Population-based study for the comorbidities and associated factors in Ménière’s disease

To date, no study has reported the objective metabolic laboratory findings worldwide or the comorbidities for Ménière’s disease (MD) using a population-based design in Asian populations. The aim of this study was to investigate the comorbidities and associated factors for MD using the Korean National Health Insurance Service database. This retrospective population-based study was conducted using a data from the National Sample Cohort database from 2009 to 2015. We only enrolled patients whose records showed a prescription for MD medicine and audiometry findings as well as an appropriate diagnostic code. We also included a matched cohort without MD who were enrolled randomly and matched for sex, age, year of diagnosis, income level, and residential area with the MD group with a ratio of 10:1. We evaluated comorbidities including autoimmune, allergic, metabolic diseases and cancer and the health screening data including general characteristics (height, weight, waist circumference, body mass index, and blood pressure), laboratory findings (fasting glucose, cholesterol, triglyceride, high-density lipoproteintryglyceride (HDL) cholesterol, low-density lipoproteintryglyceride cholesterol, hemoglobin, creatinine, aspartate aminotransferase and alanine aminotransferase, and gamma-glutamyltrans- peptidase (rGT)), and general health behaviors (smoking, alcohol, and exercise) of the MD group, and compared these characteristics with those of the MD-free control group. A total of 2,013 and 20,130 participants were included in the MD and MD-free control groups (1,640 and 15,458 for health screening data). We found the increase in incidence of allergic rhinitis and allergic asthma, decrease in systolic blood pressure, HDL cholesterol, and rGT, and less frequent alcohol consumption and less prevalent smoking in the MD group. No significant differences were observed between the groups in the incidence of autoimmune diseases, and cerebro- and cardiovascular disease as well as health screening data and objective laboratory findings. Inconsistence with published studies, the results of this study suggest that the autoimmunity and metabolic disorder, and skeletal growth might not be associated with the onset of MD. Another well-designed study for other races will be needed to the generalization of this study results.

Health screening data. Among the general characteristics studied, we found no significant differences in height, weight, body mass index (BMI), and waist circumference between the patients with MD and controls. Systolic BP was significantly lower in the patients with MD. Laboratory findings showed that HDL cholesterol and rGT levels significantly decreased in the patients with MD. We found no significant differences in fasting glucose, total cholesterol, triglyceride, LDL cholesterol, hemoglobin, creatinine, AST, and ALT levels between the patients with MD and controls. Of the general health behaviors, alcohol intake and smoking history were less frequent in the patients with MD, whereas there were no differences in exercise frequency (Table 4).

Discussion
In this large population-based study, we investigated comorbidities including autoimmune, allergic and metabolic diseases, cancer, as well as health screening data (including general characteristics, laboratory findings, and general health behaviors of patients with MD), and compared these characteristics to those of matched MD-free controls.
We found an increase in the incidence of allergic rhinitis and allergic asthma in patients with MD, similar to that in previous studies that reported a higher number of patients with a history of allergy and the prevalence of allergic diseases, a higher positive rate of the allergic skin test, and elevated Immunoglobulin E levels in patients  3,[24][25][26] . Randomized controlled studies regarding the treatment of allergy and its association with MD attacks would be helpful to achieve confirmative conclusions. We further found less frequent alcohol consumption and less prevalent smoking in patients with MD, consistent with the findings of previous studies that reported a lower OR in current smokers and alcohol drinkers 10 and low prevalence of smoking and alcohol consumption among MD cases 3 . For alcohol consumption in patients with MD, there are positive aspects, such as alcohol consumption has inhibitory effects on hypothalamic vasopressin production, leading to an increase of diuresis and reduction of endolymphatic pressure 27 . This might be a possible cause for less frequent alcohol consumption in the MD population. However, there is a negative aspect that alcohol has toxic effects on the cochlear and labyrinth 28 , therefore, clinicians should not guide their patients to consume more alcohol before more confirmatory studies are conducted. There are no studies that reported a direct association between smoking and MD; however, there is a retrospective study that reported that smoking cessation contributes to the prevention of new peripheral vestibular disorders among males 29 . Further, smoking is known to cause vasoconstriction and a decrease in blood flow. Since alcohol consumption and smoking have a positive correlation 30 , alcohol consumption might be a confounder in the association between smoking and MD.
Moreover, no significant differences in the incidence of autoimmune diseases and cerebro-and cardiovascular disease were observed between patients with MD and controls. In health screening data, body weight, waist circumference, BMI, and objective laboratory findings (fasting glucose, cholesterol, triglyceride, LDL cholesterol, AST, and ALT levels) were not significantly different between patients with MD and controls. Moreover, systolic BP, HDL cholesterol, and rGT levels were lower in patients with MD.
The exact mechanism of MD remains unclear, although several etiologies including autoimmune, allergic, genetic, traumatic, or infectious (viral) conditions have been proposed 1 . The association between autoimmunity and MD has been supported by the high prevalence of systemic autoimmune diseases in patients with MD 10,13-15 . The relationship between metabolic disorder and MD has also been suggested. Previous studies have reported that DM is associated with severe hearing loss and frequent vertigo in cases of MD 9 and higher BMI and systolic BP are related to MD 10 . Therefore, the present study results are not consistent with those of previous studies.
The primary disadvantage of administrative database-based research is the risk of inaccurate cohort identification, which can lead to selection bias. Therefore, definitions of specific diseases determine the quality of the population-based study 31 . Previous studies used the definition of diseases only based on diagnostic code, not including specific tests or treatments, used self-reported diagnosis, or did not match the control cohort with patient cohort 10,[13][14][15] . In this study, MD was defined not only based on diagnostic code but also based on the audiometry findings and medications 2 . Moreover, to select accurate comorbidities, we applied a registration program code that requires many tests or prescription history or used prescription history, refers to the previous studies. Therefore, selection bias might be the reason for the heterogeneity between the results of previous studies and this study. However, elevated levels of serum autoantibodies 16,17 , cytokines, and chemokines 12 in patients with MD have also been reported in Italian and Spain studies. Therefore, there is a possibility that the lack of significance in autoimmune diseases might be due to racial characteristics. www.nature.com/scientificreports/ A previous UK study reported that tall stature and high leg length are associated with the prevalence of MD, suggesting that early life environmental exposures may influence skeletal growth and onset of MD 8 . However, inconsistent with the previous study, height was not significantly different between the two groups in this study. This may be due to the self-reported diagnosis of MD in the previous study 8 or racial differences.
In conclusion, this study observed the comorbidities and conditions of patients with MD using a national population database, multiple diagnostic criteria, and stratified matched control group. To the best of our knowledge, this study is the first globally to observe objective laboratory findings and to compare height, weight, BMI, waist circumference, BP, and many comorbidities using population-based design in Asian countries. Some of www.nature.com/scientificreports/ the results are not consistent with those of previous studies; therefore, other well-designed studies for other races will be needed to generalize the results of this study.

Conclusions
We investigated the comorbidities including autoimmune, allergic, metabolic diseases and cancer and the health screening data including general characteristics (height, weight, waist circumference, body mass index, and blood pressure), laboratory findings (fasting glucose, cholesterol, triglyceride, HDL cholesterol, low-density LDL cholesterol, hemoglobin, creatinine, AST, ALT, and rGT), and general health behaviors (smoking, alcohol, and exercise) of patients with MD using a national population database. We found the increase in incidence of allergic rhinitis and allergic asthma, decrease in systolic blood pressure, HDL cholesterol, and rGT, and less frequent alcohol consumption and less prevalent smoking in the MD group. Inconsistence with published studies, the results of this study suggest that the autoimmunity and metabolic disorder, and skeletal growth might not be associated with the onset of MD. In this study, we applied a registration program code, used multiple diagnostic criteria, and stratified matched control group to minimize the selection bias; however, previous studies used the definition of diseases only based on diagnostic code, not including specific tests or treatments, used self-reported diagnosis, or did not match the control cohort with patient cohort. Therefore, selection bias might be the reason for the heterogeneity between the results of previous studies and this study. Another well-designed study for other races will be needed to the generalization of this study results.

Methods
Data source. In South Korea, 97% of the population is covered by national health insurance, and the remaining 3% is covered by the medical aid program. The Korean NHIS includes all claims of both these programs; thus, it can be used as a data source to identify the health status of the whole nation. Using the sub-datasets, we selected a National Sample Cohort (NSC) database (DB), which has been used in previous studies 32,33  Comorbidities based on medical treatment history. The South Korean government has enhanced benefit coverage and established a registration program for four major conditions (cancer, cardiovascular disease, cerebrovascular disease, and rare diseases). The patients must meet the diagnostic criteria related to the imaging, cultivation, genetic, histologic test, or clinical diagnosis, which is applied differently to each condition, to be registered in this program. In this study, comorbidity data were extracted using the V code from the registration program or prescription history combined with ICD-10 diagnosis code (comorbidities that cannot apply V code) to identify patients more accurately with comorbidities. We evaluated four types of comorbidities: (i) autoimmune diseases (rheumatic and non-rheumatic), (ii) allergic diseases, (iii) metabolic diseases, and (iv) cancer. The details of the diagnostic criteria are presented in Table 1. We included the comorbidities which could refer to the definition of disease according to the previous population based Korean studies. We also referred to the comorbidities in the other disorder in previous population based Korean studies 38,39 . Health screening data. All insured Koreans older than 40 years undergo a biannual health checkup supported by the NHIS, and employees older than 20 years are required to undergo health checkups once a year. We used general health screening data between 2009 and 2015 in the NHIS-NSC DB. Regarding general health screening data, weight, height, waist circumference, and blood pressure (BP) were measured. Fasting blood glucose, total cholesterol, triglyceride, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, hemoglobin, serum creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and gamma-glutamyl transferase (rGT) levels were also measured in a fasting state. Information on general health behaviors such as alcohol consumption, smoking, and exercise was obtained using self-report questionnaires 40 .