Asymmetric Suprathreshold Speech Recognition and the Telephone Ear Phenomenon

Transcription

1 J Am Acad Audiol 9 : (1998) Asymmetric Suprathreshold Speech Recognition and the Telephone Ear Phenomenon Shlomo Silman*t** James Jerger$ Robert Fanning Carol A. Silverman" ** Michele B. Emmer" Abstract We compared interaural suprathreshold speech recognition scores in 62 adults ith sensorineural hearing loss. Subjects ere tested at to sites, 25 at the Baylor College of Medicine, Houston, TX and 37 at Brooklyn College, Brooklyn, NY Ears ere categorized according to telephone usage. Results shoed, at both sites, a small but significant difference in speech understanding scores beteen the ear habitually used on the telephone and the opposite, nontelephone ear. The average speech recognition score as approximately 5 percent better on the telephone ear. Results are interpreted in relation to the theories of auditory deprivation and auditory acclimatization. Key Words : Auditory deprivation, speech recognition, telephone usage Abbreviations : NTE = ear not used on the telephone, PTA = pure-tone average, SSRS = suprathreshold speech recognition scores, TE = ear used on the telephone ate-onset auditory deprivation has been described in adults ith bilateral sensorineural hearing impairment ho have L been aided in only one ear (Silman et al, 1984 ; Gelfand et al, 1987 ; Silverman and Silman, 1990 ; Hurley, 1993). It is a phenomenon characterized by asymmetry in suprathreshold speech recognition. If only one ear has been aided, gradual deterioration in suprathreshold speech recognition occurs in the unaided ear but not in the aided ear. This deterioration has *Brooklyn College, CONY, Brooklyn, Ne York ; taudiology and Speech Pathology Service, Department of Veterans Affairs Medical Center, East Orange, Ne Jersey ; Department of Otorhinolaryngology and Communicative Sciences, Baylor College of Medicine, Houston, Texas ; "Communicative Sciences Program, Hunter College, CUNY, Ne York, Ne York ; Departments of Otolaryngology Head and Neck Surgery and Communicative Sciences, The Ne York Eye and Ear Infirmary, Ne York, Ne York ; **Ph.D Program, Graduate School and University Center, CUNY, Ne York, Ne York, "Research Foundation, CUNY, Ne York, Ne York Reprint requests : Shlomo Silman, Center for Auditory Research, Brooklyn College, CUNY, 2900 Bedford Avenue, Brooklyn, NY been attributed to interaural asymmetry in auditory stimulation resulting from long-term lack of amplification in the unaided ear. Recent clinical observations have suggested to us the possibility of another source of such asymmetry in suprathresbold speech understanding. We have observed that, in persons ith mild bilateral sensorineural loss ho have never been aided in either ear, suprathreshold speech recognition is often slightly better in the ear habitually used on the telephone. The purpose of the present study as to attempt a more systematic study of this apparent "telephone ear" (TE) effect. We sought to determine hether habitual use of the telephone in a single ear is, indeed, associated ith asymmetry in suprathreshold speech understanding. Subjects METHOD Data ere collected at to sites : the Audiology Service of The Methodist Hospital, Baylor College of Medicine, Houston, TX and Brooklyn College, Brooklyn, NY Data ere collected retrospectively at Baylor and prospectively at 380

2 Telephone Ear Effect/Silman et al Brooklyn College. There ere 25 subjects at the Baylor site and 37 subjects at the Brooklyn site. Specific inclusion criteria ere as follos : (a) reported onset of hearing impairment in adulthood consistent ith presbyacusis or noiseinduced hearing impairment ; (b) all bone-conduction thresholds ithin 10 db of their respective air-conduction thresholds ; (c) interaural difference in pure-tone average (PTA) for 500, 1000, and 2000 Hz not exceeding 5 db bilaterally ; (e) tympanometric peak pressure ithin -100 dapa for a 226-Hz probe tone bilaterally (Jerger, 1970) ; (f) contralateral acoustic-reflex thresholds for the 226-Hz probe tone consistent ith cochlear pathology bilaterally ; and (g) negative history of amplification. Of the total of 62 subjects, 34 ere men and 28 ere omen. The sex ratio as approximately equivalent at the to sites. At each site, each subject as queried as to the ear habitually used on the telephone. Potential subjects ere asked hich ear as habitually used for telephone communication. Individuals ere excluded as subjects if the telephone as alternated beteen ears. Eighteen subjects identified the right ear as the TE. The other 44 identified the left ear as the TE. Throughout the remainder of this paper, the opposite ear is designated as the nontelephone ear (NTE). Subjects ranged in age from 23 to 87 years. The average age as 66.7 years. Procedure Within the context of routine clinical audiologic evaluation, suprathreshold speech recognition scores (SSRS) ere obtained ith recorded ord lists, either the 50-ord CID W-22s presented at a sensation level of 40 db relative to the spondee recognition threshold (Brooklyn) or half-list versions of the PAL PB-50 ord lists, presented at to or three suprathreshold levels (60, 80, and 100 db SPL), to define the maximum of the performance versus intensity function (Baylor). The possibility that ear differences in SSRS could be explained by differences in effective presentation levels of the speech stimuli is countered by the fact that sensation level as held constant at the Brooklyn site, and that maximum SSRS scores ere defined at levels as high as 100 db SPL at the Houston site. At both sites, conventional pure-tone, speech, and acoustic-immittance assessment as carried out in a sound-treated audiometric test booth meeting standard criteria for background noise in audiometric rooms (ANSI, 1977). Pure-tone and speech tests ere administered through a dual-channel diagnostic audiometer. Speech audiometric materials ere delivered by means of a dual-channel magnetic tape recorder/playback system using commercially available recordings. Conventional acoustic-immittance measures ere obtained using an Amplaid, model 720 instrument (Baylor) or GSI-33 (version 2) acoustic-admittance meter (Brooklyn). All instrumentation as routinely calibrated according to ANSI standards (ANSI, 1987, 1989). Statistical significance as evaluated at an alpha error level of RESULTS F or each subject, e computed the difference beteen the SSRS of the TE and the SSRS of the opposite or NTE. Figure 1 plots this TE-NTE difference as functions of three variables : (1) subject gender (male vs female), (2) telephone ear (right ear vs left ear), and (3) site (Baylor vs Brooklyn). Across all three variables, the TE-NTE difference varied beteen 4 and 6 percent. To evaluate the possibility that the TE-NTE difference as contaminated by one or more of these three factors, e carried out a simple analysis of variance ith the difference, TE-NTE, as the dependent variable and test site, gender, and TE as the independent variables. There ere no significant interactions among the independent variables and no significant main effects. We felt justified, therefore, in collapsing the difference data across site, gender, and TE. In this total group (n = 62), the average SSRS on TE and NTE ears ere 74.1 percent and 69.0 percent, respectively. Thus, the TE-NTE difference as 5.1 percent. A paired t-test revealed Female Male Use left ear on phona Uae right ear on phone Baylor Brooklyn a DIFFERENCE IN PERCENT (TE-NTE) Figure 1 The TE effect. Average difference beteen SSRS of the TE and the opposite ear (NTE). Differences are shon for males vs females, right vs left TEs, and to different sites of data collection

3 Journal of the American Academy of Audiology/Volume 9, Number 5, October 1998 z U aw?z z~ Wv 0 <4o ao-as >a9 PTA Category in d8 HL Figure 2 Effect of degree of hearing sensitivity loss on the TE effect. The TE-NTE difference is plotted for three subgroups, categorized according to PTA. that this difference as statistically significant (t = 2.94, df = 61, p =.0046). Average PTAs for the TE and NTE ears of the total group ere 41.7 db and 41.8 db, respectively. Thus, the SSRS difference beteen TE and NTE cannot be explained by greater hearing sensitivity loss in the NTE. Visual inspection of the data suggested that there as a slight tendency for the TE-NTE difference to decrease as degree of hearing sensitivity loss increased. To evaluate this trend, e divided the total group into three subgroups according to PTA on the TE. In the subgroup of 26 subjects ith PTA less than 40 db HL, the TE-NTE difference in SSRS as 7.5 percent. In the 22 subjects ith PTA beteen 40 and 49 db HL, the difference as 3.4 percent. Finally, in the subgroup of 14 subjects ith PTA greater than 49 db HL, the difference as only 3.1 percent. This trend is graphically illustrated in Figure 2. In the first group (PTA less than 40 db HL), the TE-NTE difference as significant (t = 2.093, df = 25, p =.047), but in the second and third groups, the difference as not significant. The folloing case report illustrates ho the magnitude of the TE-NTE difference may decrease ith increasing PTA. The subject, RL, as evaluated audiologically at ages 55 and 61 years. Her initial presenting complaint as gradual, bilateral decline in hearing sensitivity over a 1- to 2-year period. RL had negative otologic, neurologic, and general medical histories and no previous history of amplification. The audiologic and acoustic-immittance findings at each evaluation revealed a bilateral, symmetric, sensorineural hearing impairment. RL habitually used the telephone in her right ear. At the initial evaluation, PTAs ere 39 db HL bilaterally. The SSRS ere 84 percent and 72 percent for the right and left ears, respectively, yielding a TE-NTE difference of 12 percent. Binaural amplification as recommended but RL declined the use of any amplification. Six years later, at age 61, both ears had PTAs of 70 db HL and both SSRS had declined to 62 percent, yielding a TE-NTE difference of 0 percent. The TE effect as present hen RL had a borderline moderate degree of hearing impairment. It then apparently dissipated hen the magnitude of hearing impairment became borderline severe and attenuated telephone usage. At the re-evaluation, RL stated that approximately 2 years earlier, she had become an infrequent telephone user (although telephone use as still habitually restricted to the right ear) because of her "embarrassment" over her difficulty ith telephone communication. DISCUSSION T hese findings suggest the folloing tentative hypothesis : in unaided adults ith mild bilateral, symmetric sensorineural hearing impairment, long-term habitual use of the telephone in a single ear results in much greater auditory stimulation of the TE than of the NTE. Such restriction of telephone use to one ear under these conditions appears to result in asymmetric SSRS, similar to the asymmetric SSRS observed in auditory deprivation from long-term monaural hearing-aid use. That is, over time, the SSRS for the NTE declines more than the SSRS for the TE, resulting in the TE effect. On the other hand, in unaided adults ith moderate or severe, bilateral, symmetric, sensorineural hearing impairment, the TE phenomenon appears to be attenuated. Perhaps the moderate or greater degree of hearing impairment adversely affects telephone communication to a degree that is sufficient to result in infrequent telephone usage. Accordingly, habitual restriction of telephone communication to a single ear does not lead to substantially greater auditory stimulation in the TE than in the NTE in the monotic condition over time. Thus, asymmetric SSRS are not observed. Some studies have demonstrated acclimatization (Gatehouse, 1992, 1993) or increase in hearing aid benefit (Cox and Alexander, 1992) in monaurally fitted adults ith bilateral, symmetric, sensorineural hearing impairment, 382

4 Telephone Ear Effect/Silman et al as evidenced by slightly but significantly improved SSRS in the aided but not unaided condition during the 3- to 4-month postfitting period. It is relevant to ask hether these effects are based on the same phenomenon underlying the telephone ear effect. Gatehouse (1993) and Cox et al (1996) have noted that their acclimatization and hearing aid effects emerged only after sufficient high-frequency amplification had been provided. But telephones are manufactured ith limited frequency responses that sharply filter out high frequencies above approximately 3300 Hz hen the telephone exchange is involved (personal communication, Harry Levitt). Furthermore, the subjects in the present study typically had donardly sloping audiometric configurations. Therefore, restricted high-frequency responses for telephones as ell as unaided subjects ith sloping audiometric configurations probably resulted in lack of substantial high-frequency stimulation. Thus, the present observations on telephone use seem inconsistent ith the apparent need for high-frequency stimulation posited by Gatehouse and Cox et al. It may be, hoever, (Gatehouse, personal communication) that improvement in speech identification ability occurs in a manner that is closely coupled to the frequency regions for hich audibility has been materially improved by the provision of either amplification or telephone stimulation. Gatehouse (1993) dras a parallel, in this connection, ith the ork on neural plasticity of auditory cortex reported by Rencanzone et al (1993). Within this conceptual frameork, the present results are not inconsistent ith the previous observations of improved performance only after highfrequency stimulation. It is of further interest to note the results of a study of dichotic listening in 144 Greek telephone operators by Kapranis and Tzavaras (1993). These investigators found that telephone operators ho coupled their earpiece to the left ear had a significantly smaller right-ear advantage, on a dichotic digits test, than operators ho coupled the earpiece to the right ear. Kapranis and Tzavaras concluded that the extraordinary verbal stimulation of the left ear over many years had affected functional cerebral organization in such a ay that the right-ear-to-lefthemisphere pathay no longer enjoyed its normal advantage over the left-ear-to-righthemisphere pathay in the dichotic paradigm. Whether this effect resulted from acclimatization of the left ear or deprivation of the right ear is open to question, but the effect of prolonged, high-intensity verbal stimulation of one ear over many years as striking. An alternative explanation of the TE effect is that asymmetry of speech understanding ability is the cause, rather than the effect, of the choice of ear to be used on the telephone. In this scenario, inherent asymmetry in speech understanding dictates hich ear the individual ill choose for habitual telephone usage. The TE effect derives from this inherent asymmetry. The individual chooses, as the ear to be used habitually on the telephone, the ear ith intrinsically better speech recognition ability. Arguing against this possibility, hoever, is the undoubted fact that, in right-handed individuals, keeping the right hand free for riting is a strong determinant of hich hand is used to hold the telephone receiver. In the present sample, for example, 44 of the 62 subjects (71%) held the receiver in the left hand. Another counter argument is illustrated in Figure 2. As degree of loss increases, the effect tends to diminish. If the choice of ear to be used on the telephone resulted from intrinsic asymmetry in speech understanding ability, such asymmetry ought to remain constant as loss increases. Further research on a larger sample is needed to substantiate the TE phenomenon. Research also is needed to investigate the possible relations beteen monaural versus binaural amplification and frequent versus infrequent habitual telephone use that is restricted to one ear. Acknoledgments. This project as partially supported by VA grant #C0845R. The authors ould like to thank the folloing for their contribution for data collection and comments : Michael Bergen, Olga Lis, and Pamela Marx. REFERENCES American National Standards Institute. (1977). Criteria for Permissible Maximum Ambient Noise During Audiometric Testing. (ANSI S ). Ne York : ANSI. American National Standards Institute. (1987). Specifications for Instruments to Measure Aural Acoustic Impedance and Admittance. (ANSI ). Ne York: ANSI. American National Standards Institute. (1989). Specifications for Audiometers. (ANSI S ). Ne York : ANSI. Cox RM, Alexander GC. (1992). Maturation of hearing aid benefit : objective and subjective measurements. Ear Hear 15 :

5 Journal of the American Academy of Audiology/Volume 9, Number 5, October 1998 Cox RM, Alexander GC, Taylor IM, Gray GA. (1996). Benefit acclimatization in elderly hearing aid users. J Am Acad Audiol 7: Gatehouse S. (1992). The time course and magnitude of perceptual acclimatization to frequency responses : evidence from monaural fitting of hearing aids. JAcoust Soc Am 92: Gatehouse S. (1993). Role of perceptual acclimatization in the selection of frequency responses for hearing aids. J Am Acad Audiol 4: Gelfand SA, Silman S, Ross L. (1987). Long-term effects of monaural, binaural and no amplification in subjects ith bilateral hearing loss. Scand Audiol 16 : Hurley RM. (1993). Monaural hearing-aid effect : case presentations. J Am Acad Audiol 4: Jerger J. (1970). Clinical experience ith impedance audiometry. Arch Otolaryngol 92 : Kapranis G, TzavarasA. (1993). Neuropsychological consequences of the exposure of a group of telephonists to unusual auditory stimuli. Percept Mot Skills 76 : Recanzone GH, Schreiner CE, Merzenich MM. (1993). Plasticity in the frequency representation of primary auditory cortex folloing discrimination training in adult ol monkeys. J Neurosci 13 : Silman S, Gelfand SA, Silverman CA. (1984). Effects of monaural versus binaural hearing aids. JAcoust Soc Am 76: Silverman CA, Silman S. (1990). Apparent auditory deprivation from monaural amplification and recovery ith binaural amplification : to case studies. J Am Acad Audiol 4: