Auditory neuropathy spectrum disorder (ANSD) refers to a disorder in which the neural conduction is impaired with spared cochlear amplification [1]. The prime struggle of ANSD is the poor speech perception, more so in adverse listening conditions. Understanding the pathophysiology and the underlying mechanisms resulting in poor speech perception has been of interest to the clinicians and the researchers alike. In one such attempt, auditory late latency responses (ALLR) have been related to their speech perception [1–4]. The results of these studies have shown perplexing results. Vanaja and Manjula [5] studied the latency and amplitude of late latency response in predicting the usefulness of hearing aids and FM devices in individuals with ANSD. Results revealed that those individuals with ANSD with higher amplitude LLR benefit better with hearing aids compared to those with absent LLR. The usefulness of an FM device did not correlate with the presence or absence of LLR. ALLR-based studies have generally shown that ALLRs are not present in all individuals with ANSD [1, 3]. Kumar and Jayaram [3] reported no significant difference in the mismatch negativity between normal and ANSD individuals. They could not find any correlation between speech perception scores and ALLRs. Hood [6] reported a client with Charcot Marie tooth disease who had good CAEP, robust OAE and poor ABR with moderate to severe hearing loss, yet did not find a benefit with hearing aids. On the contrary, Rance et al,[2] recorded ALLRs in children with ANSD and found that those with measurable speech recognition scores had the presence of good ALLRs. Narne and Vanaja [4] found a positive correlation between N1/P2 amplitude and speech identification scores. They also found a significant difference in N1/P2 amplitude between good and poor speech perception groups. The ALLRs have also been found to relate to the aided benefit in individuals with ANSD [4-7] .The possible reason for such difference in findings in individuals with ANSD may be due to the heterogeneity of the condition in terms of pathophysiolgy and the comorbid factors. In terms of the speech identification scores, the group is heterogeneous, and it warrants a study with a larger sample size to validate the findings. Further, all the studies in the literature have considered speech identification scores in quiet. Speech in noise perception (SPIN) is known to be a more sensitive measure in determining the speech perception difficulties faced by individuals with ANSD. Therefore, it was of interest to investigate the relationship between SPIN and the ALLR in this group.

Aim

The study aimed to investigate the relationship of ALLR with the speech identification scores in quiet as well as in noise, in individuals with ANSD.

Thirty-five individuals in the age range of 16-35 years, (Mean age: 23.17 years, SD: 6. 46) with a confirmed diagnosis of ANSD participated in the study. The diagnosis was made by an experienced audiologist based on the audiological findings following the criteria of Starr, Sininger, and Praat [8] and the report of the neurologist. The audiological profiles of individuals with ANSD tested are given in Table 1.  All the participants were native speakers of Kannada. Informed consent was taken from all the participants, and the test procedure used conformed to the ‘AIISH ethical guidelines for bio-behavioral research project involving human subjects' [9] dated 16-05-2013. Individuals with an associate middle ear infection and any other neurological abnormalities were excluded in the study. It was also made sure that all participants in both groups were literate and had passed secondary school examinations. The procedure involved a preliminary test followed by the experimental evaluations.

Preliminary tests

All the tests were carried out in an audiometric room. Preliminary test includes Puretone audiometry, Speech audiometry, Immittance audiometry which include tympanometry and acoustic reflexes, Otoacoustic emission and Auditory Brainstem Response based on the standard protocol to rule out any space-occupying lesion. The Audiological findings are summarized in Table 1.

Table 1: Audiological profile of individuals with ANSD.

Experimental tests

Behavioral test

The word identification scores were estimated in quiet as well as at 0 dB SNR [speech shaped noise] for each ear separately by presenting the stimulus through TDH 49 headphones routed through GSI Audiostar pro diagnostic audiometer. The standardized phonetically balanced (PB) list in Kannada [10] was the stimulus, and the words were presented at 40 dBSL (with reference to the SRT). Separate lists were used for SIS and SPIN to avoid any practice effect. Participants were asked to repeat back the PB word heard, and the responses were noted down. The raw scores thus obtained were converted into percentage scores for further use. 

Electrophysiological test

The ALLRs were recorded using Biologic Navigator pro (version 7.2.0) for 'da' stimuli of 180 msc. The stimulus was monaurally presented at 80dB nHL through insert earphones at1.1/sec rate. The responses were recorded using a vertical montage (Cz-M1 and Cz-M2 with the ground electrode at nasion). The analysis window was kept as 533 ms, and the recorded EEG was filtered between 0.1 and 100 Hz. Any sweep with an EEG activity beyond +/- 45μV was rejected and was not considered for the averaging. The responses were averaged across 200 sweeps, and the resultant averaged waveform was analyzed by two experienced audiologists. The responses were visually inspected by two experienced audiologists to identify P1, N1, P2 and N2 waves. If the wave were found to be present, the peak latency of the waves and the amplitude of the N1-P2 complex were noted down.  The waveform and spectra of the 'da' syllable used for LLR recording are shown In Figure 1. The grand average LLR obtained from the individuals with ANSD is depicted in Figure 2.

Figure 1: Waveform and spectra of the da syllable used for LLR recording.

Figure 2: grand average LLR obtained from the individuals with ANSD.

The participants were categorized into two groups based on the SPIN score obtained for the better ear. The 95% confidence interval was considered for grouping. Accordingly, the Good SPIN group had scored more than 24% (19 individuals) while the Poor SPIN group had a score of less than 10% (15 individuals) scores. The mean, SD, median and range of latency of P1, N1, P2, N2 latency and N1-P2 amplitude of each group is represented in Table 2.

Table 2: The mean, standard deviation (SD) median and range of the latency of P1, N1, P2 , N2 in ms and the N1-P2 amplitude of LLR in µv of the two groups (Good & Poor SPIN).

The difference in the latency of P1, N1, P2, N2 and the N1-P2 amplitude was compared between the two groups using Mann Whitney U test as data were non-normally distributed. The result showed that there is no significant difference in the absolute latency of each peak and the amplitude of N1-P2 between the two groups. The results are summarized in Table 3 ^$^#the individuals with ANSD were also categorized into two groups based on their SIS in quiet. The 95% confidence interval was considered for grouping. Accordingly, the Good SIS group had SIS more than 64% (16 individuals) while the poor SIS group had SIS less than 44% SIS (15 individuals).

Table 3: Result of Mann Whitney U test comparing latency and amplitude of LLR between the two groups.

 The difference in the latency of P1, N1, P2, N2 and the N1-P2 amplitude of LLR was compared between the two groups using independent sample t-test as data was normally distributed. The results showed that there is no significant difference in the latency and amplitude of LLR between the two groups, as shown in Table 4.

Table 4: Result of independent sample t-test comparing the latency and amplitude of LLR between the two groups.

Grouping of participants with ANSD was also done based on the presence and absence of LLR. There were 23 participants with LLR present and 12 participants with LLR absent. The difference in the mean SIS and SPIN for each ear was compared between the two groups using Mann Whitney U-test as the data was non-normally distributed. Results revealed no significant difference in the SIS as well as SPIN between the two groups as shown in Table 5.

Table 5: Result of Mann Whitney U test comparing latency and amplitude of LLR between the two groups.

The participants with ALLR present were also categorized into good ALLR groups and the poor ALLR groups based on the 95% confidence interval of amplitude of N1-P2 of the right ear.  Based on the N1-P2 amplitude, the Good LLR group had amplitude more than 6.67 µv [12 individuals] while the poor LLR group had N1-P2 amplitude less than 4.67 µv [8 individuals]. The differences in the SIS and SPIN between the two groups were tested using Mann Whitney U-test as the data were non-normally distributed. The results showed that there is no significant difference in the SIS [Z= -0.62, p>0.05] as well as SPIN [Z= -0.17, p>0.05] between the two groups. The relationship of PTA, SIS, SPIN and the duration of hearing loss, with the latency and amplitude of LLR was tested using Spearman correlation. The results showed no significant correlation of any of these parameters with their ALLR as shown in Table 6.

Table 6: Results of Pearson’s correlations of Amplitude and latency of LLR with PTA, duration of hearing loss SIS and SPIN.

The findings of the present study that the latency and amplitude of LLR do not correlate with their SIS and SPIN are in agreement with the previous studies [3]. In contrast to this, Narne and Vanaja [4] and Rance et al. [2] found correlation between N1-P2 of LLR and the SIS of individuals with ANSD. However, the correlation between SPIN and the amplitude of LLR was not attempted in their study. Individuals with ANSD usually experience trouble understanding speech in the presence of noise even with good speech identification in quiet. Hence the findings of a correlation of SIS and amplitude of LLR may not impart useful information reflecting their real-life difficulties. In the present study participants were also categorized as good and poor performers based on their SPIN scores as well as their SIS. However, no relationship between SIS as well as SPIN with the amplitude and latency of LLR was found in the present study. This difference in findings could be attributed to the heterogeneity of the condition and also could be due to the difference in the speech perception abilities among the individuals being tested. Vanaja and Manjula [5] reported hearing aid benefit with the individuals with ANSD whose cortical potential responses were present. Even though in the current study the correlation of ALLR with hearing aid benefit was not studied, there was no difference found in SIS and SPIN between the two groups of ANSD, divided based on the presence or absence of LLR. Further, an attempt was also made to evaluate the difference in the SIS and SPIN between groups of ANSD having good and poor LLR amplitude. On a similar note, no difference in SIS and SPIN was found between the groups divided based on the amplitude of LLR. The probable cause for the difference in the findings from previous studies could be the difference in the stimuli used or the lesser number of participants involved. Apart from this, heterogeneity in this population could also be a factor. In the present study, LLR could not be recorded for a few participants. This could be due to the difference in the synchrony at the cortical level in individuals with ANSD. Similar results were reported in previous literature [1, 3]. In the present study, the individuals with ANSD were categorized based on the presence or absence of LLR. In those with measurable LLR, no relationship between the latency and amplitude of LLR with the SIS and SPIN was observed. This could be probably due to the fact that LLR represents the detection or registration of stimulus at the cortical level and not the perception of speech. The findings that no relationship between the PTA, SIS, SPIN and duration of hearing loss with the latency and amplitude of LLR in individuals with ANSD can be also attributed to the heterogeneity of the condition. Among individuals with ANSD their PTA, SIS and SPIN does not show any correlations, which are cardinal clinical findings in the neural loss compared to the cochlear loss. A few studies have reported benefit with hearing aid in those individuals with ANSD who had cortical auditory evoked potential [4-12]. In the present study, we did not try to correlate the latency and amplitude of LLR with the benefit with a hearing aid, but there was no difference in the SIS and SPIN scores for those individuals whose LLR was recordable compared to those whose LLR was absent. Overall, there was no correlation of behavioral measure with the electrophysiological responses for individuals with ANSD in the present study. On a similar note, lack of correlation between the reaction time for behavioral tasks of stimulus differentiation and the neural P300 was found by Apeksha and Kumar [13].

The results of the study suggest that speech perception in individuals with ANSD is not related to their ALLRs. The current findings are derived from a larger sample size compared to all the previous similar studies. The relationship is investigated in different perspectives, and there is a consensus in the results obtained. ALLR depicts the registration of the stimulus at the cortex and is less dependent on the onset of the stimulus. Therefore, neither the presence of ALLR nor its characteristics when present can be taken as an index or indicator of good speech perception. Speech perception being a more complex process may not be dependent solely on the stimulus registration at the auditory cortex. The findings suggest that speech perception in individuals with ANSD cannot be predicted by the presence/absence of ALLR or its latency and amplitude characteristics.

I wish to thank our Director, All India Institute of Speech and Hearing, for allowing me to conduct the study. I extend my sincere thanks to all our participants for their patient cooperation.

Conflicts of interest

The author has no financial conflicts of interest

1. Starr A, Picton TW, Sininger Y, Hood LJ, Berlin CI. Auditory neuropathy. Brain. 1996; 741-53.
2. Rance G, Cone-Wesson BK, Wunderlich JL, Dowell R. Speech perception and cortical event related potentials in children with auditory neuropathy. Ear Hear. 2002; 23:239-53.
3. Kumar AU, Jayaram M. Behavioral and Brain Functions Auditory processing in individuals with auditory neuropathy. 2005; 8:1-8.
4. Narne VK, Vanaja C. Speech identification and cortical potentials in individuals with auditory neuropathy. Behav Brain Funct. 2008; 4:15.
5. Vanaja CS, Manjula P. LLR as a measure of benefit derived from hearing devices with auditory dys-synchrony. Infirst conference on Auditory Neuropathy Department of Speech pathology and Audiology, National Institute of Mental Health and Neuro sciences. 2004; 136-146.
6. Hood LJ. A Review of Objective Methods of Evaluating Auditory Neural Pathways. Laryngoscope 1999; 109:1745-1748.
7. Alvarenga KF, Amorim RB, Agostinho-Pesse RS, Costa OA, Nascimento LT, Bevilacqua MC. Speech perception and cortical auditory evoked potentials in cochlear implant users with auditory neuropathy spectrum disorders. Int J Pediatr Otorhinolaryngol. 2012;76:1332-1338
8. Starr A, Sininger YS, Pratt H. The Varieties of Auditory Neuropathy. J Basic Clin Physiol Pharmacol. 2000; 11:215-230.
9. Venkatesan S. Ethical Guidelines for Bio-behavioral Research Involving Human Subjects. All India Inst Speech Hear. 2009; 1-23.
10. Yathiraj A, Vijayalakshmi CS. phonemically balanced wordlist in Kannada. University of Mysore. 2005.
11. Yuvaraj P, Thomas A, Sabu A, Mannarukrishnaiah J. Resolving Auditory Neuropathy Spectrum Disorder in a Case with Leigh’s Disease. Journal of All India Institute of Speech and Hearing. 2016; 35:14-18.
12. Mathai JP, Yathiraj A. Performance-intensity function and aided improvement in individuals with late-onset auditory neuropathy spectrum disorder. Ear and hearing. 2017 Mar 1;38:109-117
13. Apeksha K, Kumar UA. P300 in individuals with auditory neuropathy spectrum disorder. Journal of Indian Speech Language & Hearing Association. 2017; 31:23.