Professional Documents
Culture Documents
net/publication/337585059
CITATIONS READS
5 435
2 authors, including:
Elliott H. Berger
Berger Acoustical Consulting LLC
107 PUBLICATIONS 1,479 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Elliott H. Berger on 14 February 2020.
(Received 11 April 2019; revised 3 August 2019; accepted 8 August 2019; published online 27
November 2019)
Acoustic trauma (AT) is permanent hearing loss after a single noise exposure. A few human cases
resulting from continuous, i.e., nonimpulsive noise, have been reported as reviewed by Ward
[(1991). “Hearing loss from noise and music,” presented at Audio Engineering Society, New York,
October 4–8]. This paper updates that review by examining 11 cases in nine reports, from 1950 to
2006, with the intention of determining minimum exposures that may cause AT, including the
potential risk of exposure to noise from magnetic resonance imaging machines. Diffuse-field
related levels above 120 dBA for 10 s or more, or above 130 dBA for 2–3 s (values well above
OSHA’s unprotected exposure limits), can lead to AT. These cases appear to represent a susceptible
fraction of the population, because much more intense exposures (e.g., 130 dBA for 32 min) have
been tolerated by groups of volunteers who suffered only temporary threshold shifts. AT from con-
tinuous noise is unlikely to occur in OSHA-compliant hearing conservation programs, and probably
rare enough in the general civilian population that clinical trials of drugs aimed at treating it are
unlikely to be practical. AT from impulse noise, such as gunfire, which is specifically not the topic
of the current work, is more amenable to clinical trials, especially in military settings.
C 2019 Acoustical Society of America. https://doi.org/10.1121/1.5132712
V
J. Acoust. Soc. Am. 146 (5), November 2019 0001-4966/2019/146(5)/3873/6/$30.00 C 2019 Acoustical Society of America
V 3873
Prior to the 1980s, human TTS studies often used expo-
sures that were much more intense than exposures leading to
AT in case reports. In these studies groups of subjects sus-
tained only TTS, highlighting the variation in susceptibility to
AT. Selected TTS studies from both the reference list of
Ward (1991) and our own files were reviewed to illustrate
this phenomenon; no formal search was done for TTS studies.
In recent years, a new term, “acoustic shock,” has entered
the medical literature. A PubMed search in April 2019 yielded
only ten relevant references, generally defining this term as a
syndrome of persistent otalgia, hyperacusis, tinnitus, and/or
vestibular problems. This syndrome was reported primarily in
clusters of call center workers in Australia, New Zealand, the
United Kingdom, and Denmark who had experienced unex-
FIG. 1. Reproduction of Fig. 1 from Ward (1991), showing his four safe
pected loud sounds. The syndrome is considered by many
exposures (encircled points) and his five unsafe exposures, together with a authors to be primarily psychogenic (Hooper, 2014). We did
dashed line representing all exposures equal in energy to 1 min of 135 dB, not address this condition, limiting our review to cases report-
and a dotted line connecting “the four weakest individual exposures that ing measurable hearing loss after single noise exposures.
produced permanent loss in a small but otherwise unknown fraction of those
exposed. Presumably any exposure above and to the right of this line can be Because the case reports expressed sound levels in dif-
considered hazardous to the most susceptible ears.” Note that Ward’s time ferent ways [diffuse-field levels, dB hearing level (HL), cou-
axis extends beyond the range covered in Figs. 2–4, specifically starting at pler levels, and manikin levels], all noise exposures were
less than 1 s to include the SN data point (simulated automotive airbag expo-
sure) which was excluded from the current analysis since its duration was
converted to diffuse-field related levels (dBA) to make them
only 0.4 s. comparable to levels measured in occupational hearing con-
servation programs (HCPs) that represent the classical stud-
at 137 dB from a cordless telephone required only 2 s of expo- ies upon which accepted damage-risk criteria are based.
sure to cause AT (Orchik et al., 1987). The goal of this paper These conversions relied on ISO 11904-2 (2004) and on
is to review cases of human AT from continuous noise, updat- tables provided by Bentler and Pavlovic (1989, 1992);
ing Ward’s (1991) report by adding in more recent case detailed computations are available on request. For AT cases
reports, to estimate the levels and durations that can cause AT arising from acoustic reflex testing, it was sometimes unclear
in susceptible people. Implications of these data for clinical which of the test tones (500, 1000, 2000, or 4000 Hz) had
trials of interventions aimed at treating AT are discussed. caused the hearing loss; the assumption was made that the
tone with the highest A-weighted sound level had been
II. METHODS responsible.
The literature search began with papers cited by Ward III. RESULTS
(1991) and over 10 000 papers in the authors’ files (over
6000 for E.H.B., over 10 000 for R.A.D., undoubtedly Several TTS studies (i.e., no resultant AT) with very
including duplicates). The reference lists of relevant papers intense exposures and the 11 informative cases of AT (three
that we identified as “interesting” were also searched. A of the data points are from a single study) are summarized in
PubMed search in January 2019 used the following terms: Table I and Fig. 2. Three columns in Table I show sound lev-
els in different ways: L refers to levels as recorded in the
“acoustic trauma” OR (hearing loss AND (noise OR underlying papers, LDF refers to levels that have been cor-
music) AND (sudden OR acute)). rected to estimate diffuse-field levels, and LDF,A refers to
levels that have been corrected to estimate A-weighted dif-
This search yielded 577 hits, only 11 of which seemed likely fuse-field levels. In addition, 8-h equivalent exposures are
to be relevant based on title and abstract. The Defense shown for diffuse-field levels, both LDF,eq8h and LDF,Aeq8h.
Technical Information Center database was also searched, TTS-only studies included between 10 and 40 subjects
and email queries were sent to 47 colleagues who were (89 in total), with the exception of Raleigh et al. (1963),
known to be interested in NIHL and hearing conservation which was a report of two workers trapped without hearing
(most are published authors in these fields). protection for 2 hours in a workplace where a steam line had
Potentially relevant papers were reviewed by both ruptured. In that report, a diffuse-field level of 136 dBA
authors; nine papers were considered informative because caused no AT.
they provided data on exposures (level and duration) and AT cases were documented after exposures such as a
audiometric outcomes (pure-tone threshold shifts of at least diffuse-field related level of 121 dBA for 8 min (Davis et al.,
10 dB, at least 1 week post-exposure, compared with pre- 1950), and 123 dBA for 10 s (Arriaga and Luxford, 1993),
exposure audiograms, or in cases of unilateral exposure com- which were less intense than those tolerated by subject
parisons to the thresholds of the contralateral ear). Three of groups in TTS studies. An apparent outlier was the report of
the nine papers were cited by Ward (1991), four were in the Elonka (1986) with a diffuse-field related level of 111 dBA
authors’ files, and the PubMed search yielded two more. for 10 s; this was a Letter to the Editor showing only
3874 J. Acoust. Soc. Am. 146 (5), November 2019 Elliott H. Berger and Robert A. Dobie
TABLE I. TTS and AT cases and associated exposures. The L (dB) values are as reported in the papers cited; the diffuse-field values (LDF and LA,DF) are cor-
rected as discussed in text. The exposure durations (dur) are specified in seconds (s). In the frequency column “various” indicates that in different experiments
there were different tones or noise bands that were utilized, and “range” indicates the sound was a relatively broadband of noise. The first three rows of TTS
data and the first five rows of AT data were included in Ward (1991). All of the reflex text cases were on subjects that had prior significant sensorineural hear-
ing loss.
TTS Cases
Davis et al., point HD 1950 experiment speaker tones & noise various 130 130 130 1920 118 118
Eldred et al., point E 1955 experiment speaker jet noise range 135 135 135 60 108 108
Raleigh et al. 1963 accident steam line steam line rupt. range 138 138 136 7200 132 130
Schori 1976 experiment TDH-39 pink noise range 130 124 123 10 89 88
Brownsey 1973 experiment speaker pink noise range 129 127 127 30 97 97
AT Cases
Davis et al., subj WM 1950 experiment speaker tone 2000 130 130 131 60 103 104
Davis et al., subj S 1950 experiment speaker tone 4000 120 120 121 480 102 103
Davis et al., subj HD 1950 experiment speaker tone 500 140 140 137 1800 128 125
Lenarz and Gulzow 1983 reflex test audiometer tone 500 125 HL 135 132 10 100 97
Orchik et al. 1987 accident phone ring tone 750 137 135 134 2 93 92
Arriaga and Luxford 1993 reflex test audiometer tone 2000 120 HL 122 123 10 87 89
Elonka 1986 reflex test audiometer tone unstated 105 HL 111 111 10 76 76
Hunter et al. 1999 reflex test audiometer tone 1000 120 HL 126 126 20 94 94
Kung and Sataloff 2006 accident siren wail 1000 140 140 140 2 98 98
McMillan and Kileny 1994 accident bike horn bike horn 1500 143 143 144 5 105 106
Singleton et al. 1984 accident phone ring tone 750–1000 137 135 134 2.5 94 93
borderline threshold shifts 1 week after exposure (5–15 dB, exchange rate for exposures between 90 and 115 dBA.
depending on which pre-exposure test was compared to the NIOSH’s recommended exposure limit is a 3-dB exchange
post-exposure test) and lacking details such as the frequencies rate between 85 and 140 dBA. Although the OSHA and
of the test tones that were used to elicit acoustic reflexes. NIOSH curves are quite different, all exposures resulting in
Figure 2 also includes the U.S. Occupational Safety and AT exceeded these criteria, with the exception of the outlier
Health Administration (OSHA) and U. S. National Institute previously mentioned (Elonka, 1986).
for Occupational Safety and Health (NIOSH) limits. Figure 3 presents the same data as Fig. 2, with the labels
OSHA’s unprotected permissible exposure limit is for daily for the data points removed for simplification, but with the
career-long immersion in noise, and specifies no exposure addition of a curve that can be drawn from the data we
exceeding 1 s for levels above 115 dBA, with a 5-dB assembled, to indicate tentative levels and durations capable
of causing AT in the most susceptible persons as the result
J. Acoust. Soc. Am. 146 (5), November 2019 Elliott H. Berger and Robert A. Dobie 3875
of a single exposure. We emphasize the tentative nature of 2.2 dB. However, for those cases in which we had to use our
this curve. It suggests that diffuse-field related levels above best judgment to understand the author’s intent, we suspect
120 dBA for periods of 10 s or more, or above 130 dBA for that our uncertainties are at least twice that large, closer to
periods of 2–3 s (values that are well above OSHA’s unpro- approximately 5 dB.
tected exposure limits), can lead to AT. These cases appear Because 89 healthy young human volunteers sustained
to represent a susceptible fraction of the population, because only TTS after exposures that were both more intense and
much more intense exposures (e.g., 130 dBA for 32 min) longer in duration than exposures reported to cause AT (indi-
have been tolerated by groups of volunteers who suffered vidual TTS studies had between 10 and 40 subjects), it is
only temporary threshold shifts. reasonable to assume that AT cases represent more suscepti-
One may note that our tentative curve is more restrictive ble individuals. Patients suffering AT after acoustic reflex
than Ward’s “safe” curve in Fig. 1. As an example, his curve testing had substantial pre-existing hearing loss (which is
allows 135 dBA for 10 s whereas we suggest the limit for why such high levels were required to elicit acoustic
10 s is only about 122 dBA. The differences in part are reflexes); prior hearing loss due to factors besides noise
because we corrected the original reported data points to exposure could thus be a risk factor for AT.
diffuse-field levels as noted in Sec. II, and also due to the Many subjects in the studies intended to elicit only TTS
addition of new data showing risk at lower levels. sustained exceedingly large threshold shifts: up to 60 dB,
requiring more than 24 h to recover (Davis et al., 1950).
IV. DISCUSSION TTSs like these, even with no permanent threshold shift,
have been shown in animals to be associated with loss of
Case reports are, by nature, anecdotal and represent the auditory nerve function (Kujawa and Liberman, 2009) and
lowest level of medical evidence, below stronger designs cannot be confidently considered safe for humans.
such as clinical trials, cohort studies, and cross-sectional All AT cases, with the exception of an outlier previously
studies. One case report (Kung and Sataloff, 2006) appeared mentioned, had exposures in excess of those permitted with-
in a non-refereed publication and another (Elonka, 1986) out hearing protection by OSHA or recommended by
was a brief Letter to the Editor. However, it would be unethi- NIOSH. All cases involved either pure tones or narrowband
cal to perform controlled experiments designed to cause AT sounds, which are generally considered more hazardous on
in humans; thus, case reports provide the only available evi- an equal-energy basis than the broadband sounds that char-
dence for human AT from continuous noise that we were acterize most occupational noise exposures [ISO 1999-2013
able to uncover. Moreover, even in the absence of the ethical (2013)]. For these reasons clinical trials of NIHL in HCPs
issues, controlled clinical trials of the nature that might would be unlikely to yield useful numbers of cases. Some
reveal AT following short-duration exposures would of steady-state military exposures where even double hearing
necessity be limited to impulse noise from weapons (not protection can be inadequate (e.g., flight deck of aircraft car-
covered in this review) since those are the only types of rier) might be exceptions to this rule.
short-duration sounds that occur with some predictability or As discussed in this review there are few published
regularity. Such trials would likely need to occur in military reports of AT from non-occupational exposures to continu-
populations, and a recent multi-center trial of tinnitus ous noise. Some studies document exposures that are now
retraining therapy (TRT), discusses the challenges of clinical far less prevalent: TTS research no longer permits large
trials even for something as benign as TRT (Scherer et al., threshold shifts that could cause permanent damage; contem-
2018). Therefore, our review captured only cases of AT in porary cordless phones do not ring through the earpieces;
which clinicians chose to submit case reports; unreported audiologists now may decline to test acoustic reflexes at
AT cases almost certainly outnumber those we reviewed. very high levels. While actual cases are probably more
In examining the nine papers that met our criteria, one numerous than well-documented case reports, they are
of them dating from 1950 and two-thirds having been con- spread throughout the population at large. Clinical trials of
ducted over 25 yr ago, the authors noted a lack of precision therapy for rare disorders like this typically require multiple
in many of these papers with respect to specification of the sites and long accrual periods, with accordant increase in
actual transducers used, the acoustics of the test space (for cost and threats to data quality. It is questionable, in the
sound-field testing), the spectrum of the noise in the case of opinion of E.H.B. and R.A.D., whether clinical trials for AT
non-pure tone sources, and even the actual tone in question from continuous noise could be practical.
when more than one tone was used in the study. Thus, not Recent research (Campbell, 2019; Oishi and Schacht,
only did we need to adjust earphone values to diffuse-field 2011) has sought without notable success to find nutritional
related sound levels, but also had to make judgment calls or drug treatments for human NIHL and AT. An important
about other factors such as those mentioned above. pitfall in NIHL clinical trials is that hearing protection devi-
Depending on the precision of the underlying reported ces (HPDs), such as earplugs and earmuffs, are readily avail-
data, the accuracy of our estimated DF values varies. able, ethically mandatory in clinical trials, and can markedly
Computed adjustments when all the transducers, the test sig- reduce NIHL, thus necessitating large sample sizes to have a
nals, and the reported levels were sufficiently specified chance to detect small changes in hearing. Another is that
would be expected to fall within the uncertainty of the stand- NIHL develops slowly, requiring trials lasting years.
ards we followed. In the case of computations following ISO AT might seem a more attractive target for drug research,
11904-2 (2004) the expanded uncertainty (95th percentile) is because AT often results from unexpected exposures, with no
3876 J. Acoust. Soc. Am. 146 (5), November 2019 Elliott H. Berger and Robert A. Dobie
HPDs in place, and does not require long follow-up. But clini- kurtotic noises are more hazardous than ones that are not
cal trials for AT have other inherent difficulties: (Davis et al., 2012). Furthermore, the health of the subjects
in many of the studies we reviewed was not compromised in
(1) Many people suffering AT will lack pre-exposure audio-
ways that might be anticipated for those requiring an MRI.
grams, making estimation of threshold shift difficult;
The studies cited in Sec. I have reported TTSs following
(2) unexpected exposures are often difficult to replicate and
MRIs. Although no permanent shifts were reported, most of
measure;
the authors recommended use of hearing protection. We too
(3) it would be unethical to subject human subjects to expo-
recommend use of hearing protection and given the nature
sures that are likely to cause AT. Thus, clinical trials are
of our findings summarized in Fig. 4, suggest that during
typically limited to drug treatment given after AT (ani-
MRI scanning on patients wearing presumably well-fitted
mal studies have shown that many interventions prior to
hearing protection AT is unlikely to occur, but cannot be
AT can reduce hearing loss, but few are effective after
ruled out.
AT, even when promptly delivered).
Impulse noise (e.g., gunfire from recreational or military ACKNOWLEDGMENTS
exposure) may be a more important cause of AT than contin- Portions of this work were presented at the annual
uous noise, but was not considered in this review. Several meeting of the National Hearing Conservation Association,
clinical trials been reported for AT from impulse noise, usu- February 8, 2019, Grapevine, TX. Dr. Christian Giguère
ally military in origin (e.g., Pilgramm and Schumann, 1985; consulted regarding methods of estimating diffuse-field
Kopke et al., 2015). AT trials in military settings have levels from HLs, coupler levels, and mannequin levels. Dr.
accrued impressive numbers of subjects and appear feasible William Clark consulted regarding historical literature in the
for future research. It is often difficult in such trials to be fields of TTS and AT.
sure that the last exposure, after which the soldier reported
auditory symptoms, was really the cause of the threshold 1
We did not include one of Ward’s safe exposures, labeled SN in his graph
shift; instead, the threshold shift may have developed over since it was a 0.4-s burst of noise intended to simulate an airbag discharge,
days to weeks of training. Thus, results of such trials might and our definition of AT exposures requires duration of >1 s.
J. Acoust. Soc. Am. 146 (5), November 2019 Elliott H. Berger and Robert A. Dobie 3877
Jin, C., Li, H., Li, X., Wang, M., Liu, C., Guo, J., and Yang, J. (2018). Price, D. L., De Wilde, J. P., Papadaki, A. M., Curran, J. S., and Kitney, R.
“Temporary hearing threshold shift in healthy volunteers with hearing pro- I. (2001). “Investigation of acoustic noise on 15 MRI scanners from 0.2 T
tection caused by acoustic noise exposure during 3-T multisequence MR to 3 T,” J. Mag. Res. Imag. 13, 288–293.
neuroimaging,” Radiology 286(2), 602–608. Raleigh, R. L., Pringle, D., and Goddard, F. (1963). “Temporary hearing
Kirchner, D. B., Evenson, E., Dobie, R. A., Rabinowitz, P., Crawford, J., loss associated with a hydrocarbon cracking plant explosion,” J. Occup.
Kopke, R., and Hudson, T. W. (2012). “Occupational noise-induced hear- Med. 5, 499–500.
ing loss,” J. Occup. Environ. Med. 54, 106–108. Ravicz, M. E., Melcher, J. R., and Kiang, N. Y.-S. (2000). “Acoustic noise
Kopke, R., Slade, M. D., Jackson, R., Hammill, T., Fausti, S., Lonsbury- during functional magnetic resonance imaging,” J. Acoust. Soc. Am.
Martin, B., Sanderson, A., Dreisbach, L., Rabinowitz, P., Torre, P. III, and 108(4), 1683–1696.
Balough, B. (2015). “Efficacy and safety of N-acetylcysteine in prevention Robinson, D. W. (1971). Occupational Hearing Loss (Academic Press,
of noise induced hearing loss: A randomized clinical trial,” Hear. Res. London), pp. 173–178 and 253–260.
323, 40–50. Sakat, M. S., Kilik, K., and Bercin, S. (2016). “Pharmacological agents used
Kujawa, S. G., and Liberman, M. C. (2009). “Adding insult to injury: for treatment and prevention in noise-induced hearing loss,” Eur. Arch.
Cochlear nerve degeneration after ‘temporary’ noise-induced hearing Otorhinolaryngol. 273, 4089–4101.
loss,” J. Neurosci. 29, 14077–14085. Salvi, R., and Sheppard, A. (2018). “Is noise in the MR imager a significant
Kung, B., and Sataloff, R. T. (2006). “Noise-induced perilymph fistula,” Ear risk factor for hearing loss?,” Radiology 286(2), 609–610.
Nose Throat J. 85, 240–246. Scherer, R. W., Sensinger, L. D., Sierra-Irizarry, B., and Formby, C. (on
Lenarz, T., and Gulzow, J. (1983). “Acoustic inner ear trauma by behalf of the TRTT Research Group) (2018). “Lessons learned conducting
impedance measurement,” Laryngol. Rhinol. Otol. 62, 58–61 (in a multi-center trial with a military population: The Tinnitus Retraining
German). Therapy Trial,” Clin. Trials 15(5), 429–435.
McJury, M., and Shellock, F. G. (2000). “Auditory noise associated with Schori, T. R. (1976). “Evaluation of safe exposure guidelines for moderate
MR procedures: A review,” J. Mag. Res. Imag. 12, 37–45. and high intensity continuous noise,” AMRL-TR-76-99, Aerospace
McMillan, P. M., and Kileny, P. R. (1994). “Hearing loss from a bicycle Medical Research Laboratory, Wright-Patterson Air Force Base, Ohio.
horn,” J. Am. Acad. Audiol. 5, 7–9. Segal, S., Harell, M., Shahar, A., and Englender, M. (1988). “Acute acoustic
Oishi, N., and Schacht, J. (2011). “Emerging treatments for noise-induced trauma: Dynamics of hearing loss following cessation of exposure,” Am.
hearing loss,” Expert Opin. Emerg. Drugs 16, 235–245. J. Otol. 9(4), 293–298.
Orchik, D. J., Schmaier, D. R., Shea, J. J., Jr., Emmett, J. R., Moretz, W. M., Singleton, G. T., Whitaker, D. L., Keim, R. J., and Kemker, F. J. (1984).
and Shea, J. J. III (1987). “Sensorineural hearing loss in cordless telephone “Cordless telephones: A threat to hearing,” Ann. Otol. Rhinol. Laryngol.
injury,” Otolaryngol. Head Neck Surg. 96, 30–33. 93, 565–568.
Pilgramm, M., and Schumann, K. (1985). “Hyperbaric oxygen therapy for Ward, W. D. (1991). “Hearing loss from noise and music,” presented at
acute acoustic trauma,” Arch Otorhinolaryngol. 241, 247–257. Audio Engineering Society, Oct. 4–8, New York.
3878 J. Acoust. Soc. Am. 146 (5), November 2019 Elliott H. Berger and Robert A. Dobie