intelligence Scale for Children-Revised (WISC-R) (Wechsler, 1997), with no significant difference between the two groups of children. The LP group included children with clinically diagnosed reading-based learning disability (LD) and/or those with clinically diagnosed attention deficit disorder (LD/ADD). Because perceptual deficits are known to cut across diagnostic categories, children with different clinical diagnoses (LD and LD/ADD) were combined in the LP group. Pure-tone hearing thresholds for both ears were equal to or better than 25 dB HL in octave band frequencies ranging from 250 to 8000 Hz (ANSI S3.6-2004). The results of acoustic reflexes, middle ear tympanometry (ear canal volume: 0.9e2.0 cm3, static compliance: 0.3e1.5 mmho, sound pressure level: 50 dapa) (Shanks and Shohet, 2009), and word discrimination scores were within normal limits. LP children were referred by teachers, parents, or therapists; their diagnoses accorded with Wechsler Individual Achievement Test - Second Edition (WIAT-II) (Wechsler, 2001) and were confirmed by a certified clinical psychologist. The medical records of the subjects were reviewed by a child neuropsychiatrist to ensure correct diagnosis and the existence of no other diagnoses. These children had age-appropriate education and no history of special education services according to teacher reports and educational records. All children were right-handed and monolingual Persian speakers. This study was approved by the Ethics Committee of the University of Social Welfare and Rehabilitation Sciences. 2.2. Stimuli and recording parameters
Q2
Tests of click-evoked Auditory Brainstem Response (click ABR) and speech ABR (sABR) were conducted for all subjects and analyzed using the Bio-logic Navigator Pro System (Bio-logic Systems Corporation, Natus Medical Inc., Mundelein, IL). Click ABR was obtained before recording sABR. Two-channel ABRs were collected, with electrodes at Cz (noninverting), mastoids (inverting) and the forehead (ground), while impedance of all electrodes was below 5 KU and within 1.5 KU of each other. Stimuli were presented through unshielded insert earphones (ER-3A, Etymotic Research, Elk Grove Village, IL). Click ABR was recorded by 100 ms click stimulus with alternate polarity at the rate of 11.1/sec at 80 dB SPL, and using 30e3000 Hz band-pass filtering. Individual traces over 15.0 mV were removed online. Two blocks of 2000 artifact-free sweeps were collected for each subject. The sABR test stimulus used was the synthesized syllable/da/ for the duration of 40-ms as used in previous studies and obtained from the Auditory Neuroscience Lab of Nina Kraus and colleagues at Northwestern University. The stimulus consisted of an initial noise burst and formant transition between a consonant and a steady state vowel (Johnson et al., 2005). Stimuli were presented using the custom stimulus option in the standard Biologic AEP software (version 7.0.0) at a rate of 10.9/sec. Stimulus intensity level was 80 dB SPL, as measured using a 2 cm3 DB0138 coupler, Bruel & Kjaer Type 2203 sound level meter, and microphone was 1-inch in diameter. The electrodes were arranged as for click ABR. Responses were collected online by 100e 2000 Hz band-pass filter using 1024 digital sampling points over an 85.33-ms epoch. Two blocks of 2000 artifact-free sweeps were recorded in the right ear. We didn’t use any attentional task during ABR recordings. To rule out differences in state as a potential confound, our subjects stayed awake with closed eyes during testing. No formal assessment of subject state was conducted (Vander Werff and Burns, 2011). These tests were conducted in a double-walled soundtreated booth with standard noise level (ANSI S3, 1-1999, R2008) and dimmed lights.
3
2.3. Data processing for click ABR and sABR All the waves of both click ABR and sABR were identified and marked manually by 2 expert audiologists separately. Some waves were excluded from data analysis because of unreliable peak latency or inappropriate morphology. Click ABR waves of two replications were visually marked as waves I, III, V, and Vn (a negative trough after wave V). Grand mean waveforms for the click ABR of children with LP and those with NL abilities are shown in Fig. 1. For the analysis of sABR waves, four waves were identified and marked, including waves I and III and two onset waves V and A. The sABR onset waves V and A were similar to click ABR waves V and Vn (Song et al., 2006) and were marked according to the click ABR method (Hall, 1997). Therefore, wave V was considered as the biggest peak near 6 ms just before the negative slope. A negative peak was marked after the negative slope in its time expected latency. Neural synchronization to the onset of stimulation was measured by calculating the IPL of VeA, VeA peak to trough amplitude, and slope of the VeA complex. Grand mean waveforms for the sABR of children with LP and those with NL abilities are shown in Fig. 2. 2.4. Statistical analysis Normal distribution of the data was assessed using the KolmogoroveSmirnov test. Multivariate one-way analysis of variance was conducted to compare the LP and NL groups for latency and amplitude of waves I, III, V, and Vn and IPLs IeIII, IIIeV, IeV, and Ve Vn. Similar statistical analysis was performed for sABR to compare the two groups for latency and amplitude of waves I, III, V, and A and IPLs IeIII, IIIeV, IeV, and VeA. The F values, p values, and estimations of effect (partial h2) have been reported for these statistical analyses. The dependent t-test was used to compare the V-Vn slope of click ABR with the VeA slope of sABR. Pearson correlation test was used to determine the relationship between wave latencies (I, III, V and Vn) and IPLs (IeIII, IIIeV, IeV, and VeVn) for click ABR with wave latencies (I, III, V and A) and IPLs (IeIII, IIIeV, IeV and Ve A) for sABR in both the groups. Because the sample size was large enough to consider it as normally distributed, approximately a 95% confidence interval (CI) could be calculated for the difference between the two correlation coefficients for each wave component of click ABR and sABR in the two groups (Cohen and Cohen, 1983). All statistical analyses were done using SPSS Statistics 18.0 at p < 0.05. 3. Results 3.1. Comparison of click ABR findings between groups The mean and SD for latencies and amplitudes of waves and IPLs of click ABR for both the groups are shown in Table 1. Significant differences in wave latencies (Pillai trace ¼ 0.130, F ¼ 3.733, p ¼ 0.028) and IPLs (Pillai trace ¼ 0.148, F ¼ 3.179, p ¼ 0.018) were observed between the two groups. Wave latencies III, V, and Vn and IPL V-Vn are longer in children with LP than those with NL abilities. There was no significant difference between the two groups with respect to wave amplitudes (Pillai trace ¼ 0.029, F ¼ 0.425, p ¼ 0.830) and the slope of V-Vn (t ¼ 1.359, p ¼ 0.178). The slope of VeVn was similar in both the groups (NL children: mean ¼ 0.717, SD ¼ 0.162, LP children: mean: 0.733, SD ¼ 0.361). 3.2. Comparison of sABR findings between groups The mean and SD for latencies and amplitudes of waves and IPLs of sABR for both the groups are shown in Table 2. There were significant differences in wave latencies (Pillai trace ¼ 0.171, F ¼ 3.649,
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Q3 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 4/8
4
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
Fig. 1. Grand mean waveforms for the click ABR. Mean waveforms are shown for children with LP (black line) and NL children (gray line). Click ABR peaks I, III, V, and Vn are labeled.
Fig. 2. Grand mean waveforms for the sABR. Mean waveforms are shown for children with LP (black line) and NL children (gray line). sABR peaks I, III, V, and A are labeled.
Table 1 Wave latencies, amplitudes and inter-peak latencies of click ABR for the two groups. NL children Mean
SD
LP children N
Latencies (ms) I 1.539 0.109 34 III 3.669 0.115 34 V 5.434 0.130 34 Vn 6.201 0.114 33 Inter-Peak Latencies (ms) IeIII 2.153 0.105 34 IIIeV 1.765 0.115 34 IeV 3.878 0.140 34 VeVn 0.639 0.091 33 Amplitudes (mV) I 0.219 0.049 34 III 0.248 0.054 34 V 0.213 0.041 34 Vn 0.303 0.073 33 VeVn 0.549 0.127 33
Statistical results
SD
N
F
p
Partial h2
1.576 3.766 5.547 6.363
0.077 0.192 0.220 0.202
48 49 49 48
3.049 6.792 6.988 1.093
0.085 0.011
HEARES6804_proof ■ 10 May 2014 ■ 1/8
Hearing Research xxx (2014) 1e8
Contents lists available at ScienceDirect
Hearing Research journal homepage: www.elsevier.com/locate/heares
Research paper
Brainstem response to speech and non-speech stimuli in children with learning problems Q5
Saeed Malayeri a, *, Yones Lotfi a, Seyed Abdollah Moossavi a, Reza Rostami b, Soghrat Faghihzadeh c a b c
Department of Audiology, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran Department of Psychology, Tehran University, Tehran, Iran Department of Social Medicine, Zanjan University of Medical Sciences, Tehran, Iran
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 September 2013 Received in revised form 30 March 2014 Accepted 23 April 2014 Available online xxx
Neuronal firing synchronization is critical for recording auditory responses from the brainstem. Recent studies have shown that both click and/da/synthetic syllable (speech) stimuli perform well in evoking neuronal synchronization at the brainstem level. In the present study, brainstem responses to click and speech stimuli were compared between children with learning problems (LP) and those with normal learning (NL) abilities. The study included 49 children with LP and 34 children with NL. Auditory brainstem response (ABR) to 100-ms click stimulus and speech ABR (sABR) to/da/40-ms stimulus were tested in these children. Wave latencies III, V, and Vn and inter-peak latency (IPL) VeVn in click ABR and wave latencies I, V, and A and IPL VeA in sABR were significantly longer in children with LP than children with NL. Except IPL of IeIII, a significant positive correlation was observed between click ABR and sABR wave latencies and IPLs in children with NL; this correlation was weaker or not observed in children with LP. In this regard, the difference between correlation coefficients of wave latencies I, III, and V and IPLs I eV and VeVn/VeA was significant in the two groups. Deficits in auditory processing timing in children with LP may have probably affected ABR for both click and speech stimuli. This finding emphasizes the possibility of shared connections between processing timing for speech and non-speech stimuli in auditory brainstem pathways. Weak or no correlation between click and speech ABR parameters in children with LP may have a clinical relevance and may be effectively used for objective diagnoses after confirming its sufficient sensitivity and specificity and demonstrating its acceptable validity with more scientific evidence. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Click-evoked auditory brainstem response (ABR) is the most commonly used auditory evoked potential to assess data
List of the abbreviations: ABR, Auditory brainstem response; ADHD, Attention deficit hyperactivity disorder; AEP, Auditory evoked potential; ANSD, Auditory neuropathy spectrum disorder; ANSI, American national standard institute; CI, Confidence interval; FFR, Frequency following response; IPL, Inter peak latency; LP, Learning problem; MANOVA, Multivariate analysis of the variance; NL, Normal learning; sABR, Speech ABR; SD, Standard deviation; TD, Typically-developed; WISC-R, Wechsler intelligence scale for children - revised; WIAT-II, Wechsler individual achievement test - second edition * Corresponding author. Evin St, Daneshjo St, Koudakyar St, Tehran, Iran. Tel./ fax: þ98 21 22180066. E-mail addresses: [email protected], [email protected] (S. Malayeri), [email protected] (Y. Lotfi), [email protected] (S.A. Moossavi), rezaros@ gmail.com (R. Rostami), [email protected] (S. Faghihzadeh).
processing in auditory neural pathways in the first 10 ms after stimulus presentation. This response comprises 7 waves with positive and negative peaks that can be used to estimate auditory threshold. This response can also be used for otoneurologic assessment, particularly to diagnose possible lesions along the auditory nerve and/or in auditory brainstem pathways (Burkard and Don, 2007; Hall, 2007; Roeser et al., 2007). The brainstem is a structure with high neural synchronization, and the effects of its different pathologies usually manifest as changes in the absolute latency, inter-peak latency (IPL), and/or amplitude of waves in the click ABR. The reliability of results for auditory threshold estimation and their applications in differential diagnosis of sensory-neural hearing impairments are the most important reasons for the widespread use of click ABR along with other diagnostic tests in clinical practice (Hall, 2007). Over the past two decades, use of more complex stimuli such as speech sounds or music that challenges the brainstem has been the
http://dx.doi.org/10.1016/j.heares.2014.04.009 0378-5955/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
1 2 3 4 5 6 7 8 9 10 11 12 Q1 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 2/8
2
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
focus of study (Banai et al., 2007). Researchers have studied brainstem responses to complex stimuli and the impact of training, experience, and intervention on such responses. Effects of these factors on brainstem responses to complex stimuli have been investigated in individuals with varied disorders. The findings of most of these studies are interesting in terms of displaying capabilities and functions of the auditory brainstem. In addition, most authors are of the opinion that ABR to 40 ms/da/synthetic syllable or speech ABR (sABR) can be used an objective and non-invasive tool to study the role of the brainstem in processing complex auditory stimuli such as speech in individuals with developmental and learning difficulties (Skoe and Kraus, 2010). Furthermore, many studies have reported that most children with learning problems (LPs) and attention deficit disorders are likely to have a potential impairment in the neural encoding of complex acoustical signals compared with typically developed (TD) children with normal learning (NL) abilities (Banai et al., 2009; Cunningham et al., 2001; Hornickel et al., 2011; Kraus et al., 1996; King et al., 2002; Song et al., 2006, 2008; Wible et al., 2004). By definition, children with LPs are those who cannot be effectively taught and expected to progress when regular education programs, even in the presence of support, are implemented. They need to be “pulled out” into special settings where they can receive remedial services (Will, 1986). By the time these children reach school-going age, much of the critical time for language development has already passed. Therefore, it is not surprising that these children lag behind their peers not only in language learning and speech but also in various other skills. A substantial percentage of these children exhibit difficulties in perceiving and discriminating between auditory stimuli, including speech sounds (King et al., 2002; Mody et al., 1997; Tallal and Stark, 1981; Wible et al., 2004). Studies have shown a relationship between impairment of speech recognition tasks and deficits in cortical and brainstem auditory processing times (King et al., 2002). The present study examines neural brainstem responses to click and speech stimuli in children with LPs and compares them with the results in children with NL abilities. Approximately 9% children in the United States have reading and learning problems (King et al., 2002). In recent decades, several studies have tried to investigate the different levels of auditory neural pathways, which fail to process auditory stimuli, in such children. For instance, some studies have reported that abnormal cortical processing may be responsible for the inability to process speech (Kraus et al., 1996; Lachmann et al., 2005; Paul et al., 2006) and non-speech stimuli (Baldeweg et al., 1999; Kujala et al., 2006; Nagarajan et al., 1999). Despite the heterogeneous nature of LPs and the use of different research methods, the findings of most previously mentioned studies corroborate the theory that deficits in cortical processing of auditory stimuli are responsible for most LPs in children (Song et al., 2008). Electrophysiological studies measuring auditory processing times at the brainstem level have reported no differences in click ABR between children with LP and those with NL abilities (Grontved et al., 1988; Jerger et al., 1987; Lauter and Wood, 1993; Mason and Mellor, 1984; McAnally and Stein, 1997; Purdy et al., 2002; Tait et al., 1983). In contrast, some defects in speech stimuli processing timing, such as abnormalities in wave latencies or in the spectral content of the frequency following response (FFR) have been observed in sABR recordings (Cunningham et al., 2001; King et al., 2002; Wible et al., 2004). The sABR to the/da/synthetic syllable gives information about time-locked responses that are either transient or sustained. Transient responses are the result of transient, non-periodic characteristics of the stimulus, whereas sustained time-locked responses are due to sustained, periodic characteristics of the stimulus (Kraus and Nicol, 2005; Skoe and Kraus, 2010). While most
studies investigating sABR in children with LPs have focused on the abovementioned responses, a few studies have tried to study waves with latencies lower than wave V. Song et al. (2006) examined children with language-based LP and those with NL abilities and observed a relationship between wave latencies in click ABR and sABR. The study found no significant differences between the two groups of children for click ABR in the presence or absence of noise. However, longer latencies for waves V and A increased V-A duration, resulted in shallower V-A slope, and delayed wave III latency for sABR in the absence of noise. The result for sABR in the presence of noise was not reported in this study because earlier work (Banai et al., 2005; Cunningham et al., 2001; Johnson et al., 2005; King et al., 2002; Russo et al., 2004; Wible et al., 2004) demonstrated that the/da/-evoked response sufficiently provided means to objectively identify children at a risk of LP. Moreover, Russo et al. (2004) reported that the onset waves of the/da/-evoked response in the presence of noise are not reliably identifiable. The children participating in the study were divided into two groups- children with normal sABR and children with abnormal sABR (Banai et al., 2005), and the correlation between wave V latency for click ABR and the four measures of VeA onset response for sABR were examined (latencies of waves V and A, VeA duration, VeA slope). In the normal sABR group, a significant correlation was observed between wave V latency in click ABR and latency of wave V and A in sABR. However, these findings were not consistent with those seen in the abnormal sABR group (Song et al., 2006). Song et al. (2008) investigated the hypothesis that defects in auditory processing timing in children with LP originate either from the rostral level of the brainstem or from deficits in sensory encoding at the lower levels of neural auditory pathways. The study concluded that a significant difference existed in the rostral components of sABR (waves V and A) between children with LP and those with NL abilities. The authors of the study reported that wave I of sABR did not emerge reliably in either group. Moreover, no significant differences were observed in wave III between the two groups. They concluded that the defects in the auditory processing timing in children with LP with abnormal sABR resulted from the corticofugal modulation of subcortical activity. Based on the potential shared connections in timing neural encodings between speech and non-speech stimuli at the brainstem level and the probable effects of corticofugal modulation or top-down processing of the subcortical activity, the present study compared the timing and amplitude components of click ABR and sABR between children with LPs and those with NL abilities. The relationship between matched components of click ABR and sABR was also investigated. Findings of this study corroborate the results of previous studies regarding auditory processing timing at the brainstem level and the possible relationship between processing timing of speech and non-speech stimuli. 2. Methods 2.1. Participants The study was conducted at Rofeydeh Hospital. Eighty-three children aged 8e12 years were shortlisted to participate in the study. Among these, 49 children (33 boys, mean age: 9.88 1.42 yr) had a history of LPs and the remaining 34 children (20 boys, mean age: 9.92 1.51 yr) were typically developed with NL abilities and no history of educational problems or learning difficulties at school. All the study participants were patients at the Atiyeh Comprehensive Psychiatric Center. Children with a history of neurological problems, medical diseases, major affective disorders, or schizophrenia and those under systemic treatment were excluded from the study. All the
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 3/8
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
children had an IQ between 90 and 115 on Wechsler Intelligence Scale for Children-Revised (WISC-R) (Wechsler, 1997), with no significant difference between the two groups of children. The LP group included children with clinically diagnosed reading-based learning disability (LD) and/or those with clinically diagnosed attention deficit disorder (LD/ADD). Because perceptual deficits are known to cut across diagnostic categories, children with different clinical diagnoses (LD and LD/ADD) were combined in the LP group. Pure-tone hearing thresholds for both ears were equal to or better than 25 dB HL in octave band frequencies ranging from 250 to 8000 Hz (ANSI S3.6-2004). The results of acoustic reflexes, middle ear tympanometry (ear canal volume: 0.9e2.0 cm3, static compliance: 0.3e1.5 mmho, sound pressure level: 50 dapa) (Shanks and Shohet, 2009), and word discrimination scores were within normal limits. LP children were referred by teachers, parents, or therapists; their diagnoses accorded with Wechsler Individual Achievement Test - Second Edition (WIAT-II) (Wechsler, 2001) and were confirmed by a certified clinical psychologist. The medical records of the subjects were reviewed by a child neuropsychiatrist to ensure correct diagnosis and the existence of no other diagnoses. These children had age-appropriate education and no history of special education services according to teacher reports and educational records. All children were right-handed and monolingual Persian speakers. This study was approved by the Ethics Committee of the University of Social Welfare and Rehabilitation Sciences. 2.2. Stimuli and recording parameters
Q2
Tests of click-evoked Auditory Brainstem Response (click ABR) and speech ABR (sABR) were conducted for all subjects and analyzed using the Bio-logic Navigator Pro System (Bio-logic Systems Corporation, Natus Medical Inc., Mundelein, IL). Click ABR was obtained before recording sABR. Two-channel ABRs were collected, with electrodes at Cz (noninverting), mastoids (inverting) and the forehead (ground), while impedance of all electrodes was below 5 KU and within 1.5 KU of each other. Stimuli were presented through unshielded insert earphones (ER-3A, Etymotic Research, Elk Grove Village, IL). Click ABR was recorded by 100 ms click stimulus with alternate polarity at the rate of 11.1/sec at 80 dB SPL, and using 30e3000 Hz band-pass filtering. Individual traces over 15.0 mV were removed online. Two blocks of 2000 artifact-free sweeps were collected for each subject. The sABR test stimulus used was the synthesized syllable/da/ for the duration of 40-ms as used in previous studies and obtained from the Auditory Neuroscience Lab of Nina Kraus and colleagues at Northwestern University. The stimulus consisted of an initial noise burst and formant transition between a consonant and a steady state vowel (Johnson et al., 2005). Stimuli were presented using the custom stimulus option in the standard Biologic AEP software (version 7.0.0) at a rate of 10.9/sec. Stimulus intensity level was 80 dB SPL, as measured using a 2 cm3 DB0138 coupler, Bruel & Kjaer Type 2203 sound level meter, and microphone was 1-inch in diameter. The electrodes were arranged as for click ABR. Responses were collected online by 100e 2000 Hz band-pass filter using 1024 digital sampling points over an 85.33-ms epoch. Two blocks of 2000 artifact-free sweeps were recorded in the right ear. We didn’t use any attentional task during ABR recordings. To rule out differences in state as a potential confound, our subjects stayed awake with closed eyes during testing. No formal assessment of subject state was conducted (Vander Werff and Burns, 2011). These tests were conducted in a double-walled soundtreated booth with standard noise level (ANSI S3, 1-1999, R2008) and dimmed lights.
3
2.3. Data processing for click ABR and sABR All the waves of both click ABR and sABR were identified and marked manually by 2 expert audiologists separately. Some waves were excluded from data analysis because of unreliable peak latency or inappropriate morphology. Click ABR waves of two replications were visually marked as waves I, III, V, and Vn (a negative trough after wave V). Grand mean waveforms for the click ABR of children with LP and those with NL abilities are shown in Fig. 1. For the analysis of sABR waves, four waves were identified and marked, including waves I and III and two onset waves V and A. The sABR onset waves V and A were similar to click ABR waves V and Vn (Song et al., 2006) and were marked according to the click ABR method (Hall, 1997). Therefore, wave V was considered as the biggest peak near 6 ms just before the negative slope. A negative peak was marked after the negative slope in its time expected latency. Neural synchronization to the onset of stimulation was measured by calculating the IPL of VeA, VeA peak to trough amplitude, and slope of the VeA complex. Grand mean waveforms for the sABR of children with LP and those with NL abilities are shown in Fig. 2. 2.4. Statistical analysis Normal distribution of the data was assessed using the KolmogoroveSmirnov test. Multivariate one-way analysis of variance was conducted to compare the LP and NL groups for latency and amplitude of waves I, III, V, and Vn and IPLs IeIII, IIIeV, IeV, and Ve Vn. Similar statistical analysis was performed for sABR to compare the two groups for latency and amplitude of waves I, III, V, and A and IPLs IeIII, IIIeV, IeV, and VeA. The F values, p values, and estimations of effect (partial h2) have been reported for these statistical analyses. The dependent t-test was used to compare the V-Vn slope of click ABR with the VeA slope of sABR. Pearson correlation test was used to determine the relationship between wave latencies (I, III, V and Vn) and IPLs (IeIII, IIIeV, IeV, and VeVn) for click ABR with wave latencies (I, III, V and A) and IPLs (IeIII, IIIeV, IeV and Ve A) for sABR in both the groups. Because the sample size was large enough to consider it as normally distributed, approximately a 95% confidence interval (CI) could be calculated for the difference between the two correlation coefficients for each wave component of click ABR and sABR in the two groups (Cohen and Cohen, 1983). All statistical analyses were done using SPSS Statistics 18.0 at p < 0.05. 3. Results 3.1. Comparison of click ABR findings between groups The mean and SD for latencies and amplitudes of waves and IPLs of click ABR for both the groups are shown in Table 1. Significant differences in wave latencies (Pillai trace ¼ 0.130, F ¼ 3.733, p ¼ 0.028) and IPLs (Pillai trace ¼ 0.148, F ¼ 3.179, p ¼ 0.018) were observed between the two groups. Wave latencies III, V, and Vn and IPL V-Vn are longer in children with LP than those with NL abilities. There was no significant difference between the two groups with respect to wave amplitudes (Pillai trace ¼ 0.029, F ¼ 0.425, p ¼ 0.830) and the slope of V-Vn (t ¼ 1.359, p ¼ 0.178). The slope of VeVn was similar in both the groups (NL children: mean ¼ 0.717, SD ¼ 0.162, LP children: mean: 0.733, SD ¼ 0.361). 3.2. Comparison of sABR findings between groups The mean and SD for latencies and amplitudes of waves and IPLs of sABR for both the groups are shown in Table 2. There were significant differences in wave latencies (Pillai trace ¼ 0.171, F ¼ 3.649,
Please cite this article in press as: Malayeri, S., et al., Brainstem response to speech and non-speech stimuli in children with learning problems, Hearing Research (2014), http://dx.doi.org/10.1016/j.heares.2014.04.009
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Q3 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
HEARES6804_proof ■ 10 May 2014 ■ 4/8
4
S. Malayeri et al. / Hearing Research xxx (2014) 1e8
Fig. 1. Grand mean waveforms for the click ABR. Mean waveforms are shown for children with LP (black line) and NL children (gray line). Click ABR peaks I, III, V, and Vn are labeled.
Fig. 2. Grand mean waveforms for the sABR. Mean waveforms are shown for children with LP (black line) and NL children (gray line). sABR peaks I, III, V, and A are labeled.
Table 1 Wave latencies, amplitudes and inter-peak latencies of click ABR for the two groups. NL children Mean
SD
LP children N
Latencies (ms) I 1.539 0.109 34 III 3.669 0.115 34 V 5.434 0.130 34 Vn 6.201 0.114 33 Inter-Peak Latencies (ms) IeIII 2.153 0.105 34 IIIeV 1.765 0.115 34 IeV 3.878 0.140 34 VeVn 0.639 0.091 33 Amplitudes (mV) I 0.219 0.049 34 III 0.248 0.054 34 V 0.213 0.041 34 Vn 0.303 0.073 33 VeVn 0.549 0.127 33
Statistical results
SD
N
F
p
Partial h2
1.576 3.766 5.547 6.363
0.077 0.192 0.220 0.202
48 49 49 48
3.049 6.792 6.988 1.093
0.085 0.011
