Diabetes-Related Changes in Auditory Brainstem Responses

Abstract

Objectives/Hypothesis

Determine effects on auditory brainstem response (ABR) of diabetes mellitus (DM) severity.

Study Design

A cross-sectional study investigating DM severity and ABR in military Veteran subjects with (166) and without (138) DM and with no more than moderate hearing loss.

Methods

Subjects were classified by three age tertiles (<50, 50–56, and 57+). DM severity was classified as insulin-dependent (IDDM), non–insulin-dependent (NIDDM), or no DM. Other DM measures included serum glucose, HbA1c, and several DM-related complications. ABR measures included wave I, III, and V latencies; I–III, III–V, and I–V latency intervals; and wave V amplitude; for each ear at three repetition rates (11, 51, and 71 clicks/second), and both polarities. Outcomes were stratified by age tertile and adjusted for pure tone threshold at 3 kHz. Repeated measures multivariate analysis of covariance modeled the ABR response at each repetition rate for DM severity (main effect) and hearing at 3 kHz (covariate). Modeled contrasts between ABR variables in subjects with and without DM were examined.

Results

Significant differences existed between no DM and IDDM groups in the younger tertile only. Adjusting for threshold at 3 kHz had minimal effect. Self-reported noise exposure was not related to ABR differences, but HbA1c and poor circulation were.

Conclusions

IDDM is associated with an increased wave V latency, wave I–V interval, and reduced wave V amplitude among Veterans under 50 years. Results were related to several DM complications.

INTRODUCTION

Diabetes mellitus (DM) produces complications of vascular and neurologic malfunction in those with the disorder. We have reported that military Veterans with DM have a greater hearing deficit than age-matched Veterans without DM, noted primarily among younger Veterans.^1-3 Others have reported that patients with DM have hearing loss greater than those without.^4-6 Bainbridge et al. found that among 5,140 participants of the National Health and Nutrition Examination Survey (NHANES), those with DM evidenced greater hearing loss than those without.⁷ Participants aged 20 to 49 years had the greatest prevalence disparity of high-frequency impairment of mild or greater severity in the better ear.

Though a number of studies have attempted to identify the source of the pathophysiology of diabetes-associated hearing loss, neither the location(s) of the lesions, nor the mechanism(s) of the hearing deficit in those with DM has been established. Kakarlapudi et al. examined electronic medical records of over 12,000 DM patients and over 53,000 age-matched controls from a US Department of Veterans Affairs database and reported that sensorineural hearing loss was more common in patients with DM, which correlated with the degree of elevated serum creatinine.⁸ Fukushima et al. have described DM-associated pathology changes within the cochlea, which included thickened vessels of the stria vascularis, atrophy of the stria vascularis, and loss of outer hair cells.⁹ Hirose suggested that the severity of DM or the serum glucose level may be related to hearing loss.¹⁰ Few studies have examined hearing loss and severity of DM, as assessed by treatment required for patients to manage their DM (oral hypoglycemic agents vs. insulin), or through measures of DM complications (neuropathy, retinal angiopathy, nephropathy), or of acute or intermediate measures of control (serum glucose, glycosylated hemoglobin, or HbA1c).

Numerous authors have reported auditory brainstem response (ABR) latencies indicative of increased central conduction times within the lower auditory pathways in subjects with type 1 and/or type 2 DM.^2,11-13 No studies have examined the association of DM with measured functionality along the entire auditory pathway, including cognitive function.

Most studies of diabetes and hearing have employed rather small numbers of participants. Some previous studies have been primarily of older subjects, whereas those of relatively young subjects have been of type 1 DM patients almost exclusively. Studies with large numbers that have included DM patients of all ages have not included tests of the auditory system beyond establishing pure tone thresholds for the usual clinical audiogram frequencies.

We undertook a cross-sectional study of Veterans with and without DM to assess possible differences in the auditory system between the two groups, including peripheral and central auditory function and cognitive measures. We planned the current study to include DM patients primarily of type 2 to include younger adults, include tests of the entire auditory system, have a sample size sufficient to distinguish clinically significant differences between those with and without DM, examine the effect of DM severity, and identify any relationship between DM-related hearing changes and other DM-related complications. This report addresses our assessment of the ABR, controlled for hearing thresholds, among those participants.

MATERIALS AND METHODS

Subjects

A detailed description of the methods used has been published elsewhere.³ Briefly, patients up to age 71 and receiving treatment through the Portland VA Medical Center were eligible for the study. Exclusions included those wearing a hearing aid; those undergoing treatment for cancer, multiple sclerosis, or other neurologic diseases; and those with dementia or communication difficulties. Among willing participants we further excluded subjects with a hearing loss (HL) in both ears exceeding 40 dB HL at 2 kHz or 70 dB HL at 4 kHz, and those with a clinically significant conductive component (air-bone gap >10 dB) or evidence of active otitis externa. From among patients not so excluded, we recruited 166 patients with DM. Because we focused on younger patients, some older patients were randomly excluded. We similarly recruited 138 subjects without DM from among those who met the inclusion criteria from the random sample. Of initially identified eligible patients, nearly 85% either declined (about 83%) or were excluded upon examination (2%) so the remaining 15% of participants represent a mostly self-selected group meeting study criteria. Demographic questionnaire data for two subjects (one with and one without DM) were missing. They were excluded from the analysis, leaving 302 subjects for study.

Data Collection

Subjects completed an initial questionnaire on demographics and selected medical items, which included questions associated with DM duration, treatment history, and complications, such as circulatory or sensory symptoms in hands or feet. For one measure of foot neuropathy we constructed a composite scale from the subjects’ responses to questions about numbness, tingling, or burning sensations in their feet. The foot index score ranged from 0 (not bothered by sensations) to 3 (bothered by three unpleasant sensations). Similar indices were constructed for hand and foot circulation and hand neuropathy. The neuropathy and circulation indices were used as potential indicators of small vessel or neural types of DM complications. An audiometric test of both ears was conducted including pure tone thresholds from 0.25 to 14 kHz. Blood glucose and HbA1c levels were determined via finger stick. Foot neuropathy tests were conducted using the standard method performed in an annual DM exam using a 10-g nylon monofilament. From participants’ medical charts we captured certain recent values for laboratory tests, including urinary microalbumin and results of clinical exams, such as retinal findings.

Pure tone threshold testing

Briefly, pure tone air conduction thresholds were obtained for pulsed tones. Frequencies tested were octave frequencies from 0.25 to 8 kHz, and the interoctave frequencies 1.5, 3, and 6 kHz. Subjects also underwent extended high-frequency testing (10 kHz, 12.5 kHz, and 14 kHz). When no response was obtained for pure tone thresholds, the value recorded was arbitrarily set to the maximum output by the equipment at the test frequency and that imputed value was included in analyses.

ABR measurement

Auditory brainstem response (ABR) testing was performed using SmartEP Version 3.95 (Intelligent Hearing Systems, Miami, FL). The stimuli were broadband clicks of 1 μs duration, presented at one intensity (110 dB peak equivalent sound pressure level, which was approximately 78 dB HL), two polarities (rarefaction, condensation) and three rates (11, 51, and 71 clicks/second) in each ear monaurally. The ABR was recorded differentially using a silver/silver chloride surface electrode secured to the ipsilateral mastoid for the active (inverting) channel. A surface electrode attached to the high forehead (Fz) served as the reference (noninverting) electrode and a surface electrode attached to the brow (Fpz) served as the ground. Efforts were made to maintain electrode impedances below 2 kΩ. The recorded signal was amplified (100,000), filtered (0.3–3.0 kHz), and digitized for subsequent analysis. The ABRs were collected as two samples of 1024 runs for each stimulus polarity and rate. The averages of the two samples were compared to determine whether ABR morphology was replicable. Replicable samples were then combined for a total of 2048 runs per scored ABR waveform.

Data Management

Initial data cleaning, variable conversions, data reduction, and preliminary analyses were done using Statistical Package for the Social Sciences version 13 (SPSS Inc., Chicago, IL). Definitive analyses were performed using SAS software version 9.1.3 (SAS Institute Inc., Cary, NC).

Diabetes severity groups

We categorized participants according to whether they had DM and whether they used insulin, creating a DM severity measure with 137 (45.4%) in the no DM group, 88 (29.1%) in the non–insulin-dependent diabetes mellitus (NIDDM) group, and 77 (25.5%) in the insulin-dependent diabetes mellitus (IDDM) group. To confirm the measure’s ability to correctly classify severity, other items related to DM severity were compared across the DM severity groups. One such item was the composite scale computed to measure foot neuropathy; others were retinopathy and microalbuminuria.

Age groups

We divided the total participants into nearly equal age tertiles. The youngest tertile (26–49 years) contained 38.1% of the participants; the middle (50–56 years) and the oldest tertiles (57–71 years) contained 31.5% and 30.5%, respectively. When auditory measures were found to be associated with DM severity, we examined tertile subgroups to determine if the association was distributed equally in our population or was centered in an age group.

Noise exposure

We created a noise index from questionnaire data collecting the history of noise exposure and use of hearing protection in a variety of military, occupational, and recreational settings, and past sudden intense noise exposures. Index values from the four queried sources of noise were combined into a total noise exposure scale ranging from 8 (no noise exposure) to 40 (frequent noise exposure without hearing protection) in all exposure categories.

Analysis

The main focus of the analysis was to examine ABR measures by DM severity, while controlling for age group, hearing, and prior noise exposure, and to determine the relative contribution of DM (and DM severity) to ABR differences compared to those without DM. A secondary aim was to determine if ABR differences were correlated with any traditional DM complications.

Diabetes severity (no DM, NIDDM, and IDDM) constituted the primary independent variable for the primary analysis. Using an analysis of variance (ANOVA) we examined whether or not the mean values for the ABR measures differed by DM severity and age group. Because similar effects were seen using rarefaction and condensation polarities, effects of DM on the ABR were almost exclusively confined to the youngest tertile, and because previously published hearing results were also mostly confined to this age group,³ rarefaction data from the youngest tertile were further examined.

We fitted a repeated measures multivariate analysis of covariance (MANCOVA) model to the ABR response vector (wave I latency, wave III latency, wave V latency, and wave V amplitude) at each click repetition rate (11, 51, and 71 clicks/second), using DM severity as a categorical main effect. Pure tone thresholds were available for all subjects in both ears, and thresholds at 3 kHz for the respective ear were included as a continuous covariate in each model. Self-reported noise exposure did not differ by DM severity and so was not adjusted for in these models.

Interactions between diabetes and pure tone threshold at 3 kHz never showed statistical significance in these analyses. The MANCOVA framework allowed us to compute wave I–III and wave I–IV latency intervals directly from parameters of the fitted model. Correlation across ABR measures on the same subject was modeled using a random patient-level intercept, and correlation among measurements on the same ear was modeled by assuming an unstructured covariance among repeated measures. Thus, this framework also allowed us to adjust for correlation among ABR measures within and between ears on the same subject.

We examined contrasts (mean differences) between ABR variables in subjects with and without DM. We also estimated contrasts expressing the mean ABR measure among NIDDM or IDDM groups minus the mean value among those with no DM. Contrasts were derived directly from linear combinations of the estimated parameters in the fitted MANCOVA model. The accuracy of model tests and contrasts were assessed by constructing normal probability plots and histograms of the transformed residuals for each of the fitted models, as well as boxplots of the transformed residuals by click rate and DM severity group. To examine the relative contribution of DM severity to ABR measures, we examined ABR contrasts both with and without adjusting for pure tone thresholds at 3 kHz.

We addressed our secondary aim (to examine whether ABR measures were associated with DM complications) using ABR measures best associated with IDDM. These were used as outcomes in a linear regression using only the no DM and IDDM groups in the model. Predictor variables were: HbA1c, retinopathy (each eye separately), microalbuminuria levels designated as ‘‘high’’ by the clinical laboratory, self-reported hypertension, and each of the four indices of extremity neuropathy or poor circulation.

RESULTS

Subject Characteristics

Characteristics of the 302 participants are shown in Table I. Male participants predominated (88.7%) and did not differ significantly among the three DM severity categories. Age was similarly distributed among the three groups, but education was higher in the IDDM group. Mean blood glucose and HbA1c, measures that reflect how well patients are able to manage their DM, differed significantly according to DM severity for all age groups. Self-reported noise exposure was high in this sample of Veteran patients, but did not differ significantly by DM severity.

TABLE I.

Characteristics of Study Participants.

Characteristic	No Diabetes		Diabetes Mellitus				Sig
			NIDDM		IDDM
	No.	%	No.	%	No.	%
Participants	137	100.0	88	100.0	77	100.0	.172**
Age Tertile
26–49 [n=115]	49	35.8	35	39.8	31	40.3
50–56 [n=95]	38	27.7	27	30.7	30	39.0
57–71 [92]	50	36.5	26	29.5	16	20.8
Gender
Male	116	84.7	81	92.0	71	92.2	.125**
Female	21	15.3	7	8.0	6	7.8
Mean blood glucose (mg/dL)	107.0		176.0		223.0		<.001§
SD	18.8		92.0		117.7
Missing	3	2.2	1	1.1	3	3.4
Mean HbA1c	5.35		7.12		7.94		<.001§
SD	.48		1.78		2.03
Missing	6	4.4	1	1.1	3	3.4
Mean duration of diabetes (yr)	NA		5.45		11.83		<.001§
SD	NA		3.83		6.64
Missing	NA		3	3.4	5	6.5
Education							.03**
Some high school or less	7	5.1	6	6.8	5	6.5
Completed high school	23	16.8	22	25.0	8	10.4
Post-high school	59	43.0	39	44.3	47	61.0
Completed college	48	35.0	19	21.6	17	22.1
Missing	0	0.0	2	2.3	0	0.0
Noise exposure history
Military noise index mean	6.84		6.99		7.11		.483§
Civil noise index mean	6.26		6.39		6.42		.735§
Recreational noise index mean	6.12		6.33		6.16		.563§
Sudden noise index mean	6.42		6.40		6.56		.692§
Total indices mean	25.7		26.27		26.32		.504§
Missing	9	6.6	6	6.8	6	8.9

^**Significance of distribution differences calculated by summary chi-square.

^§Significance of mean differences calculated by 1-way analysis of variance.

Sig = significance; SD = standard deviation; NA = not applicable.

The DM severity measure was validated by comparing other factors known to be related to DM severity across the three DM severity groups for each age tertile. The mean values for retinopathy, peripheral neuropathy, and nephropathy as judged by the presence of micro-albuminuria, were significantly higher in the IDDM group than in the no DM group (Table II). For the NIDDM group, only microalbuminuria mean values were significantly higher. Duration of DM was significantly longer for the IDDM group in each age tertile.

TABLE II.

Comparison of Mean Diabetes Mellitus Complications by Diabetes Status and Insulin Dependence.

	No Diabetes	NIDDM	IDDM
Retinopathy score right eye (range 0–5)**
Youngest tertile	0.00	0.00	0.71*
Middle tertile	0.00	0.19	0.73†
Oldest tertile	0.00	0.08	0.56†
Retinopathy score left eye (range 0–5)**
Youngest tertile	0.00	0.00	0.71†
Middle tertile	0.00	0.19	0.77†
Oldest tertile	0.00	0.04	0.44†
Microalbuminuria level high (% high)§
Youngest tertile†	0.0%	29.0%	44.0%
Middle tertile†	2.6%	34.8%	20.7%
Oldest tertile†	0.0%	45.8%	57.1%
Feet numb, tingling or burning index (range 0–3)**
Youngest tertile	0.68	1.03	1.48*
Middle tertile	0.92	1.44	1.93*
Oldest tertile	0.60	0.63	1.00‡

^**Significance calculated using analysis of variance comparison to no diabetes group (Bonferroni correction for multiple comparisons within tertile).

^§Significance calculated using chi-square test of independence among diabetes groups.

^‡P ≤.05.

^*P ≤.01.

^†P ≤.001.

NIDDM = non–insulin-dependent diabetes mellitus; IDDM = insulin-dependent diabetes mellitus.

Pure Tone Thresholds by Diabetes Severity, Age, and Test Frequencies

As previously reported,³ in the youngest tertile the NIDDM group demonstrated significantly higher mean thresholds than those with no DM across all test frequencies, including at 3 kHz. There was no significant difference in the threshold between the no DM group and the IDDM group at 3 kHz. The adjusted pure tone threshold pattern is shown in Figure 1, illustrating audiogram contrasts for the DM severity groups based on the model derived from all participants (n = 302) and both ears (n = 604 ears).

Fig. 1.

Differences in mean pure tone threshold between subjects without diabetes and those with non–insulin-dependent diabetes mellitus (NIDDM) or insulin-dependent diabetes mellitus (IDDM) for the youngest age tertile, by frequency. The reference line at 0 is equivalent to no mean difference from those without diabetes. Filled circles signify a statistically significant difference (P < .05). Reprinted with permission from Austin et al. Diabetes-related changes in hearing. Laryngoscope 2009;119:1788–1796.

Table III provides unadjusted pure tone threshold results for the youngest tertile, stratified by DM severity and ear.

TABLE III.

Mean Pure Tone Thresholds by Diabetes Severity, Youngest Age Tertile (27–49 Years).

Pure Tone Threshold	Right Ear (n= 115)
	No Diabetes (n=49)		NIDDM (n=35)		IDDM (n=31)
	Mean	SD	Mean	SD	Mean	SD
250	8.57	6.46	11.43	6.25	13.71*	7.41
500	10.20	6.29	16.29†	7.61	17.26†	6.69
1000	10.82	5.81	16.14†	6.54	14.52‡	5.82
1500	10.82	6.48	16.14‡	7.48	13.55	6.73
2000	12.86	7.30	17.29‡	7.70	15.16	9.08
3000	13.88	6.87	21.14*	13.07	18.23	11.30
4000	16.43	15.68	24.43‡	17.14	18.87	11.95
6000	19.29	15.84	24.00	18.30	20.00	13.35
8000	16.94	17.91	22.43	18.96	21.45	17.14
10000**	24.29	21.16	36.14‡	23.67	37.90‡	20.07
12500**	31.73	22.65	45.00‡	27.03	45.81‡	19.07
14000**	41.73	22.58	51.42	26.89	56.13‡	17.69
	Left Ear (n= 115)
	No Diabetes (n=49)		NIDDM (n=35)		IDDM (n=31)
Pure Tone Threshold	Mean	SD	Mean	SD	Mean	SD
250	9.49	6.31	14.86†	5.88	14.19*	7.21
500	10.33	6.19	17.43†	6.23	16.13*	7.04
1000	11.43	6.29	17.57†	5.99	13.23	7.02
1500	10.31	7.67	17.57†	6.68	13.23	7.02
2000	13.47	7.45	19.43*	8.02	15.00	7.42
3000	16.53	11.65	24.00‡	13.33	17.90	11.46
4000	21.22	15.80	31.00‡	18.46	20.32	11.90
6000	21.73	15.76	29.09	19.02	20.65	12.02
8000	18.78	17.06	25.00	17.86	21.29	16.78
10000**	24.90	21.47	37.29‡	24.17	34.52	21.38
12500**	35.71	24.45	46.71	26.79	46.13	23.12
14000**	42.55	24.77	53.14	23.17	53.71	17.93

^‡P ≤.05 compared to no diabetes.

^*P ≤.01 compared to no diabetes.

^†P ≤.001 compared to no diabetes.

^**Thresholds placed at limits of equipment +5 dB.

NIDDM = non–insulin-dependent diabetes mellitus; IDDM = insulin-dependent diabetes mellitus; SD = standard deviation.

ABR Comparisons by Diabetes Severity, and Repetition Rate

Raw ABR amplitude and latency data from univariate ANOVA for the youngest tertile are presented in Table IV for 11 clicks/second. Data are for each ear, stratified by click polarity (condensation or rarefaction) and DM severity. These univariate ABR results are not adjusted for any hearing threshold. Similar ANOVA analyses for click rates of 51 and 71 clicks/second were done (data not shown). The pattern of results was similar for right and left ears and whether clicks were presented with condensation or rarefaction phase. In the youngest tertile, the ABR means in the no DM group differed from those in the IDDM group, but not the NIDDM group. Five sporadic differences between no DM and either NIDDM or IDDM occurred in the middle or oldest tertiles, but no pattern was apparent. These differences were: for the middle age tertile, left ear, at 71 clicks/ second, rarefaction, wave I–V latency, NIDDM (P = .006), and IDDM (P = .016); and for the oldest tertile, right ear, at 51 clicks/second, rarefaction polarity, latency for wave V (P =.038), and right ear at 11 clicks/ second, rarefaction, wave I–V interval, NIDDM (P = .001), and IDDM (P = .010). Because the sample size was greater using the rate of 11 clicks/second, those data are presented.

TABLE IV.

Auditory Brainstem Response by Diabetes Severity and Laterality—Repetition Rate 11/Second, Both Polarities, Youngest Age Tertile.

ABR Variable	Right Ear (n=76–105)**
	No Diabetes (n=31–45)		NIDDM (n=22–33)		IDDM (n=21–28)
	Mean	SE	Mean	SE	Mean	SE
Rarefaction
Wave I latency	1.61795	.033176	1.66897	.061434	1.72391	.042885
Wave III latency	3.83988	.035202	3.80167	.045658	3.96923‡	.043925
Wave V latency	5.70889	.046085	5.71061	.062916	5.91574‡	.052640
Wave I–V interval	4.04808	.050037	3.96161	.051571	4.17935	.062508
Wave V amplitude	.37778	.018821	.34406	.030908	.25278*	.019225
Condensation
Wave I latency	1.71757	.039764	1.75893	.064334	1.76413‡	.055942
Wave III latency	3.81307	.037003	3.81518	.046617	3.98400	.055073
Wave V latency	5.79778	.039077	5.85071	.055463	6.00600*	.065420
Wave I–V interval	4.04595	.046756	4.02857	.044240	4.20870‡	.046826
Wave V amplitude	.36849	.022461	.31594	.020925	.22140*	.020592
	Left Ear (n=76–103)**
	No Diabetes (n=37–45)		NIDDM (n=19–33)		IDDM (n=18–28)
ABR Variable	Mean	SE	Mean	SE	Mean	SE
Rarefaction
Wave I latency	1.66188	.038114	1.63600	.050272	1.76667	.042305
Wave III latency	3.86053	.034678	3.85000	.037694	3.96786	.047658
Wave V latency	5.69943	.052936	5.77500	.060253	5.97946‡	.065743
Wave I–V interval	4.04038	.055391	4.08500	.042720	4.17500	.068899
Wave V amplitude	.38041	.022994	.30239	.030326	.24396‡	.020994
Condensation
Wave I latency	1.70000	.031695	1.70208	.052817	1.75682	.057245
Wave III latency	3.87012	.036732	3.86161	.060837	3.98696	.055943
Wave V latency	5.85852	.051123	5.80833	.055012	6.05673‡	.052027
Wave I–V interval	4.11842	.039967	4.07500	.039325	4.28214	.068620
Wave V amplitude	.36777	.026042	.33013	.024115	.25377*	.017604

Bonferroni correction for multiple comparisons.

^**n varied by variable tested.

^‡P ≤.05 compared to no diabetes.

^*P ≤.01 compared to no diabetes.

ABR = auditory brainstem response; NIDDM = non–insulin-dependent diabetes mellitus; IDDM = insulin-dependent diabetes mellitus; SE = standard error.

Separate repeated measures MANCOVAs were performed for the three separate repetition rates to examine ABR measures by diabetes severity for the youngest age tertile, using 3 kHz pure tone threshold as a covariate. All produced similar results. Because of the similarity of results, we present the analysis with largest number of subjects available (i.e., both ears, 11 clicks/second). We selected the rarefaction data because unadjusted ANOVA results were similar for both polarities and we believe rarefaction data to be more comparable to other studies.

Age was not associated with ABR measures within this tertile and so was not included in the model. The adjusted ABR final models indicate that DM severity was an important determinant of ABR responses, but limited to IDDM subjects. ABR linear contrasts derived from the model results are shown in Table V for the youngest tertile at 11 clicks/second. Those with IDDM had consistently longer latencies for waves III and V, and wave interval I–V, and lower amplitude values (wave V) than those without DM. There were no significant differences between the no DM and the NIDDM groups.

TABLE V.

Difference (Contrast) Between Mean ABR Values^** for No Diabetes Less ABR Values for IDDM or NIDDM—11 Clicks/Second, Both Ears, Rarefaction, Youngest Tertile.

ABR Measures Contrasted	Contrast	SE	df	t Value	P Value
Wave I latency, no diabetes-NIDDM	−0.023	0.056	343	−0.42	.67
Wave I latency, no diabetes-IDDM	−0.102	0.058	343	−1.75	.08
Wave III latency, no diabetes-NIDDM	0.015	0.055	343	0.27	.79
Wave III latency, no diabetes-IDDM	−0.117	0.058	343	−2.03	.04
Wave V latency, no diabetes-NIDDM	−0.019	0.068	343	−0.28	.78
Wave V latency, no diabetes-IDDM	−0.230	0.071	343	−3.25	<.01
Wave V amplitude, no diabetes-NIDDM	0.044	0.029	343	1.49	.14
Wave V amplitude, no diabetes-IDDM	0.124	0.030	343	4.09	<.01
Wave I–III latency interval, no diabetes-NIDDM	0.038	0.044	343	0.86	.39
Wave I–III latency interval, no diabetes-IDDM	−0.015	0.046	343	−0.32	.75
Wave I–V latency interval, no diabetes-NIDDM	0.005	0.060	343	0.08	.94
Wave I–V latency interval, no diabetes-IDDM	−0.128	0.062	343	−2.07	.04

^**Adjusted for pure tone hearing threshold at 3 kHz.

ABR = auditory brainstem response; IDDM = insulin-dependent diabetes mellitus; NIDDM = non–insulin-dependent diabetes mellitus; SE = standard error.

Because logistic regression suggested that the contrasts between the mean ABR values of the IDDM group and the no DM group in the youngest tertile did not change much when the 3 kHz pure tone threshold was included, we built two separate repeated MANCOVA models, adjusted and unadjusted for 3 kHz pure tone threshold. We then compared the contrasts in those two models to observe the effect of adjustment. Individually adjusting for pure tone thresholds made small (<10%) differences in the no DM-IDDM contrasts. These contrast comparisons are presented in Figure 2.

Fig. 2.

Auditory brainstem response contrasts between no diabetes mellitus and non–insulin-dependent diabetes mellitus (NIDDM) or insulin-dependent diabetes mellitus (IDDM) groups, adjusted and unadjusted for 3 kHz pure tone threshold, youngest tertile, 11 repetitions/second. *Contrast value near zero.

ABR Measures and Diabetes Complications

ABR measures used as outcomes to examine our secondary aim were the wave V latency, wave V amplitude, and wave I–V interval. Duration of DM and selected DM complication variables were used as independent variables in separate linear regression models for these outcomes in analyses restricted to IDDM and no DM groups. Duration of DM was not associated with ABR measures. Further, most DM complication variables used did not predict these outcomes; however, several were consistently significant predictors. These are presented in Table VI.

TABLE VI.

Diabetes Complications Predictive^** of ABR Measures^§ Associated with IDDM Among Youngest Tertile Study Participants.

DM Complication Measure	ABR Measure
	Wave V Latency (n=61)	Wave V Amplitude (n=61)	Wave I–V Interval (n=54)
HbA1c	+	+	+
Duration of diabetes	−	−	−
Right retinopathy	−	−	−
Left retinopathy	−	−	−
Microalbumin high	−	±	−
Hypertension	−	−	−
Feet
Poor circulation index	+	+	+
Paresthesia index	−	−	−
Hands
Poor circulation index	+	+	+
Paresthesia index	−	−	−

^**Using linear regression.

^§Rep rate 11/second, right ear, rarefaction.

ABR = auditory brainstem response; DM = diabetes mellitus.

+ = statistically significant; – = borderline significant; ± = not significant.

DISCUSSION

Summary

The purpose of our study was to determine whether, among Veterans without severe hearing loss, those with DM have different auditory function than those without. We previously confirmed that those with DM have more pure tone hearing loss than those without, most marked among those under 50 years of age, particularly at the lower (0.25–1 kHz) and high (10.0, 12.5, and 14 kHz) frequencies. At 3 kHz, there was a significant difference between the no DM group and the less severe DM group (NIDDM), but no difference with the more severe (IDDM) group.

Abnormal ABRs, as measured by increased waveform peak latencies and reduced amplitudes, were also largely restricted to younger patients in our study, but, unlike the pure tone thresholds, they were confined to the IDDM group. This finding persisted when ABR results were adjusted for hearing thresholds at 3 kHz. There was not a significant effect of click rate, suggesting that increasing presentation rate did not reveal any excess deficits among patients with DM. Adjusting for pure tone thresholds at 3 kHz produced little effect in our subjects, suggesting that IDDM was the dominant predictor of the ABR abnormality, with little confounding by peripheral hearing.

Effect of diabetes severity

AUDITORY BRAINSTEM RESPONSE

Auditory nerve function, as measured by ABR wave I latency, was not impaired in patients with DM. Wave I–III interpeak latency, often taken as a measure of peripheral conduction time from the auditory nerve through the vicinity of the cochlear nucleus, was similar in patients with NIDDM and with no DM. For the IDDM group, the interval was either significantly (rarefaction polarity) or borderline significantly (condensation polarity) longer in the unadjusted measures compared with patients with no DM (Table IV), but in the adjusted model (Table V) there was no significant difference. The mean ABR wave V amplitude and latency were abnormal for all stimulus conditions examined in the IDDM group. Though we did not include the wave III–V interpeak interval in our regression models, the abnormal mean wave I–V interpeak interval in the presence of a normal wave I–III interpeak interval is taken as evidence for slowed conduction of neural responses and/or loss of neural synchrony within the central auditory system.

Because the ABR effect of IDDM for young patients is at a level above the auditory nerve and is present even when adjusted for hearing, we suggest that there are at least two mechanisms through which DM could impair auditory function: one is probably cochlear and the other within the central auditory system above the auditory nerve. The latter could signify an auditory processing disorder that would not be revealed through audiometric testing.

Other findings suggesting that the effect of DM on the cochlea and the effect on the central auditory system above the auditory nerve are independent is that in the younger age group, only NIDDM had significantly different pure tone thresholds in the middle ranges, particularly at 3 kHz, whereas for the unadjusted ABR measures only the IDDM group had significantly different mean values from the no DM group.

It is possible that there is an enhanced cochlear vulnerability to noise exposure among patients with DM that is largely obviated by insulin, though enhanced noise vulnerability would not explain the ABR results for patients with IDDM. Consistent with the hypothesis that there is a primary DM effect on the auditory pathways above the cochlea, previous examinations of ABR responses in patients with DM have reported abnormal central conduction times in the absence of hearing loss.^5,14

In addition to delays in wave V and the I-V interval, delays in the wave I-III interval are also associated with DM in some previous reports.^2,5,15 Involvement of wave I is a less consistent finding in the literature.¹³

Comparison of ABR and other diabetes complications

Our measure of DM severity was whether or not patients with DM used insulin. Approximately 47% of our subjects with DM (45% of those under 50) used insulin. More severe DM (IDDM) in the younger subjects was significantly associated with certain ABR measures arising central to the auditory nerve. This differed from our results of the hearing thresholds, in which NIDDM subjects differed from subjects without DM in the middle frequencies, but IDDM subjects did not. Another difference is that hearing thresholds were not associated with any of the usual DM measures or complications. When NIDDM subjects were excluded from the bivariate analysis, ABR wave V amplitudes were strongly negatively associated with HbA1c and poor foot circulation, and moderately so with microalbuminuria; wave V latency and wave I–V interpeak interval were both strongly associated with HbA1c levels and poor circulation indices in both hands and feet.

Limitations

Higher click rates were initially selected because faster presentation rates increase demand on neural conduction and recovery time and are known to reveal breakdowns associated with some neurological disease states. A disadvantage is that with greater breakdowns the participant results available to analyze diminish.

Our research subjects were Veterans, and of the participating subjects with DM, only one was diagnosed with type 1 DM, which is usually diagnosed in childhood. Consequently, our results apply to the more common type 2 DM, which can occur at any age, but is rare in adolescents and children.

We considered the possibility that noise-related hearing losses might account for the ABR differences between the no DM and IDDM groups, but there were no noise exposure differences between the study groups. Also, the ABR differences between the two groups were nearly as robust after controlling for hearing thresholds at 3 kHz.

CONCLUSION

Consistent ABR differences between DM patients and those without DM were not seen in ages over 50 years or among those with NIDDM. DM-associated auditory brainstem dysfunction was present among younger patients with the more severe IDDM. This differed from pure tone threshold results, which demonstrated DM-associated dysfunction in the younger group at a greater range of frequencies among those with NIDDM. Central auditory dysfunction was present when ABR measures were adjusted for hearing, indicating that this effect cannot be explained by peripheral auditory dysfunction in these IDDM patients.