Diabetes Research and Clinical Practice, 11 (1991)
17-22
17
Elsevier DIABET 0045 1
Auditory function in young patients with type 1 diabetes mellitus Jorge P. Ferrer l, Oscar Biurrun2, Juan Lorente2, Jesus I. Conget ‘, Rafael de Espafia2, Enric Esmatjes ’ and Ramon Gomis ’ ‘Endocrinoloa
and Nutrition Service and 20torhinolaryngology
Service, Hospital Clinic i Provincial, University of Barcelona
School of Medicine, Barcelona,
Spain
(Revision received 7 August 1990) (Accepted 13 August 1990)
-_____
--
Summary Comprehension of the effect of diabetes mellitus on auditory function has been hindered by the fact that previous studies have evaluated hearing function in heterogeneous groups of patients with diabetes mellitus, thus giving conflicting results. We have performed audiometric studies in 46 consecutive patients, 13 with newly diagnosed type 1 diabetes mellitus (group 1) and 33 with type 1 diabetes mellitus of more than 3 years of duration (group 2), of 14 to 40 years of age. The results were compared to an age-matched control group. Pure-tone auditory thresholds were significantly higher in all frequencies 250-8000 Hz in both groups when compared to the control subjects. Ten patients, all of which belonged to group 2, had auditory thresholds above 30 dB in at least one frequency, showing a conversational hearing loss that ranged between 11 and 44%. However, none of them referred subjective hypoacusia. Univariate analysis revealed significant associations between auditory thresholds and age, duration of disease as well as retinopathy, but not with neuropathy, HbA,, or hypoglycaemic episodes. Only age and duration of disease independently correlated with an auditory threshold using multiple regression. We conclude that type 1 diabetes mellitus can cause mild sensorineural hearing impairment which correlates with age and duration of disease. Key words: Type 1 diabetes mellitus; Hearing loss; Audiometry
Introduction Previous studies have shown contradictory results concerning auditory function involvement in __~ Correspondence to: Jorge P. Ferrer, Endocrinology and Nutrition Service, Hospital Clinic i Provincial, c/villarroel 170, Barcelona, 08036, Spain.
0168-8227/91/$03.50
0 1991 Elsevier Science Publishers
patients with diabetes mellitus [l-8]. Several reports have shown a bilateral sensorineural hearing loss [l-5] affecting mainly the high and middle frequencies [ 1,3,4] in patients with diabetes mellitus, an audiometric profile which is also characteristic of presbyacusis. Most of these studies [l-5,8] have not discriminated between type 1 and type 2 diabetes mellitus, according to modern
B.V. (Biomedical
Division)
18
criteria, and have included both young and old patients. On the other hand, at least two studies [6,7] have failed to detect significant differences in audiometric findings when only young patients with type 1 diabetes mellitus were examined. Furthermore, there has been a great deal of controversy regarding the precise relationship between auditory function and diabetic complications and metabolic control [ l-81. The purpose of this study has been to compare the auditory function of patients with type 1 diabetes mellitus as compared to healthy subjects, and to assess its possible relationship with other parameters such as age, duration of diabetes, metabolic control and chronic complications.
Subjects and Methods Patients Forty-six subjects of 14-40 years of age with type 1 diabetes mellitus, according to National Diabetes Data Group criteria [ 91, participated in
the study (Table 1). Patients were divided into two groups : group 1 comprised 13 patients whose diagnosis of type 1 diabetes mellitus had been made a maximum of 30 days before audiometric evaluation; and group 2 included 33 patients with more than 3 years of duration of diabetes. The patients constitute consecutive cases who were either admitted to the diabetes unit ward for other reasons (all patients in group 1 and 23 in group 2) or were being followed by one of the authors in the outpatient diabetes clinic of the University of Barcelona School of Medicine. None of the subjects included in the study had a history of hearing problems, ototoxic drug intake, excessive noise exposure, thyroid disease or features of Wolfram’s syndrome, and all had serum creatinine values under 115 pmol/l. All patients were ketone-free and had used no medication other than insulin during the 48 h that preceded the study. Audiometric testing was performed after obtaining informed consent from each patient. Thirty-seven healthy individuals with normal hearing of 14 to 40 years of age were used as
TABLE 1 Clinical characteristics
of patients with diabetes*
Number of patients Age (years) Age range Males/females Duration of diabetes (years): range mean + SD HbA,, (%) Number of patients (“/b) with: retinopathy neuropathy microalbuminuria proteinuria Hypoglycaemic episodes/week Total number of severe hypoglycaemic episodes * Values are expressed
as means + SD.
Group 1
Group 2
Groups
13 25.5 + 8.1 14-38 815
33 26.0 + 9.3 14-40 14119
46 25.9 f 8.9 14-40 22124
9.2 + 2.4
4-32 12.1 f 5.8 9.5 + 1.9
O-32 8.9 + 1.4 9.4 f 2.0
-
11 (33.3%) 15 (45.5%) 6 (18.2%) 3 (9.1%)
11 (23.9%) 16 (34.8%) 6 (13.0%) 3 (5.4%)
-
2.1 + 1.4
-
1.1 f 1.1
-
1 and 2
19
controls. The mean age of this group (23.16 k 9.38 years) did not differ significantly from that of groups 1 and 2. Methods Pure-tone audiometry was performed using an Amplaid 300 clinical audiometer in a sound-proof room; both air and bone conduction were tested at frequencies 250-8000 Hz and 250-4000 Hz, respectively. Results are expressed as mean airconduction auditory threshold of both ears. Hearing loss was quantified in patients with an auditory threshold above 30 dB in any frequency, employing a formula that takes into account only the auditory threshold in conversational frequencies (i.e., 500-4000 Hz) [lo]. Independent evaluation of each patient to ascertain the presence of diabetic complications was carried out using funduscopy, performed by an ophthalmologist after pupillary dilatation, sensitive and motor conduction velocity studies, mean values of urine albumin excretion measured with a radioimmunoassay method in at least two recent 24-h urine samples, serum creatinine and 24-h urinary protein measurement. Microalbuminutia was considered to be present if the mean urinary albumin excretion rate exceeded 20 pg per min, and clinical proteinuria was defined as 24-h urinary protein above 0.5 g per day [ lo]. Metabolic control was assessed by measurement of HbA,, with a high-performance liquid chromatography method (one determination in group 1, mean of the last three values within 6 months of the study in group 2). Normal values in our laboratory are 4.7-5.9%. Severe hypoglycaemia was defined as a hypoglycaemic episode that required assistance from another person to solve it. The mean number of capillary glycaemic values under 2.8 mM (50 mg/dl) per week reported by the patients were recorded. Student’s t-test, chi-square test, Pearson’s correlation coefficient and multiple regression using SPSS/PC+ statistics software pack were employed to analyze the data. Statistical significance was determined at the 0.05 level.
Results Group 1:
None of the patients with newly diagnosed type 1 diabetes mellitus had hearing loss (i.e., auditory threshold above 30 dB in at least one frequency). The auditory thresholds in these patients were slightly but significantly higher in all frequencies from 250 to 8000 Hz than in healthy controls (Fig. 1). Age and HbA,, were not significantly correlated with auditory threshold in this group of patients. Group 2: Eleven of the 33 patients with longduration type 1 diabetes mellitus showed hearing impairment according to the above definition. The hearing loss in these 11 subjects, quantified as described in Methods, was mild (mean 17.89% + 4.68 SD, range 11-24x). Thus, no patient exhibited severe hearing loss, and none of them reported subjective hypoacusia. In the 11 patients the nature of hearing loss was sensorineural-symmetrical or nearly symmetrical, affecting mainly the high frequencies (Fig. 2). When the patients in group 2 were considered as a whole, the auditory threshold was significantly higher than in the control group in all frequencies from 250 to 8000 Hz (Fig. 1). The auditory threshold was significantly positively correlated with the patients’ age at all fre500 Hz (P < 0.001) quencies : 250 Hz
Frequencies
(Hz
)
Fig. 1. Mean pure-tone auditory thresholds f SEM in patients from group 1 (a), 2 (b) and control (c). The auditory threshold in control subjects was significantly lower than in groups 1 (P < 0.01) and 2 (P < 0.0001) in all frequencies.
20 Frequences
t e
(Hz
1
50
2 60
Fig. 2. Mean pure-tone auditory thresholds k SEM in the 11patients with diabetes classified as hearing-impaired, showing a predominant involvement of high frequencies.
(P < O.OOOl), 1000 Hz (P < O.OOl), 2000 Hz (P < 0.0000 l), 4000 Hz (P < 0.0001) and 8000 Hz (P < 0.00001). The auditory threshold also showed a positive correlation with duration of diabetes in all frequencies: 250 (P < 0.05), 500 (P < O.Ol), 1000 (P < 0.01) 2000 (P < O.Ol), 4000 (P < 0.01) and 8000 Hz (I’ < 0.01). However, it did not correlate with HbA,,, number of hypoglycaemic episodes per week, or total number of severe hypoglycaemic episodes. Microalbuminutia was only significantly correlated with the auditory threshold at 8000 Hz (P < 0.05); in the same way, patients with nephropathy (microalbuminuria or clinical proteinuria) had a higher auditory threshold than those without nephropathy,
only at 8000 Hz (P < 0.005) (Table 2). Patients with retinopathy had a significantly higher auditory threshold than those without retinopathy in the frequencies 500-8000 Hz (Table 2). As expected, patients with retinopathy were significantly older (P < 0.001) and had a longer duration of disease (P < 0.0001) than those without retinopathy. The auditory threshold values between patients with and without neuropathy were not statistically different. Stepwise multiple regression analysis, considering groups 1 and 2 together in order to further reflect the effect of duration of disease, showed that age significantly correlated with the auditory threshold at all frequencies (P < 0.001, least significant P) and duration of diabetes was still independently significant at the frequencies 1000 Hz (I’< 0.02), 2000 Hz (P-C 0.02) and 8000 Hz (P = 0.05) (Table 3).
Discussion The results of this study show that type 1 diabetes mellitus affects pure-tone audiometry, even though hearing impairment is mild and subclinical. Previous reports had given conflicting results [ l-81. Kurien et al. [ 61 found that the auditory threshold was significantly higher in all frequencies in the 30 patients with type 1 and 2 diabetes mellitus of up to 50 years of age when
TABLE 2 Relationship Frequency
between pure-tone
auditory thresholds
(AT) and chronic diabetic complications* Nephropathy
Neuropathy
Retinopathy
(Hz)
250 500 1000 2000 4000 8000
with
without
19 -1_1.3 18 * 1.7 18.8 + 1 20.2 * 1.8 25 f 3.2 27.3 f 4.2
19 + 18.6 * 18.9 + 21.8 f 22.8 f 24.4 f
* Values are expressed N.S., not significant.
1.3 1.9 2 2.1 2.3 2.8
P value
with
N.S. N.S. N.S. N.S. N.S. N.S.
22.1 22.9 24 26.3 31.3 40
as means of AT in Db + SEM.
without f 2.5 + 2 f 1.8 k 2.7 & 2.8 + 6.2
18.3 f 17.3 + 17.8 + 19.9 f 22.1 * 22.6 f _~~_.
I 1.4 1.4 1.6 2.1 2.3
P value
with
N.S. N.S. N.S. N.S. N.S. 0.003
20.2 21.9 22.9 26.3 30.2 32.6
P value
without * + + + + +
2 2.8 2.1 2.1 2.8 4.8
18.4 + 16.4 f 16.8 f 19.9 f 20.6 i 22.3 f
1 1.2 1.2 1.6 2.2 2.6
N.S. 0.049 0.021 0.018 0.001 0.048
21 TABLE 3 Relationship
of auditory thresholds
with age and duration of diabetes using multiple regression Frequencies
Age Duration
of diabetes
analysis
(Hz)
250
500
1000
2000
4000
8000
0.02* N.S.
0.0005 N.S.
0.001 0.019
0.000 1 0.013
0.0011 N.S.
0.0000 0.044
* P value.
to a control group. Axelsson et al. [ 31 studied 205 patients with both insulin-dependent and noninsulin-dependent diabetes mellitus of up to 70 years of age and showed that sensorineural hearing loss was present in 9-31x of cases. Jorgensen [4] observed that 28 of 69 patients with diabetes mellitus showed hearing loss. This occurred more often in older patients and when retinopathy was present. Other workers have also shown hearing loss in heterogeneous groups of patients with diabetes mellitus [2,5]. Surprisingly, two of these studies [2,3] have shown that patients of the same age group treated with insulin tended to have a better auditory threshold than those on oral hypoglycaemic agents. On the other hand, both Sieger et al. [6] and Osterhammel et al. [ 71 failed to find any effect of diabetes mellitus on hearing function when only young insulin-dependent diabetic patients were evaluated. It could therefore be argued that possibly only type 2 diabetes mellitus or type 2 diabetes mellitus-associated conditions could affect auditory function. Our results show that young patients with type 1 diabetes mellitus may also exhibit higher auditory thresholds than healthy individuals ; discordance with other reports of type 1 diabetes mellitus patients is probably explained by the fact that these studies examined subjects of shorter duration of diabetes and younger age (e.g., all patients in the study of Sieger et al. [6] were less than 20 years old). Another author [ 81 also found that 17 insulin-dependent diabetes patients of 15 to 55 years of age, all of whom had neuropathy and microangiopathic complications, had a higher auditory threshold as compared
well as abnormal auditory evoked potentials when compared to healthy controls. According to our results and those of others [ l-31, age seems to be a major independent factor in the development of type 1 diabetes mellitusassociated hearing loss. Duration of diabetes also showed a significant correlation with the auditory threshold in many frequencies after multiple regression analysis and probably accounted for the significant effect of retinopathy observed with univariate analysis. Although other studies did not find a significant correlation of duration of diabetes with hearing loss, this is probably due to the fact that these studies did not analyze duration of diabetes mellitus as a quantitative parameter [ 1,3,4] and thus lost statistical power to detect statistically significant differences. Furthermore, most studies have included patients with type 2 diabetes mellitus [l-5], thus precluding a precise assessment of duration of disease. It is noteworthy that in our study no patient with newly diagnosed type 1 diabetes mellitus presented hypoacusia. Our study failed to show a correlation between the auditory threshold and HbA,, or number of hypoglycaemic episodes. Before accepting these results it should be stated that the group of patients we studied had been in unsatisfactory to poor control of their diabetes; thus, there were not enough patients with optimized levels of HbA,, to allow an adequate assessment of the effect of this parameter on auditory function. Furthermore, precise quantification of hypoglycaemic episodes is presently a difficult task, since it often relies on the patients’ memory.
22 Furthermore, there was no association between auditory impairment and the presence of neuropathy. Other reports which suggested an association of hearing impairment with diabetic neuropathy either did not use controls with diabetes mellitus but without neuropathy [5], or compared patients with diabetic neuropathy with other patients with diabetes that were not matched for duration of disease [S]. Microangiopathic changes in the inner ear have been observed in patients with diabetes mellitus [7], and studies with auditory-evoked potentials in patients with diabetes mellitus and no hearing impairment have shown central (brainstem) abnormalities [ 121. Since patients with retinopathy, unlike those with neuropathy, had more hearing impairment than those without it, it is tempting to hypothesize that microangiopathic lesions, peripheral or central, could be responsible for auditory impairment in type 1 diabetes mellitus. The audiometric pattern seen in our patients in consistent with both of these possible origins of auditory impairment. Further investigations, combining audiometric, histopathologic and electrophysiologic studies in the same group of patients, are needed to clarify this issue. Several considerations can be inferred from this study. In the first place, it is important to recognize that a mild symmetrical sensorineural deficit in a young patient with diabetes mellitus can be reasonably ascribed to diabetes mellitus itself, and thus does not oblige further diagnostic studies if there are no other associated findings. Moreover, although none of the patients had subjective hypoacusia, we speculate that if hearing impairment in type 1 diabetes mellitus is a function of age and duration of disease then severe presbyacusia may become a real problem with improved longevity of type 1 diabetes mellitus patients due to better treatment and prevention of diabetic complications. It remains to be
established whether long-term optimized metabolic control can influence auditory function. We conclude that mild asymptomatic sensorineural hearing loss occurs in many patients with type 1 diabetes mellitus, and that this deficit is related to the age of the patient and duration of disease.
References I Kurien, M., Thomas, K. and Bhanu, T.S. (1989) Hearing
1 L
3 4
5
6
7
8
9
10
11 12
threshold in patients with diabetes mellitus. J. Laryngol. oto1. 79, 314. Wackym, P.A. and Linthicum, F.H. (1986) Diabetes mellitus and hearing loss: clinical and histolopathologic relationships (1975) Am. J. Otol. 7, 176. Axelsson, A., Sigroth, K. and Vertes, D. (1974) Hearing in diabetics. Acta Otolaryngol. 356 (Suppl), 15. Jorgensen, M.B. and Buch, N.H. (1961) Studies on innerear function and cranial nerves in diabetics. Acta Otolaryngol. 53, 350. Friedman, S.A., Schulman, R.H. and Weiss, S. (1975) Hearing in diabetic neuropathy. Arch. Intern. Med. 135, 573. Sieger, A., White, N.H., Skinner, M.W. and Spector, G.J. (1983) Auditory function in children with diabetes mellitus. Ann. Otol. Rhinol. Laryngol. 92, 237. Osterhammel, D. and Christau, B. (1980) High frequency audiometry and stapedius muscle reflex thresholds in juvenile diabetics. Stand. Audiol. 9, 13. Eliachar, 1. and Kanter, Y. (1986) Auditory brainstem evoked potentials in insulin-dependent diabetics with and without peripheral neuropathy. Acta Otolaryngol. 102, 204. National Diabetes Data Group (1979) Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28, 1039. Mogensen, C.E. (1988) The kidney and hypertension in diabetes mellitus. Martinus Nijhoff Publishing, Boston, pp. 7-16. Katz, J. (1978) Handbook of clinical audiology. The William’s & Wilkins Company, Baltimore, pp. 98-140. Khardori, R., Soler, N.G., Good, D.C., DevlescHoward, A.B., Broughton, D. and Walbert, J. (1986) Brainstem auditory and visual evoked potentials in type 1 (insulindependent) diabetic patients. Diabetologia 29, 362.
17-22
17
Elsevier DIABET 0045 1
Auditory function in young patients with type 1 diabetes mellitus Jorge P. Ferrer l, Oscar Biurrun2, Juan Lorente2, Jesus I. Conget ‘, Rafael de Espafia2, Enric Esmatjes ’ and Ramon Gomis ’ ‘Endocrinoloa
and Nutrition Service and 20torhinolaryngology
Service, Hospital Clinic i Provincial, University of Barcelona
School of Medicine, Barcelona,
Spain
(Revision received 7 August 1990) (Accepted 13 August 1990)
-_____
--
Summary Comprehension of the effect of diabetes mellitus on auditory function has been hindered by the fact that previous studies have evaluated hearing function in heterogeneous groups of patients with diabetes mellitus, thus giving conflicting results. We have performed audiometric studies in 46 consecutive patients, 13 with newly diagnosed type 1 diabetes mellitus (group 1) and 33 with type 1 diabetes mellitus of more than 3 years of duration (group 2), of 14 to 40 years of age. The results were compared to an age-matched control group. Pure-tone auditory thresholds were significantly higher in all frequencies 250-8000 Hz in both groups when compared to the control subjects. Ten patients, all of which belonged to group 2, had auditory thresholds above 30 dB in at least one frequency, showing a conversational hearing loss that ranged between 11 and 44%. However, none of them referred subjective hypoacusia. Univariate analysis revealed significant associations between auditory thresholds and age, duration of disease as well as retinopathy, but not with neuropathy, HbA,, or hypoglycaemic episodes. Only age and duration of disease independently correlated with an auditory threshold using multiple regression. We conclude that type 1 diabetes mellitus can cause mild sensorineural hearing impairment which correlates with age and duration of disease. Key words: Type 1 diabetes mellitus; Hearing loss; Audiometry
Introduction Previous studies have shown contradictory results concerning auditory function involvement in __~ Correspondence to: Jorge P. Ferrer, Endocrinology and Nutrition Service, Hospital Clinic i Provincial, c/villarroel 170, Barcelona, 08036, Spain.
0168-8227/91/$03.50
0 1991 Elsevier Science Publishers
patients with diabetes mellitus [l-8]. Several reports have shown a bilateral sensorineural hearing loss [l-5] affecting mainly the high and middle frequencies [ 1,3,4] in patients with diabetes mellitus, an audiometric profile which is also characteristic of presbyacusis. Most of these studies [l-5,8] have not discriminated between type 1 and type 2 diabetes mellitus, according to modern
B.V. (Biomedical
Division)
18
criteria, and have included both young and old patients. On the other hand, at least two studies [6,7] have failed to detect significant differences in audiometric findings when only young patients with type 1 diabetes mellitus were examined. Furthermore, there has been a great deal of controversy regarding the precise relationship between auditory function and diabetic complications and metabolic control [ l-81. The purpose of this study has been to compare the auditory function of patients with type 1 diabetes mellitus as compared to healthy subjects, and to assess its possible relationship with other parameters such as age, duration of diabetes, metabolic control and chronic complications.
Subjects and Methods Patients Forty-six subjects of 14-40 years of age with type 1 diabetes mellitus, according to National Diabetes Data Group criteria [ 91, participated in
the study (Table 1). Patients were divided into two groups : group 1 comprised 13 patients whose diagnosis of type 1 diabetes mellitus had been made a maximum of 30 days before audiometric evaluation; and group 2 included 33 patients with more than 3 years of duration of diabetes. The patients constitute consecutive cases who were either admitted to the diabetes unit ward for other reasons (all patients in group 1 and 23 in group 2) or were being followed by one of the authors in the outpatient diabetes clinic of the University of Barcelona School of Medicine. None of the subjects included in the study had a history of hearing problems, ototoxic drug intake, excessive noise exposure, thyroid disease or features of Wolfram’s syndrome, and all had serum creatinine values under 115 pmol/l. All patients were ketone-free and had used no medication other than insulin during the 48 h that preceded the study. Audiometric testing was performed after obtaining informed consent from each patient. Thirty-seven healthy individuals with normal hearing of 14 to 40 years of age were used as
TABLE 1 Clinical characteristics
of patients with diabetes*
Number of patients Age (years) Age range Males/females Duration of diabetes (years): range mean + SD HbA,, (%) Number of patients (“/b) with: retinopathy neuropathy microalbuminuria proteinuria Hypoglycaemic episodes/week Total number of severe hypoglycaemic episodes * Values are expressed
as means + SD.
Group 1
Group 2
Groups
13 25.5 + 8.1 14-38 815
33 26.0 + 9.3 14-40 14119
46 25.9 f 8.9 14-40 22124
9.2 + 2.4
4-32 12.1 f 5.8 9.5 + 1.9
O-32 8.9 + 1.4 9.4 f 2.0
-
11 (33.3%) 15 (45.5%) 6 (18.2%) 3 (9.1%)
11 (23.9%) 16 (34.8%) 6 (13.0%) 3 (5.4%)
-
2.1 + 1.4
-
1.1 f 1.1
-
1 and 2
19
controls. The mean age of this group (23.16 k 9.38 years) did not differ significantly from that of groups 1 and 2. Methods Pure-tone audiometry was performed using an Amplaid 300 clinical audiometer in a sound-proof room; both air and bone conduction were tested at frequencies 250-8000 Hz and 250-4000 Hz, respectively. Results are expressed as mean airconduction auditory threshold of both ears. Hearing loss was quantified in patients with an auditory threshold above 30 dB in any frequency, employing a formula that takes into account only the auditory threshold in conversational frequencies (i.e., 500-4000 Hz) [lo]. Independent evaluation of each patient to ascertain the presence of diabetic complications was carried out using funduscopy, performed by an ophthalmologist after pupillary dilatation, sensitive and motor conduction velocity studies, mean values of urine albumin excretion measured with a radioimmunoassay method in at least two recent 24-h urine samples, serum creatinine and 24-h urinary protein measurement. Microalbuminutia was considered to be present if the mean urinary albumin excretion rate exceeded 20 pg per min, and clinical proteinuria was defined as 24-h urinary protein above 0.5 g per day [ lo]. Metabolic control was assessed by measurement of HbA,, with a high-performance liquid chromatography method (one determination in group 1, mean of the last three values within 6 months of the study in group 2). Normal values in our laboratory are 4.7-5.9%. Severe hypoglycaemia was defined as a hypoglycaemic episode that required assistance from another person to solve it. The mean number of capillary glycaemic values under 2.8 mM (50 mg/dl) per week reported by the patients were recorded. Student’s t-test, chi-square test, Pearson’s correlation coefficient and multiple regression using SPSS/PC+ statistics software pack were employed to analyze the data. Statistical significance was determined at the 0.05 level.
Results Group 1:
None of the patients with newly diagnosed type 1 diabetes mellitus had hearing loss (i.e., auditory threshold above 30 dB in at least one frequency). The auditory thresholds in these patients were slightly but significantly higher in all frequencies from 250 to 8000 Hz than in healthy controls (Fig. 1). Age and HbA,, were not significantly correlated with auditory threshold in this group of patients. Group 2: Eleven of the 33 patients with longduration type 1 diabetes mellitus showed hearing impairment according to the above definition. The hearing loss in these 11 subjects, quantified as described in Methods, was mild (mean 17.89% + 4.68 SD, range 11-24x). Thus, no patient exhibited severe hearing loss, and none of them reported subjective hypoacusia. In the 11 patients the nature of hearing loss was sensorineural-symmetrical or nearly symmetrical, affecting mainly the high frequencies (Fig. 2). When the patients in group 2 were considered as a whole, the auditory threshold was significantly higher than in the control group in all frequencies from 250 to 8000 Hz (Fig. 1). The auditory threshold was significantly positively correlated with the patients’ age at all fre500 Hz (P < 0.001) quencies : 250 Hz
Frequencies
(Hz
)
Fig. 1. Mean pure-tone auditory thresholds f SEM in patients from group 1 (a), 2 (b) and control (c). The auditory threshold in control subjects was significantly lower than in groups 1 (P < 0.01) and 2 (P < 0.0001) in all frequencies.
20 Frequences
t e
(Hz
1
50
2 60
Fig. 2. Mean pure-tone auditory thresholds k SEM in the 11patients with diabetes classified as hearing-impaired, showing a predominant involvement of high frequencies.
(P < O.OOOl), 1000 Hz (P < O.OOl), 2000 Hz (P < 0.0000 l), 4000 Hz (P < 0.0001) and 8000 Hz (P < 0.00001). The auditory threshold also showed a positive correlation with duration of diabetes in all frequencies: 250 (P < 0.05), 500 (P < O.Ol), 1000 (P < 0.01) 2000 (P < O.Ol), 4000 (P < 0.01) and 8000 Hz (I’ < 0.01). However, it did not correlate with HbA,,, number of hypoglycaemic episodes per week, or total number of severe hypoglycaemic episodes. Microalbuminutia was only significantly correlated with the auditory threshold at 8000 Hz (P < 0.05); in the same way, patients with nephropathy (microalbuminuria or clinical proteinuria) had a higher auditory threshold than those without nephropathy,
only at 8000 Hz (P < 0.005) (Table 2). Patients with retinopathy had a significantly higher auditory threshold than those without retinopathy in the frequencies 500-8000 Hz (Table 2). As expected, patients with retinopathy were significantly older (P < 0.001) and had a longer duration of disease (P < 0.0001) than those without retinopathy. The auditory threshold values between patients with and without neuropathy were not statistically different. Stepwise multiple regression analysis, considering groups 1 and 2 together in order to further reflect the effect of duration of disease, showed that age significantly correlated with the auditory threshold at all frequencies (P < 0.001, least significant P) and duration of diabetes was still independently significant at the frequencies 1000 Hz (I’< 0.02), 2000 Hz (P-C 0.02) and 8000 Hz (P = 0.05) (Table 3).
Discussion The results of this study show that type 1 diabetes mellitus affects pure-tone audiometry, even though hearing impairment is mild and subclinical. Previous reports had given conflicting results [ l-81. Kurien et al. [ 61 found that the auditory threshold was significantly higher in all frequencies in the 30 patients with type 1 and 2 diabetes mellitus of up to 50 years of age when
TABLE 2 Relationship Frequency
between pure-tone
auditory thresholds
(AT) and chronic diabetic complications* Nephropathy
Neuropathy
Retinopathy
(Hz)
250 500 1000 2000 4000 8000
with
without
19 -1_1.3 18 * 1.7 18.8 + 1 20.2 * 1.8 25 f 3.2 27.3 f 4.2
19 + 18.6 * 18.9 + 21.8 f 22.8 f 24.4 f
* Values are expressed N.S., not significant.
1.3 1.9 2 2.1 2.3 2.8
P value
with
N.S. N.S. N.S. N.S. N.S. N.S.
22.1 22.9 24 26.3 31.3 40
as means of AT in Db + SEM.
without f 2.5 + 2 f 1.8 k 2.7 & 2.8 + 6.2
18.3 f 17.3 + 17.8 + 19.9 f 22.1 * 22.6 f _~~_.
I 1.4 1.4 1.6 2.1 2.3
P value
with
N.S. N.S. N.S. N.S. N.S. 0.003
20.2 21.9 22.9 26.3 30.2 32.6
P value
without * + + + + +
2 2.8 2.1 2.1 2.8 4.8
18.4 + 16.4 f 16.8 f 19.9 f 20.6 i 22.3 f
1 1.2 1.2 1.6 2.2 2.6
N.S. 0.049 0.021 0.018 0.001 0.048
21 TABLE 3 Relationship
of auditory thresholds
with age and duration of diabetes using multiple regression Frequencies
Age Duration
of diabetes
analysis
(Hz)
250
500
1000
2000
4000
8000
0.02* N.S.
0.0005 N.S.
0.001 0.019
0.000 1 0.013
0.0011 N.S.
0.0000 0.044
* P value.
to a control group. Axelsson et al. [ 31 studied 205 patients with both insulin-dependent and noninsulin-dependent diabetes mellitus of up to 70 years of age and showed that sensorineural hearing loss was present in 9-31x of cases. Jorgensen [4] observed that 28 of 69 patients with diabetes mellitus showed hearing loss. This occurred more often in older patients and when retinopathy was present. Other workers have also shown hearing loss in heterogeneous groups of patients with diabetes mellitus [2,5]. Surprisingly, two of these studies [2,3] have shown that patients of the same age group treated with insulin tended to have a better auditory threshold than those on oral hypoglycaemic agents. On the other hand, both Sieger et al. [6] and Osterhammel et al. [ 71 failed to find any effect of diabetes mellitus on hearing function when only young insulin-dependent diabetic patients were evaluated. It could therefore be argued that possibly only type 2 diabetes mellitus or type 2 diabetes mellitus-associated conditions could affect auditory function. Our results show that young patients with type 1 diabetes mellitus may also exhibit higher auditory thresholds than healthy individuals ; discordance with other reports of type 1 diabetes mellitus patients is probably explained by the fact that these studies examined subjects of shorter duration of diabetes and younger age (e.g., all patients in the study of Sieger et al. [6] were less than 20 years old). Another author [ 81 also found that 17 insulin-dependent diabetes patients of 15 to 55 years of age, all of whom had neuropathy and microangiopathic complications, had a higher auditory threshold as compared
well as abnormal auditory evoked potentials when compared to healthy controls. According to our results and those of others [ l-31, age seems to be a major independent factor in the development of type 1 diabetes mellitusassociated hearing loss. Duration of diabetes also showed a significant correlation with the auditory threshold in many frequencies after multiple regression analysis and probably accounted for the significant effect of retinopathy observed with univariate analysis. Although other studies did not find a significant correlation of duration of diabetes with hearing loss, this is probably due to the fact that these studies did not analyze duration of diabetes mellitus as a quantitative parameter [ 1,3,4] and thus lost statistical power to detect statistically significant differences. Furthermore, most studies have included patients with type 2 diabetes mellitus [l-5], thus precluding a precise assessment of duration of disease. It is noteworthy that in our study no patient with newly diagnosed type 1 diabetes mellitus presented hypoacusia. Our study failed to show a correlation between the auditory threshold and HbA,, or number of hypoglycaemic episodes. Before accepting these results it should be stated that the group of patients we studied had been in unsatisfactory to poor control of their diabetes; thus, there were not enough patients with optimized levels of HbA,, to allow an adequate assessment of the effect of this parameter on auditory function. Furthermore, precise quantification of hypoglycaemic episodes is presently a difficult task, since it often relies on the patients’ memory.
22 Furthermore, there was no association between auditory impairment and the presence of neuropathy. Other reports which suggested an association of hearing impairment with diabetic neuropathy either did not use controls with diabetes mellitus but without neuropathy [5], or compared patients with diabetic neuropathy with other patients with diabetes that were not matched for duration of disease [S]. Microangiopathic changes in the inner ear have been observed in patients with diabetes mellitus [7], and studies with auditory-evoked potentials in patients with diabetes mellitus and no hearing impairment have shown central (brainstem) abnormalities [ 121. Since patients with retinopathy, unlike those with neuropathy, had more hearing impairment than those without it, it is tempting to hypothesize that microangiopathic lesions, peripheral or central, could be responsible for auditory impairment in type 1 diabetes mellitus. The audiometric pattern seen in our patients in consistent with both of these possible origins of auditory impairment. Further investigations, combining audiometric, histopathologic and electrophysiologic studies in the same group of patients, are needed to clarify this issue. Several considerations can be inferred from this study. In the first place, it is important to recognize that a mild symmetrical sensorineural deficit in a young patient with diabetes mellitus can be reasonably ascribed to diabetes mellitus itself, and thus does not oblige further diagnostic studies if there are no other associated findings. Moreover, although none of the patients had subjective hypoacusia, we speculate that if hearing impairment in type 1 diabetes mellitus is a function of age and duration of disease then severe presbyacusia may become a real problem with improved longevity of type 1 diabetes mellitus patients due to better treatment and prevention of diabetic complications. It remains to be
established whether long-term optimized metabolic control can influence auditory function. We conclude that mild asymptomatic sensorineural hearing loss occurs in many patients with type 1 diabetes mellitus, and that this deficit is related to the age of the patient and duration of disease.
References I Kurien, M., Thomas, K. and Bhanu, T.S. (1989) Hearing
1 L
3 4
5
6
7
8
9
10
11 12
threshold in patients with diabetes mellitus. J. Laryngol. oto1. 79, 314. Wackym, P.A. and Linthicum, F.H. (1986) Diabetes mellitus and hearing loss: clinical and histolopathologic relationships (1975) Am. J. Otol. 7, 176. Axelsson, A., Sigroth, K. and Vertes, D. (1974) Hearing in diabetics. Acta Otolaryngol. 356 (Suppl), 15. Jorgensen, M.B. and Buch, N.H. (1961) Studies on innerear function and cranial nerves in diabetics. Acta Otolaryngol. 53, 350. Friedman, S.A., Schulman, R.H. and Weiss, S. (1975) Hearing in diabetic neuropathy. Arch. Intern. Med. 135, 573. Sieger, A., White, N.H., Skinner, M.W. and Spector, G.J. (1983) Auditory function in children with diabetes mellitus. Ann. Otol. Rhinol. Laryngol. 92, 237. Osterhammel, D. and Christau, B. (1980) High frequency audiometry and stapedius muscle reflex thresholds in juvenile diabetics. Stand. Audiol. 9, 13. Eliachar, 1. and Kanter, Y. (1986) Auditory brainstem evoked potentials in insulin-dependent diabetics with and without peripheral neuropathy. Acta Otolaryngol. 102, 204. National Diabetes Data Group (1979) Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28, 1039. Mogensen, C.E. (1988) The kidney and hypertension in diabetes mellitus. Martinus Nijhoff Publishing, Boston, pp. 7-16. Katz, J. (1978) Handbook of clinical audiology. The William’s & Wilkins Company, Baltimore, pp. 98-140. Khardori, R., Soler, N.G., Good, D.C., DevlescHoward, A.B., Broughton, D. and Walbert, J. (1986) Brainstem auditory and visual evoked potentials in type 1 (insulindependent) diabetic patients. Diabetologia 29, 362.
