Home Ahead of print Instructions Contacts
About us Current issue Submit article Advertise  
Editorial board Archives Subscribe Login   

 Table of Contents  
Year : 2015  |  Volume : 21  |  Issue : 4  |  Page : 254-259

Latency-intensity function of speech evoked auditory brainstem responses in individuals with moderate sensory neural hearing loss

Department of Speech and Hearing, School of Allied Health Science, Manipal University, Manipal, Karnataka, India

Date of Web Publication16-Oct-2015

Correspondence Address:
G Archana
Department of Speech and Hearing, School of Allied Health Science, Manipal University, Manipal, Karnataka
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/0971-7749.165759

Rights and Permissions

Introduction: Compared to click evoked auditory brainstem response (ABR), speech evoked ABR holds an additional advantage by providing information on neural encoding of speech sounds. There is limited data available on speech ABR, both in normal and pathological ears. Aim: The present study focused on finding the latency-intensity function of speech evoked ABR in moderate sensorineural hearing loss (SNHL). Materials and Methods: Speech ABR was done using the standardized protocol on 13 ears with moderate SNHL were chosen from 20 participants. Results: Speech ABR recording on these ears produced V Peak till 60 dBnHL which was at the level of 10 dBSL. Results showed that the absolute latency of V Peak at the threshold level, that is, at 60 dBnHL was found to be 8.11 ms; 6.68 ms at 70 dBnHL; 5.96 ms at 80 dBnHL; and 5.41 ms at 90 dBnHL. Conclusion: It can be concluded that using speech as a stimulus in evoked responses, it can result in better estimation of loudness growth pattern in individuals with cochlear pathology. A phenomenon like recruitment can be explored in detail as it gives different results in comparison to click or tone evoked responses.

Keywords: Latency-intensity function, Hearing loss, Speech evoked auditory brainstem response

How to cite this article:
Archana G, Kishan M M, Kumar V, Rajashekhar B, Prakash HP. Latency-intensity function of speech evoked auditory brainstem responses in individuals with moderate sensory neural hearing loss. Indian J Otol 2015;21:254-9

How to cite this URL:
Archana G, Kishan M M, Kumar V, Rajashekhar B, Prakash HP. Latency-intensity function of speech evoked auditory brainstem responses in individuals with moderate sensory neural hearing loss. Indian J Otol [serial online] 2015 [cited 2020 Feb 17];21:254-9. Available from: http://www.indianjotol.org/text.asp?2015/21/4/254/165759

  Introduction Top

Auditory signals travel from the cochlea of the inner ear to multiple relay centers in the brainstem prior to being processed at the level of the auditory cortex. The auditory pathway has different nuclei connected to each other through the nerve fibers. The output from one nucleus becomes the input to the other. Evoked potential is one of the techniques that can be used to study the processing of the stimulus at different levels of the auditory pathway.

The auditory brainstem response (ABR) is a noninvasive measure of far-field representation of stimulus-locked, synchronous electrical events. In response to an acoustic signal, a series of potential fluctuations measured at the scalp provide information about the functional integrity of brainstem nuclei along the ascending auditory pathway, making it a widely used clinical measure of auditory function.[1],[2],[3] ABR measure is used as the gold standard to diagnose the degree and the type of hearing impairment. The click evoked ABR is used widely by clinicians when evaluating hearing and the integrity of the auditory brainstem in certain populations, such as infants or neurologically impaired patients and difficult to test population.[4] The click evoked ABR has proven to be a valuable measure for evaluating auditory function, even helping to distinguish between sensorineural hearing loss (SNHL), and conductive hearing loss.[2],[3],[5],[6] Despite its clinical utility, the ABR in response to clicks or tones does not provide information about the ability of the individual to discriminate more complex acoustic events.

An essential function of the central auditory system is the neural encoding of speech sounds. The neural encoding of the sound begins in the auditory nerve and travels to the auditory brainstem.[7] However, a nonspeech stimulus such as click or tone does not provide insights about the actual processing of speech sounds. The study of encoding of speech sound at the brainstem level is essential, since speech is a complex signal varying in many acoustic parameters over time. A complete description of how the auditory system responds to speech can only be obtained by using speech stimuli. Recently, speech evoked ABR measures have been introduced as a means to study the brainstem encoding of speech sounds.[7],[8] It has been established as a valid and reliable means to assess the integrity of the neural transmission of speech stimuli at the brainstem. Typically with a click or a tone stimulus we get an ABR which consists of mainly Peaks I, III, and V, whereas with a consonant vowel syllable (CV), the ABRs consist of responses to CV separately.

The speech evoked ABR provides discrete representations of many aspects of the acoustic structure of speech,[7] including separate neural representations of speech sound onset, phase-locking to the fundamental and formant frequencies and speech sound offset. Speech burst ABR appears to more sensitive than conventional click stimuli in potentially identifying abnormal perceptual processing attributes in subjects with SNHL.

A striking characteristic of ABRs is its sensitivity to the intensity of the acoustic stimulus. Decrease in stimulus intensity causes the amplitude of early Peaks (I, II, and III) to decrease more than that of Peak V. At low stimulus intensities, only Peak V is discernible.[9] Authors found that speech ABR recorded under lower intensity condition, demonstrated delayed latencies, reduced amplitudes of the feature peaks; moreover, with a stronger correlation among the latencies, which may be the foundation to maintain a high speech discrimination in poor acoustics circumstances.

Need for the study

Speech ABR has its wide application in differentially diagnosing auditory processing disorders. The brainstem response to speech has proven to be a mechanism for understanding the neural bases of normal attention-independent auditory function.[7],[10] Because a speech signal provides different acoustic information than a click (i.e., speech syllables are longer and contain less high frequency information compared to clicks), it provides additional information about neural encoding at the brainstem level and uncovers abnormal encoding in children with learning problems.[8],[11],[12] The speech evoked ABRs possibly provide additional diagnostic information, that is, it is hard to obtain using only clicks.[13]

In peripheral hearing loss (SNHL), speech ABR has not gained much importance, but it can be used as an objective tool for assessing speech processing in these individuals. In individuals with SNHL, temporal aspects and frequency resolutions are affected which are important in understanding speech. Click evoked ABR is not sensitive to these parameters. Its sensitivity is limited to synchronicity and degree of hearing loss and does not assess the finer aspects of speech. Hence, speech ABR has its own implications in assessing speech processing skills of brainstem level with factors such as age, hearing loss, and intensity altering the speech ABR.

Clinically, we rely more on subjective tests to evaluate the speech understanding abilities in individuals with SNHL. Due to increased variability in results, subjective tests cannot be used. As the subjective tests are dependent on the responses elicited by the subjects, it cannot be used in difficult to test population. Hence, the objective tests will be helpful in studying the speech processing in these individuals. Therefore, the data obtained by this study can be used as an objective way of accounting for speech processing at the level of the brainstem. It will be useful in testing difficult to test population and further to assess the benefit from hearing aids, in individuals with SNHL. Speech ABR has its own implications in assessing speech processing in individuals with SNHL. It is necessary to establish a relationship between speech ABR (VA) latency with respect to different intensities of stimulus in SNHL, on the premise that it will help in predicting loudness growth in SNHL individuals to speech stimulus and how it is different from click evoked ABR. In literature, not many studies focusing on L-I function in SNHL individuals exist; the current study is hence focused on the effect of intensity on latency of V of speech ABR; which enable clinicians to comment on loudness growth factors (such as recruitment) in peripheral hearing loss.


The aim of the current study is to investigate the latency-intensity function for speech evoked ABR (VA) in individuals with SNHL.


  • Determining absolute latency of V Peak for different intensities
  • Recording of latency-intensity function in SNHL.

  Materials and Methods Top

Study center

The study was carried out in the Department of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal, Karnataka, India. Convenient sampling was done for the period of December 2011–February 2012. Randomly, participants were selected from the outpatient section of Kasturba Medical College Hospital, Manipal, Karnataka, India. A consent letter was taken from all the participants. After that, they informed on the benefits and risk factors in participating in the study. 20 subjects participated in the study out of which 13 ears were selected for the testing.

Subject selection criteria

Inclusion criteria

  • Age range of 18–65 years of age
  • SNHL
  • Audiometric thresholds between 41 and 55 dBHL (moderate degree)
  • Bilateral normal middle ear function ("A" type tympanograms)
  • Flat and minimal sloping pattern (all thresholds between 250 Hz and 8 kHz not exceeding more than 10–15 dB per octaves).

Exclusion criteria

  • Past or present history of middle ear pathology
  • Sharply sloping hearing loss
  • Above 65 years of age
  • Retro cochlear lesion.


The following instruments were used in the study.

Pure tone audiometer

A calibrated dual channel Grason Stadler Inc-61; GSI-61 Clinical Audiometer was used to estimate the behavioral thresholds.

  • TDH-49 earphones
  • Radio Ear B-71 bone vibrator.

Middle ear analyzer

A calibrated Middle ear analyzer (Grason Stadler Inc-Tympstar; GSI-Tympstar) was used to assess the middle ear status.

Speech evoked auditory brainstem response

Speech evoked ABR was recorded using intelligent hearing system smart evoked potential; IHS Smart EP Software, version 3.94 (6860 SW 81st St, Miami, FL 33143, United States).

Speech evoked auditory brainstem response

Subjects were tested in a sound treated room making them lie down comfortably in an inclining chair. The electrode sites were cleaned using NuPrep-AgCl skin preparation gel to bring the impedance within acceptable limits (<5 kΩ) inter electrode impedance to 2 kΩ and a Ten20 conductive gel was used for creating a favoring condition for the electrodes to pick the responses from the skin. Speech evoked auditory brainstem audiometry was administered to the participants who were confirmed to be with moderate SNHL. IHS Smart EP Software, version 3.94 was used for recording the data. A Klatt cascade/parallel formant synthesizer was used to synthesize a 40 ms speech like/da/syllable).[11] The stimulus was constructed to include an onset burst frication during the first 10 ms, followed by 30 ms formant transitions. The stimulus contains the key acoustic phonetic information which gives information relevant to CV identity.

Statistical analysis

The collected data were subjected to statistical analysis using SPSS version 17 for windows (Chicago, Inc., IL, USA) to find out the mean and standard deviation (SD) of absolute latency with respect to different intensities and to find out the relation between them.

  Results Top

The aim of the present study was to find out the latency-intensity function of speech evoked ABR in individuals with moderate SNHL age ranging from 18 to 65 years of age. Speech ABR was elicited for different intensities from 90 dBnHL till threshold of the individual. Obtained data were tested for statistical significant difference using SPSS 17.

The mean absolute latency for different intensities from 60 to 90 dBnHL was found to be 8.11, 6.68, 5.96, and 5.41 ms, respectively. The corresponding SDs were 0.50741, 0.63258, 0.28962, and 0.29634 respectively [Table 1]. The mean amplitude corresponding to each intensity was not calculated as it showed higher variability within participants. Shift in latency from 90 to 80 dBnHL was found to be 0.5438 ms, 80 dBnHL to 70 dBnHL 0.7231 ms, 70 dBnHL to 60 dBnHL 1.4254 ms, with the maximum shift noticed from 70 to 60 dBnHL that is, 1.4254 ms.
Table 1: Mean and SD of latency of Peak V

Click here to view

Pearson's correlation coefficient showed a negative correlation between latency and intensity (r = −0.894) and was statistically significant at 0.01 levels and indicated that with every 10 dB increase in intensity, there was a significant reduction in the latency.

In the [Graph 1 [Additional file 1]], the x-axis represents stimulus intensity and the y-axis of the latency at 95% confidence interval. The red bar represents the standard error from the mean latency across different intensities. The standard error was found to be increasing when the stimulus intensity was decreased from higher to lower intensities.

The histogram [Graph 2 [Additional file 2]] looks reasonably normally distributed (the red bars shows the distribution of dependent variable, that is, latency, and the green curve, the normal distribution curve), indicating that the normality of errors assumption has probably been met. To verify this assumption, a normal P-P plot was used [Graph 3 [Additional file 3]].

The dotted red line which represents the observed probability did not deviate much from the expected probability (represented by a green line). As there is less deviation of observed probability with the expected probability, it can be concluded that obtained latency difference is due to the change in intensity.

In the [Graph 4 [Additional file 4]] red circles represent subjects participated, the green line represents mean latency of all the participants across varying intensities, and the red lines represent 95% of confidence interval. Regression analysis showed r2 = 0.799 that is, its 79% of chances that the change in latency is by the effect of stimulus intensity.

Further, linear regression was fitted to the equation, y = ax + b, where y indicates latency, "a' as a constant, "x" as intensity and "b" as regression coefficient (i.e., y = −0.088× (x) +13.14). That is, for every 10 dB decrease in intensity leads to 0.8 ms prolongation in latency.

  Discussion Top

The objectives of the present study were determining absolute latency for different intensities and recording of latency-intensity function in individuals with moderate SNHL. Speech ABR recording was done using/da/stimulus on 13 ears with moderate SNHL. The results revealed that V Peak was observed only till 60 dBnHL which was at the level of 10 dBSL approximately of all the subjects. Correlation analysis showed a significant difference in varying the stimulus intensity on latency (P = 0.01). The linear regression showed a negative correlation between the latency of V Peak and intensity of speech stimuli (r − 0.894). This indicates that as the intensity decreases, there was an increase in the latency. The slope was found to be 0.8 ms prolongation of V Peak latency for every 10 dB by a decrease in intensity.

In the present study, the mean latencies of V Peak at different intensities were as follows 5.41 ms at 90 dBnHL, 5.96 ms at 80 dBnHL, 6.68 ms at 70 dBnHL, and 8.11 ms at 60 dBnHL. On observation, at the threshold level, there was a significant prolongation and at higher intensities, it was found to be either normal or near normal. This can be probably explained by the high frequency loss in the subjects who participated in the study. ABR being an onset response is best evoked by basal fibers as they elicit responses of shorter latencies when compared to those of apical region; having high frequency loss eliminates the prospect of getting shorter latencies and apical fibers contribute to the response which produces prolonged latencies.[3],[14]

In the present study, at the threshold level, the absolute latency prolongation of V Peak could thus be justified by the fact that it might have been elicited by the apical fibers since there was high frequency cochlear loss in the participants. At higher intensities, the contribution of basal fibers to the V th Peak elicitation could have resulted in it appearing at shorter latencies in comparison with responses evoked by threshold level stimulus. Even though there is no supporting literature to this, it could be hypothesized that the active mechanism of the outer hair cell (OHCs), instrumental in providing feedback energy to the basilar membrane at lower intensities could be affected at moderate SNHL with OHCs being more vulnerable to damage and causing the pathology. Further, the residual OHCs might take longer time for the active mechanism for further conduction to take place whereas, at higher intensities, inner hair cells are directly involved so that shorter latency of Peak V occurs.

Comparing the results of the present study with latency-intensity function of speech evoked ABR in normal hearing individuals at same hearing level, it is apparent that at the threshold level of individuals with moderate SNHL (60 dBnHL), the latency of V Peak appears prolonged by 0.6 ms from the normal range. At 70 dBnHL, the mean value of V Peak latency (6.4715 ms) of the present study falls closer to the normal range (mean = 6.68 ms; SD ± 0.63). However, at higher intensities (80 dBnHL and 90 dBnHL), the mean values (5.96 ms and 5.41 ms respectively) appear earlier than the normative range by 0.3 ms.

As discussed earlier, the contribution of only apical fibers and OHC dysfunction (sensory loss) at threshold levels could have caused the prolongation of V Peak at the threshold level, which is consistent with the literature evidence.[3],[5] Early appearance of V Peak at higher intensities could be attributed to the recruitment phenomenon. Recruitment is a phenomenon of abnormal loudness growth and is usually associated with the cochlear pathology.[14] This could be further explained with one of the postulates of the center-clipping theory of recruitment[15] which focuses on the abnormal neural discharge found in recruiting ears. Although there are huge challenges posed by psychophysical data, it has been best possibly hypothesized that small increase in intensity results in an abnormal increase in the neural discharge rate and consequently shorter latencies of peaks.[14],[16] Cochlear pathology and more precisely, recruiting ears usually have rapid loudness growth which several authors have reported with psychophysical bases.[5],[15],[16] The findings of the current study are consistent with these hypotheses.

The present study concentrated on moderate SNHL. This degree of hearing loss is typically accompanied by different loudness growth pattern that is, at lower intensity, loudness is perceived lesser than normal, while at higher intensities, loudness perceived is more or equal to normal individuals; this is explained as a partial recruitment phenomenon.[14] Findings of the present study are in support of this partial recruitment phenomenon; that is, in the present study, at higher intensities, early latency, and at threshold level prolonged latency of V Peak was found.

It has also been reported that in individuals with recruitment, as the stimulus intensity increases, there is an increase in the number of units activated, particularly spill over occurs to those regions with the central frequency above the stimulus frequency [14] in comparison with normal ears. This also leads to an overall increase in the neural discharge rate. By this, it can be postulated that along with the psychophysical aspects, physiological aspects also contribute to the recruitment phenomenon.[7] In literature, the slope of L-I function using click evoked ABR was found to be 0.44 ms/dB; however, the current study reveals a much steeper slope of up to 0.8 ms using speech as stimuli. Using speech ABR, there is a steeper slope in cochlear pathology which is better in predicting the type and nature of the hearing loss. This could be attributed to the complexity of the stimulus that is, speech is a complex signal having combination of modulation signals with variation in the frequency and amplitude moment to moment; these fluctuations can magnify the phenomenon of recruitment. So using speech evoked ABR, the loudness growth can be studied in detail in cochlear pathology.

In the present study, focus was given only to the onset responses, even though it was noticed that there is a difference between speech ABR and click ABR with respect to slope; frequency following response (FFR) would give a better picture of speech processing skills. As hypothesized, transient stimulus elicits responses which are by the synchronous activity of auditory nerve to the signal. In cochlear pathology, signals can be distorted due to the lesion in cochlea and synaptic junction, when higher intensity stimulus is presented, the response elicited can be normal for transient response, even though they have difficulty in processing the signal. In such cases, the focus could be on FFR for a true representation of difficulties encountered by individuals with cochlear pathology.

  Conclusion Top

It can be concluded that using speech as a stimulus in evoked responses, it can result in better estimation of loudness growth pattern in individuals with cochlear pathology. A phenomenon like recruitment can be explored in detail as it gives different results in comparison to click or tone evoked responses.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Despland PA, Galambos R. The auditory brainstem response (ABR) is a useful diagnostic tool in the intensive care nursery. Pediatr Res 1980;14:154-8.  Back to cited text no. 1
Jacobson JT. The Auditory Brainstem Response. San Diego: College-Hill; 1985.  Back to cited text no. 2
Hood LJ. Clinical Applications of the Auditory Brainstem Response. San Diego, CA: Singular Publishing Group, Inc.; 1998.  Back to cited text no. 3
Starr A, Don M. Brainstem potentials evoked by acoustic stimuli. In: Picton TW, editor. Handbook of Electroencepholography and Clinical Neurophysiology. Amsterdam: Elsevier; 1988. p. 97-150.  Back to cited text no. 4
Hall JW 3rd. Handbook of Auditory Evoked Responses. Needham Heights, MA: Allyn and Bacon; 1992.  Back to cited text no. 5
Josey A. Auditory brainstem response in site of lesion testing. In: Katz J, editor. Handbook of Clinical Audiology. Baltimore: Williams and Wilkins; 1985. p. 534-48.  Back to cited text no. 6
Russo NM, Nicol TG, Zecker SG, Hayes EA, Kraus N. Auditory training improves neural timing in the human brainstem. Behav Brain Res 2005;156:95-103.  Back to cited text no. 7
Banai K, Nicol T, Zecker SG, Kraus N. Brainstem timing: Implications for cortical processing and literacy. J Neurosci 2005;25:9850-7.  Back to cited text no. 8
Fu QY, Liang Y. A Preliminary investigation of speech evoked auditory brainstem responses with two stimulus intensities. Bioinformatics and Biomedical Engineering; 2009. ICBBE 2009. 3rd International Conference.  Back to cited text no. 9
Kraus N, Nicol T. Brainstem origins for cortical 'what' and 'where' pathways in the auditory system. Trends Neurosci 2005;28:176-81.  Back to cited text no. 10
Cunningham J, Nicol T, Zecker SG, Bradlow A, Kraus N. Neurobiologic responses to speech in noise in children with learning problems: Deficits and strategies for improvement. Clin Neurophysiol 2001;112:758-67.  Back to cited text no. 11
King C, Warrier CM, Hayes E, Kraus N. Deficits in auditory brainstem pathway encoding of speech sounds in children with learning problems. Neurosci Lett 2002;319:111-5.  Back to cited text no. 12
Song JH, Banai K, Russo NM, Kraus N. On the relationship between speech- and nonspeech-evoked auditory brainstem responses. Audiol Neurootol 2006;11:233-41.  Back to cited text no. 13
Salvi RJ, Hamernik RP, Henderson D. Response patterns of auditory nerve fibers during temporary threshold shift. Hear Res 1983;10:37-67.  Back to cited text no. 14
Moore DR. Auditory processing disorders: Acquisition and treatment. J Commun Disord 2007;40:295-304.  Back to cited text no. 15
Killion MC, Fikret-Pasa S. The 3 types of sensorineural hearing loss: Loudness and Intelligibility considerations. Hear J 1993;46:1-4.  Back to cited text no. 16


  [Table 1]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Materials and Me...
Article Tables

 Article Access Statistics
    PDF Downloaded220    
    Comments [Add]    

Recommend this journal