Increased risk of sudden sensorineural hearing loss in patients with hepatitis virus infection

The etiology of sudden sensorineural hearing loss (SSNHL) remains unclear. Possible causes of SSNHL include vascular diseases, viral infection, and autoimmune disorders. Therefore, we investigated whether hepatitis virus infection is correlated with the risk of SSNHL. Using data from the Taiwan Longitudinal Health Insurance Database, we conducted a retrospective matched-cohort study to compare patients diagnosed with hepatitis B or C virus (HBV/HCV) infections from January 1, 2000, to December 31, 2010, (N = 170,942) with frequency-matched controls (N = 512,826) at a ratio of 1:3 by sex, age, and index year. We followed each patient until the end of 2010 and evaluated the incidence of SSNHL. At the end of the follow-up period, 647 (0.38%, 647/170,942) patients developed SSNHL in the HBV/HCV group compared with 978 (0.19%, 978/512,826) in the control groups, with a statistical significance of P < 0.001 (using the log-rank test). The incidence rate ratio of SSNHL was 5.743-fold higher in the HBV/HCV group than in the control group (283.17 vs. 49.31 per 100,000 person-years, P < 0.001). The risk of SSNHL increased with HBV/HCV infection, and an adjusted hazard ratio of 5.103 (95% CI, 4.585–5.678) was determined using Cox proportional hazards regression. This study contributes to the awareness of the increased risk of SSNHL in HBV/HCV-infected populations. Our findings suggest that an underlying viral infection contributes to the development of SSNHL.


Introduction
Sudden sensorineural hearing loss (SSNHL) is often an unexpected and traumatic experience for patients. SSNHL is defined as a sudden loss of more than 30 dB of hearing acuity in three contiguous frequencies within 72 h [1]. SSNHL is usually classified as idiopathic because the causative factor is not identified in most cases [2]. The potential causes of SSNHL include vascular disorders, viral infection, autoimmune disorders, neurological disorders, neoplasms, and ototoxic drugs [3]. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Several studies have discussed the presence of viral infection in patients with SSNHL [4][5][6], and the types of viruses implicated in such cases include mumps, cytomegalovirus, measles, rubella, varicella zoster virus, herpes simplex virus, enteroviruses, and human immunodeficiency virus [5][6][7]. Hepatitis virus infection has been correlated with SSNHL in only a small number of case reports [8][9][10][11]. Hepatitis B virus (HBV) and hepatitis C virus (HCV) infections have emerged as a global health problem, and as many as 350 million and 123 million individuals have been infected with HBV and HCV, respectively, worldwide [12,13]. In highly endemic areas such as Taiwan, the prevalence of chronic HBV infection is estimated to be between 15% and 20% [14,15], whereas the prevalence of HCV is estimated at 4%-10% in the general population [16,17]. Chronic HBV or HCV infections have been proven to increase the risk of hepatocellular carcinoma [18]. We are interested in investigating whether HBV or HCV infections have an impact on the incidence of SSNHL.
To assess the association between HBV or HCV infection and the subsequent development of SSNHL, we conducted a nationwide population-based cohort study by analyzing data from a nationwide medical database, the National Health Insurance Research Database.

Data sources
In this study, we used data from the National Health Insurance Research Database (NHIRD) to investigate the association between hepatitis virus infections, including HBV and HCV infection, and SSNHL over a 10-year period (2000-2010), from the outpatient and hospitalization Longitudinal Health Insurance Database in Taiwan. In addition, all patients aged <20 years were also excluded. A total of 683,768 participants were enrolled in this study; among these, 170,942 had HBV or HCV infection and 512,826 were controls without HBV or HCV infection who were matched by age, sex, and index year (Fig 1).

Kaplan-Meier model for cumulative risk of SSNHL in HBV/HCV infection
The cumulative incidence curve for SSNHL in the total HBV/HCV infection cohort was significantly higher than that for the comparison cohort, after adjustment for age, sex, and comorbidities (Fig 2, log-rank test; P < 0.001). In the patients with HBV/HCV infections, the risk of SSNHL increased progressively with the duration of follow-up, rather than with being limited to the immediate days after a diagnosis of HBV/HCV infection. In the subgroups of HBV and HCV infection, the same trend was independently revealed (Fig 2).
HRs and incidence of SSNHL stratified by sex, age group, and comorbidities, by using Cox regression for patients with HBV/HCV infection compared with those without HBV/HCV infection Cox proportional hazards regressions analysis revealed an increased risk of SSNHL, with an adjusted HR of 5.103 (95% CI, 4.585-5.678), in patients with HBV/HCV infection, after adjustment for sex, age, and comorbidities ( Table 2). Adjusted HRs of 4.825 (95% CI, 4.216-5.521) and 5.582 (95% CI, 4.686-6.651) for developing SSNHL were observed in male and female patients with HBV/HCV infection, respectively, compared with the control group (P < 0.001). Among all age groups, a higher adjusted HR for SSNHL was observed in the HBV/HCV group than in the control group. Regardless of the presence of the comorbidities of DM; hypertension; depression; stroke; CKD; osteoporosis; nephritis, nephrotic syndrome, and nephrosis; and hyperlipidemia, a higher adjusted HR (ranging from 2.356 to 15.114) was observed in patients with HBV/HCV infection than in the controls (P < 0.001).

Discussion
To the best of our knowledge, this is the first large-scale retrospective matched-cohort study to explore the association between SSNHL and HBV/HCV infection. The main finding of this study was that patients who were diagnosed as having HBV/HCV infections between January 1, 2000, and December 31, 2010, had a significantly higher incidence of SSNHL, with an IRR of 5.743 (P < 0.001), than did the general population without HBV/HCV infections. The prevalence of HBV/HCV infections varies from <0.5% in Western countries to 8%-25% in endemic countries in East Asia [19]. In our study, the prevalence of HBV/HCV infections was 17.4%. Compared with females, males were predominantly infected with HBV/HCV (110,150 males vs. 60,792 females). The age at diagnosis of HBV/HCV infection was usually >40 years, which was compatible with a previous report describing a mean age at diagnosis of 42 years; 60% of these were male patients [20]. In addition, our study demonstrated that the development of SSNHL was most common in patients aged !40 years, and that most patients with development of SSNHL were men (395 males vs. 252 females).
At the end of follow-up, all comorbidities were significantly differentially distributed between the HBV/HCV and control groups (Table 1). In the HVB/HCV group, a lower rate of the participants with individual comorbidity was observed for most comorbidities, whereas for depression and SLE, a higher rate of the participants was observed compared with the control group. A higher frequency of some degree of depression was reported in both hepatitis B and C patients [21]. However, our finding of a higher percentage of SLE in patients with HBV/ HCV infection was inconsistent with previous results [22]. Additional in-depth investigations are warranted to explore the codependence among these comorbidities and HBV/HCV infections through multivariate analyses.
Our data revealed that the incidence of SSNHL was approximately two-fold higher in the HBV/HCV group than in the control group, and that this difference was significant. The general estimated incidence of SSNHL varies from 5 to 20 cases per 100,000 per year [1]. In our study, the annual incidence of SSNHL was approximately 38 cases per 100,000 in the HBV/ HCV group compared with 19 cases per 100,000 in the control group. The incidence noted in our study was relatively high because the use of a nationwide and large-scale survey may have increased the number of patients treated for SSNHL in outpatient and inpatient departments.

Fig 2. Kaplan-Meier curves for the cumulative risk of SSNHL among patients aged !20 years with HBV/HCV infection (stratified by HBV or HCV alone, using the log-rank test).
https://doi.org/10.1371/journal.pone.0175266.g002 In a previous report, the annual SSNHL incidence in Taiwan ranged between 6.49 and 10.21 per 100,000; this number included only the SSNHL patients with hospital admission [23].

Risk of sudden hearing loss with hepatitis virus infection
Despite the association between HBV/HCV infection and several risk factors for SSNHL, the risk of SSNHL remained significantly higher in the HBV/HCV infection cohort, after adjustment for sex, age group, and comorbidities. The association between HBV/HCV infections and SSNHL may be because of shared risk factors. However, we can confidently claim that the increased risk of SSNHL in these patients was likely the effect of HBV/HCV infection, because the possible confounding factors for SSNHL were already substantially adjusted for in this study. Even in the subgroups of HBV or HCV infection alone, a significantly increased risk of SSNHL (with a higher adjusted HR) was observed compared with the control group. These data indicate that hepatitis virus infection has a very strong impact on the risk of SSNHL.
The underlying mechanisms linking HBV/HCV infections with SSNHL development remain unclear. We hypothesized that, in patients with HBV/BCV infections, SSNHL could occur due to an acute exacerbation of viral hepatitis and subsequent SNHL [8] or a chronic viral reaction causing chronic hearing loss [24,25]. Viruses could gain access to the inner ear via the hematogenous route and induce severe pathophysiologic changes or an immunemediated reaction [8]. HBV/HCV infections can stimulate the production of inflammatory cytokines such as tumor necrosis factor-alpha, interleukin-1, and interleukin-6, which are injurious to the cochlear hair cells [26]. In addition, hepatitis virus infection has a well-documented association with polyarteritis nodosa, which is a life-threatening necrotizing vasculitis that may result in hearing loss [24,27].
Our results demonstrate that a greatly increased risk of SSNHL was observed in patients with HBV/HCV infections, according to the data of a large population (170,942 patients with HBV/ HCV and 512,826 controls) selected from a retrospective matched-cohort comprising 1,000,000 people covered by the National Health Insurance program; the large sample size benefitted the statistical analysis. This large data resource enables us to investigate the risk factors for SSNHL in Taiwan, with an acceptable selection bias and an enhanced statistical precision.
Our research has several shortcomings. First, several potentially confounding risk factors, such as alcohol consumption, smoking, ototoxic drug effects, and noise exposure for SSNHL, were unavailable in the data resource, which may have led to a certain bias [28]. Second, the database did not provide audiometric results regarding the degree of hearing impairment or routine blood and biochemistry tests data. Third, a population-based study cannot clarify the real mechanism underlying the association between HBV/HCV infection and SSNHL because extracting cochlear tissue pathogens or detecting cochlear injury through imaging is very difficult [7,8]. In addition, patients who developed SSNHL due an ototoxic effect after antiviral drug administration for HVB/HCV infection (which has been reported in some studies) could not be excluded [10,11]. However, a case-control study revealed pegylated interferon plus ribavirin therapy does not have any impact on the hearing thresholds of patients with HCV [29]. Finally, we could not exclude the possibility of virus-unrelated hepatitis, such as alcoholrelated hepatitis, being correlated with the risk of SSNHL. Additional large-scale studies need to be performed to clarify the discrepancy between virus-induced SSNHL and hepatitisinduced SSNHL. Despite these limitations, this study contributes to the awareness of the increased risk of SSNHL in HBV/HCV-infected populations.

Conclusions
In this study, HBV/HCV infections present a clearly elevated risk for SSNHL. Regular audiometric tests are recommended for patients with HBV/HCV infection to assess their hearing ability and enable the earlier detection of SSNHL. We also suggest that HBV or HCV carriers presenting with the sudden onset of hearing loss should be examined for the possibility of acute exacerbation of chronic HBV/HCV infection.