Utility of otoacoustic emissions and olivocochlear reflex in predicting vulnerability to noise-induced inner ear damage.

AIM: The aim of the present study was to explore the possible utility of otoacoustic emissions (OAEs) and efferent system strength to determine vulnerability to noise exposure in a clinical setting.
MATERIALS AND METHODS: The study group comprised 344 volunteers who had just begun mandatory basic training as Hellenic Corps Officers Military Academy cadets. Pure-tone audiograms were obtained on both ears. Participants were also subjected to diagnostic transient-evoked otoacoustic emissions (TEOAEs). Finally, they were all tested for efferent function through the suppression of TEOAEs with contralateral noise. Following baseline evaluation, all cadets fired 10 rounds using a 7.62 mm Heckler & Koch G3A3 assault rifle while lying down in prone position. Immediately after exposure to gunfire noise and no later than 10 h, all participants completed an identical protocol for a second time, which was then repeated a third time, 30 days later.
RESULTS: The data showed that after the firing drill, 280 participants suffered a temporary threshold shift (TTS) (468 ears), while in the third evaluation conducted 30 days after exposure, 142 of these ears still presented a threshold shift compared to the baseline evaluation [permanent threshold shift (PTS) ears]. A receiver operating characteristics curve analysis showed that OAEs amplitude is predictive of future TTS and PTS. The results were slightly different for the suppression of OAEs showing only a slight trend toward significance. The curves were used to determine cut points to evaluate the likelihood of TTS/PTS for OAEs amplitude in the baseline evaluation. Decision limits yielding 71.6% sensitivity were 12.45 dB SPL with 63.8% specificity for PTS, and 50% sensitivity were 12.35 dB SPL with 68.2% specificity for TTS.
CONCLUSIONS: Interestingly, the above data yielded tentative evidence to suggest that OAEs amplitude is both sensitive and specific enough to efficiently identify participants who are particularly susceptible to hearing loss caused by impulse noise generated by firearms. Hearing conservation programs may therefore want to consider including such tests in their routine. As far as efferent strength is concerned, we feel that further research is due, before implementing the suppression of OAEs in hearing conservations programs in a similar manner.

Introduction

Excessive noise is one of the most common occupational and recreational hazards. According to the World Health Organization (WHO), about 16% of hearing loss worldwide are attributable to occupational noise exposure.

Military personnel in particular are routinely exposed to impulse noise from live-fire training. Impulse noise is a transient noise stimulus which is usually due to blast effect and the rapid expansion of gases. Gunfire may therefore be categorized as an impulse noise which can quickly result in hearing loss, making hearing protection strongly advisable to reduce its harmful effects. The US National Institute for Occupational Safety and Health criteria document states that exposure to impulse noise should not exceed 140 dBA.[1] In a 2006 study, hearing loss was the most prevalent disability due to military service in USA resulting in a striking $1.6 billion cost for rehabilitation the same year.[2] In another study, 55.8% of Belgian military personnel were reported to have suffered from hearing impairment attributable to firearm noise exposure.[3] Temporary threshold shifts (TTSs) and permanent threshold shifts (PTSs) in pure-tone audiometry are the most common methods of displaying the harmful effects that result from such noise exposure.

Occupational and military physicians have gone to great lengths to establish an objective index that would help them evaluate susceptibility to hearing loss, when exposed to noise. In this context, the ongoing research on otoacoustic emissions (OAEs) has stimulated the hope of finding such an index.[456] OAEs provide a noninvasive and sensitive method of both evaluating the outer hair cells, which are one of the primary targets of noise-induced hearing loss, and also detecting early indications of noise-induced inner ear damage before any actual evidence of hearing loss appears in standard audiometry thresholds. In some studies, normal hearing participants showed reduced OAEs,[47] yet it is unclear whether diminished OAEs are predictive of eventual noise-induced hearing loss.

Additionally, the efferent system and particular medial olivocochlear reflex (MOCR) have been hypothesized to serve as a protective mechanism to reduce receptor damage during intense acoustic exposure.

The aim of the present study was to evaluate the extent to which OAEs and MOCR can determine vulnerability to noise exposure in a clinical setting and hence possibly predict an impending hearing handicap.

Materials and Methods

The present cohort study was conducted in the Otorhinolaryngology − Head & Neck Surgery Department of the author’s institute from September 2011 to October 2016. The study group comprised 344 volunteers who had just begun the mandatory basic training as Hellenic Corps Officers Military Academy cadets. Participants were both male and female, strictly of ages 18–19 and they were all informed about the purpose of the study. Informed written consent was obtained after volunteers were assured of complete confidentiality.

History taking and initial evaluation

Participants were interviewed by a certified otorhinolaryngologist who completed a study-designed questionnaire and noted a detailed history record. Age, gender, and handedness were recorded along with specific information regarding general illness, otologic history, past noise exposure, ototoxicity, etc. Exclusion criteria were generally set as follows: age over 19 or under 17 years, history of acoustic trauma or other hearing loss, noisy leisure exposure, presence of tinnitus, history of otitis media, history of ototoxic drugs use, head injury, history of neurological or mental disease, and pregnancy or lactation.

All volunteers who met these screening criteria proceeded to otoscopy check for clear ear canals, and cerumen, if present, was removed. At this stage, participants with any otological abnormality were excluded from the study. A certified otorhinolaryngologist performed a full head and neck clinical examination and 226-Hz probe-tone tympanometry was conducted using standard clinical instrumentation (Granson-Stadler GSI-33 analyzer). Participants showing peak immittance ±50 daPa atmospheric pressure, with grossly normal amplitude and slope of the tympanogram were accepted and proceeded to audiometry and OAE testing.

Pure-tone audiometry

Pure-tone audiograms were obtained at frequencies of 0.5, 1, 2, 3, 4, 6, and 8 kHz on both ears. Only air conduction audiometry was performed using the usual procedure of 10-dB descending and 5-dB ascending steps. Audiometric testing was performed one volunteer at a time, in a double-walled sound attenuating booth using an Interacoustics AC40 clinical audiometer. Pure-tone stimuli were delivered through TDH-49 supra-aural earphones. Earphone placement was checked by the examiner and all audiograms were collected manually by qualified audiologists. The audiometer was calibrated according to the standards of the International Organization for Standardization, ISO 389 and checked every year. Listening checks were performed each day of testing. Participants showing audiometric thresholds above 20 dB at any of the assessed frequencies were classified as having hearing loss and were excluded from the study so as to only include participants without any preexisting hearing loss.

Otoacoustic emissions

Furthermore, participants were subjected to diagnostic transient-evoked otoacoustic emissions (TEOAEs) using ILO 292 DP Echoport instrument (Otodynamics Ltd., UK) with the ILO version 5 software, in a sound-attenuated chamber. TEOAEs were evoked with a 84 dB SPL click of 0.1 ms duration and 16 clicks/s repetition rate presented in nonlinear mode. TEOAEs were collected and averaged until 260 low-noise averages were obtained. The results were windowed in a 2.5 ms onset delay and 20.5 ms duration. Fourier analysis provided the signal-to-noise ratio (SNR) values for the frequency bands 1, 1.5, 2, 3, and 4 kHz. TEOAE inclusion criteria included emission reproducibility of 60% or greater and SNR of 3 dB SPL or greater for 2 or more frequency bands. If a participant failed to achieve valid OAEs, he or she was immediately retested and in case of a second failure, the participant was graded as zero amplitude. We evaluated the overall OAE amplitude (total response) instead of amplitude to particular frequency, because literature suggests[8] that when the sensitivity and tuning of the inner ear are impaired in some frequency range, fundamental aspects of TEOAE generation and propagation in the cochlea are not clear enough to preclude possible additional influences of remote cochlear places on TEOAE. Thus, acoustic trauma to the base of the cochlea can result in the loss of TEOAE energy at frequencies much lower than those affected in the audiogram possibly due to subtle damage to the stimulated region, by a wide-band stimulus.[9]

Olivocochlear reflex

Finally, volunteers were tested for efferent function through the suppression of TEOAEs with contralateral noise. We used methodology based on the protocol suggested by Prasher et al.[10] using a 60 dB SPL linear click to elicit emissions. The number of sweeps was reduced from 260 to 60 and 10 runs of 60 sweeps each were averaged alternately with and without contralateral white noise stimulus. Subsequently, two sets of five alternate buffers were combined to give an average of 300 sweeps each. Continuous white noise generated by an Interacoustics AC40 clinical audiometer was presented to the ear contralateral to the probe via TDH 49 ear phones at 60 dB HL. A measure of the degree of suppression was taken as the difference between the amplitude in decibels of the total response without contralateral noise and that with contralateral noise. Again all measurements took place in a sound attenuated chamber.

Firearm use

Following baseline evaluation all cadets fired 10 rounds using a 7.62 mm Heckler & Koch G3A3 assault rifle while lying down in a prone position. The noise generated was measured at the ear of gunner. Effective time was 1.59 ms, peak pressure was 1.31 kPa, and peak pressure level 156.3 dB. The cadets were provided with disposable foam earplugs and the drill instructors ensured both obligatory usage and proper fitting according to standard Greek Army regulations. According to the manufacturer (3M®, Bullet type earplugs), they achieved a frequency-dependent noise attenuation of 33.9–45.5 dB (noise reduction rating = 29 dB). The rest of the time both before and after the firing drill, the cadets’ activities were severely restricted, and they were not exposed to any significant levels of military or recreational noise.

Second and third evaluation

Immediately after exposure to gunfire noise and no later than 10 h, all participants completed an identical protocol for a second time, including clinical examination, 226-Hz probe-tone tympanometry, pure-tone audiometry, and TEOAEs testing. Any audiometry threshold shifts were established and considered temporary hearing insult (TTS). Finally, all participants were called back to complete a third audiological evaluation 30 days after noise exposure. Prior arrangements were made to ensure that the cadets’ training schedule for this period only included classroom courses, so that no intervening noise exposure between evaluations was possible. For further measure, at the third session, volunteers were asked to complete a detailed noise history covering the previous month. Any threshold shifts compared to the baseline evaluation were considered to mirror permanent damage and noted as PTS. A threshold shift (either TTS or PTS) was recorded whenever an average 5 dB HL or greater shift occurred in the threshold at any of the frequencies tested for either ear. All shifts were considered clinically significant.

All research was conducted in compliance with all applicable regulations of both the Ministry of Defense and the author’s Institution regulations regarding the protection of human participants during research. Normally distributed continuous variables were expressed as mean and standard deviation (SD). The Kolmogorov–Smirnov test was applied to evaluate the normality of the distributions. Differences between the 2 study groups as well as other subgroups (i.e., TTS, PTS) were evaluated by Student’s t-test or the Mann–Whitney U test, whenever more appropriate. All reported P-values are based on two-sided tests and compared to a significance level of 5%. The McNemar test was used to study paired data such as the left and right ear threshold shifts. Correlations between variables were examined by Spearman’s rho correlation coefficients. Receiver operating characteristics (ROC) curves were generated and the area under the curve (AUC) [and its 95% confidence intervals (95% CI)] was calculated to determine the best discriminating level of OAE and MOCR for predicting TTS and PTS. Optimal discrimination limits were identified at the cut-point that maximizes sensitivity and specificity. All the other data analysis was performed using the Statistical Package for the Social Sciences version 24.0 software (SPSS Inc., Chicago, Illinois). For histograms, R 3.4.1 was used.

Results

After implementing exclusion criteria, a total of 344 volunteers (688 ears) were considered eligible. The male group comprised 188 participants outnumbering the female group by 32 volunteers (female group, N = 156).

Table 1 demonstrates that after the firing drill 280 participants suffered a TTS (468 ears), while 142 of these ears still showed a threshold shift compared to baseline evaluation 30 days after exposure (PTS ears). A total of 191 and 24 participants suffered bilateral TTS and bilateral PTS, respectively. It is self-evident that no patient was found to suffer a PTS without having demonstrated a respective TTS in the second hearing evaluation. In 454 ears, PTS was found to be smaller (or none) than the original TTS, while in 14 ears, TTS and the consequent PTS were of the same value. With regard to frequency, hearing damage mainly occurred in the 3 kHz frequency region [343 ears with TTS (49.9%) and 62 ears with PTS (9.1%)]. In terms of TTS, 405 ears were affected in two or more frequencies while 57 ears showed PTS in more than one frequency. The amount of hearing damage (TTS and PTS) per frequency is shown in Figures 1 and 2.

Table 1

Permanent threshold shifts and temporary threshold shifts according to ear side (left–right) and handedness

		All participants	Left-handed participants		Right-handed participants
			Ear		Ear
			Left	Right	Left	Right
PTS ears	N	142	10	18	67	47
	%	20.6	1.44	2.61	9.73	6.82
Non-PTS ears	N	546	38	30	229	249
	%	79.4	5.53	4.37	33.3	36.2
TTS ears	N	468	37	37	196	198
	%	68	5.37	5.37	28.48	28.78
Non-TTS ears	N	220	11	11	100	98
	%	32	1.6	1.6	14.55	14.25

Figure 1

Histograms indicating the amount of temporary threshold shifts (TTS) per audiometric frequency and number of affected ears

Figure 2

Histograms indicating the amount of permanent threshold shifts (PTS) per audiometric frequency and number of affected ears

All participants were questioned for handedness to assess whether potential hearing damage would confirm the “head shadow effect.” “Head shadow effect” refers to the unilateral or asymmetric hearing loss in which hearing loss of the left ear is severe in right-handers and vice versa,[11] due to the protection of one ear because of posture during rifle shooting. Table 1 also shows PTS ears according to side and handedness. The McNemar test validated the head shadow effect which reached statistical significance for right-handed individuals (P = 0.027). In the examination of left-handed participants, no statistical significance was reached perhaps due to the small sample size, yet a distinct trend toward significance was noticed (P = 0.115). The overall percentage of left-handed individuals that suffered a TTS was 77.1%, while the corresponding percentage for right handers was 67.7%. The difference was not statistically significant (x 2 test, P = 0.184). The PTS percentage was 37.5% for the left-handed individuals and 22.6% for right-handed individuals. The difference was marginally significant (x 2 test, P = 0.042). This statistical significance most probably bears no clinical value and should be attributed to the small sample size of the left handers.

As far as OAEs were concerned, Table 2 demonstrates OAEs amplitudes in the three evaluations that were conducted. Although most TTS ears and PTS ears showed a decreased amplitude in OAEs compared to the baseline evaluation, no significant correlation could be established between OAE overall amplitude decrease and audiometric threshold shifts [Spearman’s test (Spearman’s correlation coefficient, 0.13 < ρ < 0.183)]. In other words, in no way could OAEs replace pure-tone audiometry in assessing hearing damage. These results are in line with traditional notions on OAEs as suggested by extensive literature.[12] Participants were also divided into “No shift”, “TTS,” and “PTS” and for each group, descriptive data of OAE amplitudes were summarized [Table 3]. Additionally, despite focusing on the total response, we made an effort to correlate TTS/PTS on specific frequencies with OAE amplitude to that same frequencies, to establish whether frequency-specific OAE amplitude can predict vulnerability to TTS/PTS at that frequency. Nevertheless, statistical significance was not reached in any case. One reason for this may be that many TTS/PTS occurred in the 6–8 kHz frequency region, where TEOAEs amplitude is not reliable.

Table 2

OAEs’ amplitudes and differences established during consecutive evaluations

	N	Minimum	Maximum	Mean	Standard deviation
OAE amplitude 1st evaluation (dB SPL)	683	0.00	27.90	12.21	5.02
OAE amplitude 2nd evaluation (dB SPL)	684	0.0	24.9	11.52	5.26
OAE amplitude 3rd evaluation (dB SPL)	666	0.0	26.5	11.57	5.41
Amplitude difference between 1st and 2nd evaluation (dB)	680	−7.20	9.40	0.64	1.77
Amplitude difference between 1st and 3rd evaluation (dB)	662	−8.10	14.90	0.54	2.06

Table 3

Descriptive data of OAE amplitudes and efferent effects for “No shift,” “TTS,” and “PTS” groups

		No shift group	TTS group	PTS group
Amplitude of OAEs’ − 1st evaluation (dB SPL)	Mean	13.38	12.41	9.69
	Standard deviation	4.99	4.84	4.61
	Minimum	0.0	1.90	1.90
	Maximum	27.90	25.10	25.10
Amplitude of OAEs’ − 2nd evaluation (dB SPL)	Mean	12.81	11.82	8.83
	Standard deviation	5.04	5.07	4.85
	Minimum	2.4	0.0	0.0
	Maximum	24.9	23.7	23.7
Amplitude of OAEs’ − 3rd evaluation (dB SPL)	Mean	13.16	11.79	8.76
	Standard deviation	5.12	5.14	4.64
	Minimum	2.2	0.6	1.9
	Maximum	26.5	24.1	23.9
MCOR effect − suppression of OAEs’ amplitude with contralateral noise (dB)	Mean	1.54	1.48	1.42
	Standard deviation	1.01	0.94	1.00
	Minimum	0.00	0.00	0.00
	Maximum	4.90	5.80	5.80

Regarding the olivocochlear reflex function, Table 4 summarizes OAEs amplitude with and without contralateral noise. The mean suppression of OAEs amplitude due to MOCR function was 1.49 dB SPL with a SD of 0.95. Again participants were also divided into “No shift,” “TTS,” and “PTS” and for each group, efferent effects were summarized [Table 3].

Table 4

OAEs’ amplitude with and without contralateral noise

	N	Min	Max	Mean	Standard deviation
OAEs’ amplitude without contralateral noise (dB SPL)	624	0.1	23.5	9.73	5.28
OAEs’ amplitude with contralateral noise (dB SPL)	619	−2.2	22.3	8.30	5.28
Suppression of OAEs’ amplitude with contralateral noise (dB)	619	0	5.8	1.49	0.95

Furthermore, having established that the opposite ear of the dominant hand is at a greater risk of hearing damage, according to the head shadow effect, we made an effort to correlate OAEs amplitude in baseline evaluation, with the risk of hearing damage in these ears. Mann–Whitney test established that OAE amplitude was significantly correlated with the future risk of both TTS (P = 0.092) and PTS (P < 0.001) [Figure 3]. Nevertheless, because our main aim was to investigate whether OAE amplitude could be used as a screening diagnostic test for noise vulnerability, we used a ROC curve for OAEs amplitude and future TTS or PTS in the left ear of right handers and the right ear of left handers. ROC curve analysis showed that OAEs amplitude in a baseline evaluation is predictive of TTS and PTS [Figures 4 and 5]. The AUC for TTS was 0.582 (95% CI: 0.517–0.648; P = 0.017) and for PTS was 0.745 (95% CI: 0.680–0.809; P < 0.001).

Figure 3

OAE amplitude per TTS/PTS and non-TTS/PTS (box plots)

Figure 4

ROC curves for OAEs’ amplitude (POAE) in baseline evaluation and MOCR to predict TTS

Figure 5

ROC curves for OAEs’ amplitude (POAE) in baseline evaluation and MOCR to predict PTS

Results differed, however, when evaluating whether MOCR strength can be linked to the likelihood of threshold shifts. Mann–Whitney test established that the suppression of OAEs amplitude with contralateral noise (the difference between the amplitude in decibels of the total response without contralateral noise and that with contralateral noise) was not significantly correlated with neither the risk of future TTS (P = 0.164) nor PTS (P = 0.071) [Figure 6]. Despite that correlation did not reach statistical significance, a certain trend toward significance was evident, at least as far as PTS was concerned. Again, a ROC curve analysis was used for MOCR strength and future TTS or PTS in the left ear of right handers and the right ear of the left handers. This time TTS AUC was 0.548 (95% CI: 0.480–0.615; P = 0.167) and PTS AUC was 0.566 (95% CI: 0.492–0.640; P = 0.077) [Figures 4 and 5].

Figure 6

MOCR strength per TTS/PTS and non-TTS/PTS (box plots)

The curves were used to determine cut points to evaluate the likelihood of PTS and TTS for OAEs amplitude in the baseline evaluation. Decision limits yielding 71.6% sensitivity were 12.45 dB SPL (positive if less than or equal to this value) with 63.8% specificity for PTS and 50% sensitivity were 12.35 dB SPL (positive if less than or equal to this value) with 68.2% specificity for TTS. Because ROC curve analysis did not reach significance for MOCR strength, optimal cut points were not determined.

Finally, to test whether 5 dB HL threshold shift might be considered as small and fall within normal repeated measure variation, we repeated the statistical analysis undertaken, evaluating only ≥10 dB HL shifts.

In this case, after the firing drill, 368 ears suffered a ≥10 dB HL TTS, while 96 of these ears still showed a threshold shift compared to baseline evaluation 30 days after exposure (PTS ears). When “head shadow effect” was applied, we analyzed 183 TTS and 58 PTS ears. Mann–Whitney test established that OAE amplitude was significantly correlated with the future risk of both TTS (P = 0.006) and PTS (P < 0.001).

ROC curve analysis showed that OAEs amplitude in a baseline evaluation is predictive of both TTS and PTS [Figures 7 and 8]. The AUC for TTS was 0.587 (95% CI: 0.526–0.647; P = 0.006) and for PTS was 0.678 (95% CI: 0.601–0.755; P < 0.001). The curves were used to determine cut points to evaluate the likelihood of PTS and TTS for OAEs amplitude in the baseline evaluation. Decision limits yielding 70.7% sensitivity were 12.35 dB SPL (positive if less than or equal to this value) with 57.6% specificity for PTS and 61.1% sensitivity were 12.65 dB SPL (positive if less than or equal to this value) with 57.8% specificity for TTS.

Figure 7

ROC curve for OAEs’ amplitude in baseline evaluation to predict TTS when only ≥10 dB HL shifts were considered significant

Figure 8

ROC curve for OAEs’ amplitude in baseline evaluation to predict PTS when only ≥10 dB HL shifts were considered significant

Overall, it seems that results were not fundamentally different when only ≥10 dB HL shifts were considered significant.

Discussion

Firearms produce impulse noises characterized by peak pressure level and frequency. Exposure to such noise is particularly common during military service and it is beyond doubt that even minimal exposure can result in high risk of hearing loss. Potential adverse effects on hearing have been reported even after a single gunshot, thus hearing protection is obligatory according to national and global regulations. Despite the use of hearing protectors, noise exposure can result in TTS or PTS although these shifts usually do not exceed the 20 dB HL threshold that usually defines normal hearing.[13]

In our study, PTS ears amounted to a striking 20.6% (N = 142) which translates to 118 participants. It goes without saying that because all threshold shifts were considered clinically important, the aforementioned amounts do not mirror patients that suffered an actual hearing handicap. In fact, not a single volunteer was recorded to have an audiometric threshold >20 dB HL in any of the measured frequencies after noise exposure. Yet, it is evident that although precautionary measures effectively averted any major losses of hearing, clinically evident on standard audiometry, minor cochlear damage, even with earplugs on, remained a very common incidence. Most damages occurred around the 3 kHz frequency region, just as literature suggests for firearm acoustic trauma. Our results are in line with multiple studies that suggest that military personnel are vulnerable to cochlear damage during training.[2314151617]

What seems to be the major issue, though, when dealing with groups exposed to firearm noise is which particular participants are more prone to developing hearing loss because of such intense impulse noise. Because such audiometric threshold shifts are a common occurrence and specific treatment of noise-induced damages still remains ambiguous, prevention remains the cornerstone of proper management when dealing with these acute exposures to noise. However, prevention could be far more effective if parts of the population could be classified as “vulnerable” or “resistant” according to characteristics that could be determined prior to exposure to impulse noises.

Such a classification is hardly a novel concept. Several individual characteristics have been proposed as indicators of susceptibility to noise-induced damages: age,[18] sex,[19] race,[20] medical conditions,[212223] smoking,[24] genetic predisposition, genes,[2526272829] among others. In the same context, even since the 1960s,[30] a vast number of clinical tests had been proposed[313233] to forecast such damages, yet none proved specific or sensitive enough to achieve satisfactory results in a clinical setting. Currently, no audiometric test is available to evaluate oversensitive participants to impulse noise.

On the other hand, OAEs have repeatedly shown high sensitivity and specificity in assessing cochlear damage after short exposure to noise.[3335] Taking these suggestions to the next level, Lapsley Miller et al.[6] further reported that the PTS in military personnel was predicted by baseline low-level or absent OAEs. These authors have concluded that the best predictor is TEOAEs, with risk increasing more than sixfold as the emission amplitude decreased. Additionally, after studying 317 US Marine Corps recruits before and after exposure to impulse-noise sources from the firing of weapons, Marshall et al.[5] also suggested that low-level OAEs indicate an increased risk of future hearing loss by as much as ninefold.

Our results seem to further validate the above notions. In our study, evidence suggests that indeed baseline OAEs recorded prior to exposure to gunfire noise could effectively identify participants that may suffer from a future TTS and/or PTS. In addition, because cadets had no prior exposure to significant occupational-related impulse noises, it appears that such a prediction applies not only to previously “damaged” ears, as shown by previous work,[56] but to the general population as well. If further studies confirm our suggestions, serious consideration should be given for the inclusion of OAEs as an invaluable prerequisite for all military and nonmilitary hearing conservation programs.

On the other hand, efferent function has been linked to the risk of permanent noise-induced damage in mammals such as guinea pigs,[36] rabbits,[37] and mice.[38] In particular, Zheng et al.[39] exposed chinchillas to impulse noise similar to that of firearms after having surgically sectioned the olivocochlear bundle. These authors suggested that inner hair cells of de-efferented ears were more susceptible to damage due to impulse noise than those in efferented control ears. One could therefore assume that the intersubject variability of MOCR strength could be a distinctive factor in determining different susceptibility to noise-induced hearing damages. Such links could be presumably true for humans as well, yet it is particularly difficult to study, because it would ultimately require a series of experiments in populations that would be measured for olivocochlear reflex strength and then deliberately exposing them to high noise levels. Such a methodology obviously carries significant ethical dilemmas and despite efforts to do so in favorable populations with occupational or recreational exposure to potentially harmful noise,[11404142] it has been scarcely implemented, at least as far as impulse noises are concerned. To our knowledge, there was only one effort[43] to reveal any correlation between the amount of suppression of OAEs and individual TTS susceptibility after exposure to firearm generated noise. The authors concluded that the suppression of OAEs does not predict individual susceptibility to mild TTS induced by impulse noise in humans and tried to suggest possible explanations for the missing association, but still stressed the need for further research on the topic.

Our study is the first to attempt to correlate efferent strength, as estimated by the suppression of OAEs amplitude with contralateral noise, with both TTS and PTS participants taking into account head shadow effect. Although statistical significance was not reached, we feel that results clearly suggest that further research is due, to conclusively validate or dismiss the notion that efferent system serves as a protective mechanism to noise-induced insults of the inner ear.

The significance of our study lies in the fact that both the study population and the exposure to noise were fundamentally different and remarkably coherent in comparison to previous work. All recruits were of the same age and with no previous active duty or exposure to firearms of any kind. In both of the previous studies mentioned,[56] the authors used active duty personnel of various ages and experience therefore, with undefined previous exposure to noise and possibly different amounts of preexisting cochlear damage. Likewise, both studies did not use uniform and quantifiable exposure to noise for all compared participants. That is to say that either serving in an aircraft carrier or undergoing a noisy basic training involving weapon noise exposures were considered enough to comprise a uniform “noise exposed group.” The same ambiguity applied to other similar efforts with populations exposed to harmful environments[1240] or activities.[4142] Clearly, in such obscure conditions, real exposure may have been highly diversified between participants, thus inserting bias. In our case, all participants fired 10 single rounds using the same 7.62 mm Heckler & Koch G3A3 assault rifle while lying down in the same prone position. Additionally, both before and after the firing drill, cadet activities were severely restricted, and they were not exposed to any significant levels of military or recreational noise.

We also feel that as far as statistical analysis was concerned, the main advantage is that the analysis went one step further from simply exhibiting correlation between OAEs amplitude and future hearing damage. ROC curve analysis established optimal cut-off points and calculated sensitivity and specificity enabling the prediction of patients’ developing TTS and PTS. To our knowledge, it is the first time that an effort was undertaken to investigate the actual diagnostic value (in the form of sensitivity and specificity) of a clinical test on vulnerability to acoustic injury, based on OAEs amplitude. Results suggest that sensitivity for detecting future PTS reached a rather satisfactory 71.6%, thus proving the procedure to be sensitive enough to be used as screening evaluation, especially considering the lack of alternative methods.

We should also stress that in our study, both TTS and PTS predictions were attempted. Although TTS is by definition only temporary, one should not underestimate its significance. The understanding of the damaging effects of traumatic sounds has undergone a revolution over the last few years, so that it is now understood that a TTS that does not produce a PTS as measured by pure-tone thresholds does not mean that no long-term damage was done. In many circumstances, significant TTS will produce long-term degeneration of auditory-nerve fibers which degrades hearing ability (e.g., hearing speech in noisy environments) without producing a significant change in thresholds (termed “hidden hearing loss”).[44] The new knowledge shows that traumatic sounds do much more damage than had been realized, which makes it all the more important that participants who are particularly prone to noise trauma be identified by OAE levels, or any other means.As far as efferent function methodology is concerned, it should be noted that the type of protocol used to assess efferent function established that the stimulus artifacts were reduced due to lower intensity clicks and the patients and environment effects were minimized and were similar to those with and without contralateral noise due to reducing sweeps from 260 to 60. Finally, the use of different multiple sweeps and an averaging of the data allowed us to establish adequate stability and test–retest repeatability as current research suggests.[45] However, one of the main problems with the assessment of the MOC reflex strength is that insufficient averaging is performed so that the OAE measurement SNRs are not adequate for measuring the change produce by the MOCR-induced efferent activity. So, the variation from one measurement to the next can be more than the MOC effect. This will add a random variation to the MOC measurements that will wash out any actual correlation between the MOC strength and the PTS (or TTS). A nice exposition of this issue is given in Goodman et al.,[46] which shows that to accurately measure a 1 dB MOC change, each OAE measurement must have a SNR of ∼25 dB. Probably some of the participants in this study had high enough SNRs and these helped produce the trend of positive correlation that was measured. The issue should stand as a guide for future reference, so that all future studies could consider pursuing sufficiently high SNRs, by extensive averaging. In all, to our knowledge, there is no prior reference in the English literature of a study correlating MOCR strength with resistance to overall acoustic injury (both TTS and PTS) due to impulse noise in a clinical setting.

Conclusion

Although hearing loss due to the impulse noise of firearms remains a significant risk among the military population, currently there is no clinical method that can effectively identify participants who are particularly susceptible to take it, to implement precautionary measures for the prevention of such damages.

The current data are intriguing and yield tentative evidence to suggest that OAEs amplitude is both sensitive and specific enough to be used as a predictor, and so hearing conservation programs should consider including them in their routine. As for MOCR strength, we feel that further research is due, before proceeding with the implementation of suppression of OAEs in hearing conservations programs in a similar manner.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.