1109

CORRELATION BETWEEN THE HEARING LOSS

CLASSIFICATIONS AND SPEECH RECOGNITION

Correlação entre as classificações de perdas auditivas

e o reconhecimento de fala
Willian Toledo dos Anjos(1), LudimilaLabanca(2), Luciana Macedo de Resende(3),
Letícia Pimenta Costa-Guarisco(4)

ABSTRACT

Purpose: to check the correlation between Speech Reception Threshold and Index of Speech
Recognition with mean audiometric results. Methods: we selected 241elderly patients who underwent
examinations of the pure tone audiometry and speech audiometry. As inclusion, audiometry should
have a sensorineural hearing loss. The tone thresholds for air obtained were classified according
with the following averages: Average 1 – Average of frequencies of 500, 1000 and 2000 Hz; Average
2 – Average of frequencies of 500, 1000, 2000 and 4000 Hz; Average 3 – average of frequencies of
500, 1000, 2000 and 3000 Hz; and 4 average – average of frequencies of 500, 1000, 2000, 3000
and 4000 Hz. The data were compared with Speech Reception Threshold and Index of Speech
Recognition, and treated statistically. Results: Average 1 showed higher correlation with the Speech
Reception Threshold (rho = 0.934, CI = 0.901 to 0.958; eqm = 52.2). In relation to the Index of Speech
Recognition, it was observed that the average 3 showed the highest degree of correlation with the test
(rho = – 0.768, CI = –0.807 to –0.721; eqm = 245) followed averages 2 and 4. Conclusion: for elderly
people with ski slop sensorineural hearing loss, the Speech Reception Threshold has the strongest
correlation with the average frequencies 500 Hz, 1000 Hz and 2000 Hz, while the Index of Speech
Recognition has the highest correlation with the average which include the frequencies 3000 Hz and
4000 Hz.

KEYWORDS: Hearing Loss; Presbycusis; Audiometry; Aged; Speech Intelligibility 

„„ INTRODUCTION Those afflicted by hearing impairments tend to avoid

contact, develop low self-esteem, have few friends
and limited participation in social relations3.
Speech discrimination is essential for human
communication and, in order for it to be effective, Ski slope sensorineural hearing loss is the most
unimpaired hearing is critical. Hearing losses common type to be found in clinical audiology
practice; it includes presbycusis and is frequently
negatively impact the quality of life by restricting social
related to difficulties in speech intelligibility4-6. The
interactions, especially in noisy environments1,2.
frequencies 500 Hz, 1000 Hz and 2000 Hz are
considered to be the most relevant for speech.
(1)
Federal University of Minas Gerais – UFMG, Belo Hori-
zonte, MG, Brazil. Vowels and consonants, the building blocks of
(2)
Federal University of Minas Gerais – UFMG, Belo Hori- speech, have different spectral characteristics.
zonte, MG, Brazil. Vowels are naturally more intense and carry
(3)
Speech-language pathology and Audiology, Federal Uni- acoustic energy at low frequencies (400–500 Hz),
versity of Minas Gerais – UFMG, Belo Horizonte, MG, which are favored by the range of human audibility7.
Brazil.
Consonants, in turn, are sounds having spectral
(4)
Speech-language pathology and Audiology, Federal Uni-
versity of Minas Gerais – UFMG, Belo Horizonte, MG, energy at high frequencies, above 2000 Hz, albeit
Brazil. 20–35 dB weaker than vowels8. However, speech
Conflict of interest: non-existent intelligibility depends on consonant sounds, which

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

1110  Anjos WT, Labanca L, Resende LM, Costa-Guarisco LP

contribute 60 % of it, while vowels contribute only correlation with the speech recognition threshold
40%9. Due to the spectral characteristics of these (SRT) and with the speech discrimination score
sounds and the range of human audibility, one can (SDS).
understand why individuals with hearing loss at high
frequencies have impaired speech recognition. „„ METHODS
In audiology, speech recognition ability is
measured by the tests Speech Reception Threshold The present study was approved by the Research
(SRT) and Speech Discrimination Score (SDS). Ethics Committee of the Federal University of
The SRT corresponds to the softest sound intensity Minas Gerais (UFMG) under protocol no. 155 09.
level at which an individual can recognize 50% An informed consent form was prepared explaining
of the common words given. Typically, SRT and the theme of the study, its aims and the importance
SDS values are consistent with the average of the of analyzing the audiometric tests of the individuals
hearing thresholds obtained for the speech-related selected for the study.
frequencies. The SDS evaluates speech discrimi- This comparative study based on the results of
nation using a list of monosyllables and bisyllables audiometric tests was conducted with an elderly
40 dB above the SRT thresholds10. population who received care at the Instituto Jenny
The classification of hearing losses is a widely de Andrade Faria, an annex to the UFMG Hospital
discussed topic in speech-language pathology das Clínicas.
and audiology. In Brazil, hearing losses have The study employed a convenience sample of
been predominantly classified on the basis of the all the elderly individuals with a complaint of hearing
frequencies of speech relative to the tritone average loss who underwent audiometric evaluation with
of the air-conduction thresholds for 500 Hz, 1000 measurements of pure-tone hearing thresholds for
Hz and 2000 Hz11,12. However, a number of authors air and bone conduction and speech audiometry
have advocated the use of hearing loss classifica- consisting of the SRT and SDS tests between April
tions by frequency ranges in order to include the 2011–April 2012. Some exclusion criteria were
high frequencies in composing these averages, established: individuals with mixed or conductive
especially for elderly individuals13,14. hearing loss in at least one ear, or incomplete/
The Bureau Internacional d’ Audiophonologie inconclusive tests, e.g, when pure-tone thresholds
(BIAP, 2005) recommended that the classification of over the frequencies of 500 Hz to 4000 Hz were not
hearing losses should take into account the averages obtained or speech audiometry was not performed.
of the pure-tone thresholds for air conduction at In total, 241 tests were analyzed: 153 (63.5%)
500 Hz, 1000 Hz, 2000 Hz and 4000 Hz in order to of female patients and 88 (36.5%) of male patients.
encompass the high frequencies, as these are the Ages ranged from 60 to 97 years (mean, 77.9 years).
most compromised in the pathologies of hearing15. The results of 482 ears were evaluated.
Because the frequency ranges of consonants The present study examined the pure-tone
above 2000 Hz are the chief constituents of speech hearing thresholds for air conduction obtained
intelligibility, Russo proposed, in 2009, that the through the analysis of pure-tone audiometry based
degree of hearing loss should be classified based on the following averages:
on the average audibility threshold obtained over • Average 1: tritone average of frequencies 500
the frequencies of 500 Hz to 4000 Hz13. Hz, 1000 Hz and 2000 Hz according to Davis
The classifications of hearing losses relying on and Silverman, 197011 and Lloyd and Kaplan,
the tritone average as suggested by Lloyd & Kaplan 197812;
(1978)12 and Davis & Siverman (1970)11 were found • Average 2: pure-tone average of frequencies 500
to be adequate to classify the hearing losses with a Hz, 1000 Hz, 2000 Hz and 4000 Hz according
flat audiometric configuration. However, in ski slope to Recommendation 02/1 of the Bureau
hearing losses, such as presbycusis, this classi- Internacional d’Audiophonologie (BIAP)15;
fication is not always consistent with the patients’ • Average 3: pure-tone average of frequencies
complaints, as these refer to difficulties in speech 500 Hz, 1000 Hz, 2000 Hz and 3000 Hz, and
recognition, mostly. • Average 4: pure-tone average of frequencies
Given that the major auditory function concerns 500 Hz, 1000 Hz, 2000 Hz, 3000 Hz and 4000
verbal communication and speech recognition Hz.
abilities, it is paramount that the speech tests be The data were treated statistically using the R
considered in the classification of hearing losses, software. Distribution of frequencies was performed
thus expressing the actual hearing impairment of for the categorical variable (sex). For the continuous
the patients. The aim of the present study was to variables (age, frequencies, SRT and SDS),
assess which pure-tone averages have a stronger measures of central tendency (mean and median)

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

 Hearing losses and speech recognition  1111
and variability (standard deviation, minimum and coefficient of correlation, i.e., the greater the corre-
maximum) were used. lation, the lower the MSE (lower prediction error);
The four averages and the individual frequencies consequently, the better the variable as a predictor.
500 Hz, 1000 Hz, 2000 Hz, 3000 Hz and 4000 Hz To ascertain whether the correlations obtained
were correlated with the SRT and SDS values
were statistically different, comparison of correla-
obtained in the audiometric tests in order to assess
which pure-tone average or individual frequency tions was the method adopted for hypothesis testing.
best represents speech recognition ability. To that The correlation values were compared both for the
end, Spearman’s coefficient of correlation was used, SRT and the SDS, and the results were obtained
with a confidence interval of 95%. To determine using p-values, with statistical significance defined
the strength of the correlation, the following by p<0.05.
classification scale was used16: 0–0.2: very poor
correlation; 0.21–0.4: poor correlation; 0.41–0.6:
„„ RESULTS
moderate correlation; 0.61– 0.8: good correlation,
and 0.8 –1.0: very good correlation.
In addition, the mean squared error (MSE) was The mean values obtained in the audiologic
calculated. The MSE is inversely related to the evaluation of the 482 ears can be found in Figure 1.
Intensity (dBNA)

Frequency (Hz)

Figure 1 – Means for the pure-tone thresholds by frequency of 482 ears

Table 1 shows the descriptive analysis of the 4000 Hz, Average 1, Average 2, Average 3 and
threshold averages by frequency, the four calcu- Average 4 with the SRT and SDS. In addition,
lated pure-tone averages, the SRT in dBHL, and the confidence intervals and mean squared errors were
SDS in percent values. calculated, which served to identify the analyses
Correlations were established for the individual with the highest predictive value for the SRT and
frequencies 500 Hz, 1000 Hz, 2000 Hz, 3000 Hz, SDS. The results can be seen in Table 2.

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

1112  Anjos WT, Labanca L, Resende LM, Costa-Guarisco LP

Table 1 – Descriptive analysis of the pure-tone frequencies, calculated means, mean SRT and SDS in
the study sample

Mean Median SD Minimum Maximum

Descriptive
(dB HL) (dB HL) (dB HL) (dB HL) (dB HL)
250 Hz 30.36 25 15.53 0 100
500 Hz 31.05 30 16.08 0 95
1000 Hz 33.98 35 16.85 0 100
2000 Hz 42.99 45 18.04 5 120
3000 Hz 49.42 50 19.29 0 120
4000 Hz 54.08 55 19.47 0 115
6000 Hz 67.75 70 19.47 10 120
8000 Hz 65.81 65 18.43 5 110
Average 1 36.01 35 15.50 3.33 90
Average 2 40.53 40.63 15.31 2.5 87.5
Average 3 39.36 40 15.54 2.5 88.75
Average 4 42.30 43 15.68 2 94
SRT 41.29 40 16.57 10 110
SDS* 75.99* 84* 22.66* 0* 100*
* – values in %

Table 2 – Correlations of the individual frequencies 500 Hz, 1000 Hz, 2000 Hz, 3000 Hz and 4000 Hz,
Average 1, Average 2, Average 3 and Average 4 with the SRT and the SDS

SRT SDS
Variable CI MSE CI MSE
Rho Rho
CI.inf CI.up (dBHL2) CI.inf CI.up (dBHL2)
500 Hz 0.807* 0.765 0.844 108.4 -0.569 -0.634 -0.501 322.9
1000 Hz 0.889* 0.857 0.915 71.5 -0.675** -0.720 -0.620 284.2
2000 Hz 0.841* 0.800 0.875 88.8 -0.735** -0.778 -0.688 282.1
3000 Hz 0.700 0.633 0.752 138.2 -0.696** -0.745 -0.637 323.4
4000 Hz 0.636 0.574 0.691 162.4 -0.653** -0.706 -0.589 348.1
Average 1 0.934* 0.901 0.958 52.2+ -0.734** -0.776 -0.683 251.6
Average 2 0.918* 0.887 0.941 59.0 -0.768** -0.811 -0.720 245.3
Average 3 0.922* 0.886 0.944 56.4 -0.768** -0.807 -0.721 245.0+
Average 4 0.895* 0.861 0.921 68.7 -0.773** -0.813 -0.725 251.2
Average 1- mean of frequencies 500 Hz, 1000 Hz and 2000 Hz
Average 2- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz
Average 3- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz and 3000 Hz
Average 4- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz, 3000 Hz and 4000 Hz
Rho- Spearman’s coefficient of correlation
MSE- Mean Squared Error
SRT- Speech Reception Threshold
SDS- Speech Discrimination Score
CI- Confidence Interval
CI inf- lower limit of Confidence Interval
CI up- upper limit of Confidence Interval
*very good correlation
**good correlation
+
best variable as a predictor of correlation
Spearman’s coefficient of correlation test
Mean Squared Error test

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

 Hearing losses and speech recognition  1113
In Figure 2, a matrix with hypothesis testing to assess whether the correlations found in Table 2
p-values is presented. Using that matrix, it is possible are statistically different, both for the SRT and SDS.

Average Average Average Average

500 Hz 1000 Hz 2000 Hz 3000 Hz 4000 Hz
  1 2 3 4
500 Hz 0,000 0,084 0,001 0,000 0,000 0,000 0,000 0,000
1000 Hz 0,000 0,003 0,000 0,000 0,000 0,003 0,001 0,336
2000 Hz 0,000 0,014 0,000 0,000 0,000 0,000 0,000 0,000
3000 Hz 0,003 0,278 0,053 0,001 0,000 0,000 0,000 0,000
4000 Hz 0,048 0,296 0,004 0,020 0,000 0,000 0,000 0,000
Average 1 0,000 0,000 0,479 0,108 0,011 0,001 0,003 0,000
Average 2 0,000 0,000 0,022 0,001 0,000 0,000 0,114 0,000
Average 3 0,000 0,000 0,022 0,001 0,000 0,000 0,297 0,000
Average 4 0,000 0,000 0,006 0,000 0,000 0,002 0,146 0,200

- p-value for the SDS

- p-value for the SRT
Average 1- mean of frequencies 500 Hz, 1000 Hz and 2000 Hz.
Average 2- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz.
Average 3- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz and 3000 Hz.
Average 4- mean of frequencies 500 Hz, 1000 Hz, 2000 Hz, 3000 Hz and 4000 Hz.
Hypothesis testing of the equality of the correlations of Spearman’s coefficient of correlation.
Level of significance: p<0.05.

Figure 2 – Correlation matrix with hypothesis testing p-values comparing the significance of the
correlations for the SRT and SDS

„„ DISCUSSION auditory system that arise with aging17,21. In elderly

individuals with presbyacusis, the apical coil of the
The aim of the present study was to identify which cochlea, responsible for the detection of low-pitched
pure-tone average has the greatest correlation with sounds, is more preserved compared with the basal
speech recognition. To that end, the pure-tone coil, which detects the high-pitched sounds17.
thresholds for air conduction were classified using Regarding speech discrimination, Table 1 shows
four different averages: Average 1 (500 Hz, 1000 that the mean for the SDS was 75.99%, which was
Hz and 2000 Hz), average 2 (500 Hz, 1000 Hz, expected for elderly patients22-26. Studies suggest
2000 Hz and 4000 Hz), average 3 (500 Hz, 1000 that the difficulty in speech comprehension, demon-
Hz, 2000 Hz and 3000 Hz) and average 4 (500 Hz, strated by the lower SDS, occurs as a result of the
1000 Hz, 2000 Hz, 3000 Hz and 4000 Hz). These structural and physiologic changes in the auditory
averages were correlated with the tests of speech system with aging22,23.
recognition, SRT and SDS, obtained in the audio- Table 2 shows the correlation values for the
logic evaluation. individual frequencies 500 Hz, 1000 Hz, 2000
The curve depicted in Fig. 1, which shows the Hz, 3000 Hz and 4000 Hz, Average 1, Average 2,
average pure-tone thresholds of all the study patients Average 3 and Average 4 with the SRT and SDS.
by frequency, is similar to a typical audiogram of Although all the averages showed very good corre-
elderly individuals, as expected. In this population, lation, it was found that Average 1 had greater
a ski slope configuration is commonly found, with correlation with the SRT (0.934). For that reason,
a high-frequency hearing loss and preserved Average 1 was considered to be the most adequate
low-pitch sensitivity14,17-20. Presbyacusis is defined estimator of the threshold of speech recognition16.
as a bilateral sensorineural hearing loss that is This fact was corroborated by the mean squared
more pronounced for high-pitched sounds due to error found in the correlation of Average 1 with the
the degenerative and physiologic changes in the SRT (52.2). This lower mean demonstrated the lower

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

1114  Anjos WT, Labanca L, Resende LM, Costa-Guarisco LP

prediction error of Average 1 relative to the SRT. In are more frequent in the patient’s vocabulary, which
other words, Average 1 is the best predictor of SRT increases redundancy and the chances of correct
values. Additionally, Fig. 2 shows that the correlation guesses. This makes guessing with the SRT test
of SRT and Average 1 differs statistically from the easier than with the SDS.
correlations between SRT and the other averages.
This proves that, even with such narrow confidence A recent study compared the audiologic perfor-
intervals (Table 2), the correlation of Average 1 with mance of elderly individuals based on the classifi-
the SRT can be regarded as the strongest. Thus, cations of Davis and Silverman, 197011 and on the
this result indicates that the averages of 500 Hz, Recommendation 02/1 of the Bureau Internacional
1000 Hz and 2000 Hz used to compose Average 1 d’Audiophonologie (BIAP)15. The authors reviewed
are key frequencies to estimating the SRT. This fact the records of 140 elderly individuals focusing on
can be confirmed by the analysis of the individual the clinical history and the pure-tone audiometry
frequencies in Table 2. results. Predominance of mild to moderate sensori-
The correlation values for the SDS are negative, neural hearing loss was found, with slight differences
since they are inversely proportional: as averages regarding its prevalence depending on the hearing
increase, SDS percentages decline. All the averages loss classification adopted. Based on Davis and
had good correlation with the SDS, with statisti- Silverman, 99 cases of ears with normal thresholds
cally superior correlation values for Averages 2, 3
were found, while only 66 cases were identified
and 4 (Table 2 and Fig. 2). As Fig. 2 illustrates, the
according to the BIAP recommendation. The study
correlation between Average 1 and the SDS, albeit
good, is statistically inferior. On the other hand, no concluded that the classifications achieved similar
statistically significant difference was found when results; however, the BIAP recommendation was
comparing the correlations between Averages 2 found to be more sensitive in detecting hearing
and 3; 2 and 4, and 3 and 4. This finding indicates losses in the elderly27. A study conducted in Finland
that the correlations of the SDS and Averages with 5400 volunteers aged 55–75 years demon-
2, 3 and 4, in addition to being superior, are also strated a remarkable difference between the hearing
quite similar; therefore, it is impossible to infer loss classification by the World Health Organization
which of these three averages is the best predictor. (WHO) criteria compared with the European Union
Nevertheless, these results allow us to conclude criteria. In that study, the percentage of individuals
that the frequencies 3000 Hz and 4000 Hz are
with normal hearing was greater when the WHO
important for speech recognition, since the corre-
lation of Average 1 was lower with the SDS, and classification was used28.
statistically different as compared with the other The results of the analyses enabled us to
three averages. To corroborate these findings, the conclude that speech discrimination as measured
MSE (Table 2) was lower for Average 3, followed by the SDS is influenced by the frequencies 3000
by Average 2—albeit with quite similar values. This Hz and 4000 Hz. For that reason, the importance
shows that both averages, 2 and 3, have a greater of including these frequencies in the pure-tone
predictive value regarding the SDS, i.e., the more average used for classifying the hearing loss
severe the impairment at 3000 Hz and 4000 Hz, the
should be emphasized, since one of the objectives
worse the performance in the SDS test. This finding
of this classification is to reflect a person’s hearing
is explained in the literature7,9 which indicates that
60% of speech intelligibility relies on frequencies abilities and difficulties especially with regard to
above 1000 Hz. communication.
The contribution of high frequencies to speech
recognition was more evident in the SDS than in the „„ CONCLUSION
SRT. We believe this occurred because in the SDS
test, whose purpose is to assess the percentage of
For the elderly population with ski slope senso-
speech recognition, one-syllable words are used,
as they provide fewer clues in speech and are less rineural hearing loss, the SRT is more strongly
redundant, which reduces the likelihood of correct correlated with the average of frequencies 500
guesses. In the SRT evaluation, the purpose is Hz, 1000 Hz and 2000 Hz, while the SDS shows
to assess the threshold of speech recognition; greater correlation with the average that includes
therefore, three-syllable words are used, as they frequencies 3000 Hz and 4000 Hz.

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

 Hearing losses and speech recognition  1115

RESUMO

Objetivo: verificar quais médias tonais possuem maior correlação com o Limiar de Recepção da Fala
e com o Índice de Reconhecimento da Fala. Métodos: foram selecionados 241 exames de pacientes
idosos com perda auditiva neurossensorial que realizaram audiometria tonal liminar e logoaudiome-
tria. As avaliações audiométricas foram classificados com base nos limiares tonais de via aérea das
seguinte formas: Média 1- Média das frequências de 500, 1000 e 2000 Hz; Média2-Média das frequ-
ências de 500, 1000, 2000 e 4000 Hz; Média 3 – Média das frequências de 500, 1000, 2000 e 3000 Hz
e Média 4 –Média das frequências de 500, 1000, 2000, 3000 e 4000 Hz. Os dados foram comparados
com os testesLimiar de Recepção da Fala e Índice de Reconhecimento da Fala e tratados estatistica-
mente. Resultados: a Média 1 apresentou maior valor de correlação com o Limiar de Recepção da
Fala (rho=0,934; IC=0,901 a 0,958; eqm=52,2). Em relação ao Índice de Reconhecimento da Fala, foi
observado que amédia 3 apresentou omaior grau de correlação com o teste(rho= – 0,768; IC= –0,807
a –0,721; eqm = 245) seguido das médias 2 e 4. Conclusão: para a população idosa com perda audi-
tiva neurossensorial descendente, o Limiar de Recepção de Fala possui correlação mais forte com a
média das frequências 500 Hz, 1000 Hz e 2000 Hz, enquanto o Índice de Reconhecimento de Fala
possui maior correlação com as média que incluem as frequências de 3000 Hz e 4000 Hz.

DESCRITORES: Perda Auditiva; Presbiacusia; Audiometria; Idoso; Inteligibilidade da Fala

„„ REFERENCES 10. Redondo MC, Lopes Filho OC. Testes básicos

de avaliação auditiva. In: Lopes Filho OC. Tratado
1. Magalhães ATM, Gómez MVSG. Speech de Fonoaudiologia. São Paulo: Roca; 1197. P.
discrimination index in presbycusis. Arq. Int. 83-108;
Otorrinolaringol. 2007;11(2):169-74. 11. Davis H, Silverman RS. Auditory tests and
2. Veras RP, Mattos LC. Audiologia do hearing aids. New York: Holt Rinehart and
envelhecimento: revisão da literatura e perspectivas Winston; 1970. Hearing handicap standards for
atuais. RevBrasOtorrinolaringol. 2007;73(1):128-34. hearingandmedicolegalrules; p.253-79.
3. Francelin M, Motti TF G, Morita I. As implicações 12. Lloyd L, Kaplan H. Audiometric interpretation: a
sociais da deficiência auditiva adquirida em manual of basic audiometry: Press, 1978.
adultos. Saude Soc. 2010;19(1):180-92. 13. Russo ICP, Pereira LD, Carvallo RMM,
4. Humes LE, Watson BU, Christensen LA, Cokely Anastasio ART. Encaminhamentos sobre a
CG, Halling DC, Lee L. Factors associated with classificação do grau de perda auditiva em nossa
individual differences in clinical measures of speech realidade. RevSocBrasFonoaudiol [periódico
recognition among the eldery. J. Speech Hear Res. online]. 2009;14(2):287-8.Disponível em: http://
1992;37:465-74. www.scielo.br/pdf/rsbf/v14n2/23.pdf.
5. Jerger J, Jerger S, Pirozzolo F. Correlacional 14. Baraldi GS, Almeida LC, Borges AC. Evolução
analysis of speech audiometric scores, hearing da perda auditiva no decorrer do envelhecimento.
loss, age and cognitive abilities in the elderly. Ear RevBrasOtorrinolaringol. 2007;73(1):64-70. 
Hear. 1991;12:103-8. 15. Bureau International d’ Audiophonologie (BIAP).
6. Solomon LN, Webster JC, Curtis JF. A factorial [Les recommendations] [Internet]. Liège: BIAP; 2005.
study of speech perception. J. Speech Hear. Res. Recommendation BIAP nº 02/1bis. Classification
1960;37:655-61. audiométrique des déficiencesauditives; [cited 2011
7. Russo ICP, Behlau M. Percepção da fala: análise May 15]; p.5. Available from:  http://www.biap.org/
acústica do português brasileiro. São Paulo: Lovise; recom02-1.htm. French;
1993. 16. Fonseca JS; Martins GA. Curso de Estatística.
8. Miller GA, Nicely PE. An analysis of perceptual 6ª Edição. São Paulo: Atlas; 1996;
confusions among some English consonants. 17. Corso JF. Presbycusis, hearing aids and aging.
J.Acoust.Soc.Am. 1955;27(2):338-14. Audiology. 1977;16(2):146-63.
9. Fletcher H. Speech and hearing communication. 18. Melo ADP, Castiquini EAT, Noronha-Souza AEL.
New Jersey: D.VanNostrand; 1953. Identificação de perda auditiva nos alunos que

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116

1116  Anjos WT, Labanca L, Resende LM, Costa-Guarisco LP

frequentam a Universidade Aberta à Terceira Idade. 24. Studebaker GA, Sherbecoe RL, McDaniel DM,
Salusvita. 2004;23(2):279-90. Gwaltney CA. Monosyllabic word recognition at
19. Mattos LC, Veras RP. A prevalência da perda higher-than-normal speech and noise levels. J.
auditiva em uma popu­lação de idosos da cidade Acoust. Soc. Am. 1999;105(4):2431-44.
do Rio de Janeiro: um estudo seccional. Braz J
Otorhinolaryngol. 2007;73(5):654-9. 25. Martini A, Mazzoli M, Rosignoli M, Trevisi P,
20. Russo ICP. Achados audiométricos em uma Maggi S, Enzi G, et al. Hearing in the elderly: a
população de idosos presbiacúsicos brasileiros em population study. Audiology. 2001;40(6):285-93.
função do sexo e da faixa etária. Pró-Fono R Atual 26. Kano CE, Mezzena LH, Guida HL. Estudo
Cient. 1993;5(1):8-10. comparativo da classificação do grau de perda
21. Pedalini M, Liberman P, Piranas S, Jacob W,
auditiva em idosos institucionalizados. Rev CEFAC.
Câmara J, Miniti A Análise do perfil audiológico de
2009;11(3):473-7.
idosos através de testes de função auditiva
periférica e central. Rev Soc Bras Fonoaudiol. 27. Tenório JP, Guimarães JATL, Flores NGC, Iório
1997;5(63):489-95. MCM. Comparação entre critérios de classificação
22. Felder E, Schrott, EA. Quantitative evaluation dos achados audiométricos em idosos. J. Soc. Bras.
of myelinated nerve fibres and hair cells in cochlea Fonoaudiol. 2011;23(2):114-8.
of humans with age-related high-tone hearing loss.
28. Uimonen S, Maki-Torkko E, Jounio-Ervasti
Hear. Res. 1995;91:19-32.
23. Fish I. The selective and differential vulnerability K, Sorri M. Hearing in 55 to 75 year old people in
of the auditory system. In: Wolstenholmeand GEW, northern Finland-a comparison of two classifications
Knight J (eds). Sensorioneural Hearing Loss. New of hearing impairment. Acta Otolaryngol.
York, Churchill-Livingstone, 1970. 1997;33:53-9.

Received on: October 08, 2012

Accepted on: July 26, 2013

Mailing address:
Letícia Pimenta Costa Guarisco
Rua Ouro Preto, 1275/ 04, Santo Agostinho
Belo Horizonte – MG – Brasil
CEP: 30170-041
E-mail: [email protected]

Rev. CEFAC. 2014 Jul-Ago; 16(4):1109-1116