Introduction

The COVID-19 pandemic has brought about changes in various aspects, including daily life and hospital treatment. Furthermore, the South Korean government recommended the use of masks not only for individuals exhibiting respiratory symptoms but also for healthy members of the general public as part of a nationwide effort to prevent and control the spread of infection. Such changes brought about by COVID-19 may have also impacted the prevalence and condition of skin-related diseases1. Therefore, it is necessary to compare the prevalence of skin-related ailments before and after the COVID-19 pandemic, as well as to analyze prevalence rates during the period of the pandemic2,3. For this reason, it is imperative to focus special attention on the changes in specific diseases, particularly atopic dermatitis (AD), following COVID-19. The prolonged exposure of the skin to warm and humid conditions, as experienced with mask wearing, can exacerbate AD. Among the diseases that have undergone changes after COVID-19, we are particularly interested in AD.

AD is a widely encountered condition, affecting 15–20% of adolescents1,4,5,6. Additionally, although AD is not a fatal disease with a low mortality rate, it can recur and has a negative impact on people’s health and quality of life7,8. Therefore, there is a need for further research about AD including COVID-19 pandemic. However, there has been a lack of studies analyzing and interpreting the prevalence of AD up to 2022. Since the COVID-19 pandemic occurred over a short period starting from 2020, it is essential to conduct detailed analysis including the most recent data9,10,11,12,13. Additionally, whereas previous studies lacked investigation into risk factors or their utilization of invalidated definitions of AD such as claim-based codes11, we identify risk factors influencing AD by analyzing the data including information up to 2022.

Therefore, our study aims to analyze the changing prevalence of AD before and after pandemic and to conduct an analysis considering the changed situation in 2022, including the risk factors. Based on this analysis, we aim to identify risk factors that influence AD and present a new perspective on AD after 2022, including aspects that have not been considered.

Methods

Patient selection and data collection

In this study, we used data from the Korea Youth Risk Behavior Web-based Survey (KYRBS) from 2009 to 202214,15. The KYRBS is an online survey administered to Korean adolescents in computer laboratories at their schools, and it utilizes a self-reporting method that monitors the health and behavior of Korean youths while being supervised by the Korean Disease Control and Prevention Agency (KDCA)9,16. This data includes information about behaviors and health such as smoking, drinking, physical activity, and allergic diseases. To ensure that the survey data reflects the entire Korean population, it employs sample selection methods such as stratified cluster sampling and two-step stratification16,17. And the weighting is calculated by multiplying the inverse of the sampling rate by the inverse of the response rate, and then multiplying this by the post-adjustment weighting rate16. The data collected included all adolescents aged 13–18 years with a response rate over 95% while participants with missing data were excluded. The research protocol received approval from both the Institutional Review Board of Kyung Hee University (KHUH 2022-06-042) and the KDCA. Written informed consent was obtained from all participants prior to their involvement in the study. Additionally, the KYRBS provides public access to its data, making it a valuable resource for conducting various epidemiological investigations. This research adhered to the ethical guidelines established by relevant national, and institutional review boards for human research and followed the 1975 Helsinki Declaration, as amended in 2008.

Ascertainment of AD

Our research aimed to confirm the trend in the prevalence of AD among adolescents from 2009 to 202216,18. Additionally, the objective of our study was to investigate the risk factors associated with atopic dermatitis. Patients with AD were defined by positive responses to two questions: “Have you ever been diagnosed AD by a doctor within the past 12 months?” and “Have you ever been diagnosed AD by a doctor throughout life?” In addition, the questions were used the same during the study11,19,20.

Covariates

The data was divided into time periods before and during the COVID-19 pandemic. Since the first case of COVID-19 in South Korea was reported in January 2020, the year 2020 was defined as the first year of the pandemic21,22. Before COVID-19, 2009–2019, was subdivided into intervals of 2 to 3 years to obtain a stable estimate of the prevalence of allergic diseases (2009 to 2011, 2012–2013, 2014–2015, 2016–2017, and 2018–2019). Prevalence of AD was further by stratified variables including grade, sex, region of residence (rural and urban)23,24,25,26,27,28, body mass index (BMI; underweight, normal, overweight, obese, and unknown), school performance (low, lower-middle, middle, upper-middle, and high), parent’s highest educational level (middle school graduated or under, high school graduated, university graduated or higher, and unknown), household income (lower or middle lower, middle, and middle higher or higher), alcohol consumption (non-drinker, 1–2, 3–5, 6–9, and ≥ 10 day/month), smoking status (non-smoker and smoker), stress status (none, mild, moderate, high, and severe), sadness and despair, suicidal thoughts, and suicide attempts. BMI was calculated with body height, weight, sex, and age with reference to the 2017 Korean National Growth Charts for children and adolescents29.

Statistical analysis

We included demographic characteristics of the study participants as covariates30. We analyzed secondary data from KYRBS spanning a total of 14 years, from 2009 to 2022. Only records with complete data were included in the study. Analysis was conducted using composite sampling linear and logistic regression models. The term “composite sampling linear regression” refers to a method of conducting linear regression analysis by amalgamating data obtained from multiple sampling sources31. To investigate the trends in prevalence before and during the COVID-19 pandemic, we utilized a weighted linear regression model, deriving a β coefficient along with a 95% confidence interval (CI)32. The βdiff was used to emphasize the difference between the pre-pandemic (2009–2020) and the pandemic (2020 and 2022). This method was chosen to examine changes over time and to contrast the effects before and during the pandemic. We used a binary logistic regression model to calculate the odds ratio for the 2009–2019 and 2020–2022. We also employed this method to identify risk factors of AD and calculate the ratio of odds ratio (OR) by using odds ratios from before and during the pandemic. We further determined risk factors for the years 2020, 2021 and 2022, to deepen our understanding of the pandemic’s impact33,34,35. Categorical data were presented as frequencies and percentages, whereas numerical data were presented as means along with their corresponding 95% CIs. All statistical analyses were performed using SAS version 9.4 software (SAS Institute, Cary, NC, USA) and GraphPad Prism version 9.5.0 (GraphPad, San Diego, CA, USA)35,36,37.

Ethics approval

The study protocol was approved by the Institutional Review Board of KDCA (2014-06EXP-02-P-A).

Informed consent

All participants provided written informed consent.

Results

Table 1, Tables S1 and S2 show the data collected from KYRBS. The survey included a total of 917,461 participants aged 13–18 years from 2009 to 2022. The number of participants in each period was as follows: 223,947 from 2009 to 2011, 146,621 from 2012 to 2013, 140,103 from 2014 to 2015, 127,804 from 2016 to 2017, and 117,343 from 2018 to 2019 (Fig. 1). During the pandemic, there were 54,948 participants in 2020, 54,848 in 2021, and 51,847 in 2022. In terms of participant demographics, 471,306 (51.37%) of the participants were male, 446,155 (48.63%) female, and 502,419 (54.76%) lived in rural areas. 106,900 (11.65%) were people with obese, 765,939 (83.48%) were non-drinkers and 749,053 (81.64%) non-smokers. 30,765 (3.35%) of participants reported suicide attempts while 137,537 (14.99%) reported suicidal thoughts.

Table 1 Baseline characteristics of participating adolescents in KYRBS, 2009–2022 (n = 917,461).
Full size table
Figure 1
figure 1

Study population in KYRBS, 2009–2022. KYRBS Korea Youth Risk Behavior Web-based Survey.

Full size image

Table 2 and Fig. 2 shows the weighted prevalence of AD from 2009 to 2022, further stratified against socioeconomic factors. The weighted prevalence of AD increased from 6.79% (95% CI 6.66–6.91) in 2009–2011 to 7.22% (95% CI 7.06–7.38) in 2016–2017. However, there was a slight decrease immediately preceding the pandemic, from 2018–2019 (6.89%, 95% CI 6.72–7.05) to 2022 (5.82%, 95% CI 5.60–6.04). There was no significant difference between the pre-pandemic and pandemic in prevalence (βdiff − 0.041, 95% CI − 0.124 to 0.042) showing that the prevalence decreased since the start of the pandemic. Table 3 was analyzed to more closely examine the changes in AD prevalence during the pandemic period. The prevalence consistently showed a decreasing trend. Notably, from 2021 to 2022, there was a significant decrease (weighted OR 0.93, 95% CI 0.88 to 0.98). Table 4 shows the odds ratios of the socioeconomic factors of AD. Overall, from 2009 to 2022, high school grade, female sex, urban, high BMI group, lower-middle school performance, university graduated or higher parent’s highest educational level, middle household income, current alcohol consumption, smoker, severe stress, sadness and despair, suicidal thoughts, and suicidal attempts are observed as risk factor of AD. In addition, we calculated the ratio of OR for the pandemic/pre-pandemic period to analyze vulnerable groups during the pandemic. AD vulnerable groups during the pandemic were followed by; high school grade (ratio of OR 95% CI 1.07 [1.01 to 1.12]), middle household income (ratio of OR 95% CI 1.18 [1.09 to 1.27]), ≥ 10 days alcohol consumption (ratio of OR, 95% CI 1.25 [1.05 to 1.49]), and current smoker (ratio of OR 95% CI 1.10 [1.02 to 1.19]).

Table 2 The national trend of the prevalence of AD and β-coefficients and odds ratios before and during COVID-19 pandemic, weighted % (95% CI), in the KYRBS.
Full size table
Figure 2
figure 2

Longitudinal trend and prevalence of AD in Korean adolescents, 2009–2022. AD atopic dermatitis.

Full size image
Table 3 Comparison with the previous year by the wORs for prevalence of AD and each socioeconomic factor, weighted % (95% CI).
Full size table
Table 4 Ratio of ORs for association between prevalence of AD and each socioeconomic factor, 2009–2022.
Full size table

In Table S3, risk factors such as high school grade, female sex, high BMI group, low school performance, high school graduated parent’s highest educational level, current alcohol consumption, smoker, severe stress, sadness and despair, suicidal thoughts, and suicidal attempts were identified during the pandemic in 2020–2022, respectively.

Discussion

Key results

In this study we found the prevalence of AD to be 5.82% in 2022 among adolescents aged 13 to 18 years. Analyzing the prevalence of AD is important because this allergic disease has a significant impact on the quality of life, particularly among adolescents. The prevalence of AD continued to increase prior to the pandemic. However, the most significant decline in prevalence occurred after the COVID-19 pandemic began, with a smaller decline observed during the pandemic. Examining this prevalence contributes to understanding the scope of the COVID-19 pandemic and its influences on allergic disease trends. Additionally, current alcohol consumption, smoker, severe stress, sadness and despair, suicidal thoughts, and suicidal attempts are observed as risk factors of AD. These results could also provide a new perspective on coping with AD.

Global epidemiology and mechanism

In South Korea, the first cases of COVID-19 were discovered in January 2020, and the severity of the pandemic increased over time37. One hypothesis for this phenomenon could be the reduced exposure to allergens due to the changes during COVID-19. Several significant social and political changes occurred during this period, which could have influenced the level of allergen exposure. Because the most drastic of measures included social distancing, lockdowns, closures of public places, decreased physical interaction, mask mandates, and other self-isolating actions38,39,40. Furthermore, during the COVID-19, there was a decrease in air pollution in South Korea41, which is consistent with decreased smoking and automobile traffic resulting from decreased socialization42. Air pollution has previously been linked to AD as air particulates can damage the skin barrier through oxidative stress, thus triggering immune dysregulation which can potentially increase inflammation of AD43,44. Heavy metals exposure (a factor of air pollution) has similarly been associated with allergic diseases, acting as a direct activator of Th2 cells as well as an adjuvant, increasing the effect of other allergens42. In addition, the use of face masks has further enhanced the protective effect as studies have shown that they reduce allergic diseases, particularly AD45,46. While face masks could have an inflammatory effect on facial dermatoses due to increased skin hydration, temperature, and sebum secretion, it has been found that in patients with AD, facial eczema is deceased due to covering sensitive areas46. In 2022, many COVID-19 regulations began to be rescinded, with South Korea removing the mask mandate and other social restrictions as the pandemic started to wind down. This return to pre-pandemic action could partially explain the decreased trend of AD during 202213. However, although there is a decreasing trend, it is too early to determine the impact of reducing social restrictions in 2022, so future research is required.

In this study, we found that alcohol consumption and smoking are both associated with an increased risk of AD, which is consistent with the literature9. However, in Korean adolescents, we found a minor negative association between high BMI and AD, which is in contrasts with other sources, as overweight or obese individuals have been found to have an increased risk of AD through a small but important causal relationship47. Having suicidal thoughts was the only factor that demonstrated an increased risk of AD across the pandemic, 2020–2022. However, further research is necessary to fully understand the underlying mechanism behind this association. Falling rates of AD during COVID-19 may also be due to other factors. Firstly, the common allergens for AD have been found to be changing as patients are less sensitized to them and new allergens emerge48. Secondly, in the United States, it has been postulated that falling rates of allergic disease could be due to the changes in measurements49, but the KYRBS has remained consistent in its measurement and questions. Finally, a study in Colombia found that leading up to 2019, there was an increasing utilization of healthcare resources and healthcare visits, which correlated with an increased in the incidence and prevalence of AD50. The COVID-19 pandemic saw a decrease in healthcare utilization, including for allergic disease, as fewer people reported to hospitals during COVID-19 for fear of contracting SARS-CoV-251. Furthermore, many patients with AD are treated by specialists, and data from other countries data outpatient dermatological visits during the COVID-19 pandemic, with AD being the most common reason for visits52,53. However, specific studies conducted in South Korea found no decrease in incidence in adolescent AD from 2019 to 2020, so changes in incidence may not fully explain the decreasing trend of AD during this period54.

Strengths and limitations

We utilized the KYRBS, which serves as both the greatest strength and limitation of this study. The KYRBS is the only national database for the prevalence of allergic disease since the COVID-19 pandemic, with over one million Korean adolescents included in the study. The KYRBS is also integrated with the public school system in South Korea, as it is administered to most school-age adolescents via computers with anonymous online self-reporting9. However, this poses several limitations. First, the KYRBS excludes any student not in the Korean public education system, potentially excluding a segment of the adolescent population. Second, the KYRBS is a self-reporting questionnaire, which, although considered reliable based on a kappa coefficient of 0.78–0.8055, still yields subjective data that lacks objective measures. Third, in our study, we analyzed the data without considering the sensitivity and specificity of AD. Therefore, there is a limitation that the numbers may change when considering this. However, even if we consider this limitation, it is unlikely to significantly alter the overall trend of the results. Fourth, this study has limitations due to the lack of use of validated questionnaires, impacting the reliability and accuracy of its results. This common issue in self-report studies necessitates caution in interpreting data and may restrict the generalizability of the findings. Fifth, while the results of this study show an overall decreasing trend in the number of AD patients, this may be attributed to a reduction in doctor visits during the pandemic. However, our dataset lacks information on doctor visits, making this difficult to ascertain. Finally, the data is confined to one country, South Korea. Primary healthcare is highly valued worldwide56. However, contrary to this, according to statistics from the World Health Organization, many countries suffer from poor access to primary healthcare57,58. On the other hand, South Korea boasts excellent accessibility to primary healthcare59. This disparity may be attributed to cultural differences, the presence or absence of social insurance, among other factors, which could result in the trends in other countries being different.

Conclusion

This study observed the prevalence of AD over a 14-year from 2009 to 2022, analyzing both the differences before and during the COVID-19 pandemic, as well as within the pandemic itself. The prevalence of AD increased before the pandemic, then decreased at the start of the pandemic and remained stable throughout the pandemic. This prevalence helps us understand the scope of the COVID-19 pandemic and its effects on allergic disease trends. In addition, risk factors of AD include alcohol consumption, smoking, severe stress, feelings of sadness and despair, suicidal ideation, and suicide attempts. Further research is necessary to understand whether these results are specific to the COVID-19 pandemic and whether the trend of AD will revert to pre-pandemic levels as the COVID-19 pandemic ends.