Investigation of Stress Levels before the Onset of Idiopathic Sudden Sensorineural Hearing Loss

Abstract

OBJECTIVES

We hypothesized that patients with idiopathic sudden sensorineural hearing loss (ISSHL) would have experienced more stress prior to the onset than they typically did. This study investigated stress levels in patients before the onset of ISSHL.

MATERIALS and METHODS

Forty-two patients with ISSHL were investigated. We used an original questionnaire to evaluate subjective stress levels in 1 week before onset. Serum hemoglobin A1c (HbA1c) and total cholesterol were examined to evaluate biochemical stress markers reflecting the preceding 1 to 2 months. The results on admission were compared with those at the follow-up visit.

RESULTS

Significantly more patients reported greater physical exhaustion, greater mental exhaustion, or a worse physical condition on admission than at follow-up (p<0.01, for each variable). On admission, 81% of patients reported greater than normal stress with regard to at least 1 of 3 items. The mean serum HbA1c was slightly but nonsignificantly lower at the follow-up visit (p=0.10), while the mean serum total cholesterol was significantly lower at follow-up than on admission (p<0.01).

CONCLUSION

The results indicate that patients were under a greater degree of stress before the onset of ISSHL, suggesting that stress plays a role in inducing ISSHL.

INTRODUCTION

Even though the cause of idiopathic sudden sensorineural hearing loss (ISSHL) has not been identified, several possible etiologies have been proposed, such as a vascular disorder, viral infection, or membrane breaks ^[1–3]. Stress is commonly thought to be related to the onset of ISSHL, but there have been few reports investigating the precise relationship between stress and ISSHL. The National Epidemiological Survey in Japan, conducted between 1971 and 1974, reported that 25% of 2.418 patients with ISSHL noted apparent triggers prior to the hearing loss, such as the common cold, physical exhaustion, or mental exhaustion (Results of Nationwide Epidemiological Surveys on Sudden Deafness (NESSD) reported by the Research Committee on Acute Profound Deafness of the Ministry of Health and Welfare in Japan, 1975). We believe that even more patients likely had an identifiable ISSHL trigger, because 47% of the 2,418 responses to the question about a trigger were “uninvestigated.” It was not possible to determine a trigger because of a lack of clinical information.

The widely used term stress may be ambiguous, but it can be defined as a threat, real or imagined, to the psychological or physiological integrity of an individual ^[4]. People usually have many stressors in their daily lives. The duration of stress can range from a few seconds to several years. Evaluating stress is done by investigating both stressors and a stress response to them. The latter can be evaluated subjectively using a questionnaire or objectively by evaluating the changes of biophysical measurements and biochemical indicators.

We have noted that some patients have developed ISSHL at highly stressful times or immediately thereafter. Thus, we hypothesized that patients with ISSHL experienced more stress prior to the onset of ISSHL than was typical for them. A critical problem in evaluating stress levels before the onset of ISSHL, however, is that the condition itself is highly stressful. Hence physical and psychological stress responses measured after the onset are difficult to separate from conditions leading up to the hearing loss. Many biochemical indicators of stress, such as cortisol, dehydroepiandrosterone, noradrenaline, and adrenaline levels, are influenced by ISSHL. Even when using a subjective questionnaire assessing prior stress levels, the responses could be influenced by the ISSHL itself. We therefore opted to measure longer-term biochemical indicators, hemoglobin A1c (HbA1c) and cholesterol, which are influenced by stress for 1 to 2 months. We also used a simple, original questionnaire about the patient’s physical and mental condition before the onset of ISSHL based on the presumption that the duration of stress influencing the onset of ISSHL would range from 1 week to a few months.

MATERIALS AND METHODS

Diagnostic criteria for ISSHL were as follows: (1) a sudden onset of hearing loss or progressive deterioration within 72 h; (2) hearing loss of 30 dBHL or more, over three consecutive frequencies; and (3) sensorineural hearing loss of unknown etiology. Inclusion criteria for the study were as follows: presentation within 7 days of the onset of ISSHL; being admitted to our hospital; age ≥20 years; no history of diabetes mellitus; and consent to participate in the study.

We investigated patient’s stress levels subjectively with a questionnaire and objectively with biochemical measurements. We compared the results on admission with those at a follow-up visit 3 months or more after the onset of ISSHL. The follow-up visit results were used as the control data.

A total of 97 patients with ISSHL were considered for inclusion in the study, of whom 51 did not consent to participate or did not attend their scheduled follow-up visit. Four patients turned out not to have ISSHL, either during or after treatment. Their actual diagnoses were Meniere’s disease, acoustic tumor, idiopathic bilateral sensorineural hearing loss, and noise-induced hearing loss.

Of the 42 patients in the study (Table 1), 15 were men and 27 were women, with a mean (±standard deviation) age of 55.3±17.0 years (range, 25–85 years). Their initial visit to our hospital was 3.5±1.6 days after the symptom onset.

Table 1.

Sociodemographic characteristics of subjects (n=42)

Gender	15 men/27 women
Age (years)	55.3±17.0 (25–85)
Interval between the onset and first visit to our hospital (days)	3.5±1.6 (1–7)
Arithmetic mean of hearing levels at five frequencies (250–4000 Hz) in the affected ear on admission	77.7±23.0 dB
Arithmetic mean of hearing levels at five frequencies (250–4000 Hz) in the affected ear at the follow-up	36.4±28.3 dB
Treatment before presentation to our hospital	Previously treated, 18 (12 with steroids ) Not treated, 24

The arithmetic mean of hearing levels at five frequencies (250 Hz, 500 Hz, 1,000 Hz, 2,000 Hz, and 4.000 Hz) in the affected ear was 77.7±23.0 dBHL on admission and 36.4±28.3 dB at the follow-up. Before coming to our hospital, 12 patients had been treated elsewhere with steroids, and 6 were treated with adenosine triphosphate and mecobalamin without steroids. The remaining 24 did not receive treatment of any kind before visiting our hospital. The interval between the first day of admission and the follow-up visit ranged from 90 to 388 days, with a mean of 206.3±73.4 days.

The Questionnaire

Patients were instructed on the day of admission to answer three questions regarding physical exhaustion, mental exhaustion, and physical condition. These questions were derived from the NESSD survey. Patients were asked to note their perception of these items in the week before the onset of ISSHL, compared with how they typically experienced them over the previous year (Table 2). At the follow-up visit (≥3 months after the ISSHL onset), we asked them to report on their stress for the 1 week before the visit compared with the preceding year.

Table 2.

Questionnaire regarding stress levels

We want to know about your stress levels before the onset of your hearing problems.
Please answer the following three questions by comparing your condition for 1 week before the onset of your hearing problems and your average condition over the entire past year.
1. Do you think you had greater physical exhaustion?	Yes	No
2. Do you think you had greater mental exhaustion?	Yes	No
3. How was your physical condition?	Normal	Worse

Biochemical Data

Serum HbA1c and total cholesterol (TC) levels were used to evaluate relatively long-term physical reactions to stress. Blood samples were collected on admission and at the follow-up visit when the questionnaire was readministered. We compared the mean values of serum HbA1c and TC on admission with those at the follow-up and investigated a relationship between response to the questionnaire and the biochemical data. We also assessed whether the steroid administered to the 12 patients prior to presentation at our hospital influenced serum HbA1c and TC levels on admission.

Statistical Analysis

McNemar’s chi-squared test was used to compare the ratio of patients with greater stress between on admission and at the follow-up. A paired t-test was used to compare the values of biochemical indicators. The Statistical Package for the Social Science version 15 (SPSS Inc., Chicago, IL, USA) program was used for statistical evaluation. A p<0.05 was considered statistically significant.

This study was conducted between January 2011 and December 2013, and it was approved by and performed in accordance with the ethical standards of the Ethics Committee of Kitasato University, School of Medicine. All patients provided written informed consent to participate in this study. All steps of this study were planned and conducted according to the principles outlined in the Declaration of Helsinki (2008).

RESULTS

Questionnaire

The results of the questionnaire are shown in Table 3. Of the 42 study patients, 26 (62%) reported greater physical exhaustion on admission, and 11 (26%) reported greater physical exhaustion at the follow-up visit. The number of patients with greater physical exhaustion on admission was significantly higher than that at the follow-up visit (McNemar’s chi-squared test, p=0.01).

Table 3.

Results of the questionnaire (n=42)

	On Admission	At the Follow-Up Visit	McNemar’s Chi-Square Test
Greater physical exhaustion	26/42 (61.9%)	11/42 (26.2%)	p<0.01
Greater mental exhaustion	27/42 (64.3%)	13/42 (31.0%)	p<0.01
Worse physical condition	22/42 (52.4%)	9/42 (21.4%)	p<0.01
At least one of the three items	34/42 (81.0%)	19/42 (45.2%)	p<0.01

Greater mental exhaustion was reported by 27 (64%) on admission and by 13 (31%) at the follow-up visit. Mental exhaustion was thus reported by a significantly greater number of patients on admission than at the follow-up visit (McNemar’s chi-square test, p<0.01).

A worse physical condition was reported by 22 patients (52%) on admission compared with 9 (21%) at the follow-up visit. Again, a significantly greater number of patients gave this response on admission than at the follow-up visit (McNemar’s chi-square test, p<0.01).

On admission, 81% of patients reported greater stress for at least one of the three items.

Biochemical Data

The results of biochemical testing are shown in Table 4. The mean serum HbA1c level was 5.44%±0.34% on admission and 5.38%±0.35% at the follow-up, a slight but not significant difference (paired t-test, p=0.10).

Table 4.

Biochemical data (n=42)

	On Admission	At the Follow-Up Visit	Paired t-test
Serum HbA1c (%)	5.44±0.34	5.38±0.35	p=0.10
Serum TC (mg/dL)	211.5±41.9	197.4±31.3	p<0.01

Data expressed as the mean±standard deviation. TC: total cholesterol.

The mean serum TC level was significantly higher on admission than at the follow-up visit (211.5±41.9 mg/dL vs. 197.4±31.3 mg/dL; paired t-test, p<0.01).

Relationship between Responses to the Questionnaire and Biochemical Data

Among patients denying physical exhaustion at both time points, serum TC values were significantly lower at the follow-up compared with those on admission (paired t-test, p=0.04) (Table 5). A similar significant decrease in serum TC levels was seen among patients who denied mental exhaustion at both time points (paired t-test, p=0.04) (Table 6). For those who reported mental exhaustion on admission but not at the follow-up, there was a nonsignificant decrease in mean serum TC values (paired t-test, p=0.06).

Table 5.

Relationship between the questionnaire and biochemical data (n=42)“Physical exhaustion”

		Serum HbA1c (%)			Serum TC (mg/dL)
		Admission	Follow-Up	Paired t-test	Admission	Follow-Up	Paired t-test
Greater Physical Exhaustion	“Greater” at both visits (n=8)	5.48±0.18	5.41±0.32	p=0.58	218.3±57.9	186.0±21.8	p=0.17
	Admission “greater” Follow-up “no” (n=18)	5.45±0.42	5.39±0.42	p=0.27	204.7±41.7	198.4±33.1	p=0.27
	Admission “no” Follow-up “greater” (n=3)	5.17±0.06	5.17±0.21	p=1	197.3±27.6	190.7±28.0	p=0.06
	“No” at both visits (n=13)	5.47±0.32	5.41±0.32	p=0.22	220.1±35.0	204.5±35.1	p=0.04

Data expressed as the mean±standard deviation. TC: total cholesterol.

Table 6.

Relationship between the questionnaire and biochemical data (n=42) “Mental Exhaustion”

		Serum HbA1c (%)			Serum TC (mg/dL)
		Admission	Follow-Up	Paired t-test	Admission	Follow-Up	Paired t-test
Greater Mental Exhaustion	“Greater” at both visits (n=13)	5.43±0.38	5.36±0.36	p=0.23	205.0±35.0	20.8±27.6	p=0.51
	Admission “greater” Follow-up “no” (n=14)	5.48±0.35	5.40±0.37	p=0.17	222.1±55.5	195.4±38.2	p=0.06
	Admission “no” Follow-up “greater” (n=0)	-	-	-	-	-	-
	“No” at both visits (n=15)	5.41±0.31	5.38±0.34	p=0.64	207.3±32.5	196.3±29.0	p=0.04

Data expressed as the mean±standard deviation. TC: total cholesterol.

For those who reported a normal physical condition at both time points, the mean serum TC values decreased significantly at the follow-up (paired t-test, p=0.03) (Table 7). There was also a significant decrease in the mean serum HbA1c values among those who reported a worse physical condition on admission, but who said it was normal at follow-up (paired t-test, p=0.02).

Table 7.

Relationship between the questionnaire and biochemical data (n=42) “Physical Condition”

		Serum HbA1c (%)			Serum TC (mg/dL)
		Admission	Follow-Up	Paired t-test	Admission	Follow-Up	Paired t-test
Worse Physical Condition	“Normal” at both visits (n=17)	5.38±0.31	5.37±0.33	p=0.92	210.5±30.0	198.4±26.9	p=0.03
	Admission “normal” Follow-up “worse” (n=3)	5.37±0.38	5.37±0.25	p=1	201.0±25.9	186.7±31.9	p=0.27
	Admission “worse” Follow-up “normal” (n=16)	5.50±0.41	5.36±0.42	p=0.02	203.4±46.3	196.6±39.9	p=0.35
	“Worse” at both visits (n=6)	5.50±0.18	5.47±0.30	p=0.66	241.5±58.8	202.0±20.4	p=0.20

Data expressed as the mean±standard deviation. TC: total cholesterol.

Effect of Previous Treatment on Biochemical Data

We investigated serum HbA1c and TC values on admission in the 12 patients treated with steroids before they were seen at our hospital. Their mean serum HbA1c level was 5.48%±0.31%, and the mean serum TC level was 211.1±53.2 mg/dL. These values did not differ significantly from those of the 30 patients not treated with steroids (HbA1c 5.47%±0.36%, p=0.92; TC 215.6±43.3 mg/dL, p=0.76; t-test).

DISCUSSION

The causes of ISSHL remain unknown, but some etiologic hypotheses have been advanced, including circulatory disorders, viral infections, and membrane breaks ^[1–3]. It has been widely accepted that stress contributes to ISSHL, but there are few reports investigating the relationship between stress and ISSHL. This may be due to the difficulty in evaluating stress levels before the onset of ISSHL. Possible stressors include many factors, which can be both external or internal. External stressors may include physical or chemical conditions, such as heat, cold, chemicals, noise, and drugs; biological stress, such as over-exercise, infection, and lack of sleep; or social or mental stress, such as anxiety, fear, mental tension, and human relationships. Internal stress may include illnesses, pain, or bleeding. The extent to which a person experiences stress is determined both by the stressor itself and by the individual’s response to it ^[5].

We hypothesized that patients with ISSHL would have experienced more stress before or at the onset of hearing loss than usual. Neuser et al. ^[6] reported that patients with ISSHL had greater psychological distress and more stressful life events, as measured by the Minnesota Multiphasic Personality Inventory (MMPI) test and the Inventory of Stressful Life-Events (ILE). Fowler ^[7] reported that 90% of patients with ISSHL had mental problems, and over 70% of patients had a psychosomatic disorder.

Merchant et al. ^[8] and Adams et al. ^[9] demonstrated a relationship between ISSHL and stress based on the “stress response hypothesis.” They hypothesized that ISSHL may be the result of the pathologic activation of cellular stress pathways involving the nuclear factor kappa B within the cochlea. They investigated this hypothesis using findings in the temporal bone in patients with ISSHL. They also demonstrated that general exposure to stress could activate this cellular reaction.

In this study, we used our own simple questionnaire to evaluate the degree of exhaustion and physical condition before the onset of ISSHL. There are many questionnaires asking about stressors, the response to stressors, or how one coped with a stressor. For example, the Sense of Coherence, MMPI, and ILE have been widely used ^{[6, 10, 11]}. For this study, we hypothesized that patients with ISSHL would have had more stress prior to the onset of ISSHL than they typically experienced. We felt that the conventional questionnaires could not be adapted to this purpose. In addition, it seemed difficult to exclude the effects of the ISSHL itself on responses to those instruments. Our goal was to assess stress levels prior to the onset of ISSHL, even though hearing loss had already occurred by the time the patients were seen. Therefore, we made the questions very simple, prompting the patient to compare how they perceived the items at specified periods before the onset of ISSHL. We believed this approach would exclude the influence of the stress from the ISSHL per se.

In this study, patients more frequently reported greater mental or physical exhaustion and physical disorders before the onset of ISSHL than at the follow-up visit. The majority of patients (81%) reported at least one of those stressful conditions. This proportion was remarkably high compared with the 25% reported in NESSD. We consider the results of this prospective study to be more plausible than those from NESSD.

Physiologic reactions to stress can be evaluated using biochemical data. Under stressful conditions, neurotransmitters and neuroregulators are released from the limbic system, stimulating the hypothalamus to release corticotropin-releasing hormone, which in turn stimulates the pituitary gland to release adrenocorticotropic hormone, which stimulates the adrenal cortex to release cortisol. Among other actions, cortisol promotes gluconeogenesis in the liver. Blood sugar combines with serum hemoglobin, which is measured as the percentage of the HbA1 proteins that are glycated. HbA1c is the best indicator of blood sugar levels over the preceding several months. The half-life of hemoglobin is approximately 120 days, and the half-life of serum HbA1c is approximately half of that (approximately 1–2 months). The HbA1c value therefore reflects blood sugar levels in the preceding 1–2 months ^[12].

Although cortisol, dehydroepiandrosterone, noradrenaline, and adrenaline are well-known indicators of stress, these values can fluctuate immediately or within hours. Therefore, in patients presenting with ISSHL, levels of these substances were likely to have been influenced by the ISSHL alone. Since HbA1c values reflect the average blood sugar levels over the previous 1–2 months, HbA1c can also be used as a marker of stress levels for the same period ^[12–15]. We considered that the effect of ISSHL alone on the HbA1c would be negligible. We also found that the use of steroids immediately before measurement of HbA1c did not affect the levels. There have been several reports using HbA1c as a stress indicator. Netterstrom et al. ^[13] reported that HbA1c values were higher during the exam week than they were 4 months later in 23 medical students. However, personal lifestyle and genetic factors can also influence the HbA1c levels. Our study compared the HbA1c levels on admission with those at the follow-up visit, allowing each patient to act as their own control.

We found that mean serum HbA1c levels on admission were slightly higher than those at the follow-up, but the difference was not significant. It is important to keep in mind that the p-value was relatively small and that there would have been a possibility of Type 2 error.

We also investigated serum TC levels. The secretion of cortisol is promoted from the adrenal cortex by stressors, as mentioned earlier. Serum-free fatty acid levels increase in response to glucocorticoids, and the liver increases its TC-producing capacity. Therefore, serum TC can also be used to indicate a stress response ^{[4, 15, 16]}. The timing of increases in TC with stress is similar to that of HbA1c. In this study, serum TC values on admission were significantly higher than those at the follow-up. This indicates that patients had experienced a higher than usual level of stress for several months before the onset of ISSHL.

We also investigated the relationship between responses to the questionnaire and the serum HbA1c or TC levels. Overall, the relationship between these factors was not strong. One primary reason for this result was that the questionnaire evaluated stress a week before the onset, whereas the HbA1c and TC levels reflected stress in the 1 to 2 months before the onset, a substantial difference in timing. We conducted a preliminary investigation of our questionnaire before this study, creating two questionnaires asking about differing length of time, one for 1 week and the other for 1 month. The correlations between answers for 1 week and for 1 month were 0.39 for physical exhaustion, 0.80 for mental exhaustion, and 0.61 for physical condition (Spearman’s rank correlation, data not published). The correlation for physical exhaustion was relatively low, while that for mental exhaustion was high. For this investigation, we selected the questionnaire asking about 1 week because we considered that the stress level closer to onset would more likely contribute to ISSHL, and it seemed more appropriate to compare 1 week rather than 1 month with the average of the preceding 1 year. Consequently, the results of the questionnaire reflected a shorter time period than did the biochemical data, particularly for physical exhaustion. However, there was a large difference in TC values for those who reported mental exhaustion on admission but not at the follow-up. This finding might indicate a relatively strong relationship between the subjective perception of mental exhaustion and the objective evidence indicated by the TC level. Interestingly, for the group reporting no mental exhaustion on either admission or at the follow-up, the mean serum TC values were significantly higher on admission than at the follow-up. This suggests that serum TC values detect a physiological response to stress that could not be detected subjectively with the questionnaire.

We should consider the possibility that patients might take steps to avoid stress after developing ISSHL, which could have influenced the results of this investigation. The mean interval between the first examination and the follow-up visit was approximately 6 months. Although we cannot exclude this possibility, we also assume that many patients resumed their usual lifestyle, such that stress levels at the follow-up would be similar to those they were generally used to.

CONCLUSION

The results of this investigation showed that many patients subjectively perceived greater exhaustion or a worse physical condition a week before the onset of ISSHL than it was usual for them. Patients had relatively higher biochemical stress responses for 1 to 2 months before the onset of ISSHL than usual. We therefore conclude that the patients we studied were under a greater degree of stress before the onset of ISSHL, indicating that stress may play an important role in the onset of ISSHL.