Abstract
-
Objectives
- This study was performed to investigate seasonal variation in the incidence of vestibular neuritis (VN) without recent steroid treatments using nationwide health insurance data. The aim of the study is to elucidate whether seasonal trends can inform optimized diagnostic and treatment strategies for VN.
-
Methods
- A retrospective analysis was conducted using data from the Health Insurance Review and Assessment Service (HIRA) in South Korea from 2007 to 2022. Patients diagnosed with VN were identified using specific operational criteria. Seasonal trends were evaluated by analyzing monthly, quarterly, and seasonal variations in VN incidence, stratified by age and sex.
-
Results
- Out of 237,673 VN patients identified, our analysis revealed significant seasonal variations in incidence, with a notable decline during winter months—especially in February—and an increase during the spring. These patterns were consistent across sex and age groups.
-
Conclusions
- This nationwide study demonstrates that VN exhibits distinct seasonal variations that have significant implications for clinical practice. These results indicate a potential influence of seasonal factors on the occurrence of VN and contribute to more efficient allocation of healthcare resources. Future prospective studies are warranted to further elucidate the mechanisms behind these seasonal differences.
-
Keywords: Vestibular neuronitis; Prevalence; Big data; Health insurance; Seasons
INTRODUCTION
Seasonal variation in disease incidence is a well-established phenomenon in medicine. Numerous studies have demonstrated that both infectious and non-infectious conditions exhibit seasonal fluctuations, influenced by environmental factors, climatic conditions, and human behavior [1,2]. For example, viral respiratory infections, allergic disorders, and cardiovascular events often peak during specific times of the year, underscoring the impact of external variables such as temperature, humidity, and air quality on disease dynamics [3].
Vestibular neuritis (VN) is characterized by an onset of vertigo, nausea, and imbalance, typically in the absence of auditory deficits or central neurological signs [4]. Traditionally viewed as a sporadic event, emerging evidence suggests that the incidence of VN may also follow a seasonal pattern. Seasonal trends in VN could be attributed to variations in viral infections, which have been implicated as potential etiological agents, as well as other environmental triggers that exacerbate inflammation of the vestibular nerve [5,6]. Recognizing these patterns is clinically important because it may provide insights into the underlying pathophysiology of VN and guide more targeted therapeutic strategies.
Furthermore, understanding the seasonality of VN is essential for effective public health planning and resource allocation. Anticipating seasonal peaks enables healthcare systems to optimize staffing, manage hospital capacities, and tailor preventive measures to mitigate the impact during high-incidence periods [7,8].
The Health Insurance Review and Assessment Service (HIRA) is South Korea’s health insurance management agency, operating as a national public corporation. South Korea’s National Health Insurance covers approximately 98% of the population and has a wide coverage scope. HIRA’s main tasks include monitoring the contents billed to health insurance. Therefore, the data includes information such as sex, age, diagnosis codes (International Classification of Diseases 10th Revision, ICD-10), medical specialties, costs, surgical procedures, test histories, prescription histories, hospital classifications, inpatient/outpatient status, and institutional locations of individuals who utilize South Korea’s National Health Insurance.
This study utilizes nationwide data from the HIRA to systematically evaluate the seasonal variations in VN. By elucidating these patterns, we aim to provide evidence-based recommendations for optimizing treatment protocols and informing health policy decisions.
METHODS
Ethics Statement
The study was authorized by the Institutional Review Board of Ulsan University Hospital (No. UUH 2023-07-030).
Data Collection and Study Population
This study utilized data from HIRA. Study subjects were included if they had VN codes in their billing records from 2007 to 2022. However, individuals under 20 years old or with a history of steroid prescriptions within 6 months of the diagnosis date were excluded. A total of 2,724,753 patients were included based on these criteria, with additional selection through operational definitions (Fig. 1). Because steroid use is associated with suppression of the immune response and increases the risk of viral infections, which may affect the frequency of VN in our study, we excluded patients who had received steroid treatment within 6 months prior to diagnosis.
The included ICD-10 codes are three in total, with only the code for VN, H81.2, being included. Patients who visited the hospital only once within a month were excluded, as it could indicate misdiagnosis or unclear symptoms. Therefore, only patients with two or more visits within a month were included. The prescription of five types of drugs (antiemetic, benzodiazepines, antihistamines, steroids, and antiplatelet) and the presence of a caloric test were included in the criteria, as they are expected to significantly influence the purity of VN patients. Patients were considered VN patients if they had a prescription history or caloric test history. Ultimately, the study included 237,673 subjects.
Experimental Design
This study aims to (1) assess the current status of diagnosis, treatment, and management of VN in South Korea and (2) identify the seasonal characteristics of VN patients (Fig. 1). For (1), the distribution of patients was examined based on sex, age, drug/treatment, inpatient/outpatient status, cost, and hospital type. Additionally, to identify the seasonal characteristics of VN patients, the distribution of patients was examined according to variables with time-series elements. VN patients were analyzed on a monthly, seasonal, and quarterly basis. Quarters were divided into January–March (Q1), April–June (Q2), July–September (Q3), and October–December (Q4), while seasons were defined as spring (March–May), summer (June–August), autumn (September–November), and winter (December–February). For (2), the analysis was conducted based on sex, age, drug/treatment, inpatient/outpatient status, cost, and hospital type according to steroid prescriptions.
Age was analyzed in two ways: categorized into 5-year intervals (nominal age) and maintained as a continuous variable (continuous age) with 1-year increments for analysis. In particular, the research team has doubts about whether there are more VN patients among females than males. They aimed to confirm this by dividing the analysis into inpatients/outpatients and hospital types, and to understand how this phenomenon affects steroid prescriptions. All variables related to monetary amounts were converted to US dollars ($).
Categorical variables were analyzed using chi-square tests, while continuous variables were analyzed using t-tests. Percentages were reported for categorical variables, and 95% confidence intervals were reported for continuous variables. Every statistical analysis was carried out using SAS ver. 9.4 (SAS Institute Inc.).
RESULTS
Clinical Characteristics of vestibular neuritis from 2007 to 2022
From 2007 to 2022, a total of 237,673 patients diagnosed with VN in South Korea were identified (Table 1).
A total of 237,673 patients were classified according to each factor. The p-value for each factor can be interpreted depending on the variable. Variables such as drug and caloric tests confirm that within the VN patient group, there is a difference in distribution between those who received a prescription for that variable and those who did not. Sex, nominal age, patient type, and types of hospital confirm the difference in distribution according to the detailed classes included in the variables. Continuous variables were compared between the average of the overall population and the VN patient group, confirming that there were differences between each group.
VN patients were approximately twice as common among females compared to males, with the highest number of patients in the 45–74-year age group. Drugs were prescribed in the following order: antihistamines, benzodiazepines, and steroids. About 9% of all VN patients received a caloric test, with outpatient visits being five times more common than inpatient visits. When comparing males and females by hospital type, females were overwhelmingly more represented in hospitals and clinics. In particular, 45% of females received a VN diagnosis at a clinic. The average age was 54.76 years, and out of the total hospital cost of $64.7, the insurance-covered cost was $44.9 (Table 1).
Vestibular neuritis patient analysis for seasonality
While exploring the prevalence of VN patients, a seasonality was discovered. The number of patients was calculated based on overall patients, sex, and age (categorized by 65 years) (Fig. 2). It was represented for all months, but only January and July are shown. A decrease in the number of patients was mainly observed at the beginning of each year, which was consistent regardless of sex and age. The increasing trend of patients appeared to increase until 2011, then the growth rate declined.
To confirm the seasonality, patients were categorized by season and quarter and calculated accordingly. Fig. 3 was presented divided by sex and age. In the Fig. 3A fbased on seasons, about 15 underpicks were found, indicating periodicity. In the Fig. 3B based on quarters, about 10 underpicks were found, and it was relatively difficult to identify periodicity.
Each year was averaged by month/season for simplification (Fig. 4). When examining the distribution by month averaged for each year based on sex and age, a decrease was observed from November to April, with a particularly noticeable decrease in February. Decreases were observed in all factors, with no particular factor showing a greater decrease (Fig. 4). When examining each factor by season, decreases were observed from spring to winter in all factors, with a greater decrease in winter (Fig. 4).
Overall, VN incidence decreases in winter (particularly February) and increases in summer–fall. The average incidence rate of each month compared to the previous month was –5.07% in January, –5.31% in February, 9.76% in March, 0.55% in April, 3.09% in May, –1.78% in June, 2.16% in July, –2.42% in August, 3.69% in September, –0.07% in October, –3.23% in November, and –0.41% in December. There were periodic increases and decreases, but the decreases in January and February and the increases in March were clear, and the differences in the monthly incidence rates were statistically significant (p<0.0001).
DISCUSSION
Our study identified distinct seasonal variations in the incidence of VN, with a noticeable decline in cases during the winter months—particularly in February—and a corresponding rise in the spring [4,5]. This seasonal pattern was consistent across sex and age groups, suggesting that external environmental or infectious factors may influence the onset of VN. Overall, these findings enhance our understanding of VN’s epidemiology and suggest that seasonality could be a key factor in its pathogenesis. It demonstrates clear seasonal variations in the incidence of VN, with a notable decrease in patient numbers from November to April—most prominently in February—and a relative increase in the early months of the year. This seasonal pattern was consistent across sex and age groups, indicating that intrinsic or extrinsic seasonal factors may influence the onset of VN. These findings stand in contrast to a recent retrospective study [8] that evaluated 248 VN patients over a 5-year period, which reported no statistically significant monthly or seasonal differences in VN incidence across various. The discrepancy between our results and those of the prior study may be attributed to several factors. Our analysis is based on a nationwide dataset, encompassing over 237,000 VN patients from 2007 to 2022, thereby offering substantially greater statistical power and generalizability compared to the single-institution study with only 248 patients. The large-scale nature of our data likely enabled us to detect subtle seasonal trends that might be obscured in smaller samples. While the earlier study employed standard statistical tests on a limited dataset, our approach incorporated detailed monthly and seasonal averaging, as well as subgroup analyses by sex and age. This refined methodology provided a more nuanced understanding of the temporal patterns in VN incidence. The nationwide data from South Korea’s HIRA reflects a comprehensive capture of VN cases across diverse healthcare settings. In contrast, the smaller study may have been subject to local practice patterns or referral biases that limited its ability to reveal seasonal variations [9,10]. In addition, we excluded patients who were taking steroids for previous medical conditions to better understand the seasonality of the VN. Additionally, to address discrepancies with previous studies, we compared our findings with those of Koors et al. [7] and Adamec et al. [6], which reported no significant seasonal variation in VN. Koors et al. [7] analyzed 52 cases over 3 years in Virginia, USA, and while a single year (2009–2010) showed seasonal variation, this was not consistent across years or in the overall dataset. Adamec et al. [6] conducted a population-based study in two Croatian cities with 79 cases and found no evidence of seasonality, regardless of age group. Several factors may explain the differing results. First, our study used a nationwide health insurance database covering over 98% of South Korea’s population, resulting in a substantially larger sample size and broader geographical representation, including diverse climates. In contrast, the previous studies were single-center and conducted in limited regions with smaller populations, which may limit their generalizability. Second, diagnostic methodologies differ. Koors et al. [7] relied on clinical diagnoses without mandatory confirmatory testing, and Adamec et al. [6] included caloric tests or vestibular evoked myogenic potentials within 7 days of symptom onset. Our study utilized ICD-10 codes from health insurance claims data, which are based on physician diagnoses made according to national clinical guidelines, adding a level of standardization but lacking granular clinical detail. Lastly, climatic differences could contribute to varying patterns. Korea’s distinct four-season climate with pronounced temperature fluctuations may influence viral activity or environmental factors differently compared to Croatia’s continental climate and Virginia’s subtropical climate. Considering these differences in data scale, study settings, and methodologies, our results suggest that seasonal variation in VN incidence may emerge in large, population-based datasets reflecting broader environmental and climatic conditions, whereas smaller, region-specific studies may not detect such patterns.
The observed seasonal decrease during winter months in our study may have several implications. It could reflect lower rates of viral infections—traditionally implicated in VN pathogenesis—during these periods, or perhaps indicate a shift in the dominant etiological factors, such as a relative increase in vascular or autoimmune contributions during other seasons [11]. Additionally, possible factors such as immune system modulation, increased indoor activity, or reduced hospital visits (e.g., due to holidays like the Lunar New Year) could have influenced the results. This insight is particularly important for clinical practice: understanding these seasonal trends could lead to more targeted therapeutic strategies, such as optimizing the timing of steroid therapy and resource allocation during periods of higher VN incidence.
We support the hypothesis that seasonal factors play a significant role in the incidence of VN. By contrasting our robust nationwide analysis with prior smaller-scale studies that reported non-significant seasonal differences, we underscore the importance of large-scale epidemiological investigations to uncover subtle temporal trends. These insights not only enhance our understanding of VN’s multifactorial etiology but also have practical implications for guiding treatment protocols and public health policy, potentially leading to improved patient outcomes and more efficient use of healthcare resources [12].
The identification of seasonal variations in VN has important clinical implications. Recognizing that VN incidence declines during the winter and increases in the early months of the year could aid clinicians in refining their diagnostic criteria and treatment protocols. For instance, during peak incidence periods, clinicians might adopt a higher index of suspicion for VN in patients presenting with acute vertigo, thereby expediting diagnosis and intervention. Additionally, understanding these seasonal trends may inform the timing and use of steroid therapy, potentially allowing for more targeted and cost-effective treatment strategies that align with the underlying seasonal patterns.
Utilizing data from HIRA represents a significant strength of our study, as it covers approximately 98% of the South Korean population [13]. This comprehensive coverage ensures that our findings are both robust and generalizable across diverse clinical settings. The nationwide scope of HIRA data enables accurate estimation of VN incidence and facilitates the detection of subtle seasonal variations that might be overlooked in smaller, localized studies. Such high-quality administrative data not only underpins our analysis but also provides valuable insights that can be directly applied to public health planning and resource allocation [14].
Despite its strengths, our study has certain limitations. The reliance on administrative data from HIRA may lead to potential misclassification or incomplete capture of clinical details, and the lack of inclusion of additional vestibular tests—such as video head impulse tests not covered by insurance—limits the depth of our clinical correlations [15]. Moreover, while our findings suggest seasonal trends, the observational nature of the study precludes definitive conclusions about causality. There are statistical shortcomings in our study to assess seasonal factors. Future research should focus on prospective, multicenter studies that integrate detailed clinical evaluations with seasonal data, thereby further elucidating the mechanisms behind VN’s seasonal variations and optimizing treatment protocols accordingly. Additionally, it is necessary to secure more reliable testing power by introducing a time-series model. HIRA data covers most of the medical use of Korean citizens. Therefore, the trend of decreasing VN patients in February every year may have been affected by reasons such as hospital business days and holidays due to holidays. However, this study did not reflect the seasonal trend and holiday distribution analyzed by grouping by 3 months as they were not concentrated in February. In follow-up studies, we look forward to upgrading the model and controlling bias more accurately.
Our study demonstrates that VN exhibits significant seasonal variations, with a notable decrease in cases during the winter months and an increase in the spring. Utilizing the comprehensive nationwide HIRA database, which covers nearly 98% of the South Korean population, has allowed us to identify subtle temporal trends that smaller-scale studies may have overlooked. These findings have important clinical implications, suggesting that the timing of both diagnosis and treatment, particularly the use of steroid therapy, could be optimized in alignment with seasonal patterns. Despite the limitations inherent to retrospective analyses and administrative data, our results underscore the multifactorial nature of VN etiology, potentially involving environmental, infectious, and vascular components. Future prospective and multicenter studies are warranted to further elucidate the underlying mechanisms of these seasonal variations and to refine patient management strategies accordingly.
ARTICLE INFORMATION
-
Funding/Support
This work was supported by the Ulsan University Hospital Research Grant (UUH-2023-08).
-
Conflicts of Interest
Ji-Yun Park is the Editor-in-Chief and Young Jun Seo is the Associate Editor of Research in Vestibular Science. They were not involved in the review process of this article. All authors have no other conflicts of interest to declare.
-
Availability of Data and Materials
The data sets generated and/or analyzed during the current study cannot be made public due to the Personal Information Protection Act and Korean regulations governing patient confidentiality. However, the data can be obtained from the corresponding author upon reasonable request and with permission from the Health Insurance Review and Assessment Service (HIRA). Access to these data sets can be requested through this URL (https://opendata.hira.or.kr/).
-
Authors’ Contributions
Conceptualization, Methodology, Validation: All authors; Data curation, Formal analysis, Investigation, Resources, Software, Visualization: CYY; Funding acquisition: JYP; Project administration, Supervision: JYP, YJS; Writing–original draft: CYY; Writing–review and editing: All authors.
All authors read and approved the final manuscript.
Fig. 1.Flow of study and participants selection. HIRA, Health Insurance Review and Assessment Service; VN, vestibular neuritis.
Fig. 2.Monthly vestibular neuritis (VN) patients by sex and age from 2007 to 2022. Arrows indicate periodic underpicks.
Fig. 3.Number of patients with vestibular neuritis (VN) by sex and age from 2007 to 2022. (A) Seasonal VN patients. (B) Quarterly VN patients. Arrows indicate periodic underpicks.
Fig. 4.Simplified monthly and seasonal vestibular neuritis (VN) patients by sex and age. (A) Monthly average VN patients. (B) Seasonal average VN patients.
Table 1.Clinical characteristics of vestibular neuritis from 2007 to 2022
|
Characteristic |
No. of patients |
p-value |
|
Total |
237,673 (100) |
- |
|
Sex |
|
|
|
Male |
81,737 (34.39) |
<0.0001 |
|
Female |
155,936 (65.61) |
|
|
Nominal age (yr) |
|
|
|
20–24 |
4,074 (1.71) |
<0.0001 |
|
25–29 |
6,572 (2.77) |
|
|
30–34 |
9,861 (4.15) |
|
|
35–39 |
14,250 (6.00) |
|
|
40–44 |
17,942 (7.55) |
|
|
45–49 |
22,753 (9.57) |
|
|
50–54 |
27,673 (11.64) |
|
|
55–59 |
29,508 (12.42) |
|
|
60–64 |
28,189 (11.86) |
|
|
65–69 |
26,397 (11.11) |
|
|
70–74 |
22,455 (9.45) |
|
|
75–79 |
16,653 (7.01) |
|
|
80–84 |
8,217 (3.46) |
|
|
85–89 |
2,620 (1.10) |
|
|
90–94 |
465 (0.20) |
|
|
≥95 |
44 (0.02) |
|
|
Drug |
|
|
|
Antiemetic |
14,128 (5.94) |
<0.0001 |
|
Benzodiazepines |
81,667 (34.36) |
<0.0001 |
|
Antihistamine |
165,154 (69.49) |
<0.0001 |
|
Antiplatelet |
9,699 (4.08) |
<0.0001 |
|
Steroid |
23,235 (9.78) |
<0.0001 |
|
Caloric test |
20,824 (8.76) |
<0.0001 |
|
Patient’s type |
|
|
|
Inpatients |
44,540 (18.74) |
<0.0001 |
|
Outpatients |
197,486 (83.09) |
|
|
Types of hospital |
|
|
|
Male |
|
|
|
Tertiary |
7,505 (3.16) |
<0.0001 |
|
General hospital |
20,052 (8.44) |
|
|
Hospital |
4,705 (1.98) |
|
|
Clinic |
48,011 (20.20) |
|
|
Pharmacy |
1,464 (0.62) |
|
|
Female |
|
|
|
Tertiary |
9,212 (3.88) |
<0.0001 |
|
General hospital |
28,198 (11.86) |
|
|
Hospital |
8,247 (3.47) |
|
|
Clinic |
107,441 (45.21) |
|
|
Pharmacy |
2,838 (1.19) |
|
|
Continuous variable |
|
|
|
Age (yr) |
54.76 (54.64–54.89) |
<0.0001 |
|
Hospital costs ($) |
64.7 (63.7–65.8) |
<0.0001 |
|
Cost of patient-paid ($) |
19.7 (19.4–20.1) |
<0.0001 |
|
Cost of insurance-paid ($) |
44.9 (44.2–45.7) |
<0.0001 |
REFERENCES
- 1. Baloh RW. Clinical practice: vestibular neuritis. N Engl J Med 2003;348:1027–1032. ArticlePubMed
- 2. Goddard JC, Fayad JN. Vestibular neuritis. Otolaryngol Clin North Am 2011;44:361–365. ArticlePubMed
- 3. Pace WD, Dickinson LM, Staton EW. Seasonal variation in diagnoses and visits to family physicians. Ann Fam Med 2004;2:411–417. ArticlePubMedPMC
- 4. Jeong J, Youk TM, Jung HT, Choi HS. Seasonal variation in peripheral vestibular disorders based on Korean population data. Laryngoscope Investig Otolaryngol 2024;9:e1254. ArticlePubMedPMC
- 5. Seidel DU, Park JJ, Sesterhenn AM, Kostev K. Demographic data and seasonal variation in peripheral vestibular disorders in ENT practices in Germany. J Vestib Res 2019;29:181–190. ArticlePubMed
- 6. Adamec I, Krbot Skorić M, Handžić J, Habek M. Incidence, seasonality and comorbidity in vestibular neuritis. Neurol Sci 2015;36:91–95. ArticlePubMedPDF
- 7. Koors PD, Thacker LR, Coelho DH. Investigation of seasonal variability of vestibular neuronitis. J Laryngol Otol 2013;127:968–971. ArticlePubMed
- 8. Jeong J, Nam Y, Oh J, Choi HS. Monthly and seasonal variations in vestibular neuritis. Medicine (Baltimore) 2022;101:e29787. ArticlePubMedPMC
- 9. Yoon CY, Kong TH, Lee J, Seo YJ, Ahn JJ. Epidemiology of idiopathic sudden sensorineural hearing loss in the era of big data. Eur Arch Otorhinolaryngol 2023;280:2181–2190. ArticlePubMedPDF
- 10. Yoon CY, Hong S, Park JY, Seo YJ. Quantifying the prevalence of acute vestibular neuritis through big data analysis. Res Vestib Sci 2023;22:68–76. Article
- 11. Oh H, Yoon CY, Lee J, Seo YJ, Chung W, Park MK. Cost-effectiveness of hearing aids in South Korea: a multistate Markov model analysis. Clin Exp Otorhinolaryngol 2025 Jan 21 [Epub]. https://doi.org/10.21053/ceo.2024.00255Article
- 12. Kwak C, Byun Y, You S, et al. Understanding standard procedure in auditory brainstem response: importance of normative data. J Audiol Otol 2024;28:243–251. ArticlePubMedPMCPDF
- 13. Lee J, Yoon CY, Lee J, Kong TH, Seo YJ. Clinical characteristics of the "Gap" between the prevalence and incidence of hearing loss using National Health Insurance Service data. PLoS One 2024;19:e0299478. ArticlePubMedPMC
- 14. Jin IK, Lee D, Jeong Y, et al. Trends in distributions of hearing threshold levels by ages: a comparison of the ISO 7029 and newly available country-specific data. J Audiol Otol 2024;28:1–9. ArticlePubMedPMCPDF
- 15. Lee J, Yoon CY, Lee J, Kong TH, Oh SH, Seo YJ. A situational analysis of ear and hearing care in South Korea using WHO ear and Hearing Care Situation Analysis tool. Front Public Health 2023;11:1215556. ArticlePubMedPMC
Citations
Citations to this article as recorded by
KBS
, Young Jun Seo1,2
PubReader
ePub Link
Cite
XML Download
