Noise Health Home 

[Download PDF]
Year : 2011  |  Volume : 13  |  Issue : 55  |  Page : 423--431

Noise-induced tinnitus: A comparison between four clinical groups without apparent hearing loss

Ann-Cathrine Lindblad1, Björn Hagerman1, Ulf Rosenhall2,  
1 Department of Clinical Science, Intervention and Technology, Division of Ear, Nose and Throat Diseases, Unit of Technical and Experimental Audiology, Karolinska Institutet, Stockholm, Sweden
2 Department of Clinical Science, Intervention and Technology, Division of Ear, Nose and Throat Diseases, Karolinska Institutet and Department of Audiology, Karolinska University Hospital, Stockholm, Sweden

Correspondence Address:
Ann-Cathrine Lindblad
Technical Audiology, KI, M45, Karolinska/Huddinge, SE-141 86 Stockholm


The number of people with normal hearing thresholds seeking medical help for tinnitus and other hearing problems is increasing. For diagnostic purposes, existence/nonexistence of lesions or combinations of lesions in the inner ear not reflected in the audiogram was evaluated with advanced hearing tests applied to tinnitus patients with certain backgrounds, including noise exposure. For forty-six patients with pronounced tinnitus, and other symptoms, tentative diagnoses were established, including judgments of the influence of four causative factors: (1) acoustic trauma, (2) music, (3) suspected hereditary, and (4) nonauditory, for example, stress or muscular tension. They were analyzed with a test battery sensitive to lesions involving the outer hair cells, damage from impulse noise, and dysfunction of the efferent system. There were significant differences in test results between groups with individuals with the same most likely causative factor. Most patients claiming acoustic trauma had a specific type of result, 'hyper-PMTF' (psychoacoustical modulation transfer function), and abnormal test results of the efferent system. Everyone in the hereditary group had dysfunction of the efferent system. All patients working with music, except one, had some abnormality, but without specific pattern. The nonauditory group mostly had normal test results. The investigation shows that it is possible to diagnose minor cochlear lesions as well as dysfunction of the efferent system, which might be causing the tinnitus. Those abnormalities could not be detected with routine audiological tests. Malfunctioning caused by impulse noise is an obvious example of this. These findings facilitate choice of treatment, rehabilitation programs, and medicolegal decisions.

How to cite this article:
Lindblad AC, Hagerman B, Rosenhall U. Noise-induced tinnitus: A comparison between four clinical groups without apparent hearing loss.Noise Health 2011;13:423-431

How to cite this URL:
Lindblad AC, Hagerman B, Rosenhall U. Noise-induced tinnitus: A comparison between four clinical groups without apparent hearing loss. Noise Health [serial online] 2011 [cited 2020 Jul 16 ];13:423-431
Available from:

Full Text


Tinnitus following occupational noise exposure has been correlated to noise-induced hearing loss, with negligible occurrence in noise-exposed workers with normal hearing. [1],[2] However, other occupational situations have been reported in which tinnitus is a promintent symptom even with no or minor hearing impairment. One important example is musical professions [3] and another is military officers and servicemen who have been exposed to acute acoustic trauma. [4],[5]

Tinnitus is most often correlated to hearing impairment, but many tinnitus patients have no or only minor hearing loss. Savastano [6] studied tinnitus patients and reported that 57% had hearing deficits and 43% had normal hearing (pure tone average 0.5-8 kHz, <20 dB). Wiberg et al. [7] reported that young tinnitus patients (<20 years of age) had, on an average, normal hearing thresholds in the frequency range from 500 to 8000 Hz. The hearing thresholds were slightly elevated in a group of tinnitus patients of working age and still more elevated (above all in the high frequencies) in retired persons with tinnitus.

Observations about tinnitus in patients with normal pure tone audiometry raise the question about the role of the peripheral auditory system for triggering tinnitus. A significant problem is that pure tone audiometry and also other audiological tests in current clinical use are poor indicators of minor cochlear dysfunction. The rapid development of auditory physiology during the last decades provides a possibility to study very early changes of the physiological function and of the micromechanics of the cochlea. Such changes occur before pure tone thresholds are noticeably affected. [8],[9],[10]

The aim of the present study was to investigate whether some auditory physiological and psychoacoustical tests are sufficiently sensitive to detect minute cochlear lesions that cannot be diagnosed by routine clinical audiological tests. Four different clinical study groups were selected, all of them comprising individuals with disabling, persistent tinnitus, but with no or only minor pure tone threshold elevations. Two of these groups were selected since they represent aspects of occupational noise-induced tinnitus; and for comparison, two other groups were included.

 Study Groups


All subjects in the study had consulted the services of the Department of Hearing and Balance, Karolinska University Hospital, Stockholm, for persistent tinnitus. One inclusion factor was normal hearing, or at most minor hearing loss, measured with pure tone audiometry; and the pure tone average across the frequencies 500, 1000, and 2000 Hz was ≤20 dB hearing level (HL). Narrow mid-frequency dips (≤25 dB HL) were permitted. A mild, high-frequency sensorineural hearing loss was also accepted (≤40 dB HL at 4000 Hz). The high-frequency threshold average (3000, 4000, 6000 Hz) had to be ≤35 dB HL. Conductive hearing loss, even very mild, was an exclusion criterion. Only patients with normal otomicroscopic findings were accepted. Patients with middle ear problems, or other otological diseases, were excluded. To reduce confounding effects of aging, an upper age limit of 60 years was chosen. [11]

Four clinical study groups, based on etiology, were selected for the study. Participants of two of the groups had been exposed to noise that could have caused the tinnitus: (1) subjects with acute, intense acoustic trauma as the most likely causative factor and (2) professional and semiprofessional musicians, with music as the most likely causative factor. For comparison, two other causative groups were tested: (3) persons with suspected hereditary tinnitus and (4) persons judged to have nonauditory, cross-modal tinnitus. [12]

The intention was to study a series of consecutive patients. However, that was not possible since about half of the proposed participants abstained, many of them because they were afraid to be exposed to sound even of moderate intensity. Forty-six patients fulfilled the inclusion criteria and accepted the invitation to participate in the study. Of these patients, 17 were women and 29 men (mean age: 35 years; SD: 13 years; range: 18-59 years).

Etiological classification

All patients were examined, on a regular visit to the clinic, by a doctor specialized in ENT diseases and audiological medicine (in most cases one of the authors, U.R.). An ENT examination including otomicroscopy was performed, and a thorough, structured medical history was taken. The impact of the four causative factors was judged for each patient on an arbitrary scale: 0 = no influence; 1 = mild influence; 2 = marked influence; 3 = very strong influence. [Table 1] gives detailed information of the use of the scale. Many patients had more than one causative factor with judged influence larger than 0. Each patient was assigned to one of the causative groups mentioned above, and each of the causative groups consisted of those individuals who had their highest judged value for the corresponding causative factor [Table 1].{Table 1}

Acoustic trauma group

The acoustic trauma group included 16 subjects (12 men and 4 women) who had been exposed to sudden, intense noise. All the cases reported that the symptoms started immediately after the exposure, and none had experienced any problems before the incident. In 11 cases, the exposure was an impulse noise. Seven were military conscripts: 3 cases exposed to fine-caliber firearms and 4 cases to heavy-caliber guns. They had met with shooting accidents either because of ill-fitting ear protectors or because of service of weapons without ear protectors. The other four impulse noise exposures were blows from a hammer in a factory, a bus tire exploding at a bus stop, explosions during a chemistry lesson, and from firecrackers. The remaining 5 cases had been exposed to intense, stationary noise at 115-130 dB sound pressure level (SPL), lasting for several minutes, for example, a burglar alarm, activated by mistake, in a military rock shelter. The military impulse noise cases were tested about 3 weeks after the exposure as well as after 2-3 months. The other cases were tested 1-9 years after the exposure. The mean and median ages were 24 and 30 years, respectively.

Music group

The music group included 10 cases (4 women and 6 men), all of them working professionally with music, some of them full time (6 cases) and some part time along with other occupations (4 cases). The group included various professions: drummer, contrabassist, violinist, singer, choir leader, music teacher, and sound engineer. Hyperacusis and tinnitus were about equally common as the main symptoms. Stress and muscular tension were commonly reported in this group. This group was the most homogeneous regarding age, with the youngest member of the group being 28 years old. Both mean and median ages were 43 years.

Hereditary group

The hereditary group consisted of 9 cases (2 women and 7 men) with suspected hereditary origin of the symptoms. The diagnosis was based on a family history of tinnitus and hearing loss (parents, children, siblings, etc.). One case had a grandmother and an aunt who had tinnitus and hearing loss. Six cases had 2 or more relatives and 4 had 1 relative with auditory symptoms. Four of the cases belonged to the same family, a father and his three sons. Minor mid- or high-frequency dips were present in 7 cases. Seven patients had nonauditory symptoms, often mild, the most common being stress and muscular tension. The mean and median ages of the hereditary group were 35 and 37 years, respectively.

Nonauditory group

A nonauditory group of 11 patients (7 women and 4 men) without any family history of hearing problems and no apparent ototraumatic events was included. Hyperacusis and tinnitus were almost equally common as the main symptoms. The symptoms defining nonauditory factors were muscular tension, neck pain, tension headache, odontological problems, intense stress, anxiety, depression, dizziness, chronic pain, and chronic fatigue. Only nonauditory factors that preceded the auditory symptoms were accepted to avoid contamination of secondary stress and muscular tension caused by auditory problems. Some of the patients played music, but none of them was a professional musician. The mean and median ages of this group were 38 and 34 years, respectively. This group, presumably with little or no evidence of cochlear damage, can serve as a clinical reference group. All patients gave their informed consent to partipate in the study, which was approved by the local ethical committee.


The patients were examined with an auditory test battery. All patients were tested with pure tone audiometry and also fixed-frequency Békésy audiometry. For the purpose of detecting minute lesions affecting the cochlea and its regulatory system, two tests were selected: Psychoacoustical modulation transfer function (PMTF) and transient evoked otoacoustic emissions (TEOAEs) with and without contralateral noise. The test time for these tests was 2-3 h.

Psychoacoustical modulation transfer function

The active nonlinear process in the cochlea is mediated by the outer hair cells (OHCs) and facilitates the perception of the complex sound patterns in speech. These patterns are characterized by rapid sound variations combined with slow modulations caused by speech syllables, words, and intonation. A measurement termed the PMTF reflects the functioning of the inner ear when handling slow intensity variations such as those of speech. PMTF measures the thresholds of brief tones placed at the peaks and in the valleys of a fully, sinusoidally intensity-modulated, octave-band noise at various sound pressure levels. [13] The test has been shown to measure more subtle qualities of hearing, and there is evidence that it reflects hair cell function, above all that of the OHCs. [14]

PMTF measurements

The measurements were performed on the ear with the most pronounced symptoms with the brief tone, 4 ms, and the octave-band filtered noise at 4000 Hz, and with a modulation frequency of 10 Hz. The noise levels were 35-85 dB SPL in steps of 10 dB. Initially, the threshold for the brief tone was measured without noise to familiarize the subject with the brief tone.

Typical PMTF results

[Figure 1] shows a few stylized, but characteristic PMTF curves.{Figure 1}

The normal PMTF curves are nonlinear, with a maximum signal-to-noise-ratio (S/N) for the peak threshold, occurring at a noise level of about 55 dB SPL. (The term peak threshold is used for the threshold of the brief tone when the tone is placed at the peak of the noise. The term valley threshold is used when the tone is placed in the valley of the noise.) For the valley threshold, there is a corresponding maximum at about 65 dB SPL [Figure 1]a.

For a sensorineural hearing loss of cochlear origin, the nonlinearity is weaker. Both maxima have lower S/N and they occur at higher noise levels [Figure 1]b. This type of pattern indicates reduced nonlinearity.

A second type of abnormal PMTF pattern is presented in [Figure 1]c. The S/N maxima of the peak and valley threshold curves have markedly increased amplitudes and occur at the same, low noise level (35-45 dB SPL). The peak and valley curves are almost identical, which implicates that the affected ear can hardly take advantage of the silent interval around the brief tone in the valley. The term "hyper-PMTF" was coined for this pattern.

There are also intermediate varieties between normal and hyper-PMTF. A mildly abnormal variety is shown in [Figure 1]d. Like in the hyper-PMTF, the S/N maxima of the peak and valley curves are positioned at the same noise level, but here they occur at a normal noise level (>45 dB SPL), and they are not as high as in the hyper-PMTF. The level dependence for peak and valley curves is the same. At every noise level used, the S/N for the peak threshold is a roughly constant number of decibels higher than the S/N for the valley threshold.

It is evident from the examples above that there are two main types of abnormal PMTF results: The positions of the maximum peak and valley thresholds can occur at higher or lower than normal noise levels (horisontal position in graph), and the corresponding S/N maxima are then either lower or higher than normal (vertical position in graph). Therefore, the only practical way of analysis is to use a qualitative method to define certain typical patterns, for example, the hyper-PMTF. This method will be described later.

Transient evoked otoacoustic emissions with and without contralateral noise

Otoacoustic emissions (OAEs) reflect the function of the OHCs. TEOAE recordings with and without contralateral noise can be used to study the efferent system - the medial olivocochlear (MOC) system. This system is constituted by neurons in the medial olivary complex and efferent neurons passing from the vestibular nerve to the cochlea. [15] The system modulates the cochlear function by acting on the OHCs. It has been suggested that the MOC system might influence the ear's resistance to noise trauma. [16],[17],[18] In the latter study, animals with the weakest MOC control suffered the most damage when exposed to traumatizing sounds. It is also interesting to note that Veuillet et al. [19] showed results indicating that military officers with good suppression measured after acoustic trauma had a better recovery than officers with less or no suppression.

Janssen et al. [20] tested ears affected with sensorineural hearing loss with tinnitus. About half of them had decreased distortion product otoacoustic emission (DPOAE/DP) levels, which is considered normal for sensorineural hearing loss, and the other half had increased DP levels, hypothesized to be generated by cochlear hyperactivity that could cause both the abnormally high DP level and the tinnitus. Similarly, we found both increased and decreased TEOAE responses in a noise susceptibility project with young conscripts (unpublished): Groups exposed to continuous noise had low TEOAE amplitudes. In contrast, a group who had been exposed to strong, but not extremely intense, impulse sounds showed increased TEOAE responses (the correlated part) and intensified chaotic activity (uncorrelated responses to stimuli, mostly regarded as noise). Obviously, this phenomenon may complicate the interpretation of what is a normal TEOAE response.

Transient evoked otoacoustic emission measurements

For the measurements, clicks with the duration of 80 μs were repeated with a frequency of 50 Hz. The standard nonlinear mode was used to enhance those components in the response, which have a nonlinear dependence on the stimulus level, and to suppress the linear components in the response. [21] To accomplish this, the polarity of every fourth click is reversed and the sound pressure of the click is increased by a factor of 3. The acoustical responses from 1000 clicks are averaged, after removal of the primary click by time-windowing technique. Half of the click responses are averaged in one buffer, the other half of the responses are averaged in a second buffer. The stimulus level is specified as so-called peak equivalent sound pressure level, peSPL. Clicks at 75 and 85 dB peSPL were used with and without contralateral broadband noise at 70 dB SPL. The RMS values in 1000-Hz-bands for the TEOAE response (the response that is common to the two buffers, i.e., correlated to the clicks), over the interval of measurement, were used in the analyses. Also, the uncorrelated response (the difference between the contents of the two buffers) was analyzed in 1000-Hz-bands. The bands were centered around 1000, 1500, 2000, 3000, 4000, 5000, 6000, 7000, and 8000 Hz. To obtain the suppression in those bands, the differences between the decibel values of responses without contralateral noise and the decibel values of the responses with contralateral noise were calculated. Thus, a suppression value is positive when the response decreases and it is negative when it increases with contralateral noise.


Two test systems were used. Each of them consisted of a Tucker-Davis Technologies System III with RP2.1 processing unit with 24-bits A/D and D/A. The TDT systems were controlled by personal computers. Circumaural earphones, Sennheiser HDA 200, were used for the psychoacoustical measurements. The probe system used for measuring OAEs was of type ER-10C from Etymotic Research.

Classification of measured functions

The aim of the analysis was to find out how the four causative factors might induce abnormal function of the inner ear and/or the efferent system. With only 46 test subjects in this first analysis, we had to limit the number of variables in the outcome of the measurements. One reasonable way to do that was to use visual judgements to present the results. The first author, A.C.L., working in the research laboratory situated in another part of the city, and blind to the judgements of causative factors, made all interpretations using experience from earlier test subjects and patients of various ages.

Three dependent variables were studied. Like the causative factors, these variables were classified on four-graded scales, from no abnormality at all (0) to a strongly marked abnormality (3).

Outer hair cells function

The quality of the OHC function was judged from TEOAE levels and from the nonlinearity function of PMTF. ([Figure 1]a shows a normal nonlinearity and [Figure 1]b shows a reduced nonlinearity.) The OHC function was classified as follows: 0 = normal, no abnormality; 1 = slight abnormality (mild decrease of the PMTF nonlinearity and/or weakened TEOAE responses); 2 = marked abnormality (strongly reduced PMTF nonlinearity function [Figure 1]b and weakened TEOAE responses); 3 = strongly marked abnormality (very strongly reduced PMTF nonlinearity and affected TEOAE levels).

Psychoacoustical modulation transfer function positioning

Experience suggests that a variable describing the PMTF positioning (i.e., the positioning of the maxima of the peak and valley threshold curves in the PMTF results) could be useful. The hyper-PMTF would then be the most extreme abnormality. The PMTF positioning was classified as follows: 0 = normal [Figure 1]a 1 = slight abnormality (see [Figure 1]d; identical positioning of maxima for peak and valley threshold curves at noise levels higher than 45 dB SPL, or the maximum of the valley threshold curve falls at a lower noise level than that of the peak threshold curve); 2 = marked abnormality (irregularly shaped hyper-PMTF); 3 = strongly marked abnormality ([Figure 1]c; hyper-PMTF with high maxima at 45 dB SPL noise level or lower).

Efferent regulatory function - the medial olivocochlear function

The MOC function is presumed to be expressed by the suppression effect on TEOAEs when contralateral noise is presented. It sometimes happens that a person, of any age, has increasing reponses with contralateral noise in very specific and individual frequency bands. This lack of effectiveness of the efferent system has been measured on some tinnitus sufferers. [22] General statistical analyses over all frequency bands, separately or as a total, would hide such individual abnormalities. Regarding the suppression of TEOAEs with contralateral noise, the classifications were as follows: 0 = normal; 1 = slight abnormality (decreased suppression or no suppression); 2 = marked abnormality (some increase, instead of decrease, in TEOAE responses with contralateral noise, i.e., negative suppression); 3 = strongly marked abnormality (strong negative suppression). For some ears with extremely strong uncorrelated activity in the TEOAE measurements, this deficiency in regulation was taken into account in the classification.

Statistical analyses

The statistical analysis was performed with Statistica, release 7. Analyses of variance (ANOVAs) were performed for the four etiological groups, which were based on the most likely causative factor, versus OHC function, MOC function, and PMTF positioning. A corresponding analysis was done for groups versus high-frequency threshold average (3000, 4000, 6000 Hz). Correlations between age or pure tone thresholds, especially the high-frequency threshold average (3000, 4000, 6000 Hz), and abnormal function of the inner ear or efferent system were also looked at. Multiple analyses of variance (MANOVAs) could be performed for auditory test results versus individual causatives factors only if the causative factors were two-graded, that is, "No" - no or only mild influence, or "Yes" - marked or very strong influence. Tukey's honestly significant difference (HSD) test was used when needed.


Judged causative factors and auditory test results in etiological groups

The prevalence of judged causative factors as well as the outcome of the auditory tests in the four patient groups are presented in [Table 2] and [Figure 2]. Because there were significant correlations between some pure tone thresholds and the judged OHC and MOC functions, the means and standard deviations of the pure tone thresholds at 3000, 4000, and 6000 Hz are included in the table. It is apparent that the most abnormal results of the audiological tests were seen in two patient groups: the acute noise trauma group and the hereditary group. The prevalence of marked or strongly marked abnormalities in test results was lower in the music group, and almost absent in the nonauditory group.{Table 2}{Figure 2}

About half of the test subjects, especially those with the worst hearing thresholds, presented slight OHC abnormalities manifested in PMTF curves with decreased nonlinearity and/or low TEOAE levels, but no patient in this study, considering people with normal or nearly normal hearing thresholds, had the highest degree of OHC abnormality. In contrast, there were 17 subjects with a high degree of abnormality regarding the positioning of the PMTF-curve maxima, like the hyper-PMTF, the extreme example in [Figure 1]c. For most of them, acoustic trauma had been judged to be the most likely causative factor.

Almost all hereditary cases (~90%) had a high degree of MOC abnormality. The same was the case for ~70% of the subjects in the trauma group, 50% in the music group, and ~30% in the nonauditory group.

The PMTF positioning had no significant correlation to hearing thresholds or any of the other measured functions. OHC and MOC functions were significantly correlated to each other (P<0.01) and to the high-frequency threshold average (P<0.01). There was no significant influence of age.

Group differences in auditory test results

ANOVAs were performed for the four clinical groups versus OHC function, PMTF positioning, and MOC function, with threshold average (3000, 4000, 6000 Hz) as a covariate. A following Tukey's HSD test showed a few significant and characteristic differences between the etiological groups (P<0.05). [Table 3] shows an overview of the significant differences between groups.{Table 3}

Regarding the PMTF positioning, the acoustic trauma group presented the most abnormal test results, and this group was significantly different from the other three groups: hereditary (P=0.0023), nonauditory (P=0.0023), and music (P=0.016) groups.

Regarding the MOC function, the hereditary group had the most abnormal test results, which were significantly different from those of the nonauditory group (P=0.0011). Also, the results of the acoustic trauma group were significantly different from those of the nonauditory group (P=0.0032). The results of the music group fell in between.

There was a significant difference (P=0.032) in the OHC function between the acoustic trauma group and the nonauditory group. However, the OHC damage was not very pronounced in any of the groups.

The nonauditory group had a better high-frequency threshold average (3000, 4000, 6000 Hz), which differed significantly from that of the hereditary group (P=0.018), and there was a trend for a difference from the acoustic trauma group too (P=0.092). (This was analyzed with a one-way ANOVA with the said threshold-average as independent variable and group as factor, and with a subsequent Tukey's HSD test.) The age of the participants had no influence on the analyses by group.

Individual causative factors and auditory test results

MANOVAs for auditory test results versus individual causative factors could not be performed for the four-graded scales. However, MANOVAs (plus Tukey's HSD tests) with causative factors only two-graded (i.e., "No" - no or only mild influence, or "Yes" - marked or very strong influence) showed some significant differences. Individuals judged to have been markedly or very strongly influenced by acoustic trauma showed worse test results in all three auditory tests - PMTF positioning (P=0.00019), MOC (P=0.028), and OHCs (P=0.0034) - than those judged not at all or only mildly influenced. There was also a significant difference regarding the causative factor nonauditory with better PMTF positioning (P=0.0044) for subjects judged to have at least marked influence of nonauditory problems than for those without or only mild such problems. (Very few judged to have nonauditory problems were judged to have been influenced by acoustic trauma, and very few judged to have been influenced by acoustic trauma were judged to have nonauditory problems.) Regarding the causative factor hereditary, the subjects supposed to have marked or worse hereditary influence had significantly better PMTF positioning (P=0.017) and worse OHCs (P=0.017) than those judged without or with only mild hereditary influence. The average MOC result was worse for the suspected heriditary cases than for others, but not significantly so. There were no significant differences in auditory test results for the causative factor music.


Tinnitus is a symptom, not a diagnosis. In order to be able to manage tinnitus patients properly, it is important to diagnose separate tinnitus entities. This applies especially to normal-hearing patients with tinnitus, in whom only limited diagnostic information about the state of the cochlea can be extracted from the pure tone audiogram. Nevertheless, cochlear diagnostics are regularly based on demonstration of pure tone threshold elevations. Tests based on auditory physiological principles offer a possibility to sharpen the diagnostic capability, and the hypothesis of this study is that a carefully designed physiological test battery is a means to achieve this goal.

The tinnitus patients of this study were selected to represent four different causative groups, two of them related to noise exposure. They were diagnosed according to a structured medical history, and on other clinical observations not related to any of the physiological principles applied in the test battery. This allows a validation of these physiological tests.

For the patients with tinnitus that appeared after short, intense sound exposure, the acoustic trauma group, it can be assumed that they should have some damage affecting the function of the inner ear. The test results support this assumption. All patients who were judged to have suffered acoustic trauma had at least two abnormalities within the test battery. The most common marked abnormalities affected PMTF positioning plus MOC function. However, two of them had strongly affected MOC function but fairly normal PMTF positioning. The measurements of OHC function were in most cases slightly abnormal. These findings suggest a combination of inner ear damage (in spite of normal or slightly abnormal pure tone audiograms) and disturbed efferent modulation of the cochlea. The results of the acoustic trauma group are in concordance with other reports. [9],[10],[19],[23] In the study by Attias et al. involving military personnel, [9] DPOAEs could be measured on ears with hearing thresholds as high as 75 dB HL, which suggests severe damage to inner hair cells (IHCs). It has been shown that normal or near-normal thresholds in animals do not exclude damage first to IHC cochlear nerve terminals and later to IHCs. [24],[25] A large percentage of IHC afferent synapses could be damaged despite normal thresholds. Moreover, Attias et al. mention that ipsilateral enhancement of OAEs could stem from decreased MOC efferent feedback to the OHCs because of IHC damage and reduced input from the afferent arm. This agrees well with our findings of extreme TEOAE responses (or, through contralateral regulation, increase instead of decrease of TEOAEs when contralateral noise was presented) for some subjects and also with the findings of Attias et al. [8] in tinnitus patients with normal hearing thresholds.

The group of musicians presented ambiguous results regarding the mechanisms causing the symptom. All but one had cochlear dysfunction of some kind, for example the two of them who had fully developed abnormal PMTF positioning (hyper-PMTF). The statistical analysis of this small group does not give much information regarding the possible presence of minor but typical cochlear dysfunction. This is not surprising, since there is a large variety of individual mixes of types of music exposure and other circumstances for people working with music or in rooms where music is played. A larger study in progress takes this into account.

In the nine cases, for whom heredity was judged as the main causative factor, all but one had a high degree of abnormality in efferent function, and the statistical analyses showed that the main characteristic of the hereditary group was abnormal MOC function. The MOC function was mainly judged from the contralateral suppression of TEOAEs. Especially suspected hereditary cases had strongly abnormal suppression (strongly increased response) in individual combinations of frequency bands. Overall values of suppression or prechosen frequency band do not necessarily reveal such abnormalities. One may ponder if a defect MOC, at least in a young person, might signal a hereditary defect, and that such a dysfynction may sensitize the cochlea to noise damage. The latter assumption is strenghtened by the existence of a significant correlation between the results of the MOC and OHC measurements in the study. However, a larger patient material is necessary to confirm these assumptions. To the best of our knowledge, this is the first report on a type of hereditary tinnitus, related to abnormal MOC function, and possibly resulting in decreased resistance to noise damage.

The nonauditory group represents patients with tinnitus with, presumably, unaffected cochlear function. [12],[26],[27] With a couple of exceptions, with test results hinting at acoustic trauma, the outcome of the tests in this group showed little evidence of abnormal function. The statistical analysis shows that factors triggering tinnitus in the nonauditory group should preferably be searched for outside the peripheral auditory system.

Two other possible factors - the age of the participants and the presence of audiometric dips - did not predict the presence of inner ear dysfunction or abnormality affecting the efferent system.

The outcome of the test battery shows that PMTF positioning and MOC function suggests abnormalities in the group where these are most likely to appear - the acoustic trauma group. Moreover, the MOC function appears to be impaired in this group. Especially the PMTF positioning test appears to provide important diagnostic information even after relatively limited noise exposure, causing tinnitus but not pure tone threshold elevations. The test of the OHC function (TEOAEs and PMTF nonlinearity) is less important for this patient category. Still the results deviate significantly between the the nonauditory group and the acoustic trauma group [Figure 2]. The same test pattern appears in an individual analysis of no or mild influence versus marked/very strong influence of causative factors (e.g., exposure to noise). This analysis shows significantly poorer OHC function for individuals with acoustic trauma, but also for individuals with suspected heredity as a risk factor, than for individuals without these risk factors.

In this study the subjects were categorized into groups according to most likely causative factor. However, most patients had more than one judged causative factor. For some patients the results of the hearing tests point toward causative factors that were judged possible, but not judged as the most likely one. This indicates the diagnostic value of the test battery. However, it is plausible that more than one causative factor has caused a patient's tinnitus. Thus, there might be an uncertainty in interpretation of test results and a difficulty to find a one-to-one relationship between causative factor and test outcomes. The music group, at least as a group, might be an example of that.

The present study indicates that it is possible to reach a diagnostic capability of, for example, noise-induced hearing loss far beyond what is possible with pure tone audiometry only. Such a refined capacity offers possibilities to support medicolegal assessments of, for example, noise-induced hearing loss, to select appropriate treatment programs and to improve counseling for the patients.


The results of the present study suggest that the diagnostic capability can be improved by using modern physiological methods. Tinnitus and related symptoms can be caused by a broad variety of unrelated triggering factors. We found distinct cochlear lesions especially in cases where acoustic trauma or heredity are possible causative factors. A new entity of hereditary tinnitus with abnormal MOC function, and possibly increased sensititvity to noise, is described.


The study was supported by grants from Swedish Skandia Life, The Foundation Tysta Skolan, the Federation of Hard of Hearing People (HRF), and the Swedish Council for Working Life and Social Research. We express our gratitude also to Ann Johansson and Helena Norell, who performed the measurements, and to our test subjects. Support from Åke Olofsson was invaluable to the project.


1Rubak T, Kock S, Koefoed-Nielsen B, Lund SP, Bonde JP, Kolstad HA. The risk of tinnitus following occupational noise exposure in workers with hearing loss or normal hearing. Int J Audiol 2008;47:109-14.
2Dias A, Cordeiro R. Association between hearing loss level and degree of discomfort introduced by tinnitus in workers exposed to noise. Braz J Otorhinolaryngol 2008;74:876-83.
3Zhao F, Manchaiah VK, French D, Price SM. Music exposure and hearing disorders: An overview. Int J Audiol 2010;49:54-64.
4Mrena R, Savolainen S, Kuokkanen JT, Ylikoski J. Characteristics of tinnitus induced by acute acoustic trauma: A long-term follow-up. Audiol Neurootol 2002;7:122-30.
5Mrena R, Savolainen S, Kiukaanniemi H, Ylikoski J, Mäkitie AA. The effect of tightened hearing protection regulations on military noise-induced tinnitus. Int J Audiol 2009;48:394-400.
6Savastano M. Tinnitus with or without hearing loss: Are its characteristics different? Eur Arch Otorhinolaryngol 2008;265:1295-300.
7Wiberg A, Jansson G, Johansson M, Hellström P. Tinnitus and treatment in three Swedish populations: Children, employed and retired. Fremantle, Australia: Seventh International Tinnitus Seminar; 2002.
8Attias J, Furst M, Furman V, Reshef I, Horowitz G, Bresloff I. Noise-induced otoacoustic emission loss with or without hearing loss. Ear Hear 1995;16:612-8.
9Attias J, Bresloff I, Reshef I, Horowitz G, Furman V. Evaluating noise induced hearing loss with distortion product otoacoustic emissions. Br J Audiol 1998;32:39-46.
10Granjeiro RC, Kehrle HM, Bezerra RL, Almeida VF, Sampaio AL, Oliveira CA. Transient and distortion product evoked oto-acoustic emissions in normal hearing patients with and without tinnitus. Otolaryngol Head Neck Surg 2008;138:502-6.
11Kim S, Frisina DR, Frisina RD. Effects of age on contralateral suppression of distortion product otoacoustic emissions in human listeners with normal hearing. Audiol Neurootol 2002;7:348-57.
12Cacace AT. Expanding the biological basis of tinnitus: Crossmodal origins and the role of neuroplasticity. Hear Res 2003;175:112-32.
13Lindblad A-C, Hagerman B, Olofsson Å. Tone thresholds in modulated noise. I. Level dependence and relations to SRT in noise for normal-hearing subjects. Stockholm: Karolinska Institutet; 1992.
14Lindblad A-C, Hagerman B. Hearing tests for selection of sonar operators. ACUSTICA - Acta acustica 1999;85:870-6.
15Collet L, Kemp DT, Veuillet E, Duclaux R, Moulin A, Morgon A. Effect of contralateral auditory stimuli on active cochlear micro-mechanical properties in human subjects. Hear Res 1990;43:251-61.
16Zheng XY, Henderson D, Hu BH, Ding DL, McFadden SL. The influence of the cochlear efferent system on chronic acoustic trauma. Hear Res 1997;107:147-59.
17Zheng XY, Henderson D, McFadden SL, Hu BH. The role of the cochlear efferent system in acquired resistance to noise-induced hearing loss. Hear Res 1997;104:191-203.
18Maison SF, Liberman MC. Predicting vulnerability to acoustic injury with a noninvasive assay of olivocochlear reflex strength. J Neurosci 2000;20:4701-7.
19Veuillet E, Martin V, Suc B, Vesson JF, Morgon A, Collet L. Otoacoustic emissions and medial olivocochlear suppression during auditory recovery from acoustic trauma in humans. Acta Otolaryngol 2001;121:278-83.
20Janssen T, Kummer P, Arnold W. Growth behavior of the 2 f1-f2 distortion product otoacoustic emission in tinnitus. J Acoust Soc Am 1998;103:3418-30.
21Kemp DT, Bray P, Alexander L, Brown AM. Acoustic emission cochleography--practical aspects. Scand Audiol Suppl 1986;25:71-95.
22Chéry-Croze S, Collet L, Morgon A. Medial olivo-cochlear system and tinnitus. Acta Otolaryngol 1993;113:285-90.
23Desai A, Reed D, Cheyne A, Richards S, Prasher D. Absence of otoacoustic emissions in subjects with normal audiometric thresholds implies exposure to noise. Noise Health 1999;1:58-65.
24El-Badry MM, McFadden SL. Electrophysiological correlates of progressive sensorineural pathology in carboplatin-treated chinchillas. Brain Res 2007;1134:122-30.
25Kujawa SG, Liberman MC. Adding insult to injury: Cochlear nerve degeneration after "temporary" noise-induced hearing loss. J Neurosci 2009;29:14077-85.
26Shore S, Zhou J, Koehler S. Neural mechanisms underlying somatic tinnitus. Prog Brain Res 2007;166:107-23.
27Dehmel S, Cui YL, Shore SE. Cross-modal interactions of auditory and somatic inputs in the brainstem and midbrain and their imbalance in tinnitus and deafness. Am J Audiol 2008;17: S193-209.