Noise Health Home 

[Download PDF]
Year : 2017  |  Volume : 19  |  Issue : 91  |  Page : 278--285

Using auditory steady-state responses for measuring hearing protector occlusion effect

Olivier Valentin, Frédéric Laville 
 Department of Mechanical Engineering, École de technologie supérieure, Université du Québec, Montréal, Québec, Canada

Correspondence Address:
Olivier Valentin
École de Technologie Supérieure, Département de Génie Mécanique, 1100 rue Notre-Dame Ouest, Montréal (QC), H3C 1K3


Introduction: The currently available methods for measuring the occlusion effect (OE) of hearing protection devices (HPDs) have limitations. Objective microphonic measurements do not assess bone-conducted sounds directly transmitted to the cochlea. Psychophysical measurements at threshold are biased due to the low-frequency masking effects from test participants’ physiological noise and the variability of measurements based on subjective responses. An auditory steady-state responses (ASSRs) procedure is used as a technique that might overcome these limitations. Participants and Methods: Pure-tone stimuli (250 and 500 Hz), with amplitude modulated at 40 Hz, were presented to twelve adults with normal hearing through a bone vibrator at three levels in 10-dB steps. The following two conditions were assessed: the unoccluded ear canal and occluded ear canal. ASSR amplitude data as a function of the stimulation level were linearized using least-square regressions. The ASSR-based “physiological” OE was then calculated as the average difference between the two measurements. Results: A significant statistical difference was found between the average threshold-based psychophysical OE and the average ASSR-based OE. Conclusion: This study successfully ascertained that it is possible to objectively measure the OE of HPD using ASSRs collected on the same participant both with and without protectors.

How to cite this article:
Valentin O, Laville F. Using auditory steady-state responses for measuring hearing protector occlusion effect.Noise Health 2017;19:278-285

How to cite this URL:
Valentin O, Laville F. Using auditory steady-state responses for measuring hearing protector occlusion effect. Noise Health [serial online] 2017 [cited 2022 Oct 5 ];19:278-285
Available from:

Full Text


According to the World Health Organization, worldwide hearing loss estimates increased from 120 million people in 1995[1] to 250 million in 2004.[2] In the United States, the National Institute for Occupational Safety and Health evaluated that 22 million workers are affected by occupational noise-induced hearing loss.[3] Wearing hearing protection devices (HPDs) is the most common solution to protect workers from noise exposure. This is so because it is often difficult, for technological or economic reasons, to reduce noise at its source.

Unfortunately, the use of HPDs has not been rigorously adopted. The occlusion effect (OE) is one of the reasons often given to justify the nonuse of HPD. The occlusion of the ear canal induces a modification of the wearer’s voice perception and creates a discomfort that sometimes brings people to remove their HPD, which raises safety and usability concerns.[4] Reducing the OE induced by wearing HPDs has the potential to increase the auditory comfort of HPDs and could help preventing occupational hearing loss in the future. Many studies on different OE quantifications have been reported in the literature. Most of these studies used a bone vibrator source for the experimental work.[5],[6],[7],[8],[9],[10] A more limited number of studies used the voice as source.[6],[7],[11],[12],[13] All these studies used objective or subjective methods, which have the following limitations: objective measurements using a microphone do not assess bone-conducted sounds directly transmitted to the cochlea, and psychophysical measurements at threshold are biased due to the low-frequency masking effects from test participants’ physiological noise, in addition to the variability of measurement due to subjective responses.

No account was found in the literature regarding the use of electroencephalography (EEG) to measure the OE. This methodology could possibly overcome the limitations of the methods traditionally used to evaluate the OE. This study aims to examine the use of auditory steady-state responses (ASSRs) as a method for measuring the OE induced by wearing HPD by adapting the methodology used in a previous work for the measurement of HPD attenuation.[14] As mentioned in this previous work on the use of ASSR for the measurement of HPD attenuation, ASSR techniques have been largely developed to overcome the various difficulties in the assessment of hearing thresholds in very young children, older patients, or those with cognitive deficits or behavioral disorders.[15],[16] ASSRs are electrophysiological responses, recorded from the human scalp, and often evoked by one or more carrier frequencies that are amplitude modulated at a specific frequency (Fm). In practice, when a participant is exposed to such a stimulus, the spectral power of the EEG frequency spectrum of the participant that is related to the stimulus will be manifested at Fm and may also appear at its harmonics.[17] An interesting characteristic of ASSR is that its amplitudes increase with the level of the stimulus. This increase is not always linear with the stimulation level; the curve for the variation in ASSR amplitude has shown two distinct regions with two different slopes.[18],[19]

Historically, ASSRs were first recorded at modulation frequencies near 40 Hz. Because 40-Hz ASSRs are sensitive to arousal affects, a large number of studies using ASSRs for threshold estimation have concentrated on modulation frequencies in the 70–110 Hz range. In the case of adults who are alert/awake, using the 40-Hz modulation frequency provides higher ASSR amplitudes and better signal-to-noise ratios, which may enable the ASSRs to be detected more rapidly.[20],[21] Consequently, this 40-Hz modulation frequency was used in this study to limit the experimental time. To further limit the experimental time, only the following two carrier frequencies were used: 250 and 500 Hz. These carrier frequencies were chosen in the low-frequency range (below 1000 Hz), wherein the greatest positive OE was expected. ASSR results were then compared to the results obtained with a subjective psychophysical method.

 Participants and Methods


Twelve men aged 21–27 years and thresholds below 20-dB HL (from 125 to 8000 Hz) were assessed. The study was reviewed and approved by the Ethics Committee of the École de technologie supérieure (ÉTS) of Montréal. Informed consent was obtained from all participants prior to their entrance into the study.

Hearing protector

The hearing protectors chosen for this study were 3M™ polyurethane foam earplugs [Figure 1]. These hearing protectors are well suited for this experiment because they are easy to insert thanks to their conical shape and because the circumference of the plug can easily be marked with a felt-tip pen, which helps to control the insertion depth.{Figure 1}

During the entire duration of the experiments, the earplugs were systematically inserted by the experimenter. The insertion depth was 10 mm, because such a situation is representative of users wearing earplugs with a nonoptimal insertion depth.

Experimental procedure

A typical experiment involved two tasks. The first step consisted of performing the psychophysical measurements, and the second step consisted of performing the ASSR recordings. During all these measurements, volunteers sat in a comfortable ergonomic chair inside a double-walled audiometric booth, whose ambient noise level did not exceed the octave band levels specified in ISO 8253-1:2010(F).

Task 1: Threshold-based psychophysical measurements (subjective method) – 20 min

For both frequencies, the threshold-based psychophysical OE for each participant was computed from six auditory threshold measurements (three with and three without earplugs) during a single visit to the laboratory. Tests’ order was counterbalanced across participants (as presented in [Table 1]) to reduce the influence of the stimulations’ order. The hearing protectors were refitted by the experimenter for each trial. The test signals consisted of bone-conducted pure tones at 250 and 500 Hz presented via a Radioear™ B-81 bone vibrator, which was coupled to the forehead by an elastic headband with approximately 400–450 g of force. For both stimuli, the level of the first presentation was 30 dB HL. The level of each succeeding presentation was determined by the preceding response. After each failure to respond to a signal, the level was increased in 1-dB steps until the first response occurred. After the response, the intensity was decreased 10 dB, and another ascending series was started. The minimum number of responses needed to determine each bone conducted threshold was three responses out of five presentations at a single level.{Table 1}

Task 2: Auditory steady-state response-based physiological measurements (objective method) – 39 min

The ASSR-based physiological OE has been calculated as the average difference between the normal and occluded conditions. The procedure is illustrated in [Figure 2]. The three points corresponding to experimental values recorded with earplugs are projected parallel to the x-axis on the regression line (D) obtained with the three experimental points recorded without earplug, and the three corresponding abscissae are calculated. The difference between the two abscissae is calculated for each point with earplugs. The same calculation is computed with the three experimental values recorded without earplug on the regression line (D′) obtained with the three experimental points recorded with earplugs. The mean of these six values is taken as the ASSR-based physiological OE.{Figure 2}

ASSR recordings were first collected for the occluded condition (with earplugs) and then for the unoccluded condition. Because a participant’s arousal state can modulate the amplitude of the 40-Hz responses, the participants were asked to stay awake and watch a subtitled movie during the recordings of these responses. During the experiment, visual contact was maintained with the volunteer through a glass window of the audiometric booth. Between each recording, the experimenter asked the participants if they were okay and/or needed anything. In the middle of the experiment, coffee or tea was provided to any participant who indicated that she/he felt drowsy.

The LabVIEW™ based “MASTER SYSTEM” (Rotman Research Institute, software[22] was used for both ASSR stimulus generation and response recording/evaluation. The frequencies of both the carrier and modulation signals were adjusted so that an integer number of cycles occurred within each recording section. This allowed the individual sections to be linked together without acoustic artifact and to be interchangeable during artifact rejection.[23] For example, a modulation frequency of 40 Hz was adjusted to 40.039 Hz. For simplicity, the frequencies will henceforth be reported to the nearest integer value.

The stimuli consisted of pure tones, and amplitude modulated with a depth of 100% and frequency modulated with a depth of 25%. The carrier frequencies of 250 and 500 Hz were modulated at 40 Hz and individually presented to each participant. Each stimulus was amplified using an Interacoustic™ audiometer (model AC40) and presented via the same bone vibrator system as the one used in Task 1. The stimuli were calibrated in reference equivalent threshold force levels in dB re:1 μN, corresponding to 0 dB HL for the forehead (ANSI S3.6-2004), using a Larson Davis™ Model 3000+ sound level meter and a Larson Davis™ Model AMC493B artificial mastoid. Because amplitudes do not always increase linearly with the stimulation level,[18],[19] ASSRs were recorded at three stimulation levels to improve precision for the computation of ASSR-based OE. Consequently, stimuli were presented at 20-, 30-, and 40-dB HL in the 500-Hz experimental conditions. In the 250-Hz experimental conditions, stimuli were presented at 0-, 10-, and 20-dB HL (i.e., 20 dB lower) to prevent the bone vibrator to reach its operation limits and to avoid distortion of the stimulation signals. To reduce the effects of stimulus artifact that may be present in the EEG when recording ASSR, stimuli polarity was inverted for half of the 12 ASSR recordings made for each stimulus condition.[24],[25]

The setup used for the recording of ASSR is illustrated in [Figure 3]. ASSRs were collected from an electrode placed at vertex (Cz) using an electrode on the back of the neck (below the hairline) as reference. An electrode on the clavicle served as ground. All interelectrode impedances were below 3 kΩ at 10 Hz. ASSRs were filtered using a band-pass filter of 0.3–300 Hz (12 dB/octave) and amplified 50,000 times (10,000× in GRASS Technologies LP511 AC Amplifier; 5× in National Instruments USB-6259 BNC). All data were collected using an A/D conversion rate of 1000 samples per second. In each recording, individual data epochs of 1024 points each were collected and linked together into sweeps (16 epochs per sweep for lasting 16.384 s each). A typical ASSR recording required 144 sweeps across 39 min, that is, two frequencies (250 and 500 Hz) × six different stimulation levels (three unoccluded and three occluded) × two recordings of six sweeps. To minimize artifacts and other interference, epochs containing signals exceeding ±90 μV were rejected.{Figure 3}

The ASSR sweeps of data were averaged in the time domain and then analyzed online in the frequency domain using FFT. Amplitudes were expressed in nanovolts (nV). An F-ratio statistic was calculated by the MASTER system, which estimated the probability that the amplitude of the ASSR was significantly different from the average amplitude of the background noise in adjacent frequencies (within ±60 bins of the modulation frequency). A response was considered to be present if P < 0.05.


[Figure 4] (for 250 Hz) and [Figure 5] (for 500 Hz) present the individual plots of ASSR amplitudes as a function of stimulation level. [Figure 6] (for 250 Hz) and [Figure 7] (for 500 Hz) show the comparison between the objective ASSR-based physiological OEs and the subjective threshold-based psychophysical OEs. [Table 2] summarizes the OE values obtained with both methods.{Table 2}{Figure 4}{Figure 5}{Figure 6}{Figure 7}


As seen in the individual graphs [Figure 5], ASSRs’ amplitude increased as the stimulus level increased (except for participant S7 at 500 Hz at the highest stimulation level, for both conditions).

Results in [Table 2] and [Figure 7] indicate that, at 250 and 500 Hz, the ASSR-based physiological OEs seem to be significantly different from the threshold-based psychophysical values. This observation was confirmed by Wilcoxon tests, computed by using R 3.1.0 with packages MASS 7.3-35,[26] which reject the null hypothesis at a 5% significance level (P value < 0.05 at 250 and 500 Hz). A two one-sided test procedure[27] also concluded that the average ASSR-based physiological and threshold-based psychophysical OEs were not equivalent at ±5 dB (P value >0.05 at 250 and 500 Hz).

At 500 Hz, for 83.3% of the participants, the ASSR-based physiological OE was higher than the threshold-based psychophysical OE. This observation is in agreement with what was expected, because below 1 kHz, the masking effect induced by the physiological noise is a source of bias, which brings to an overestimation of the occluded thresholds at these frequencies.

However, at 250 Hz, for 75% of the participants, results are in total contradiction with what was expected: at this frequency, the ASSR-based physiological OE is smaller than the threshold-based psychophysical OE. Several reasons may explain these results. First, at 250 Hz, the Radioear™ B-81 bone vibrator reached the operation limits, and distortion effects were present in the stimulation signals at levels as low as 30 dB HL. Second, at this frequency, the vibrator “vibrates stronger” and consequently, ASSRs measured at this frequency without a hearing protector were not only generated by the stimulation propagated through the bones but also by an aerial emission above the threshold of hearing. However, when the participants wear hearing protectors, this aerial stimulation is greatly attenuated, and the ASSR are mainly generated by the stimulation through the bones. Therefore, the ASSR-based physiological OEs at 250 Hz are underestimated, because the ASSRs measured without hearing protectors are overestimated.

 Conclusion and Recommendations

The currently available methods for OE measurement have limitations. Objective measurements using a microphone do not assess bone-conducted sounds directly transmitted to the cochlea. Psychophysical measurements at threshold are biased due to the low-frequency masking effects from test participant’s physiological noise and contain a variability of measurement due to subjective responses.

Although ASSRs have been adapted in the past for measuring the threshold to bone-conduction stimuli[28],[29] or for investigating the time course of the maturation of bone-conduction hearing sensitivity,[30] no study has considered using ASSRs as a method to evaluate the OE induced by wearing HPD. This study sought to ascertain whether it is possible to objectively measure the OE of HPD using ASSRs collected in the same participant both with and without protectors, which has been adapted from the methodology used in previous work for the measurement of HPD attenuation.[14] The results showed that ASSRs can be used to objectively measure the OE induced by wearing earplugs; mean results obtained at 500 Hz on a group of 12 participants reported a significant difference between the ASSR-based physiological OE and the threshold-based psychophysical OE, which is in agreement with our initial hypothesis that the psychophysical approach underestimates the OE in the low-frequency range. However, this hypothesis could not be verified at 250 Hz due to problems with the bone vibrator performances at very low frequencies. Recommendation 2, hereafter, addresses this issue.

The following several recommendations can be formulated based on this feasibility study:ASSR-based physiological OEs reported in this feasibility study have not been compared with microphonic-based OEs. The recommendation is therefore to utilize microphone in real ear (MIRE) tests to provide an objective measurement of OE with less variability than subjective estimates provided by using a threshold-based psychophysical method.Results at 250 Hz are in total contradiction with expected results because of the following reasons: (i) at 250 Hz, the Radioear™ B-81 bone vibrator is reaching the operation limits and the stimulation signals are distorting, and (ii) the ASSR-based physiological OEs are underestimated because the ASSRs measured without hearing protectors are overestimated. Additional measurements using, for example, a carrier frequency of 315 Hz could be used to verify the validity of these assumptions. Furthermore, several solutions exist to overcome these low-frequency problems. The Radioear™ B-81 bone vibrator could be replaced by another bone vibrator or a vibration testing system more reliable and less noisy in the low-frequency range than the B-81. Another solution would be to attenuate the aerial radiation from the B-81 which occurs during the unoccluded measurements by having the participants wear earmuffs with a large enough volume, so that negligible OE would be induced by them.Due to the somewhat long duration of the experiment, the present feasibility was conducted at only two carrier frequencies which are expected to be influenced by the low-frequency masking effect: 250 and 500 Hz. The use of broadband noises would provide results over a wider frequency range, in addition to also reducing the total duration of the ASSR measurement, because it would not be necessary to discretize the 125–1000 Hz frequency range with several stimuli.ASSR-based physiological OE has been calculated as the average difference between the normal and occluded conditions using the linear least-square regression of ASSR amplitude data. Each linear regression has been computed on a set of three points (one per stimulation levels), because amplitudes do not always increase linearly with the stimulation level. Individual results presented in this study suggest that computing these linear regressions should use a minimum of three data points, because, for a same participant, there may be significant differences between the slopes of the occluded and nonoccluded stimulation levels versus amplitude data. A more extensive study with more than three points is needed to find out whether using more points would lead to more robust estimates of OE for individual participants.

Finally, as mentioned in the case of attenuation measurements using ASSRs,[14] using this method to measure individual OEs would require a thorough investigation of the intraindividual variability of each measurement.


The authors wish to express their appreciation to NSERC (Natural Sciences and Engineering Research Council of Canada) for their support of this project and to 3M Company (Saint Paul, Minnesota, USA) for providing the foam earplugs.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1World Health Organization. In: Berglund B, Lindvall T, Schwela DH, editors. Guidelines for Community Noise. WHO; 1999.
2Smith A. The fifteenth most serious health problem in the WHO perspective. Presentation to IFHOH World Congress. Helsinki, July 2004.
3Masterson EA, Tak S, Themann CL, Wall DK, Groenewold MR, Deddens JA et al. Prevalence of hearing loss in the United States by industry. Am J Ind Med 2013;56:670-81.
4Killion NC. The hollow voice occlusion effect. Proceedings of 13th Danavox Symposium, vol. 3; 1988. p. 231-41.
5Von Békésy G. Experiments in Hearing, 1st ed. New York: McGraw Hill; 1960.
6Lundh P. Sound Pressure in the Ear with Vented and Unvented Earmould. Technical Report No. 28-8-1. Copenhagen, Denmark: Oticon Electronics A/S; 1986.
7Wimmer VH. The occlusion effect from earmoulds. Hear Instrum 1986;37:56-8.
8Stenfelt S, Wild T, Hato N, Goode RL. Factors contributing to bone conduction: The outer ear. J Acoust Soc Am 2003;113:902-13.
9Stenfelt S, Reinfeldt S. A model of the occlusion effect with bone-conducted stimulation. Int J Audiol 2007;46:595-608.
10Reinfeldt S, Stenfelt S, Good T, Hakansson B. Examination of bone-conducted transmission from sound field excitation measured by thresholds, ear-canal sound pressure, and skull vibrations. J Acoust Soc Am 2007;121:1576-87.
11Hansen MO. Occlusion Effects: Part 1: Hearing Aid Users Experiences of the Occlusion Effect Compared to the Real Ear Sound Level. Technical Report No. 71. Denmark: Technical University of Denmark; 1997.
12Dillon H. Hearing Aids. New York: Thieme; 2000.
13Lecocq C, Laville F, Gargour C. Subjective quantification of earplug occlusion effect using external acoustical excitation of the mouth cavity. J Acoust Soc Am 2010;128:763-70.
14Valentin O, John MS, Laville F. Using auditory steady-state responses for measuring hearing protector attenuation. Noise Health 2017;19:1-9.
15Cone-Wesson B, Dowell RC, Tomlin D, Rance G, Ming WJ. The auditory steady-state response: Comparisons with the auditory brainstem response. J Am Acad Audiol 2002;13:173-83.
16Picton TW, John MS, Dimitrijevic A, Purcell D. Human auditory steady-state responses. Int J Audiol 2003;42:177-219.
17John MS, Dimitrijevic A, Picton TW. Efficient stimuli for evoking auditory steady-state responses. Ear Hear 2003;24:406-23.
18Lins OG, Picton PE, Picton TW, Champagne SC, Durieux-Smith A. Auditory steady-state responses to tones amplitude-modulated at 80-110 Hz. J Acoust Soc Am 1995;97:3051-63.
19Menard M, Gallego S, Berger-Vachon C, Collet L, Thai-Van H. Relationship between loudness growth function and auditory steady-state. Hear Res 2008;235:105-13.
20Dimitrijevic A, John MS, Van Roon P, Picton TW. Human auditory steady state responses to tones independently modulated in both frequency and amplitude. Ear Hear 2001;22:100-11.
21Van Maanen A, Stapells DR. Comparison of multiple auditory steady-state responses (80 versus 40 Hz) and slow cortical potentials for threshold estimation in hearing-impaired adults. Int J Audiol 2005;44:613-24.
22John MS, Picton TW. MASTER: A Windows program for recording multiple auditory steady-state responses. Comput Methods Programs Biomed 2000;61:125-50.
23John MS, Lins OG, Boucher BL, Picton TW. Multiple auditory steady-state responses (MASTER): Stimulus and recording parameters. Audiology 1998;37:59-82.
24Hall JW. Handbook of Auditory Evoked Responses. Needham Heights: Allyn & Bacon; 1992.
25Picton TW, John MS. Avoiding electromagnetic artifacts when recording auditory steady-state responses. J Am Acad Audiol 2004;15:541-54.
26R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2014. ISBN 3-900051-07-0. Available from: [Last accessed on 2014 Apr 22].
27Schuirmann DJ. A comparison of the two one-sided tests procedure and the power approach for assessing the equivalence of average bioavailability. J Pharmacokinet Biopharm 1987;15:657-80.
28Small SA, Stapells DR. Normal brief-tone bone-conduction behavioral thresholds using the B-71 Transducer: Three occlusion conditions. J Am Acad Audiol 2003;14:556-62.
29Small SA, Stapells DR. Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing. Ear Hear 2006;27:219-28.
30Small SA, Stapells DR. Maturation of bone conduction multiple auditory steady-state responses. Int J Audiol 2008;47:476-88.