A total of 70 consecutive patients with moderately severe-to-profound SSNHL diagnosed at our hospital between July 2018 and December 2020 were prospectively enrolled. All participants provided written informed consent for their inclusion in the database and the use of their data for research purposes. The study protocol was approved and implemented according to the ethical standards of the Institutional Ethics Committee of the Shanghai Jiao Tong University Affiliated Sixth People's Hospital [2018-KY-036(K)]. The progress was conducted in accordance with the spirit of the Helsinki Declaration. Patients included in the study visited the hospital for the first time within 7 days after the onset of moderately severe-to-profound SSNHL. All participants underwent standard laboratory tests and audiological diagnostic procedures. Included patients were those with hearing loss at all frequencies and mean pure tone audiometry (PTA) across 0.25–8 kHz ≥50 dB. Those excluded had acute inflammatory conditions, obstructive sleep apnea, connective tissue diseases, abnormal ear examination findings, a previous history of thyroid disorder, chronic otitis media, a history of acoustic trauma or otologic surgery, conductive hearing loss, or had used ototoxic medications. Patients with malignant disease, psychiatric conditions, dementia, hepatitis B or C, or other major comorbidities (heart failure; stroke; and severe hepatic, pulmonary, or renal dysfunction) were also excluded. A group of 100 sex- and age-matched controls without any disease at regular health check-ups was used for comparison. The exclusion criteria were the same as those for the SSNHL group.

The baseline characteristics included age, sex, height, weight, body mass index (BMI), and blood pressure on admission (systolic blood pressure and diastolic blood pressure). The clinical characteristics included affected side, accompanying symptoms (including tinnitus, vertigo, ear fullness), history of hypertension and diabetes, time to treatment, and hearing level on admission. All hearing assessments were performed in standard shielding rooms and PTA was performed for both air and bone conduction at 0.125, 0.25, 0.5, 1, 2, 4, and 8 kHz before and after 7 days' course of systemic treatment. The hearing loss of each individual was calculated by averaging the PTA value of damaged frequencies after onset. The extent of hearing recovery is calculated using PTA after onset minus PTA after treatment. All patients underwent temporal bone computed tomography or inner ear magnetic resonance imaging, ensuring no ear structural abnormality and tumors were found. Upon admission, thyroid function tests of TT3, TT4, FT3, FT4, and TSH levels were performed using blood samples obtained from the antecubital veins of all patients between 6 and 7 a.m. after an overnight fast. The thyroid function test was performed again post-treatment and its difference (post-minus pre) from the first test was used to evaluate the changes in thyroid-related hormone.

Once SSNHL was diagnosed, patients were hospitalized for 1 week. All the patients underwent comprehensive treatment, including treatment with steroids and batroxobin following the 2015 China guideline for the diagnosis and treatment of sudden deafness (6). The comprehensive treatment consisted of intravenous injection of Prednisone (1 mg/kg/day) for 3–5 days, followed by a reduced dosage for the remaining days according to the hearing improvement, and intravenous batroxobin (10U batroxobin for the first time and then reduced to 5U batroxobin, once every other day, 1–3 times in total according to the level of fibrinogen). Patients that experienced hearing recovery were divided into the effective group (PTA of impaired frequencies which improved more than or equal to 15 dB, or back to normal/unaffected ear) and the ineffective group (PTA of impaired frequencies which improved <15 dB).

Statistical analyses were performed using SPSS for Windows version 22.0 (IBM Corp., Armonk, NY). Data on quantitative variables are presented as mean ± standard deviation and qualitative variables as numbers (percentage). The Chi-squared test was used for categorical variables. The independent samples t-test was used to compare continuous variables. Linear correlation was performed to assess the association between thyroid-related hormone hearing loss and hearing recovery. ROC curve analysis was used to assess the relationship between thyroid-related hormone and the occurrence of moderately severe-to-profound SSNHL. Binary logistic regression models were used to estimate the odds ratios (OR) and 95% confidence intervals (CI) for the correlation between thyroid-related hormone and the occurrence and outcome of moderately severe-to-profound SSNHL. The collinearity of all continuous variables was examined before performing the logistic regression using the variance inflation factor. P < 0.05 was considered significant for all tests. The figures were generated using GraphPad Prism 7.0 for Windows (GraphPad Software Inc., CA).

The baseline characteristics of participants are summarized in Table 1. About half of the patients (n = 38, 54.29%) were male and the mean age of the patients was 51.00 ± 15.89 years. There were no differences in age, sex distribution, height, weight, BMI, and blood pressure on admission between the two groups. From the results, the thyroid-related hormone levels of the moderately severe-to-profound SSNHL group were significantly lower than those of the control group (TT3, TT4, FT3, and TSH; all P < 0.05) as shown in Table 1. There is a known association between hypertension and diabetes with altered thyroid function (23, 24). We further compared the thyroid parameters of moderately severe-to-profound SSNHL with or without hypertension and diabetes. It was found that there were no significant differences in moderately severe-to-profound SSNHL subgroups (all P > 0.05) (as shown in Supplementary Tables 1, 2). This indicated that thyroid parameters between moderately severe-to-profound SSNHL group and control group may not be influenced by the clinical history of diabetes and hypertension.

Patients that experienced hearing recovery were divided into effective group (n = 30) and ineffective group (n = 40). As shown in Table 2, the baseline characteristics of the two groups had no significant differences (all P > 0.05). No significant differences were found between the two groups in terms of the affected side, accompanying symptoms such as tinnitus and ear fullness, and history of hypertension and diabetes (all P > 0.05). However, the number of patients with vertigo in the ineffective group was significantly higher than that in the effective group (P < 0.05). The time to treatment and the hearing loss level of the ineffective group were 4.95 ± 1.54 days and 82.11 ± 12.50 dBHL, respectively, and both were significantly higher than that in the effective group (3.57 ± 1.98 days and 75.43 ± 12.40 dBHL) (P < 0.05). There were no significant differences in thyroid-related hormone levels (all P > 0.05), however, changes in TT3, TT4, FT3, and TSH levels were significantly higher in the effective group than in the ineffective group (all P < 0.05).

We performed a linear correlation analysis and drew the scatterplots of thyroid-related hormone levels vs. severity of hearing loss (Figure 1) and found no association between moderately severe-to-profound SSNHL and TSH when hearing loss was treated as a continuous variable. The ROC curve analysis (Figure 2) revealed that TT3 level ≤ 1.27 nmol/L (sensitivity, 84.00%; specificity, 57.14%), TT4 level ≤ 96.89 nmol/L (sensitivity, 50.001%; specificity, 75.71%), FT3 level ≤ 4.00 pmol/L (sensitivity, 86.00%; specificity, 65.71%), FT4 level ≤ 13.18 pmol/L (sensitivity, 93.00%; specificity, 12.86%), and TSH level ≤ 2.37 mIU/L (sensitivity, 80.00%; specificity, 88.57%) were the most powerful predictors of moderately severe-to-profound SSNHL. The areas under the curve were 0.737 (95% CI, 0.658–0.817), 0.636 (95% CI, 0.553–0.719), 0.837 (95% CI, 0.777–0.897), 0.458 (95% CI, 0.368–0.548), and 0.903 (95% CI, 0.858–0.948) for TT3, TT4, FT3, FT4, and TSH, respectively. FT3 and TSH levels showed good predictive efficacies for the occurrence of moderately severe-to-profound SSNHL. In the univariate logistic regression analysis, the ORs for occurrence outcome of disease with parameters are presented in Table 3. With unadjusted ORs of 0.093 (95% CI, 0.044–0.195, P < 0.05) and 0.245 (95% CI, 0.155–0.389, P < 0.05), FT3 and TSH levels showed a strong association with the occurrence of moderately severe-to-profound SSNHL. After adjusting for all other significant predictors, FT3 and TSH levels remained independent occurrence predictors with adjusted ORs of 0.064 (95% CI, 0.016–0.255, P < 0.05) and 0.270 (95% CI, 0.157–0.462, P < 0.05).

In the linear correlation analysis and scatterplots of the elevation of thyroid-related hormone levels vs. hearing recovery (Figure 3), an association was found between moderately severe-to-profound SSNHL and TSH when hearing loss was treated as a continuous variable. The elevation of TSH level (R = 0.707, P < 0.05) after treatment predicted a favorable outcome of hearing recovery. In the univariate logistic regression analysis, the ORs for the outcome of moderately severe-to-profound SSNHL with parameters are presented in Table 4. We found parameters that include vertigo, time to treatment, hearing level, and the change in TT3, FT3, and TSH levels to show strong associations with the treatment outcome. Given that cardiovascular risk factors such as hypertension and diabetes may be associated with SSNHL, we performed a multivariate logistic regression model adjusted using these two factors along with those that were significant in the univariate analysis. Using the variance inflation factor, all parameters were deemed non-collinear and could be used in further analysis. It was found that the change in TSH level remained the independent outcome predictor with an adjusted OR of 1.456 (95% CI, 1.069–1.983, P < 0.05).

In this study, although still within normal ranges, we found that the levels of thyroid-related hormones were significantly lower in patients with moderately severe-to-profound SSNHL than those of the control group (Table 1). Analyzing the relationship between thyroid-related hormones and SSNHL through ROC curve suggested that the AUC for FT3 diagnosis of SSNHL is 0.837, with a cut-off point at 4.00 pmol/L (sensitivity, 86.00%; specificity, 65.71%). In terms of TSH, diagnosis for SSNHL is 0.903, with a cut-off point at 2.37 mIU/L (sensitivity, 80.00%; specificity, 88.57%). This shows that early detection of FT3 and TSH are valuable for the diagnosis of SSNHL (Figure 2). Lower FT3 and TSH levels in the early stages of moderately severe-to-profound SSNHL were independent predictors of the occurrence of moderately severe-to-profound SSNHL (Table 3).

The mechanisms behind our findings may be understood from several established research. Hypothyroidism is related to the severity of atherosclerosis risk factors (hypertension, hyperlipemia, hyperhomocysteinemia) and may be a risk factor of atherothrombotic myocardial infarction (15). Low triiodothyronine (T3) level also plays a role in vascular diseases, including complications after brain tumor surgery, respiratory failure, and acute cardiovascular events (31–33). Given that the blood supply of the inner ear is mainly dependent on end arterioles, the function of the inner ear is greatly affected by ischemia, and ischemia may lead to hearing loss at all frequencies when thrombosis occurs in the peripheral arterioles (34).

In regard to the association of poor outcomes of moderately severe-to-profound SSNHL with low serum thyroid hormone levels in patients, one possible mechanism is secondary neuronal damage after moderately severe-to-profound SSNHL. Wang et al. (35) reported that a decrease in TSH levels and an increase in the basal metabolic rate may lead to a higher risk of post-stroke fatigue; this could lead to the production of excess reactive oxygen species and free radicals, resulting in neurotoxicity (36). In addition, thyroid-related hormones and their derivatives have a great influence on the repair of injured neurons (37, 38). Sadana et al. (39) reported that T3 can reduce infarct and edema in a focal ischemia model. Therefore, we speculate that patients with hypothyroidism may experience lower neuroprotection and worsened secondary damage, leading to a poor outcome.

Furthermore, a growing body of research found that thyroid-related hormone levels are greatly associated with the secretion of many neurotrophic factors, including nerve growth factors (40, 41). Therefore, we can infer that patients with hypothyroidism may experience inhibition of endogenous neuron repair systems, leading to poorer functional outcomes. The elevation of TSH levels after treatment predicts a favorable outcome of hearing recovery (Figure 3), evident from the greater increase in TT3, TT4, FT3, and TSH levels for the effective group than that in the ineffective group (Table 2), while TSH level remained an independent outcome predictor (Table 4). Tamura et al. (42) found that steroid can not only improve thyroid function but also lead to transient decrease in thyroid hormone and an increase in reverse thyroid hormone. Most patients presented an upward trend of thyroid-related hormone levels after glucocorticoid treatment. This may explain why some patients had thyroid hormone increase after glucocorticoid treatment and thus a better prognosis.

Our study has some limitations. First, being a preliminary study, the repeatability of the results needs to be verified in more prospective studies to determine their stability and effectiveness. Second, although the thyroid hormones were correlated with hypertension and diabetes in previous reports (23, 24), this correlation was not significant in our study. This could be due to the small sample size, where the association between thyroid-related hormone levels and cardiovascular status could not be identified. Third, the study has not proven a causal relationship between moderately severe-to-profound SSNHL and thyroid-related hormone levels, thus a randomized controlled trial is needed for further confirmation.