HR indicates hazard ratio; SSHNL, sudden sensorineural hearing loss.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Kim J, Hong JY, Kim D. Association of Sudden Sensorineural Hearing Loss With Risk of Cardiocerebrovascular Disease: A Study Using Data From the Korea National Health Insurance Service. JAMA Otolaryngol Head Neck Surg. 2018;144(2):129–135. doi:10.1001/jamaoto.2017.2569
Is sudden sensorineural hearing loss associated with cardiocerebrovascular disease?
In this propensity score–matched cohort study using 2002 to 2013 Korea National Health Insurance Service data, sudden sensorineural hearing loss was associated with a significant increased incidence of cardiocerebrovascular disease.
Clinicians may consider the risk of developing cardiocerebrovascular disease when patients receive a diagnosis of sudden sensorineural hearing loss.
The interruption of vascular supply to the cochlea has been proposed as a major etiological factor for sudden sensorineural hearing loss (SSNHL), and several risk factors for cardiocerebrovascular disease (CCVD) are associated with SSNHL, including heavy smoking, alcohol consumption, and thromboembolic events. However, the link between SSNHL and CCVD has not been fully evaluated.
To investigate the association between SSNHL and CCVD.
Design, Setting, and Participants
A retrospective propensity score–matched cohort study was conducted using a nationwide representative sample from the National Sample Cohort 2002 through 2013 data from the Korea National Health Insurance Service. The SSNHL group (n = 154) included certain patients who were diagnosed with SSNHL between January 2003 and December 2005. The comparison group was selected (4 patients for every 1 patient with SSNHL; n = 616) using propensity score matching, according to sociodemographic factors and the year of enrollment. Each patient was monitored until 2013.
Main Outcomes and Measures
Survival analysis, the log-rank test, and Cox proportional hazards regression models were used to calculate the incidence, survival rate, and hazard ratio of CCVD for each group.
Among the 770 patients, 385 (50.0%) were female and 370 (48.1%) were aged between 45 and 64 years. Of the total study population, 66 patients developed CCVD, such as stroke and acute myocardial infarction, during the 11-year follow-up period: 18 patients in the SSNHL group (incidence, 13.5 cases per 1000 person-years) and 48 from the comparison group (incidence, 7.5 cases per 1000 person-years). After adjustment for other factors, the hazard ratio of CCVD during the 11-year follow-up period was 2.18 times (95% CI, 1.20-3.96) greater for patients with SSNHL. An increased risk of stroke was associated with SSNHL (HR, 2.02; 95% CI, 1.16-3.51); however, there was no relation between SSNHL and risk of myocardial infarction (HR, 1.18; 95% CI, 0.25-5.50).
Conclusions and Relevance
This observational study using nationwide data suggests that SSNHL is associated with an increased incidence of CCVD, specifically stroke. Therefore, patient surveillance for signs of CCVD should be considered for patients who receive a diagnosis of SSNHL.
Sudden sensorineural hearing loss (SSNHL) is considered an emerging disorder in the field of neurotology, defined as sensorineural hearing loss of at least 30 dB in 3 sequential frequencies with onset over 3 days or less.1 Although it has a high spontaneous recovery rate (32%-65%), the possible causes and pathogenesis of SSNHL remain unknown.1 For this reason, SSNHL remains one of the most controversial and challenging issues in the field of neurotology. To date, previous studies have proposed 4 underlying possible causes of SSNHL, including vascular disorders, viral or bacterial infections, inner ear membrane rupture, and autoimmune diseases.1-4 However, there is no conclusive evidence for any of these hypotheses.
Among the possible causes, vascular involvement in the pathogenesis of SSNHL has been favored during the past few decades; the abrupt onset of this disorder, like acute myocardial infarction (AMI) and cerebral stroke, may correlate with a vascular event, such as a minor infarction of the cochlea. Furthermore, in agreement with the association between SSNHL and cardiovascular risk factors, various studies have indicated that thrombophilic vascular dysfunction may be an important factor in the pathogenesis of SSNHL.5-7 For example, atherosclerosis is a well-established risk factor for vascular occlusion of large arteries in coronary, cerebral, and peripheral regions; it is conceivable that it may play a similar role in the impairment of cochlear perfusion within terminal capillary beds. Quiz Ref IDIn addition, a recent study that analyzed Taiwan National Health Insurance data found that SSNHL was associated with a significant increase in the risk of stroke during the 5 years of follow-up after diagnosis. This finding suggests that SSNHL may be an early indicator of stroke.8 Furthermore, another nationwide cohort study found that patients with stroke had a higher risk of subsequent SSNHL.9 It is clear from these studies that there is a link between SSNHL and cerebrovascular disease; however, the relationship between SSNHL and cardiocerebrovascular disease (CCVD) still has yet to be fully elucidated.
To further investigate the relationship between SSNHL and CCVD, we investigated the association between SSNHL and the prospective risk of stroke and AMI, using a nationwide representative sample from the National Sample Cohort 2002 to 2013 data provided by the Korea National Health Insurance Service (KNHIS-NSC) in South Korea. This nationwide population-based data set allowed us to trace all medical service utilization history for more than 1 million South Koreans and provided a unique opportunity to examine the association between SSNHL and the risk of CCVD, while adjusting for clinical and demographic factors. From these data, we were able to identify a significant link between SSNHL and CCVD. This finding has important implications for clinicians, and supports preventative measures for CCVD in patients who receive a diagnosis of SSNHL.
This study was reviewed and approved by the Institutional Review Board of Hallym Medical University Chuncheon Sacred Hospital (Chuncheon, Korea). Written informed consent was waived for the present study because it used deidentified database data. All study conduct adhered to the tenets of the Declaration of Helsinki. This study used KNHIS-NSC data (NHIS-2016-2-208) created by the National Health Insurance Service (NHIS).
South Korea has maintained a nationwide health insurance system since 1963 under the Korean National Health Insurance Service (KNHIS), and nearly all of the data in the health system are centralized in large databases. The KNHIS controls all medical costs among beneficiaries, medical facilities, and the government. Almost all medical data, including diagnostic codes, procedures, prescription drugs, and personal information, are included in the KNHIS database. In the KNHIS, the Korean Classification of Diseases (KCD), which is a system similar to the International Classification of Diseases, is used as a system of diagnostic practice codes. No patient’s health care records are duplicated or omitted because all South Korean residents receive a unique identification number at birth. Quiz Ref IDThe present study used the database of KNHIS-NSC 2002 to 2013, which comprised 1 025 340 nationally representative random individuals, accounting for approximately 2.2% of the South Korean population in 2002. Stratified random sampling was performed using 1476 strata, by age (18 groups), sex (2 groups), and income level (41 groups: 40 for health insurance beneficiaries and 1 for medical aid) among the South Korean population of 46 million in 2002.
The SSNHL group included all patients who received inpatient and outpatient care for an initial diagnosis of SSNHL between January 2002 and December 2005. A positive SNNHL diagnosis satisfied all of following conditions: (1) the patient received a diagnosis under KCD code H9120, H9121, H9129, or H810; (2) the patient underwent pure tone audiometry; and (3) the patient was prescribed steroid medication. Patients were excluded under the following criteria: (1) diagnosis of stroke (KCD codes I60-I63) or AMI (KCD code I21) between 2002 and 2005; (2) death as a result of any cause between 2002 and 2005; (3) previous diagnosis of SSNHL prior to 2002. The comparison group was selected (4 patients for every 1 SSNHL patient) using propensity score matching, according to age, sex, residential area, household income, and comorbidities. Finally, 154 eligible patients with SSNHL and 616 patients in the comparison group were enrolled. Each patient was observed until 2013, and occurrence of stroke or AMI recorded.
Details of patients’ age, sex, residence, household income, and comorbidities were obtained from the database. To evaluate the risk of stroke or AMI by SSNHL according to age, the study population was divided into 3 age groups (<45, 45-64, >64 years). We also assessed the effects of household income, according to 3 income groups (low, 0%-30%; middle, 30%-70%; high, 70%-100% of the median), and residential area, according to 3 groups based on home residence population density (Seoul, the largest metropolitan region in South Korea; second area, other metropolitan cities in South Korea; third area, small cities and rural areas). We analyzed comorbidities, including hypertension (KCD code I10, corresponding to International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code 401, essential hypertension), diabetes mellitus (KCD code E10-E14, corresponding to ICD-9-CM code 250, diabetes mellitus), and chronic renal failure (KCD code N18, corresponding to ICD-9-CM code 585, chronic kidney disease), which are all known risk factors for stroke and AMI. We defined the presence of comorbidities as any diagnoses of these codes between 2003 and 2005, prior to the diagnosis of stroke or AMI. The operational definitions of study end points were all-cause mortality, or the incidence of stroke or AMI. Patients who had no events and were alive on December 31, 2013, were censored after this time point. The SSNHL and comparison group risks of stroke or AMI were compared as person-years at risk, which were defined as the duration between either the date of SSNHL diagnosis or January 1, 2003 (for the comparison group), and the patient’s respective end point.
Incidence rates per 1000 person-years for stroke and AMI were obtained by dividing the number of patients who had incidents of AMI or stroke by person-years at risk. The overall stroke- or AMI-free survival rate was determined using the Kaplan-Meier survival curve for the observation period. For identifying whether SSNHL increased the risk of occurrence of stroke or AMI, we calculated the hazard ratio (HR) and 95% confidence interval, adjusted for the other predictor variables, using Cox proportional hazard regression analyses. All statistical analyses were performed using R, version 3.3.1 (R Foundation for Statistical Computing), based on a significance level of .05.
There were no significant differences between SSNHL and comparison groups according to sex, age, residential area, household income, or comorbidities, which indicates that group matching was performed appropriately. Details of the study population and group characteristics are presented in Table 1. The Kaplan-Meier survival curves with log-rank tests for the 11-year follow-up are presented in the Figure. The log-rank tests indicated that the patients with SSNHL developed CCVD more frequently than the comparison group (HR, 2.18; 95% CI, 1.20-3.96) (Figure, A). On subgroup analysis, the SSNHL group had significantly lower 11-year stroke-free survival rates compared with the comparison group (unadjusted HR, 2.01; 95% CI, 1.12-3.64) (Figure, B), whereas there was no significant difference in the 11-year AMI-free survival rates between the 2 groups (unadjusted HR, 1.07; 95% CI, 0.23-5.05) (Figure, C).
In the present study, we examined a total of 7750.5 person-years, including 6421.2 person-years in the comparison group and 1329.3 person-years in the SSNHL group. Our results showed that the incidence of CCVD was 13.5 cases per 1000 person-years for the SSNHL group, compared with 7.5 cases per 1000 person-years for the comparison group (unadjusted HR, 1.84; 95% CI, 1.06-3.18) (Table 2). Therefore, the SSNHL group showed an increased incidence of CCVD of 6 cases per 1000 person-years relative to the comparison group.
We analyzed the HR for development of CCVD during the 11-year follow-up period using univariate and multivariate Cox regression models; these findings are presented in Table 2. Quiz Ref IDAfter adjustment for sociodemographic factors (sex, age, residential area, household income) and comorbidities, we observed that SSNHL had a significant association with the prospective development of CCVD (HR, 2.18; 95% CI, 1.20-3.96). In addition, we found that increasing age was significantly associated with the prospective development of CCVD (45-64 years: HR, 7.00; 95% CI, 2.09-23.49; and >64 years: HR, 11.12; 95% CI, 2.96-41.76). Patient comorbidities were also significantly related to the prospective development of CCVD (HR, 4.08; 95% CI, 2.26-7.37). No other factors significantly influenced the development of CCVD.
We further examined the association between SSNHL and stroke or AMI development. Details of the adjusted HRs for stroke by cohort, based on multivariate Cox regression analysis, are provided in Table 3. After adjustment for the other variables, the HR of stroke during the 11-year follow-up period was 2.02 times (95% CI, 1.16-3.51) greater for the patients with SSNHL, compared with those of the comparison group. In addition, we found that increasing age and comorbidities were significantly associated with the development of stroke (Table 3). However, there was no significant relationship between AMI and all variables, including SSNHL (Table 4).
Sudden sensorineural hearing loss is an acute dysfunction of the inner ear that affects men and women approximately equally, with a peak age-related incidence occurring between 50 and 60 years.10 Although the exact mechanism of SSNHL development remains unclear, interruption of the vascular supply to the cochlea is thought to be a contributing factor.4,11-13Quiz Ref ID The inner ear’s blood supply relies on the labyrinthine artery alone, with the absence of collateral arterial blood flow; this vasculature, and therefore cochlea function, could be highly vulnerable to ischemic events.13 In support of this hypothesis, 1 meta-analysis reported that patients with SSNHL presented with magnetic resonance imaging findings similar to those of patients undergoing acute ischemic events, such as stroke or AMI.4 Therefore, it would be prudent for physicians treating patients with SSNHL to be aware of this association, particularly considering the high mortality rates of these diseases.
To date, several nationwide representative sample studies have been published using data from the KNHIS-NSC.14-17 In the present longitudinal study, we examined the association between SSNHL and CCVD in 770 sociodemographically matched individuals extracted from a nationwide 11-year longitudinal cohort database of 1 025 340 South Korean patients. Interestingly, we found that patients with SSNHL had a significant increased risk of CCVD incidence, with an HR of 2.18 after adjustment for sociodemographic factors and the presence of comorbidities.
Interestingly, we found that SSNHL resulted in a significant increased risk of the occurrence of stroke. Previous studies investigating the link between SSNHL and stroke have produced conflicting results. For example, 2 cohort studies in Taiwan investigated whether SSNHL was a risk factor for the development of stroke.8,18 One cohort study consisted of a study population obtained from the Taiwan National Health Insurance Research Database, and the other one was a hospital-based retrospective cohort study. Both studies included a 5-year follow-up period. The national cohort study found a 1.6-times greater risk of stroke in patients with SSNHL, compared with the comparison group. However, there was no direct evidence to support the same hypothesis from the hospital-based retrospective cohort study. The results of this clinical study also contradict our finding that SSNHL significantly increased the risk of stroke incidence (HR, 2.02; 95% CI, 1.16-3.51); possibly, such shorter follow-up periods or insufficiently representative samples generated from a single center are not suitable for investigating this association.
Our cohort also demonstrated that patients with SSNHL had a 1.18-times (95% CI, 0.25-5.50) increased HR for AMI during the 11-year follow-up period; however, this HR was not significantly higher compared with the comparison group. Quiz Ref IDPreviously, 2 cohort studies with 5-year follow-up periods have reported a significant association between AMI and SSNHL using data from the Taiwan National Health Insurance Research Database.19,20 One cohort study found that the odds ratio for SSNHL in patients with AMI was 1.50 times that of the comparison group.19 Another study showed that patients with newly diagnosed SSNHL had a 1.39-fold higher incidence of AMI, compared with patients without a diagnosis of SSNHL.20
This study design based on a national health database has several unique strengths. First, we used a large national population-based database, which enables us to effectively analyze all stroke or AMI incidents. Second, our cohort had a relatively long follow-up period (11 years). Third, the inclusion criteria of this study are based on an established diagnostic code, with the additional requirements of pure tone examination for diagnosis and the same treatment regimen (steroid prescription). Because the KNHIS-NSC database contains all information related to SSNHL diagnosis, including examinations, procedures, and medications, we considered all of the enrolled patients in this study to have been treated appropriately. For this reason, our findings suggest that SSNHL increases the risk of CCVD, regardless of patients receiving proper treatment for SSNHL. Finally, a prior study for validation of KNHIS-NSC data found similar prevalence of 20 major diseases for each year; therefore, the reliability of the KNHIS-NSC data was defined as “fair to good.”17
However, this database has some notable limitations. First, we were unable access other health data, including body mass index, lipid profiles, and information regarding behavioral risk factors, such as smoking or alcohol consumption. Therefore, these possibly confounding factors could not be controlled in this study. Second, the diagnosis of SSNHL was dependent on the KCD diagnostic code, which might be less accurate than those obtained from a medical record that, for example, included medical history, imaging, or audiometry results. Finally, the mortality-related data were lacking in our registry, so we could not investigate whether SSNHL leads to an increase in mortality due to the increased risk of developing CCVD. Thus, our results should be interpreted as an association rather than a cause-specific mortality. Future clinical studies, which investigate a wider range of factors and diagnostic criteria, will provide additional evidence for the link between SSNHL and CCVD.
The present study investigated a possible link between SSNHL and the prospective development of CCVD. Interestingly, we observed that patients with SSNHL had a higher risk of CCVD during the 11-year follow-up period. This finding suggests that SSNHL may be a risk factor for the development of CCVD; therefore, clinicians should consider patients with SSNHL to be at an increased risk of developing CCVD, and take specific precautions to reduce their risk of stroke.
Corresponding Author: Dong-Kyu Kim, MD, PhD, Department of Otorhinolaryngology–Head and Neck Surgery, Chuncheon Sacred Heart Hospital, Hallym University College of Medicine (200-704), 77, Sakju-ro, Chuncheon-si, Gangwon-do, Republic of Korea (email@example.com).
Accepted for Publication: October 3, 2017.
Published Online: December 21, 2017. doi:10.1001/jamaoto.2017.2569
Author Contributions: Dr D.-K. Kim had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Hong, D.-K. Kim.
Drafting of the manuscript: D.-K. Kim.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hong.
Obtained funding: D.-K. Kim.
Administrative, technical, or material support: D.-K. Kim.
Study supervision: J.-Y. Kim, D.-K. Kim.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This study was supported by the Basic Science Research Program through the National Research Foundation of Korea funded by the Korean Ministry of Science, ICT and Future Planning (NRF-2017M3A9E8033231).
Role of the Funder/Sponsor: The National Research Foundation of Korea had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: