Recently,Recently, I’veI’ve been trying to orga- evaluations, and they are truly im- nize some of the columns and articles pressive, the information and insights RI’veI’ve written overover the past ten years.years. provided by the simple audiogram As I was looking through them, it be- can still provide the most pertinent came apparent that I’ve neglected to information to explain the behavioral discuss what is perhaps the implications of a hearing loss. most important hearing di- Perhaps the most important in- mension of all, the simple sight of all is an appreciation of how The audiogram. specifi c audiograms impact upon the In reality, however, perception of certain speech sounds. the “simple” audiogram, Without including speech in the Audiogram: and particularly its im- equation, it is simply not possible to Audiogram: plications, is not quite so intelligibly discuss the audiogram. simple. Even though just This, after all, is the signal we are about everybody who re- most interested in hearing (not to Explanation ceives a hearing aid has his minimize the specifi c needs of certain or her hearing tested with groups of people for other types of a pure-tone audiometer, sounds, such as musicians). and not everybody receives a comprehensive explanation Figure One Audiogram — of exactly what the results The “Speech Banana” Signifi cance mean and what the impli- cations are for them. The audiogram of a fairly typical And even for those who audiogram can be seen in Figure 1. do, at a time when prospec- (My thanks to Brad Ingrao for creat- By Mark Ross tive hearing aid purchasers ing these fi gures for me.) Let’s fi rst go are being inundated with through the fundamentals. Frequency new information, anxious (or pitch) is depicted on the horizon- about the test results, and worried tal axis, from low frequencies on the about the cost involved, much of this left (250 Hz) to high frequencies on The information and explanation will be forgotten or mis- the right (8,000 Hz). For some spe- construed by the time several weeks cifi c purposes, it may be useful to test insights provided by the have passed. I know that I’m often be- 125 Hz as well as some frequencies simple audiogram can still mused when I hear what some people higher than 8,000 Hz (for example, a recount to me what they think their progressive hearing loss above 8,000 provide the most pertinent audiologist told them about their Hz can occur with certain type of oto- information to explain the audiogram and its implications. I’m toxic medications). sure this was not what they were told, The amount of hearing loss is shown behavioral implications of but it is what they remember. And in on the vertical axis with the higher a hearing loss. Here is a this context, that is the only relevant numbers indicating a greater degree consideration. of hearing loss. A symbol on this axis clear explanation of how to For audiologists, the audiogram (red circle for the right ear and blue understand your audiogram. is such a basic dimension and so cross for the left) is a measure of the self-evident that its explanation, person’s hearing threshold at this fre- after the thousandth time, is often quency; i.e., the loudness (intensity) unconsciously truncated and simpli- point where sound is just audible. fi ed. They may go through it by rote Thresholds from zero to 15 dB are memory, unaware that their client considered to be within the normal simply isn’t getting it. For people who hearing range. After that point, people may not even have heard of a pure- will usually begin to display some tone audiometer before they stepped communication diffi culties because of into the audiologist’s offi ce, this is a the elevated hearing thresholds. The quite understandable. To expect them higher the number, the greater the im- to fully understand how the pattern of pact of the hearing loss (referring only their hearing loss impacts upon their to the unaided condition). listening behavior is not very realistic. The 100 dB point should not be Yet, in spite of the recent advances confused with a 100 percent hearing made in diagnostic audiological loss, that is a total lack of hearing.
May/June 2004 29 Hearing sensations do continue past ceding and following consonants, are this point, with some audiometers actually secondary cues that can be extending this vertical range to 120 dB. helpful for understanding speech. In short, the audiogram is a chart of Indeed, the ability to take advan- a person’s hearing loss, frequency by tage of these cues may help explain frequency. why some people seem to understand Now note the shaded area extending speech much better than one would across the audiogram (the so-called predict on the basis of their audio- “speech banana”). This is a general, but gram. Still, there are certain anchor static, representation of the acoustical locations for all speech sounds, as speech energy across frequency. Some displayed on the speech banana, and speech sounds, such as vowels, are these can offer valuable insights in predominantly composed of low fre- understanding the effects of differing quency energy, with less of their power hearing losses. Figure One Audiogram: in the higher frequencies. Other speech Looking at the audiogram again, The “Speech Banana” sounds, in particular the voiceless con- and relating it to the speech banana, A fairly typical audiometric pattern sonants, are the reverse. They have most it is apparent that a person with this for older people. of their energy in the high frequencies hearing loss will hear more of the and less in the lower frequencies. lower frequency speech sounds than Only the shaded area of the speech the higher ones. Indeed, some of the banana above the threshold curve is au- higher frequency sounds, such as the dible; portions below will not be per- /s/ sound (the most frequently used ceived. In spite of portrayals to the con- phoneme in the English language) trary on some charts, no speech sound will barely be heard at all. Will this can be pinpointed to a specifi c point person be able to understand normal on the audiogram; all of them spread speech without a hearing aid? Yes, but their energy somewhat along some seg- only with some diffi culty and then ment of the frequency spectrum. only if the talker is close by and raises In addition to their energy locations his or her voice slightly. on the frequency axis, speech sounds There are, unfortunately, many also vary in intensity. Some are natu- people with audiograms similar to this rally weaker (like the voiceless /th/ who do not, for one reason or another, sound, the weakest sound in the Eng- wear hearing aids. One often sees this Figure 2 Audiogram: lish language) and some are naturally type of audiogram in older people The “Ski Slope” more intense (like the vowel /aw/, the whose hearing loss just “crept up” on A severe-to-profound high-frequency strongest). When these are spoken them and who are still not fully aware hearing loss (“ski-slope” curve). in a normal fashion, as in the word of the diffi culty they are causing them- “thaw” there is a 30 dB intensity range selves, their family and friends. between the two phonemes. The in- Conversing with such individuals tensity of all other speech sounds in is a strain on everyone’s part. Often, the English language fall within these they’ll complain that they can “hear” bounds (note, however, that this does the talker but cannot “understand” not take into consideration infl ected what is being said. This is, indeed, a speech, which will move the entire very frequent complaint. Looking at speech banana up and down). the audiogram, we can understand As already noted, the speech ba- somewhat why this should occur: nana is a static representation of they are not hearing the full range of speech sounds. In reality, however, high frequencies where many of the conversational speech is a dynamic consonants have their predominant and time-varying event, with one energy. And we know from many speech sound (phoneme) following years of research, that consonants another in a rapid succession. contribute more to the understand- Figure 3 Audiogram: The specifi c acoustical composition ing of speech than vowels. Without Behavioral Implications of speech sounds is partially shaped a hearing aid, this person would of Hearing Loss by the preceding and following be hearing primarily low frequency Two very different ears sounds in an utterance. These specifi c vowels energy and some part of the with the same pure-tone average. effects, such as the unique modifi ca- voiced consonants, but little or no tions in vowels made by different pre- portions of the crucial voiceless con-
30 Hearing Loss sonant sounds (such as /s/, /t/, and Figure 2 Audiogram — aid (and even with one) this person /k/) . The “Ski Slope” needs to depend upon visual cues I don’t want to give the impression (speechreading) to supplement the that the perception of vowels is unim- This audiogram in Figure 2 is often information received through hearing. portant — on the contrary, it is and described as a “ski-slope” hearing Some people, because of their superior can be very important — only that by loss. Somebody with this type of hear- ability to utilize some of the second- not actually hearing some of the con- ing loss actually hears better at 250 ary cues in a speech signal referred sonants, people with hearing loss have Hz than the one whose audiogram is to above, or their superior linguistic to struggle to fi ll in the acoustical gaps shown in Figure 1, but much worse at ability, may do better than these com- they produce in the stream of speech. 1,000 Hz and higher. Looking at the ments would suggest. But there are Those individuals with superior lin- speech banana, it is apparent that while limits to even the keenest brain and guistic skills would perform this task a great deal of low frequency vowel the most developed auditory integra- better than those with lesser skills. energy will be perceived, practically no tion ability; at some point, people do Often, people with this type of loss high-frequency consonants would be. need to detect speech energy in order are not aware of the sounds that they An individual with this audiogram to understand a spoken message. And don’t hear. They defi ne what they do would have even more diffi culty in this person is missing just too much of hear as “normal” — as it is, for them. understanding speech than the one the speech signal to expect easy com- In truth, they may be hanging on whose audiogram is shown in Figure munication in any oral exchange. to speech comprehension with their 1. These people depend upon what- It is usual to describe the extent of a fi nger tips; any further distortion in ever low frequency energy they do hearing loss with verbal labels. Thus, the speech signal, such as someone receive for comprehending speech. someone may be described as having talking rapidly or in a foreign ac- The presence of noise and reverbera- mild, moderate, severe or profound cent, or in the presence of noise and tion (not exactly an uncommon oc- hearing loss, or some combination of reverberation, and they lose it. And currence in our society!) would have a terms (“mild to moderate” or “severe this audiogram depicts only a gradual disproportionate effect on them, since to profound”). This is an easy and high-frequency hearing loss; were this it would mask the only speech energy shorthand way of labeling the severity person’s audiogram something like they are able to receive. of a hearing loss. While there is merit the one shown in Figure 2, problems Complaints of “hearing” but not and often necessity to this practice, with speech perception would be “understanding” would occur even there is also the danger of oversimpli- even more severe. more frequently. Without a hearing fi cation. Or perhaps someone’s hear-
May/June 2004 31 ing loss is described by employing a hearing loss, it is doubtful that any ed on an audiogram and that is the single fi gure (derived usually from two would exhibit exactly the same discomfort threshold. (But whether averaging the hearing losses at 500 Hz, audiogram for both their ears. There included or not, it is a dimension 1,000 Hz, and 2000 Hz). The numbers would be differences in degree of that every hearing aid dispenser has themselves then serve as the basis for hearing loss at the different frequen- to take into consideration when fi t- the verbal label that is applied. For cies as well as divergences between ting a hearing aid.) This is the loudest example people with an average 40 the ears. People with bilaterally sym- sound that an individual is willing to dB hearing loss are considered to have metrical hearing losses (that is, with tolerate. When measured with tones just a moderate hearing loss. But such similar hearing losses in both ears) across frequencies and compared a shorthand label does not refl ect the may have quite different auditory ex- to the hearing thresholds (i.e., the pattern of a hearing loss, a pattern, as periences than people with bilaterally audiogram) at these same frequen- we have seen, that may lead to more in- asymmetrical hearing losses. cies, the difference between them is sights into the behavioral implications While most people exhibit a gradual- the usable “listening area.” (It is also of a hearing loss. For example, consider ly sloping hearing loss across frequency referred to as the “dynamic range”.) the audiograms shown in Figure 3. (such as in Figure 1), some people It is the auditory area — the “target” have audiograms that are very atypical. — within which a hearing aid is ex- Figure 3 — Behavioral This would include people with rising pected to provide amplifi ed sound. Inplications of Hearing Loss thresholds with frequency as well as Sounds delivered by the hearing aid those with perfectly normal hearing below the threshold of hearing would The average hearing loss is the same at the very low and high frequencies, not be audible; those exceeding the for both ears in the audiogram shown but with moderately or severe hearing threshold of discomfort would not be in Figure 3. In the left ear, the hearing losses in the mid frequencies. The vari- tolerated (the hearing aid user would thresholds at 500 Hz, 1,000 Hz, and eties are almost (but not quite) endless either turn the volume control down 2,000 Hz are 0 dB, 40 dB, and 80 dB and many of these differences have be- or simply remove the hearing aid). respectively. Their average is 40 dB. In havioral implications. One has to look The dilemma of the threshold of the right ear, the hearing loss at each at the curve and not just some average discomfort for people with hearing of the three frequencies is the same, fi gure or verbal label. loss is that it is often the same as also resulting in an average of 40 dB. It is also true that some people with that observed with normally hear- But it should be apparent, just from very similar audiograms “hear” quite ing individuals (perhaps 90 dB or so a visual inspection, that these two ears differently. While an audiogram can across frequency). But because their would perceive speech quite different- help explain much of the auditory be- threshold of hearing is elevated, but ly. Quite clearly this example demon- havior of a person, it does not explain not their threshold of discomfort, this strates that the shorthand description it all. We know that there are other di- means that people with hearing loss of a hearing loss with a number or a mensions to auditory experiences that generally have a reduced listening verbal label can be quite deceptive. cannot be explained by the audiogram. area (or narrow dynamic range). What I have often found useful in Still, the audiogram is a good place to For example, if someone’s thresh- explaining the need to understand the begin when trying to understand some- olds at 1,000 Hz were 50 dB and audiogram is commenting that it is one’s speech perception problems. their threshold of discomfort were possible for someone to be completely Hearing losses greater than about 70 90 dB, their dynamic range would deaf and have completely normal dB may produce qualitatively different be 40 dB at this frequency. This is an hearing in the same ear at the same effects than losses less than this. adequate listening space within which time. Looking at this audiogram, we The reason for this phenomenon is to deliver amplifi ed speech sounds. can state that this person has perfectly that more severe hearing losses usually If, however, the loss were 75 dB and normal hearing in the left ear, but this involve different structures within the the tolerance level at 95 dB, then the is only true at 250 Hz and 500 Hz. cochlea (i.e., more inner hair cell dam- resulting dynamic range would only Or one could assert that he is com- age) and that these different structures be 20 dB. In this instance, it would be pletely deaf in the left ear, as again in- will impact upon the cochlea’s ability more of a challenge to package sound deed he is (but only at 4,000 Hz and to separate incoming speech sounds within this more restricted dynamic higher). Both statements are correct into their various frequency compo- range (this is where advanced tech- and both describe the left ear of this nents. A hearing loss may also involve nology can be very helpful, as in hear- person. All it would take to resolve central auditory pathways and these, ing aids with fast acting automatic this verbal conundrum would be an too, may affect speech perception gain control circuits). understanding of what the audiogram quite severely. Nevertheless, in spite Finally, it is important for people actually signifi es (degree of hearing of all these caveats, the audiogram is to be familiar with the details of their loss across frequency). still the most fundamental auditory audiogram so that they can track any These three examples do not illus- dimension of all. changes with time. Hearing loss, par- trate the full variety of possible audio- There is another dimension of hear- ticularly adult onset hearing losses, grams. In a room full of people with ing loss that is sometimes also includ- may get gradually worse. (Anybody
32 Hearing Loss experiencing rapid changes in their auditory thresholds should check with an otolaryngologist as soon as pos- sible.) After a while, these changes will likely have behavioral implications that may require reprogramming one’s hearing aid (if only a few frequencies are involved) or changing to another hearing aid if the deterioration extends across frequency. It is a good idea for people to keep copies of all the audiometric tests ad- ministered to them so that compari- sons can be made (and to be sure that they are dated correctly). In brief, the audiogram is perhaps the most important indictor of one’s hearing function, the particulars of which everybody who has a hearing loss should be aware of.
Mark Ross, Ph.D., is an audiologist and associate at the Rehabilitation En- gineering Research Cen- ter (RERC) at Gallaudet University in Washing- ton, D.C. He dates his emergence into the fi eld of audiology to the time he attended the Army Aural Rehabili- tation program as a patient in 1952. He received his doctoral degree from Stanford University and taught at the University of Connecticut and worked as a clinical audiologist at the Newington Children’s Hospital. Dr. Ross is the former director of research and training at the League for the Hard of Hearing and has served on the boards of SHHH and the International Federation of Hard of Hearing People. He has written a regular column on develop- ments in research and technology in the Journal since 1990. He and his wife, Helen, live in Storrs, Connecticut.
To fi nd more Dr. Ross articles on technol- ogy for consumers, go to: www.hearingresearch.org/ross.htm
“This article is supported in part by GRANT #H133E030006 from the U.S.
Department of Education, NIDRR, to Gallaudet University. The opin- ions expressed herein are those of the author and do not necessarily refl ect those of the Department of Education.”
May/June 2004 33