Mental Health Impacts of COVID-19 Pandemic by Gender in South Korea: Links to Job Loss and Childcare

Round 1

Reviewer 1 Report

Dear authors,

your manuscript offers a valuable and empirically grounded examination of how the COVID-19 pandemic impacted mental health across gender lines in South Korea.

The use of a nationally representative dataset, combined with a retrospective cohort design and robust statistical modeling, forms a solid foundation for a study of this scope and ambition. The clarity of the analytic approach and the attention to gendered vulnerabilities are commendable. However, several aspects of the manuscript would benefit from greater depth, clarification, or refinement to fully realize its potential contribution to the literature.

First and foremost, the definition and operationalization of key variables could benefit from further elaboration. The authors use changes in health insurance status as a proxy for employment disruption, which is an innovative and data-driven approach. Still, more context is needed for international readers who may not be familiar with South Korea’s health insurance classification. Could some of these transitions reflect voluntary shifts, for example, early retirement, transition to caregiving, or freelance rather than involuntary job loss?

Similarly, the use of diagnostic codes from outpatient visits to define depression introduces the possibility of detection bias. For instance, could it be that women or higher-income individuals are more likely to seek mental health care and thus more likely to be diagnosed?

Moreover, while the study effectively demonstrates the statistical association between pandemic-related stressors and depression, it remains somewhat descriptive in its interpretation of gender disparities.

Another notable limitation is the absence of mediating or moderating variables that are highly relevant in the context of mental health during a crisis. The manuscript would benefit from acknowledging that variables such as social support, coping mechanisms, or workplace flexibility likely play a significant role in shaping mental health outcomes. Although such data may not be available within the NHID framework, the authors could explicitly recognize this limitation and perhaps suggest directions for future research that could incorporate more subjective or qualitative measures.

In addition, while the statistical analyses are thorough, the presentation of tables and figures could be improved for clarity and reader accessibility.

Table 1, in particular, is dense and could be reformatted to better highlight gender-based comparisons.

Figure 2 is referenced as a key result for the subgroup aged 20–40 but is not visible in the provided document.

Finally, the limitations section, while present, is relatively brief given the complexity of the data and design. The major revisions suggested here are intended to elevate the manuscript to the level of its underlying potential. Once addressed, these enhancements will ensure that the article speaks to the human realities of gendered mental health experiences in the wake of a global crisis.

The use of a nationally representative dataset (NHID) and the focus on gendered experiences of depression during the COVID-19 pandemic are both timely and methodologically solid. That said, a number of elements within the manuscript would benefit from further clarification, elaboration, or refinement to enhance both the scientific rigor and readability of the work. The operationalization of job loss based on changes in health insurance status is innovative but requires clearer explanation.

Please clarify whether such transitions may also capture voluntary role changes.

Are there any mechanisms to distinguish involuntary unemployment from other status changes?

Similarly, the criteria for identifying newly diagnosed depression based on outpatient claims (ICD codes F32, F33, etc.) should be expanded. Were all diagnoses provided by mental health professionals, or could general practitioners also enter these codes?

Consider discussing how access to services or diagnostic behaviour may differ by gender or income, potentially introducing detection bias. The text mentions that 43,574 individuals with prior depression were excluded from the 340,942 eligible employees.

However, it may be helpful to include these details directly in the flowchart (Figure 1) and clearly describe in the main text what criteria determined inclusion in the “final” 297,368 cohort. Additionally, clarify whether the follow-up period (2020–2022) allowed for sufficient observation time for all participants. Were all subjects followed for the same length of time.

Moreover, the use of multivariable Cox regression is appropriate. However, please clarify whether the proportional hazards assumption was tested and satisfied. The authors observe a U-shaped relationship between age and depression risk. While this is a meaningful finding, a figure showing predicted risk across the age spectrum might help readers better visualize the trend. In Table 2 and Table 3, the confidence intervals are provided but the layout is dense.

Consider simplifying these tables, potentially by splitting men and women into separate visuals, or shading significant results. The manuscript presents several important gender-specific findings, for example, that job loss increased depression risk more for women than men, and that marriage appears protective only for women.

These findings are rich in interpretive potential but are currently discussed only briefly. Could cultural expectations around caregiving, household roles, or emotional labor in Korea help explain these trends? Are there comparable findings in other East Asian or OECD countries that support or contrast with these results?

Furthermore, the inclusion of child age as a proxy for caregiving intensity is a strong feature of this study. However, it would be helpful to expand on how this variable was operationalized. For instance, were individuals with multiple children in different age groups categorized differently?

The Discussion section is well-organized, but it could benefit from deeper theoretical framing. For instance, reference to frameworks such as the Gender Role Strain Paradigm or Work-Family Conflict Theory could contextualize why women experienced greater mental health risks under similar external stressors.

Additionally, the role of teleworking or digital access could be mentioned even if data are unavailable, as this was a significant moderator of work–care stress globally.

Tables are rich in detail but may overwhelm readers in their current format. Figure 2 is referenced but not visible in the version reviewed. Ensure it is included and clearly labelled in the final version.

The abstract is well-composed but could benefit from a closing sentence that more directly summarizes the key policy message. With the incorporation of the above clarifications and modest narrative refinements, it has the potential to make a substantial contribution to the literature on gendered mental health inequalities during public health emergencies.

Author Response

1. First and foremost, the definition and operationalization of key variables could benefit from further elaboration. The authors use changes in health insurance status as a proxy for employment disruption, which is an innovative and data-driven approach. Still, more context is needed for international readers who may not be familiar with South Korea’s health insurance classification. Could some of these transitions reflect voluntary shifts, for example, early retirement, transition to caregiving, or freelance rather than involuntary job loss?

 

Thank you for the valuable comment. The dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. This limitation has now been explicitly acknowledged in the revised manuscript’s conclusion of study limitations. (page 12)

 

“In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects.”

 

2. Similarly, the use of diagnostic codes from outpatient visits to define depression introduces the possibility of detection bias. For instance, could it be that women or higher-income individuals are more likely to seek mental health care and thus more likely to be diagnosed?

 

Thank you for your insightful comment. We agree that the use of outpatient diagnostic codes may introduce detection bias, particularly if healthcare utilization varies by gender or income level. This concern is acknowledged in the Discussion and Conclusion sections of the manuscript:

 

[Discussion] (page 11)

 

“Income level, measured through health insurance premium quartiles, was significantly correlated with depression risk. Individuals in the highest income quartiles exhibited increased depression rates, a trend also observed in previous studies. This may reflect multiple factors, including financial uncertainty among asset-holders and business owners, as well as higher healthcare utilization among high-income individuals, leading to more frequent depression diagnoses [33].”

 

[Conclusion] (page 12)

 

“Fourth, our definition of depression, based on the use of healthcare services, could introduce detection bias.”

 

3. Moreover, while the study effectively demonstrates the statistical association between pandemic-related stressors and depression, it remains somewhat descriptive in its interpretation of gender disparities.

 

Thank you for the valuable comment. In response, we have added a more interpretive explanation grounded in theoretical frameworks and mechanisms behind the observed gender disparities.

 

“This gender-specific vulnerability may be attributed to heightened job search stress and anxiety over financial exhaustion, particularly as female-dominated occupations, such as those in the service sector, were disproportionately affected by the pandemic shock [13].” (page 8)

 

“The observed gender disparities in mental health outcomes during the pandemic likely stem from entrenched social norms and labor market structures that place a disproportionate caregiving burden on women. Cultural expectations and gender roles commonly designate women as primary caregivers, intensifying their stress and workload—especially during crises when caregiving demands escalate. These patterns are consistent with the Gender Role Strain Paradigm [31], which posits that psychological strain arises when individuals are unable or unwilling to meet socially prescribed gender expectations. Additionally, the findings align with the Work-Family Conflict Theory [32], which suggests that conflicting demands from work and family roles can lead to emotional distress, particularly in the absence of adequate institutional or social support.” (page 11)”

 

4. Another notable limitation is the absence of mediating or moderating variables that are highly relevant in the context of mental health during a crisis. The manuscript would benefit from acknowledging that variables such as social support, coping mechanisms, or workplace flexibility likely play a significant role in shaping mental health outcomes. Although such data may not be available within the NHID framework, the authors could explicitly recognize this limitation and perhaps suggest directions for future research that could incorporate more subjective or qualitative measures.

 

Thank you for your thoughtful comment. As suggested, we have newly added the conclusion to explicitly acknowledge the limitation regarding the absence of mediating or moderating variables. Specifically, the following sentence has been added:

“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35].”

 

5. In addition, while the statistical analyses are thorough, the presentation of tables and figures could be improved for clarity and reader accessibility. Table 1, in particular, is dense and could be reformatted to better highlight gender-based comparisons. Figure 2 is referenced as a key result for the subgroup aged 20–40 but is not visible in the provided document.

 

Thank you for the valuable comment. Based on your insight, we have revised figures and tables to be more informative and readable. Specifically, we have added p-value in Table 1 to highlight gender-based comparison.

Table 1. Socio-demographic Characteristics and Mental Health (n=297,368).

Note: p-values indicate the statistical significance of differences between male and female groups for each variable.

 

  We have revised the figure by improving the graphical quality and adding explicit numerical labels for the HR values. The updated version now ensures better readability and more effective communication of subgroup differences by gender

6. Finally, the limitations section, while present, is relatively brief given the complexity of the data and design.

Thank you for your comment. In response, we have substantially revised and expanded the limitations in conclusion section to reflect the complexity of the data and design. The revised text now reads as follows: (page 12)

“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.”

 

Detailed comments

 

7. Please clarify whether such transitions may also capture voluntary role changes. Are there any mechanisms to distinguish involuntary unemployment from other status changes?

 

Thank you for your comment. As addressed in our previous response, the transition in health insurance type used to define job loss may also capture voluntary employment changes. Unfortunately, our dataset does not allow us to distinguish between voluntary and involuntary job status changes, which we now explicitly acknowledge as a limitation in the revised manuscript. (page 12)

“In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects.”

 

8. Similarly, the criteria for identifying newly diagnosed depression based on outpatient claims (ICD codes F32, F33, etc.) should be expanded. Were all diagnoses provided by mental health professionals, or could general practitioners also enter these codes?

 

Thank you for your insightful comment. We have revised the manuscript to clarify both the scope of diagnostic codes and the providers who enter them. The following paragraph has been added to the manuscript: (page 5)

"The relevant disease codes, based on the Korean Standard Classification of Diseases (KCD-8), include F32 (Depressive episode), F33 (Recurrent depressive disorder), F34 (Persistent mood [affective] disorders), F38.1 (Other recurrent mood [affective] disorders), F40 (Phobic anxiety disorders), F41 (Other anxiety disorders), F42 (Obsessive-compulsive disorder), and F43 (Reaction to severe stress and adjustment disorders) [26]. These codes represent a range of mental health conditions commonly associated with depression, anxiety, and stress-related disorders. Similar diagnostic codes have been employed in previous studies using administrative data to assess mental health outcomes [27, 28]. Building on this approach, the present study included a broader set of related codes to more comprehensively examine the mental health impacts of pandemic-related stressors."

 

As for your question regarding who enters these codes: In the NHID system, diagnostic codes are entered by healthcare providers when submitting claims for reimbursement. These may include both mental health specialists and general practitioners, reflecting standard clinical practice in South Korea. We have clarified this point in the manuscript to enhance transparency.

Additionally, the following text was added to the data description section: (page 3)

:

“The National Health Information Database (NHID), developed and maintained by the National Health Insurance Service (NHIS), is a nationwide administrative health dataset that includes medical claims data, insurance eligibility, and demographic information for approximately 98% of the Korean population [25]. As enrollment in the NHIS is mandatory for all residents of South Korea, the NHID serves as the most comprehensive and representative national data source for analyzing healthcare utilization and socio-demographic factors. The database is constructed from healthcare provider claims submitted for reimbursement, allowing for detailed tracking of medical diagnoses, procedures, prescriptions, and service utilization. Its use of administrative health data enables objective measurement of clinical outcomes and minimizes recall bias common in self-reported surveys [25].

).”

 

9. Consider discussing how access to services or diagnostic behaviour may differ by gender or income, potentially introducing detection bias. The text mentions that 43,574 individuals with prior depression were excluded from the 340,942 eligible employees.

 

Thank you for your thoughtful comment. We agree that gender and income-related differences in access to mental health services may introduce detection bias. We have addressed this issue in the manuscript as follows:

First, as shown in Table 1, we conducted statistical tests to compare key variables across gender. The results revealed significant gender differences in age, income level, marital status, and job loss during the pandemic (all p-values < 0.0001 except for residential region). These differences highlight the potential for gender-based disparities in healthcare-seeking behavior and access to services, which may affect the likelihood of depression diagnoses.

Second, as mentioned in the discussion section, we explicitly acknowledged the possibility of detection bias due to varying healthcare utilization patterns. Higher-income individuals may be more likely to seek care and thus more likely to be diagnosed with depression, even if underlying prevalence is similar.

 

10. However, it may be helpful to include these details directly in the flowchart (Figure 1) and clearly describe in the main text what criteria determined inclusion in the “final” 297,368 cohort. Additionally, clarify whether the follow-up period (2020–2022) allowed for sufficient observation time for all participants. Were all subjects followed for the same length of time.

 

Thank you for your helpful comment. we have addressed the follow-up period in the revised manuscript. Specifically, we added the following sentence to demonstrate that the duration of follow-up was sufficient to capture meaningful health outcomes:

“Regarding follow-up outcomes, the COVID-period mortality rate among the sample ranged from 0.11% to 0.16%, with 314 individuals dying in 2020, 385 in 2021, and 482 in 2022.” (page 4)

 

These figures indicate that the three-year follow-up period allowed for adequate observation of mortality outcomes within the study population.”

.”

11. Moreover, the use of multivariable Cox regression is appropriate. However, please clarify whether the proportional hazards assumption was tested and satisfied.

 

Thank you for your important comment. As suggested, we assessed the proportional hazards assumption using Schoenfeld residuals. The residual plots demonstrated consistent time patterns across all covariates, indicating that the assumption was met. We have added the following sentence to the Methods section (page 6):

 

“We assessed the proportional hazards assumption using Schoenfeld residuals, and the residual plots showed consistent time patterns across all covariates, supporting the validity of the assumption.”

 

In addition, the test results are uploaded in the supplementary file

 

12. The authors observe a U-shaped relationship between age and depression risk. While this is a meaningful finding, a figure showing predicted risk across the age spectrum might help readers better visualize the trend. In Table 2 and Table 3, the confidence intervals are provided but the layout is dense. Consider simplifying these tables, potentially by splitting men and women into separate visuals, or shading significant results. The manuscript presents several important gender-specific findings, for example, that job loss increased depression risk more for women than men, and that marriage appears protective only for women.

 

Thank you for your helpful suggestions to enhance clarity and readability. To improve the visibility of key gender-specific findings in Tables 2 and 3, we have highlighted variables that showed significant gender differences—such as income, job loss, and presence of young children—using bold formatting. Additionally, we standardized the layout by consistently placing the reference group at the top of each variable category, thereby improving comparability and overall readability

Table 2. Association between Socio-demographic Factors and the Incidence of Depression in the Total Sample.

Note: (1) Job loss model includes employment status change as main independent variable.
(2) Childcare burden model includes family structure and presence of children by age group.
(3) Combined model includes both employment status and family structure variables.

 

Table 3. Association between Socio-demographic Factors and the Incidence of Depression by Gender Subgroup.

 

 

13. These findings are rich in interpretive potential but are currently discussed only briefly. Could cultural expectations around caregiving, household roles, or emotional labor in Korea help explain these trends? Are there comparable findings in other East Asian or OECD countries that support or contrast with these results?

Thank you for this insightful comment. In response, we have expanded the discussion and added a new paragraph in the Conclusion section (page 12) to provide further interpretation based on cultural expectations and cross-country comparisons. The revised text reads:

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

 

14. Furthermore, the inclusion of child age as a proxy for caregiving intensity is a strong feature of this study. However, it would be helpful to expand on how this variable was operationalized. For instance, were individuals with multiple children in different age groups categorized differently?

 

Thank you for this important comment. As suggested, we have clarified how the childcare burden variables were defined in the revised manuscript (page 5). The revised text reads:

“The childcare burden measures only count whether an individual has any child in the age groups of interest rather than the number of children he or she has. It is also notable that the categorization is not exclusive, that is, an individual who has children aged 7~9 also fall into the group having children aged under 18.”

 

15. The Discussion section is well-organized, but it could benefit from deeper theoretical framing. For instance, reference to frameworks such as the Gender Role Strain Paradigm or Work-Family Conflict Theory could contextualize why women experienced greater mental health risks under similar external stressors.

 

Thank you for your valuable comment. In response, we have incorporated theoretical perspectives to deepen the interpretation of the gender disparities observed in our findings. Specifically, we now discuss how the Gender Role Strain Paradigm and Work-Family Conflict Theory offer explanatory frameworks for understanding why women may have faced greater mental health burdens during the pandemic. The revised Discussion section includes the following text: (page 11)

"The observed gender disparities in mental health outcomes during the pandemic presumably stem from entrenched social norms and labor market structures that disproportionately allocate caregiving responsibilities to women. Cultural expectations and gender roles often position women as primary caregivers, significantly increasing their stress and workload, particularly during periods of crisis when caregiving demands surge. These findings align with the Gender Role Strain Paradigm [31], which suggests that socially prescribed gender roles can generate psychological strain when individuals are unable or unwilling to meet those expectations. In addition, the patterns observed in this study reflect the principles of Work-Family Conflict Theory [32], which posits that competing demands from work and family roles can result in psychological distress, particularly when institutional or social support is lacking.

 

16. Additionally, the role of teleworking or digital access could be mentioned even if data are unavailable, as this was a significant moderator of work–care stress globally.

Thank you for this valuable suggestion. Based on your insight, we have added a paragraph in the Discussion section addressing the role of telecommuting and flexible work arrangements as follows (page 11):

“Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic.”

 

17. Tables are rich in detail but may overwhelm readers in their current format. Figure 2 is referenced but not visible in the version reviewed. Ensure it is included and clearly labelled in the final version.

Thank you for your helpful comment. As noted, we have revised Figure 2 by enhancing its graphical quality and ensuring it is clearly labeled and visible in the revised manuscript. We also added explicit numerical labels for hazard ratio (HR) values to improve interpretability.

 

18. The abstract is well-composed but could benefit from a closing sentence that more directly summarizes the key policy message. With the incorporation of the above clarifications and modest narrative refinements, it has the potential to make a substantial contribution to the literature on gendered mental health inequalities during public health emergencies

Thank you for your insightful suggestions. In response, we have revised the final sentence of the abstract to more clearly summarize the core policy implications of our findings (page 1).

“"The results underscore the urgent need for gender-sensitive public health policies that expand childcare support, institutionalize flexible work arrangements such as telecommuting, and enhance access to targeted mental health services to reduce pandemic-induced gender disparities in mental health."

 

Author Response File: Author Response.docx

Reviewer 2 Report

This is an interesting and important contribution to the mental health field, the COVID-19 pandemic, and gender in South Korea. Nevertheless, some significant changes would improve the overall chances of it being published:

1. Abstract: Readers would benefit from a more structured abstract. Please inform: objectives, methods (participant, measurement instruments, etc.), results, and implications.
2. The introduction section is impoverished and needs to be rewritten: authors must provide more information on the circumstances of the pandemic in South Korea, gender disparities in mental health in the country, and more data regarding other studies worldwide.
3. Data collection must be detailed.
4. Please justify the choice of independent variables. Why these and not others?
5. Table 1 must include results for inferential analyses for each variable.
6. Gender and social factors must be included in the discussion.
7. The study lacks references with which these results must be compared.
8. The discussion section is poor and needs further details. Also, it must include a limitation section and an implications section.

Best wishes.

N/A

Author Response

1. Abstract: Readers would benefit from a more structured abstract. Please inform: objectives, methods (participant, measurement instruments, etc.), results, and implications.

 

Thank you for this valuable comment. We revised the abstract to incorporate the suggested elements—objectives, methods (participants, measurement instruments, etc.), results, and implications—as follows: (page 1)

 

“This study investigates the impact of the COVID-19 pandemic on clinically diagnosed depression in South Korea, focusing on gender disparities and structural risk factors such as job loss and childcare burden. Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression. Multivariable Cox proportional hazard regression revealed that women had significantly higher risks of depression than men, particularly among those in their 20s to 40s, those who experienced job loss, had children aged 7–9, or belonged to high-income groups. These findings suggest that the intersection of employment instability and caregiving responsibilities disproportionately affected women’s mental health during the pandemic. The results underscore the urgent need for gender-sensitive public health policies that expand childcare support, institutionalize flexible work arrangements such as telecommuting, and enhance access to targeted mental health services to reduce pandemic-induced gender disparities in mental health.”

 

2. The introduction section is impoverished and needs to be rewritten: authors must provide more information on the circumstances of the pandemic in South Korea, gender disparities in mental health in the country, and more data regarding other studies worldwide.

 

Thank you for this insightful comment. In response, we added a dedicated Background section to complement the Introduction. This new section provides detailed information on (1) the circumstances of the pandemic in South Korea, (2) gender disparities in mental health outcomes, and (3) relevant findings from international studies. The revised content is as follows: (page 2~3)

 

“2. Background

In South Korea, the COVID-19 pandemic led to one of the longest periods of school closure among major economies. Beginning with the first confirmed case in January 2020, the government implemented a series of partial and full school shutdowns aimed at preventing in-school transmission. The start of the spring semester in 2020 was delayed multiple times, and even after resumption, in-person classes remained restricted until the first half of 2022. According to UNESCO (United Nations Educational, Scientific and Cultural Organization), South Korea experienced a total of 76 weeks of partial or full school closures from March 2020 to October 2021—surpassing durations observed in the United States (71 weeks), Germany (38), the United Kingdom (27), China (27), and Japan (11) [19].

These prolonged disruptions to childcare and education services substantially increased the caregiving burden on households, particularly for working mothers. An international survey conducted by Boston Consulting Group in 2020 across five countries—the US, UK, Germany, Italy, and France—found that parents nearly doubled their weekly hours spent on childcare and domestic work during the pandemic [20]. Similarly, a study conducted in Ireland found that the surge in homeschooling brought on by school closures was associated with increased negative emotions, highlighting the emotional strain parents—especially mothers—faced during this period [21]. In Korea, although nationally representative time-use data for this period is unavailable, a 2020 survey of parents with children under the age of nine showed that 69.7% of mothers experienced increased stress related to caregiving, compared to 54.1% of fathers. This suggests a heightened caregiving burden disproportionately affecting women [22].

The economic consequences of the pandemic further widened existing gender inequalities. According to an analysis by the Bank of Korea, women were more likely to experience employment disruptions than men—particularly among those with young children. Korean women were also more likely to work in face-to-face service sectors, limiting their ability to work remotely. Additionally, they were more frequently employed in temporary or daily-wage positions, with 21.7% of women in such jobs compared to 10.3% of men, making them more susceptible to job loss and economic insecurity [23].

These structural vulnerabilities translated into mental health challenges. A national survey conducted by the Ministry of Health and Welfare in 2022 found that individuals who experienced income loss during the pandemic were nearly twice as likely to be at risk of depression (22.1%) compared to those whose income remained stable (11.5%) [24]. When combined with increased caregiving responsibilities and limited employment flexibility, these factors may have exacerbated mental health burdens, particularly for working mothers”.

 

3. Data collection must be detailed.

Thank you for this valuable suggestion. We have enhanced the Data Collection section to provide a clearer and more informative description of the National Health Insurance Database (NHID) as follows (page 3)

 

“The National Health Information Database (NHID), developed and maintained by the National Health Insurance Service (NHIS), is a nationwide administrative health dataset that includes medical claims data, insurance eligibility, and demographic information for approximately 98% of the Korean population [25]. As enrollment in the NHIS is mandatory for all residents of South Korea, the NHID serves as the most comprehensive and representative national data source for analyzing healthcare utilization and socio-demographic factors. The database is constructed from healthcare provider claims submitted for reimbursement, allowing for detailed tracking of medical diagnoses, procedures, prescriptions, and service utilization. Its use of administrative health data enables objective measurement of clinical outcomes and minimizes recall bias common in self-reported surveys [25]. “

 

4. Please justify the choice of independent variables. Why these and not others?

 

Thank you for your valuable comment. We have included a more detailed explanation for the independent variables emphasizing the gender disproportionate aspects of job security and childcare burden during the COVID-19 pandemic. (page 3)

“We focus on the two economic stressors, job loss and childcare burden as key variables to explain the gender disparity in mental health outcomes in the pandemic era.

The impact of COVID-19 on employment was most severe in the service sectors, which predominantly encompass female-friendly jobs. The sector heterogeneity implies disproportionate strain on females including intensified job seeking challenges and heightened financial burdens beyond the job loss itself.

The operation of childcare and educational institutions was temporarily halted as part of social distancing measures to minimize infections in communal facilities. Unexpected childcare hours increased due to the operational restrictions of childcare and educational institutions. Some individuals faced the dilemma of choosing between child-care and occupational activities, except for workers who could flexibly adjust labor supply through telecommuting or flexible work arrangements. South Korea experienced notably prolonged school closures compared to many other countries, significantly amplifying caregiving responsibilities for households with young children, particularly those in early elementary grades. It also exhibits one of the widest gender gaps in household labor among developed countries, especially pronounced in caregiving roles. These contextual factors offer a favorable setting to analyze how pandemic-related intensification of childcare responsibilities disproportionately impacted women's mental health”.

 

 

 

5. Table 1 must include results for inferential analyses for each variable.

 

Thank you for your helpful comment. We have revised Table 1 to include the results of inferential analyses for each variable. Specifically, we added p-values to assess the statistical significance of differences between male and female groups. These values are now presented in the rightmost column of Table 1, as requested. (page 7)

 

6. Gender and social factors must be included in the discussion.

 

Thank you for pointing this out. In response to your suggestion, we have explicitly incorporated a discussion of gender and social factors in the revised manuscript. Specifically, we added the following passage to the Discussion section: (page 11)

 

“The observed gender disparities in mental health outcomes during the pandemic presumably stem from entrenched social norms and labor market structures that disproportionately allocate caregiving responsibilities to women. Cultural expectations and gender roles often position women as primary caregivers, significantly increasing their stress and workload, particularly during periods of crisis when caregiving demands surge. These findings align with the Gender Role Strain Paradigm [31], which suggests that socially prescribed gender roles can generate psychological strain when individuals are unable or unwilling to meet those expectations. In addition, the patterns observed in this study reflect the principles of Work-Family Conflict Theory [32], which posits that competing demands from work and family roles can result in psychological distress, particularly when institutional or social support is lacking.”

 

7. The study lacks references with which these results must be compared.

 

Thank you for your constructive comment. In response, we have revised the Conclusion section to explicitly compare our findings with existing international literature, as follows: (page 12)

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

 

8. The discussion section is poor and needs further details. Also, it must include a limitation section and an implications section.

 

Thank you for your thoughtful comment. In response, we elaborated on the discussion section to include more detailed policy implications, and we expanded the conclusion section to address the study’s limitations as follows:

 

[Revised Discussion Section] (page 11)
“These findings underscore the urgent need for gender-sensitive policy responses to pandemic-induced mental health inequalities. International research has consistently emphasized that women disproportionately bore the burden of unpaid caregiving during COVID-19, contributing to psychological distress and labor market detachment [15, 35]. To address these disparities, governments should prioritize expanding affordable childcare services and increasing public investment in care infrastructure [36].

Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic. Finally, targeted mental health interventions, including subsidized psychological services for high-risk groups (e.g., single mothers, high-income professionals facing burnout), are essential to build long-term resilience.”

 

[Revised Conclusion Section (page 12)
“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results”.

 

 

Author Response File: Author Response.docx

Reviewer 3 Report

highlight the research gap and originality in abstract

• state the research method in abstract
• further explain why include depression only but not other mental health problems
• add citation for multivariable Cox proportional hazard regression analyses
• elaborate the advantage of the research method proposed Economic stressors, particularly job loss, had a substantial impact, with women who lost their jobs showing a 22% higher likelihood of depression, compared to 15% for men.
• discuss the reason why there are gender disparities in mental health outcomes. This suggests that while parenthood may offer emotional benefits, the added caregiving responsibilities of young children during the pandemic created significant stress for many parents.
• discuss with previous literature
• what are the academic, practical and policy contributions?
• since your study focuses on Korea solely, it may face limited generalisability

The paper looks like AI is involved, check if this journal allows AI usage.

Author Response

Major comments

highlight the research gap and originality in abstract

1. state the research method in abstract

 

Thank you for your comment. In response, we have revised the abstract to better emphasize the research gap, originality, and methodology, as follows: (page 1)

 

" Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression. Multivariable Cox proportional hazard regression revealed that women had significantly higher risks of depression than men, particularly among those in their 20s to 40s, those who experienced job loss, had children aged 7–9, or belonged to high-income groups."

 

2. further explain why include depression only but not other mental health problems

 

 

Thank you for your comment. We have added a concise explanation in the Introduction section to justify the use of depression as the primary mental health outcome, as follows: (page 2)

 

“Depression was selected as the primary mental health outcome because it is one of the most prevalent and disabling conditions globally and is strongly associated with key pandemic-related stressors such as job loss and increased caregiving burden [16]. Depression can escalate to severe outcomes such as suicide, making early detection and treatment a priority for public health prevention [17]. Among various mental health conditions, depression is also more consistently recorded in administrative health data due to its clear diagnostic coding and treatment pathways, allowing for reliable identification in large-scale datasets [18]”

 

3. add citation for multivariable Cox proportional hazard regression analyses

 

Thank you for your comment. We have added a citation to support the use of multivariable Cox proportional hazard regression analysis, as follows: (page 6)

 

“multivariable Cox proportional hazard regression analyses were performed to estimate the hazard ratios (HRs) and 95% confidence intervals [29].”

 

29. Ramaswami, S., Mohan, P., & Ghoshal, U. C. (2021). Survival analysis: A primer for the clinician scientists. Indian Journal of Gastroenterology, 40(5), 541–549. https://doi.org/10.1007/s12664-021-01199-8

 

 

4. elaborate the advantage of the research method proposed Economic stressors, particularly job loss, had a substantial impact, with women who lost their jobs showing a 22% higher likelihood of depression, compared to 15% for men.

 

Thank you for your valuable comment. To address your suggestion, we have included a paragraph explaining the advantages of our research method in capturing the effects of economic stressors, particularly job loss, by clearly distinguishing gender-specific impacts.

(page 5)

 

““We focus on the two economic stressors, job loss and childcare burden as key variables to explain the gender disparity in mental health outcomes in the pandemic era.

The impact of Covid-19 on employment was most severe in the service sectors, which predominantly encompass female-friendly jobs. The sector heterogeneity implies disproportionate strain on females including intensified job seeking challenges and heightened financial burdens beyond the job loss itself.”

 

5. discuss the reason why there are gender disparities in mental health outcomes. This suggests that while parenthood may offer emotional benefits, the added caregiving responsibilities of young children during the pandemic created significant stress for many parents.

 

Thank you for your suggestion. Based on your insight, we have added implications of gender disparities in the mental health outcome of caregiving burdens for the lower elementary grade children in the discussion chapter. (page 11)

“The observed gender disparities in mental health outcomes during the pandemic presumably stem from entrenched social norms and labor market structures that disproportionately allocate caregiving responsibilities to women. Cultural expectations and gender roles often position women as primary caregivers, significantly increasing their stress and workload, particularly during periods of crisis when caregiving demands surge. These findings align with the Gender Role Strain Paradigm [31], which suggests that socially prescribed gender roles can generate psychological strain when individuals are unable or unwilling to meet those expectations. In addition, the patterns observed in this study reflect the principles of Work-Family Conflict Theory [32], which posits that competing demands from work and family roles can result in psychological distress, particularly when institutional or social support is lacking.

 

 

6. discuss with previous literature

 

Thank you for your helpful comment. We have newly added a Conclusion section to incorporate comparisons with previous literature, as follows:

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

7. what are the academic, practical and policy contributions?

 

Thank you for your comment. We have addressed the academic, practical, and policy contributions across the Discussion and Conclusion sections as follows:

 

 

 

[Revised Discussion Section] (page 11)
“These findings underscore the urgent need for gender-sensitive policy responses to pandemic-induced mental health inequalities. International research has consistently emphasized that women disproportionately bore the burden of unpaid caregiving during COVID-19, contributing to psychological distress and labor market detachment [15, 35]. To address these disparities, governments should prioritize expanding affordable childcare services and increasing public investment in care infrastructure [36].

Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic. Finally, targeted mental health interventions, including subsidized psychological services for high-risk groups (e.g., single mothers, high-income professionals facing burnout), are essential to build long-term resilience.”

 

[Revised Conclusion Section (page 12)
“A notable strength of this study lies in its use of objective, clinically diagnosed mental health outcomes, rather than self-reported survey data. This enhances the validity of its findings and distinguishes it from many prior studies relying on subjective measures of psychological distress [35, 40].

 

 

8. since your study focuses on Korea solely, it may face limited generalisability

 

Thank you for your valuable comment. We have newly written the Conclusion section and incorporated your suggestion regarding generalizability, as follows: (page 12)

" In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.

 

9. The paper looks like AI is involved, check if this journal allows AI usage.

 

We confirm that AI tools were used solely for the purpose of improving English clarity and logical flow during the writing process. As instructed in the journal’s submission guidelines, this was disclosed at the time of submission. No AI-generated content was used for data analysis, interpretation, or conceptual development of the study.

Author Response File: Author Response.docx

Reviewer 4 Report

The manuscript   is fall in the scope and aims of the Journal COVID-MDPI

The quality and the scientific sound of the paper  are not so okay

There are no recommendations or limitations for this study

There is no conclusion

There is no simple summary as all MDPI journals

There is no ethical approval statement

There are no authors’ contributions

There are no acknowledgements

What is /are the creativity of this study

Abstract :

There are no highlights –why ?

There is no graphical abstract

Abstract is very short—expand your data

Abstract should be divided into backgrounds/aims/methods/results and conclusion

LN/11—public health implications—describe in detail

LN/15-16—describe why ???

LN/18-19—rewrite and be more concise

LN/23—add South Korea/Viral diseases/Pandemic and socio-demographic criteria to the keywords

Introduction

LN/27/29/37/42/59---add references

LN/29—OECD—detailed then abbreviate

What is /are the differences between pandemics/parademics and epidemics

LN/60---cross-sectional surveys--/self-reported—clarify role/each

The introduction is y very long –why ? rewrite it gain and be more concise  

Aims should be more clarified

Novelty needs to be more highlighted

Material and Methods

LN/76-89—add reference

NHIS—detailed then abbreviate

The most  descriptive methodologies are without references

Is there no reference for statistical analysis ??

There is no plan for the study area

M& Methods----It is very long and not well organized

Rewrite it again

What about the ethical approval statement

 Results

What about the clinical  symptoms of the diseased  patients  

What about the mortality percentage

What about the PM changes

Describe the other available methods for diagnosis and vaccination

How did the people go back to work after pandemic closure

It is very long

Rewrite it again

Discussion

Authors should discuss the results obtained with those of the previous investigators’ results

Rewrite it again

Conclusion

There is no conclusion –why ?

References

Some cited references need to be more updated

LN/290—delete (April.15) etc., apply for all

All references should be rewritten

Some cited references with missing data—recheck

As volume/issue/pages/number—all available—so no need for the link(s)—apply for all

Comments for author File: Comments.pdf

Author Response

Major comments

 

1. There are no recommendations or limitations for this study

 

Thank you for your thoughtful comment. In response, we elaborated on the discussion section to include more detailed policy recommendation, and we expanded the conclusion section to address the study’s limitations as follows:

 

[Revised Discussion Section] (page 11)
“These findings underscore the urgent need for gender-sensitive policy responses to pandemic-induced mental health inequalities. International research has consistently emphasized that women disproportionately bore the burden of unpaid caregiving during COVID-19, contributing to psychological distress and labor market detachment [15, 35]. To address these disparities, governments should prioritize expanding affordable childcare services and increasing public investment in care infrastructure [36].

Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic. Finally, targeted mental health interventions, including subsidized psychological services for high-risk groups (e.g., single mothers, high-income professionals facing burnout), are essential to build long-term resilience.”

 

[Revised Conclusion Section (page 12)
“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results”.

 

 

2. There is no conclusion

 

Thank you for your important comment. In response, we have added a dedicated Conclusion section following the Discussion to summarize the main findings, highlight policy implications, and acknowledge key limitations and directions for future research. The newly added text reads as follows

“6. Conclusion

This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].

A notable strength of this study lies in its use of objective, clinically diagnosed mental health outcomes, rather than self-reported survey data. This enhances the validity of its findings and distinguishes it from many prior studies relying on subjective measures of psychological distress [35, 40].

Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.

Despite these limitations, this Korea-based study makes an important empirical contribution to understanding pandemic-induced mental health inequalities. The findings offer evidence to inform gender-sensitive and equity-oriented public health responses.”

 

 

3. There is no simple summary as all MDPI journals

 

Thank you for your helpful comment. In response, we have added in conclusion section.

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups.”

 

4. There is no ethical approval statement

Thank you for pointing this out. In response, we have added the ethical approval statement in the Back Matter section of the manuscript, as follows:

“Institutional Review Board Statement: The study protocol was reviewed and exempted by the Institutional Review Board of Incheon National University (IRB No. 7007971-202204-002), as it involves analysis of anonymized secondary data.”

 

5. There are no authors’ contributions

Thank you for pointing this out. In response, we have added authors’ contributions in the Back Matter section of the manuscript, as follows:

 

“Author Contributions: Conceptualization, LEE S.J., WEE H.S., JUNG S.H., LEE J.M.; Data curation, LEE S.J.; Formal analysis, LEE S.J.; Methodology, LEE S.J.; Project administration, JUNG S.H.; Writing—original draft preparation, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.; Writing—review and editing, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.”

 

6. There are no acknowledgements

 

Thank you for pointing this out. In response, we have added Acknowledgments in the Back Matter section of the manuscript, as follows:

 

“Acknowledgments: This study was conducted using data provided by the National Health Insurance Service (NHIS) of Korea. The research management number assigned by the NHIS is NHIS-2024-1-525.”

 

7. What is /are the creativity of this study

 

Thank you for your important question. The main contribution and originality of this study lie in its use of large-scale, objective administrative health data to examine pandemic-related mental health outcomes. Furthermore, the study explores key social factors—such as job loss, childcare burden, and gender differences—providing a more nuanced understanding of mental health disparities during the COVID-19 pandemic. These points have been reflected in the revised Abstract as follows:

“Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression.”

 

Detailed comments

8. Abstract : There are no highlights –why ? There is no graphical abstract Abstract is very short—expand your data, Abstract should be divided into backgrounds/aims/methods/results and conclusion, LN/11—public health implications—describe in detail, LN/15-16—describe why ???LN/18-19—rewrite and be more concise

Thank you for the comment. We have rewritten the abstract upon your suggestion.

“This study investigates the impact of the COVID-19 pandemic on clinically diagnosed depression in South Korea, focusing on gender disparities and structural risk factors such as job loss and childcare burden. Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression. Multivariable Cox proportional hazard regression revealed that women had significantly higher risks of depression than men, particularly among those in their 20s to 40s, those who experienced job loss, had children aged 7–9, or belonged to high-income groups. These findings suggest that the intersection of employment instability and caregiving responsibilities disproportionately affected women’s mental health during the pandemic. The results underscore the urgent need for gender-sensitive public health policies that expand childcare support, institutionalize flexible work arrangements such as telecommuting, and enhance access to targeted mental health services to reduce pandemic-induced gender disparities in mental health.”

 

 

9. LN/23—add South Korea/Viral diseases/Pandemic and socio-demographic criteria to the keywords

 

Thank you for the comment. As you suggested, I have update keywords as following:

“Keywords:  Mental health; Depression; Gender disparities; Job loss; Childcare burden; South Korea; Viral diseases; Pandemic and socio-demographic criteria”

 

 

10. Introduction

LN/27/29/37/42/59---add references

LN/29—OECD—detailed then abbreviate

What is /are the differences between pandemics/parademics and epidemics

LN/60---cross-sectional surveys--/self-reported—clarify role/each

The introduction is y very long –why ? rewrite it gain and be more concise  

Aims should be more clarified

Novelty needs to be more highlighted

 

Thank you for your detailed and insightful comments regarding the Introduction. In response, we have substantially revised the introduction and added a new “Background” section to improve clarity, structure, and conciseness. The revised Introduction now focuses more directly on the research context, rationale, and aims, while the Background section provides supporting national context regarding school closures, caregiving burden, and economic impact

“1. Introduction

The COVID-19 pandemic has posed unprecedented challenges to mental health worldwide, leading to a substantial rise in related issues across the globe [1]. Data comparing the pre-pandemic period to 2020 shows a notable surge in depression prevalence across 15 major OECD (Organisation for Economic Co-operation and Development) countries, with most experiencing more than a twofold increase. South Korea emerged with the highest recorded level of depression at 36.8% and was thus the leading nation [2].

Women’s mental health has emerged as a critical concern during the pandemic, with studies showing they are disproportionately affected compared to men [1,3]. Investigations in countries including the US, Canada, the UK, Italy, China, and Chile have reported more pronounced adverse effects on women’s mental health, which emphasizes a rise in depression, anxiety, and distress rates [3-8]. In South Korea, the research on gender differences in mental health during COVID-19 has yielded varied results. Earlier studies that investigated mental health shortly after the outbreak (beyond day 55) found no significant gender differences [9], but subsequent analyses have revealed disparities between men and women and especially at certain age groups [10, 11].

Two key factors often cited as contributors to the gender gap in mental health during COVID-19 are employment disruptions and childcare burdens. First, the pandemic has severely affected women’s participation in the labor market, which led to deteriorated mental health outcomes for females [4, 10]. In contrast to earlier economic recessions, which primarily impacted male-dominated industries such as manufacturing, the COVID-19 crisis has disproportionately struck women’s employment, especially in the service sector [12]. Data from South Korea in 2020 show a marked decline in the employment rate of married women compared with married men during the pandemic [13], which highlights a change in the way economic crises influence gender-specific employment and mental health. Furthermore, the increased burden of childcare, which is intensified by the shutdown of daycare centers and schools, has placed a disproportionate strain on women, who typically assume the childcare role, creating a significant source of stress [4, 10, 14, 15].

Existing research frequently relies on self-reported, cross-sectional surveys, which can be biased by individuals' subjective interpretations of their mental health symptoms. To address these limitations, our study utilizes observed data from the National Health Insurance Database (NHID) of South Korea, encompassing healthcare utilization and socioeconomic factors such as income level, employment status, and family structure. We focus on the gender-specific mental health impacts of the COVID-19 pandemic, especially targeting a cohort of insurance subscribers with employee health insurance status in 2019, right before the pandemic. Depression was selected as the primary mental health outcome because it is one of the most prevalent and disabling conditions globally and is strongly associated with key pandemic-related stressors such as job loss and increased caregiving burden [16]. Depression can escalate to severe outcomes such as suicide, making early detection and treatment a priority for public health prevention [17]. Among various mental health conditions, depression is also more consistently recorded in administrative health data due to its clear diagnostic coding and treatment pathways, allowing for reliable identification in large-scale datasets [18]. Our retrospective cohort study aims to identify risk factors for newly diagnosed cases of depression in the employed population during the pandemic. Based on previous literature, we hypothesize that working women, particularly those who have lost jobs or have young children, have been more adversely affected [1, 3-8, 10-11, 14, 15l. This research has important policy implications because it supports the development of more effective health and economic interventions for vulnerable populations concerning mental health, which enhances preparedness for future health crises.

 

2. Background

In South Korea, the COVID-19 pandemic led to one of the longest periods of school closure among major economies. Beginning with the first confirmed case in January 2020, the government implemented a series of partial and full school shutdowns aimed at preventing in-school transmission. The start of the spring semester in 2020 was delayed multiple times, and even after resumption, in-person classes remained restricted until the first half of 2022. According to UNESCO (United Nations Educational, Scientific and Cultural Organization), South Korea experienced a total of 76 weeks of partial or full school closures from March 2020 to October 2021—surpassing durations observed in the United States (71 weeks), Germany (38), the United Kingdom (27), China (27), and Japan (11) [19].

These prolonged disruptions to childcare and education services substantially increased the caregiving burden on households, particularly for working mothers. An international survey conducted by Boston Consulting Group in 2020 across five countries—the US, UK, Germany, Italy, and France—found that parents nearly doubled their weekly hours spent on childcare and domestic work during the pandemic [20]. Similarly, a study conducted in Ireland found that the surge in homeschooling brought on by school closures was associated with increased negative emotions, highlighting the emotional strain parents—especially mothers—faced during this period [21]. In Korea, although nationally representative time-use data for this period is unavailable, a 2020 survey of parents with children under the age of nine showed that 69.7% of mothers experienced increased stress related to caregiving, compared to 54.1% of fathers. This suggests a heightened caregiving burden disproportionately affecting women [22].

The economic consequences of the pandemic further widened existing gender inequalities. According to an analysis by the Bank of Korea, women were more likely to experience employment disruptions than men—particularly among those with young children. Korean women were also more likely to work in face-to-face service sectors, limiting their ability to work remotely. Additionally, they were more frequently employed in temporary or daily-wage positions, with 21.7% of women in such jobs compared to 10.3% of men, making them more susceptible to job loss and economic insecurity [23].

These structural vulnerabilities translated into mental health challenges. A national survey conducted by the Ministry of Health and Welfare in 2022 found that individuals who experienced income loss during the pandemic were nearly twice as likely to be at risk of depression (22.1%) compared to those whose income remained stable (11.5%) [24]. When combined with increased caregiving responsibilities and limited employment flexibility, these factors may have exacerbated mental health burdens, particularly for working mothers.”

 

 

11. Material and Methods

LN/76-89—add reference

NHIS—detailed then abbreviate

The most descriptive methodologies are without references

Is there no reference for statistical analysis ??

There is no plan for the study area

M& Methods----It is very long and not well organized

Rewrite it again

 

Thank you for your helpful suggestions regarding the Materials and Methods section. In response, we have added a citation describing the National Health Information Database (NHID), which provides context on its structure, representativeness, and suitability for health services research [Kim et al., 2022]. We have also included an appropriate reference for the statistical methodology used in the study, specifically the Cox proportional hazards regression analysis, which served as the primary analytical approach [Ramaswami et al., 2021

“The National Health Information Database (NHID), developed and maintained by the National Health Insurance Service (NHIS), is a nationwide administrative health dataset that includes medical claims data, insurance eligibility, and demographic information for approximately 98% of the Korean population [25]. As enrollment in the NHIS is mandatory for all residents of South Korea, the NHID serves as the most comprehensive and representative national data source for analyzing healthcare utilization and socio-demographic factors. The database is constructed from healthcare provider claims submitted for reimbursement, allowing for detailed tracking of medical diagnoses, procedures, prescriptions, and service utilization. Its use of administrative health data enables objective measurement of clinical outcomes and minimizes recall bias common in self-reported surveys [25].”

 

“multivariable Cox proportional hazard regression analyses were performed to estimate the hazard ratios (HRs) and 95% confidence intervals [29].”

Kim, M. K., Han, K., & Lee, S. H. (2022). Current trends of big data research using the Korean National Health Information Database. Diabetes & Metabolism Journal, 46(4), 552–563. https://doi.org/10.4093/dmj.2022.0193

Ramaswami, S., Mohan, P., & Ghoshal, U. C. (2021). Survival analysis: A primer for the clinician scientists. Indian Journal of Gastroenterology, 40(5), 541–549. https://doi.org/10.1007/s12664-021-01199-8

 

 

Results

12. What about the clinical symptoms of the diseased patients  

 

Thank you for your comment. We have now specified the clinical diagnostic categories associated with the mental health conditions examined in the study

“The relevant disease codes, based on the Korean Standard Classification of Diseases (KCD-8), include F32 (Depressive episode), F33 (Recurrent depressive disorder), F34 (Persistent mood [affective] disorders), F38.1 (Other recurrent mood [affective] disorders), F40 (Phobic anxiety disorders), F41 (Other anxiety disorders), F42 (Obsessive-compulsive disorder), and F43 (Reaction to severe stress and adjustment disorders)”

13. What about the mortality percentage

 

Thank you for your comment. We have added information on mortality outcomes during the follow-up period.

“Regarding follow-up outcomes, the COVID-period mortality rate among the sample ranged from 0.11% to 0.16%, with 314 individuals dying in 2020, 385 in 2021, and 482 in 2022.”

 

14. What about the PM changes

Describe the other available methods for diagnosis and vaccination

How did the people go back to work after pandemic closure

It is very long

Rewrite it again

 

Thank you for your comments. The questions regarding changes in PM, alternative methods for diagnosis and vaccination, and the return-to-work process after pandemic closures are important public health considerations. However, these aspects were not directly analyzed in our study due to data limitations. Our analysis relied on administrative health data that did not include detailed information on vaccination types, diagnostic testing strategies, or individual-level return-to-work pathways. As a result, we did not include these topics in the main text to maintain the focus and integrity of our analysis.

 

 

15. Discussion

Authors should discuss the results obtained with those of the previous investigators’ results

Rewrite it again

Conclusion

There is no conclusion –why ?

 

Thank you for your helpful comment. We have newly added a Conclusion section to incorporate comparisons with previous literature, as follows:

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

 

16. References

Some cited references need to be more updated

LN/290—delete (April.15) etc., apply for all

 

 

All references should be rewritten

Some cited references with missing data—recheck

As volume/issue/pages/number—all available—so no need for the link(s)—apply for all

 

 

Thank you for your thorough feedback regarding the references. In response, we have made the following revisions:

1. Updated References: We reviewed all cited sources and updated several references to include more recent and relevant literature.
2. Formatting Corrections: All references have been revised to conform to the required journal style, including removal of unnecessary date notations such as “(April 15)” and hyperlinks. Volume, issue, page numbers, and DOIs have been added wherever applicable.
3. Consistency Check: We carefully cross-checked all in-text citations

 

Author Response File: Author Response.docx

Reviewer 5 Report

This study investigated the impact of the novel Coronavirus (SARS-CoV-2) on mental health, with a focus on gender disparities in depression risk and public health. Data were obtained from the National Health Insurance Database of South Korea and analysed using multivariable Cox proportional hazard regression. The results suggest that women are more susceptible to depression when they lose their job than men are, particularly if they have childcare responsibilities. These findings emphasise the importance of implementing gender-sensitive public health strategies, such as targeted financial aid and childcare support.

The manuscript is well written and comprehensible. The method is suitable and the variables are clearly explained. It contributes to the focus on the effects of the pandemic on gender, as well as providing suggestions for policymakers.

I would like to offer some comments for the author(s) to consider.

* Consider whether to divide the introduction into an introduction and a background section, or a literature review. Furthermore, consider adding the conclusion after the discussion, including the limitations and future research in the conclusion (now presented in the discussion section).

* The employment status (on page 4, in row 118) should be explained more clearly in line with the referencing used in the other sections of the text (i.e. ‘job less experiences’).

* On page 2, in row 69, please provide some evidence when you refer to the previous literature (“Based on previous literature”).

* To me, it is not entirely clear why you have focused on only 2% of the sample population. I suggest justifying your decision to focus on only 2% of the population more deeply.

* I suggest adding descriptions of values (1), (2) and (3) at the bottom of Table 2. While these values are explained in the text, it would be better to include the explanations in the table itself.

* Specifying the disease codes (F32, F33, F34, …, F38.1) could help others who wish to replicate the research in a different context.

* In Table 2, the reference category appears in different positions (sometimes first, sometimes last). Consider whether it should always be the first category examined for each variable.

* When describing Table 1 (pages 4–5, lines 149–162), use the order of the categories reported in the table as a guide. In the text, you first describe age, then marital status, job loss and childcare burden, income, and finally depression. The order of the variables in the table is as follows: age, insurance premium, marital status, job loss, childcare burden and depression. Consider whether it would be better to align the two ‘things.’

* On page 4, rows 161-162 the sentence ‘suggesting that the pandemic’s mental health impact was more significant for women.’ does not convince me, since the difference is not significant.

* On page 4, rows 161–162, the sentence ‘suggesting that the pandemic's mental health impact was more significant for women’. This does not convince me so much, since the difference is not so high.

* Evaluate whether you need to add the p-value to the bottom of Table 3. This should also be done for Figure 2.

Author Response

1. I would like to offer some comments for the author(s) to consider.

* Consider whether to divide the introduction into an introduction and a background section, or a literature review. Furthermore, consider adding the conclusion after the discussion, including the limitations and future research in the conclusion (now presented in the discussion section).

 

Thank you for your suggestion. As recommended, we have reorganized the manuscript by separating the Background and Conclusion sections as independent sections. The Background section now follows the Introduction and outlines (1) the pandemic context in South Korea, (2) gender disparities in mental health, and (3) relevant international findings. The Conclusion section has been newly created to present the study’s limitations and recommendations for future research.

“2. Background

In South Korea, the COVID-19 pandemic led to one of the longest periods of school closure among major economies. Beginning with the first confirmed case in January 2020, the government implemented a series of partial and full school shutdowns aimed at preventing in-school transmission. The start of the spring semester in 2020 was delayed multiple times, and even after resumption, in-person classes remained restricted until the first half of 2022. According to UNESCO (United Nations Educational, Scientific and Cultural Organization), South Korea experienced a total of 76 weeks of partial or full school closures from March 2020 to October 2021—surpassing durations observed in the United States (71 weeks), Germany (38), the United Kingdom (27), China (27), and Japan (11) [19].

These prolonged disruptions to childcare and education services substantially increased the caregiving burden on households, particularly for working mothers. An international survey conducted by Boston Consulting Group in 2020 across five countries—the US, UK, Germany, Italy, and France—found that parents nearly doubled their weekly hours spent on childcare and domestic work during the pandemic [20]. Similarly, a study conducted in Ireland found that the surge in homeschooling brought on by school closures was associated with increased negative emotions, highlighting the emotional strain parents—especially mothers—faced during this period [21]. In Korea, although nationally representative time-use data for this period is unavailable, a 2020 survey of parents with children under the age of nine showed that 69.7% of mothers experienced increased stress related to caregiving, compared to 54.1% of fathers. This suggests a heightened caregiving burden disproportionately affecting women [22].

The economic consequences of the pandemic further widened existing gender inequalities. According to an analysis by the Bank of Korea, women were more likely to experience employment disruptions than men—particularly among those with young children. Korean women were also more likely to work in face-to-face service sectors, limiting their ability to work remotely. Additionally, they were more frequently employed in temporary or daily-wage positions, with 21.7% of women in such jobs compared to 10.3% of men, making them more susceptible to job loss and economic insecurity [23].

These structural vulnerabilities translated into mental health challenges. A national survey conducted by the Ministry of Health and Welfare in 2022 found that individuals who experienced income loss during the pandemic were nearly twice as likely to be at risk of depression (22.1%) compared to those whose income remained stable (11.5%) [24]. When combined with increased caregiving responsibilities and limited employment flexibility, these factors may have exacerbated mental health burdens, particularly for working mothers.

 

“6. Conclusion

This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].

A notable strength of this study lies in its use of objective, clinically diagnosed mental health outcomes, rather than self-reported survey data. This enhances the validity of its findings and distinguishes it from many prior studies relying on subjective measures of psychological distress [35, 40].

Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.

Despite these limitations, this Korea-based study makes an important empirical contribution to understanding pandemic-induced mental health inequalities. The findings offer evidence to inform gender-sensitive and equity-oriented public health responses.”

 

 

2. The employment status (on page 4, in row 118) should be explained more clearly in line with the referencing used in the other sections of the text (i.e. ‘job less experiences’).

 

Thank you for your comment. The reference to employment status on page 5 has been deleted, as it was mistakenly retained after the variable was excluded from the final analysis due to lack of significance.

 

 

3. On page 2, in row 69, please provide some evidence when you refer to the previous literature (“Based on previous literature”).

* To me, it is not entirely clear why you have focused on only 2% of the sample population. I suggest justifying your decision to focus on only 2% of the population more deeply.

 

Thank you for your comment. According to data provision regulations, the National Health Insurance Service (NHIS) offers researchers a 2% stratified random sample of the total Korean population as a standard dataset, following a strict de-identification process. As this sample is both stratified and randomly selected, it preserves the key characteristics of the full population, allowing for generalizable and robust analysis.

 

5.  I suggest adding descriptions of values (1), (2) and (3) at the bottom of Table 2. While these values are explained in the text, it would be better to include the explanations in the table itself.

 

Thank you for your suggestion. We have added explanatory notes at the bottom of Table 2 to clarify the meanings of (1), (2), and (3) as follows:

Notes:
(1) Job loss model includes employment status change as main independent variable.
(2) Childcare burden model includes family structure and presence of children by age group.
(3) Combined model includes both employment status and family structure variables.

 

6. Specifying the disease codes (F32, F33, F34, …, F38.1) could help others who wish to replicate the research in a different context.

 

Thank you for your comment. We have revised the manuscript to enhance replicability, as follows: (page 5)

 

“The relevant disease codes, based on the Korean Standard Classification of Diseases (KCD-8), include F32 (Depressive episode), F33 (Recurrent depressive disorder), F34 (Persistent mood [affective] disorders), F38.1 (Other recurrent mood [affective] disorders), F40 (Phobic anxiety disorders), F41 (Other anxiety disorders), F42 (Obsessive-compulsive disorder), and F43 (Reaction to severe stress and adjustment disorders) [26]. These codes represent common mental health conditions linked to depression, anxiety, and stress. Similar codes have been used in prior studies analyzing administrative data [27, 28]. In line with this approach, our study adopted a broader set of diagnoses to comprehensively capture pandemic-related mental health outcomes.

 

7.  In Table 2, the reference category appears in different positions (sometimes first, sometimes last). Consider whether it should always be the first category examined for each variable.

Thank you for your helpful suggestion. In response, we have revised Tables 2 and 3 so that the reference group now consistently appears as the first category for each variable.

 

 

8. When describing Table 1 (pages 4–5, lines 149–162), use the order of the categories reported in the table as a guide. In the text, you first describe age, then marital status, job loss and childcare burden, income, and finally depression. The order of the variables in the table is as follows: age, insurance premium, marital status, job loss, childcare burden and depression. Consider whether it would be better to align the two ‘things.’

 

Thank you for your suggestion. In response, we have revised the description of Table 1 to follow the order of variables as presented in the table. The updated text is as follows: (page 6)

"Table 1 presents the socio-demographic characteristics of 297,368 adults, comprising 178,350 men (60%) and 119,018 women (40%). Among women, 20.2% were aged 20–29 compared to 12.5% of men. Men dominated the 30–69 age groups, while women accounted for 1.3% in the ≥70 group compared to 2.4% of men.
Income disparities were evident, with women more concentrated in the lowest income quartiles (1st: 34.3%, 2nd: 30.9%) compared to men (1st: 16.0%, 2nd: 20.3%). Marital status data showed that 63.9% of participants were married.
Women experienced job loss more frequently (20.8%) than men (17.2%), reflecting greater employment instability during the pandemic. A substantial proportion of both men and women had children. Specifically, 35.6% of men and 32.1% of women had at least one child under the age of 18. Additionally, 9.1% of men and 7.0% of women had a child aged 7–9, indicating that a notable share of the sample faced childcare responsibilities during the pandemic, including challenges related to homeschooling during school closures.
The incidence of newly diagnosed depression during the pandemic was slightly higher among women (3.3%) than men (3.1%), suggesting that the pandemic's mental health impact was more significant for women."

 

8. On page 4, rows 161-162 the sentence ‘suggesting that the pandemic’s mental health impact was more significant for women.’ does not convince me, since the difference is not significant.

* On page 4, rows 161–162, the sentence ‘suggesting that the pandemic's mental health impact was more significant for women’. This does not convince me so much, since the difference is not so high.

 

Thank you for pointing this out. To address your concern, we performed a statistical test to examine the gender difference in the incidence of newly diagnosed depression. The p-value for the gender difference was found to be <0.0001, indicating a statistically significant difference between men and women. We have now added this p-value to Table 1 (last row, rightmost column) and revised the corresponding sentence in the main text to reflect that the difference, while numerically small, is statistically significant.

 

9. Evaluate whether you need to add the p-value to the bottom of Table 3. This should also be done for Figure 2.

 

Thank you for your comment. As hazard ratios are typically interpreted based on whether their 95% confidence intervals include 1, we believe it is not necessary to add p-values separately. This approach is consistent with reporting standards in epidemiological and clinical research, where confidence intervals are preferred over p-values for assessing the statistical significance and precision of hazard ratios.

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Thank you for implementing all the important changes as requested. I believe the article is now more robust and ready for publication. Best wishes.

Thank you for implementing all the important changes as requested. I believe the article is now more robust and ready for publication. Best wishes.

Reviewer 4 Report

Major comments

 

1. There are no recommendations or limitations for this study

 

Thank you for your thoughtful comment. In response, we elaborated on the discussion section to include more detailed policy recommendation, and we expanded the conclusion section to address the study’s limitations as follows:

 

[Revised Discussion Section] (page 11)
“These findings underscore the urgent need for gender-sensitive policy responses to pandemic-induced mental health inequalities. International research has consistently emphasized that women disproportionately bore the burden of unpaid caregiving during COVID-19, contributing to psychological distress and labor market detachment [15, 35]. To address these disparities, governments should prioritize expanding affordable childcare services and increasing public investment in care infrastructure [36].

Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic. Finally, targeted mental health interventions, including subsidized psychological services for high-risk groups (e.g., single mothers, high-income professionals facing burnout), are essential to build long-term resilience.”

 

[Revised Conclusion Section (page 12)
“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results”.

 

 

2. There is no conclusion

 

Thank you for your important comment. In response, we have added a dedicated Conclusion section following the Discussion to summarize the main findings, highlight policy implications, and acknowledge key limitations and directions for future research. The newly added text reads as follows

“6. Conclusion

This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].

A notable strength of this study lies in its use of objective, clinically diagnosed mental health outcomes, rather than self-reported survey data. This enhances the validity of its findings and distinguishes it from many prior studies relying on subjective measures of psychological distress [35, 40].

Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.

Despite these limitations, this Korea-based study makes an important empirical contribution to understanding pandemic-induced mental health inequalities. The findings offer evidence to inform gender-sensitive and equity-oriented public health responses.”

 

 

3. There is no simple summary as all MDPI journals

 

Thank you for your helpful comment. In response, we have added in conclusion section.

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups.”

 

4. There is no ethical approval statement

Thank you for pointing this out. In response, we have added the ethical approval statement in the Back Matter section of the manuscript, as follows:

“Institutional Review Board Statement: The study protocol was reviewed and exempted by the Institutional Review Board of Incheon National University (IRB No. 7007971-202204-002), as it involves analysis of anonymized secondary data.”

 

5. There are no authors’ contributions

Thank you for pointing this out. In response, we have added authors’ contributions in the Back Matter section of the manuscript, as follows:

 

“Author Contributions: Conceptualization, LEE S.J., WEE H.S., JUNG S.H., LEE J.M.; Data curation, LEE S.J.; Formal analysis, LEE S.J.; Methodology, LEE S.J.; Project administration, JUNG S.H.; Writing—original draft preparation, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.; Writing—review and editing, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.”

 

6. There are no acknowledgements

 

Thank you for pointing this out. In response, we have added Acknowledgments in the Back Matter section of the manuscript, as follows:

 

“Acknowledgments: This study was conducted using data provided by the National Health Insurance Service (NHIS) of Korea. The research management number assigned by the NHIS is NHIS-2024-1-525.”

 

7. What is /are the creativity of this study

 

Thank you for your important question. The main contribution and originality of this study lie in its use of large-scale, objective administrative health data to examine pandemic-related mental health outcomes. Furthermore, the study explores key social factors—such as job loss, childcare burden, and gender differences—providing a more nuanced understanding of mental health disparities during the COVID-19 pandemic. These points have been reflected in the revised Abstract as follows:

“Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression.”

 

Detailed comments

8. Abstract : There are no highlights –why ? There is no graphical abstract Abstract is very short—expand your data, Abstract should be divided into backgrounds/aims/methods/results and conclusion, LN/11—public health implications—describe in detail, LN/15-16—describe why ???LN/18-19—rewrite and be more concise

Thank you for the comment. We have rewritten the abstract upon your suggestion.

“This study investigates the impact of the COVID-19 pandemic on clinically diagnosed depression in South Korea, focusing on gender disparities and structural risk factors such as job loss and childcare burden. Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression. Multivariable Cox proportional hazard regression revealed that women had significantly higher risks of depression than men, particularly among those in their 20s to 40s, those who experienced job loss, had children aged 7–9, or belonged to high-income groups. These findings suggest that the intersection of employment instability and caregiving responsibilities disproportionately affected women’s mental health during the pandemic. The results underscore the urgent need for gender-sensitive public health policies that expand childcare support, institutionalize flexible work arrangements such as telecommuting, and enhance access to targeted mental health services to reduce pandemic-induced gender disparities in mental health.”

 

 

9. LN/23—add South Korea/Viral diseases/Pandemic and socio-demographic criteria to the keywords

 

Thank you for the comment. As you suggested, I have update keywords as following:

“Keywords:  Mental health; Depression; Gender disparities; Job loss; Childcare burden; South Korea; Viral diseases; Pandemic and socio-demographic criteria”

 

 

10. Introduction

LN/27/29/37/42/59---add references

LN/29—OECD—detailed then abbreviate

What is /are the differences between pandemics/parademics and epidemics

LN/60---cross-sectional surveys--/self-reported—clarify role/each

The introduction is y very long –why ? rewrite it gain and be more concise  

Aims should be more clarified

Novelty needs to be more highlighted

 

Thank you for your detailed and insightful comments regarding the Introduction. In response, we have substantially revised the introduction and added a new “Background” section to improve clarity, structure, and conciseness. The revised Introduction now focuses more directly on the research context, rationale, and aims, while the Background section provides supporting national context regarding school closures, caregiving burden, and economic impact

“1. Introduction

The COVID-19 pandemic has posed unprecedented challenges to mental health worldwide, leading to a substantial rise in related issues across the globe [1]. Data comparing the pre-pandemic period to 2020 shows a notable surge in depression prevalence across 15 major OECD (Organisation for Economic Co-operation and Development) countries, with most experiencing more than a twofold increase. South Korea emerged with the highest recorded level of depression at 36.8% and was thus the leading nation [2].

Women’s mental health has emerged as a critical concern during the pandemic, with studies showing they are disproportionately affected compared to men [1,3]. Investigations in countries including the US, Canada, the UK, Italy, China, and Chile have reported more pronounced adverse effects on women’s mental health, which emphasizes a rise in depression, anxiety, and distress rates [3-8]. In South Korea, the research on gender differences in mental health during COVID-19 has yielded varied results. Earlier studies that investigated mental health shortly after the outbreak (beyond day 55) found no significant gender differences [9], but subsequent analyses have revealed disparities between men and women and especially at certain age groups [10, 11].

Two key factors often cited as contributors to the gender gap in mental health during COVID-19 are employment disruptions and childcare burdens. First, the pandemic has severely affected women’s participation in the labor market, which led to deteriorated mental health outcomes for females [4, 10]. In contrast to earlier economic recessions, which primarily impacted male-dominated industries such as manufacturing, the COVID-19 crisis has disproportionately struck women’s employment, especially in the service sector [12]. Data from South Korea in 2020 show a marked decline in the employment rate of married women compared with married men during the pandemic [13], which highlights a change in the way economic crises influence gender-specific employment and mental health. Furthermore, the increased burden of childcare, which is intensified by the shutdown of daycare centers and schools, has placed a disproportionate strain on women, who typically assume the childcare role, creating a significant source of stress [4, 10, 14, 15].

Existing research frequently relies on self-reported, cross-sectional surveys, which can be biased by individuals' subjective interpretations of their mental health symptoms. To address these limitations, our study utilizes observed data from the National Health Insurance Database (NHID) of South Korea, encompassing healthcare utilization and socioeconomic factors such as income level, employment status, and family structure. We focus on the gender-specific mental health impacts of the COVID-19 pandemic, especially targeting a cohort of insurance subscribers with employee health insurance status in 2019, right before the pandemic. Depression was selected as the primary mental health outcome because it is one of the most prevalent and disabling conditions globally and is strongly associated with key pandemic-related stressors such as job loss and increased caregiving burden [16]. Depression can escalate to severe outcomes such as suicide, making early detection and treatment a priority for public health prevention [17]. Among various mental health conditions, depression is also more consistently recorded in administrative health data due to its clear diagnostic coding and treatment pathways, allowing for reliable identification in large-scale datasets [18]. Our retrospective cohort study aims to identify risk factors for newly diagnosed cases of depression in the employed population during the pandemic. Based on previous literature, we hypothesize that working women, particularly those who have lost jobs or have young children, have been more adversely affected [1, 3-8, 10-11, 14, 15l. This research has important policy implications because it supports the development of more effective health and economic interventions for vulnerable populations concerning mental health, which enhances preparedness for future health crises.

 

2. Background

In South Korea, the COVID-19 pandemic led to one of the longest periods of school closure among major economies. Beginning with the first confirmed case in January 2020, the government implemented a series of partial and full school shutdowns aimed at preventing in-school transmission. The start of the spring semester in 2020 was delayed multiple times, and even after resumption, in-person classes remained restricted until the first half of 2022. According to UNESCO (United Nations Educational, Scientific and Cultural Organization), South Korea experienced a total of 76 weeks of partial or full school closures from March 2020 to October 2021—surpassing durations observed in the United States (71 weeks), Germany (38), the United Kingdom (27), China (27), and Japan (11) [19].

These prolonged disruptions to childcare and education services substantially increased the caregiving burden on households, particularly for working mothers. An international survey conducted by Boston Consulting Group in 2020 across five countries—the US, UK, Germany, Italy, and France—found that parents nearly doubled their weekly hours spent on childcare and domestic work during the pandemic [20]. Similarly, a study conducted in Ireland found that the surge in homeschooling brought on by school closures was associated with increased negative emotions, highlighting the emotional strain parents—especially mothers—faced during this period [21]. In Korea, although nationally representative time-use data for this period is unavailable, a 2020 survey of parents with children under the age of nine showed that 69.7% of mothers experienced increased stress related to caregiving, compared to 54.1% of fathers. This suggests a heightened caregiving burden disproportionately affecting women [22].

The economic consequences of the pandemic further widened existing gender inequalities. According to an analysis by the Bank of Korea, women were more likely to experience employment disruptions than men—particularly among those with young children. Korean women were also more likely to work in face-to-face service sectors, limiting their ability to work remotely. Additionally, they were more frequently employed in temporary or daily-wage positions, with 21.7% of women in such jobs compared to 10.3% of men, making them more susceptible to job loss and economic insecurity [23].

These structural vulnerabilities translated into mental health challenges. A national survey conducted by the Ministry of Health and Welfare in 2022 found that individuals who experienced income loss during the pandemic were nearly twice as likely to be at risk of depression (22.1%) compared to those whose income remained stable (11.5%) [24]. When combined with increased caregiving responsibilities and limited employment flexibility, these factors may have exacerbated mental health burdens, particularly for working mothers.”

 

 

11. Material and Methods

LN/76-89—add reference

NHIS—detailed then abbreviate

The most descriptive methodologies are without references

Is there no reference for statistical analysis ??

There is no plan for the study area

M& Methods----It is very long and not well organized

Rewrite it again

 

Thank you for your helpful suggestions regarding the Materials and Methods section. In response, we have added a citation describing the National Health Information Database (NHID), which provides context on its structure, representativeness, and suitability for health services research [Kim et al., 2022]. We have also included an appropriate reference for the statistical methodology used in the study, specifically the Cox proportional hazards regression analysis, which served as the primary analytical approach [Ramaswami et al., 2021

“The National Health Information Database (NHID), developed and maintained by the National Health Insurance Service (NHIS), is a nationwide administrative health dataset that includes medical claims data, insurance eligibility, and demographic information for approximately 98% of the Korean population [25]. As enrollment in the NHIS is mandatory for all residents of South Korea, the NHID serves as the most comprehensive and representative national data source for analyzing healthcare utilization and socio-demographic factors. The database is constructed from healthcare provider claims submitted for reimbursement, allowing for detailed tracking of medical diagnoses, procedures, prescriptions, and service utilization. Its use of administrative health data enables objective measurement of clinical outcomes and minimizes recall bias common in self-reported surveys [25].”

 

“multivariable Cox proportional hazard regression analyses were performed to estimate the hazard ratios (HRs) and 95% confidence intervals [29].”

Kim, M. K., Han, K., & Lee, S. H. (2022). Current trends of big data research using the Korean National Health Information Database. Diabetes & Metabolism Journal, 46(4), 552–563. https://doi.org/10.4093/dmj.2022.0193

Ramaswami, S., Mohan, P., & Ghoshal, U. C. (2021). Survival analysis: A primer for the clinician scientists. Indian Journal of Gastroenterology, 40(5), 541–549. https://doi.org/10.1007/s12664-021-01199-8

 

 

Results

12. What about the clinical symptoms of the diseased patients  

 

Thank you for your comment. We have now specified the clinical diagnostic categories associated with the mental health conditions examined in the study

“The relevant disease codes, based on the Korean Standard Classification of Diseases (KCD-8), include F32 (Depressive episode), F33 (Recurrent depressive disorder), F34 (Persistent mood [affective] disorders), F38.1 (Other recurrent mood [affective] disorders), F40 (Phobic anxiety disorders), F41 (Other anxiety disorders), F42 (Obsessive-compulsive disorder), and F43 (Reaction to severe stress and adjustment disorders)”

13. What about the mortality percentage

 

Thank you for your comment. We have added information on mortality outcomes during the follow-up period.

“Regarding follow-up outcomes, the COVID-period mortality rate among the sample ranged from 0.11% to 0.16%, with 314 individuals dying in 2020, 385 in 2021, and 482 in 2022.”

 

14. What about the PM changes

Describe the other available methods for diagnosis and vaccination

How did the people go back to work after pandemic closure

It is very long

Rewrite it again

 

Thank you for your comments. The questions regarding changes in PM, alternative methods for diagnosis and vaccination, and the return-to-work process after pandemic closures are important public health considerations. However, these aspects were not directly analyzed in our study due to data limitations. Our analysis relied on administrative health data that did not include detailed information on vaccination types, diagnostic testing strategies, or individual-level return-to-work pathways. As a result, we did not include these topics in the main text to maintain the focus and integrity of our analysis.

 

 

15. Discussion

Authors should discuss the results obtained with those of the previous investigators’ results

Rewrite it again

Conclusion

There is no conclusion –why ?

 

Thank you for your helpful comment. We have newly added a Conclusion section to incorporate comparisons with previous literature, as follows:

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

 

16. References

Some cited references need to be more updated

LN/290—delete (April.15) etc., apply for all

 

 

All references should be rewritten

Some cited references with missing data—recheck

As volume/issue/pages/number—all available—so no need for the link(s)—apply for all

 

 

Thank you for your thorough feedback regarding the references. In response, we have made the following revisions:

1. Updated References: We reviewed all cited sources and updated several references to include more recent and relevant literature.
2. Formatting Corrections: All references have been revised to conform to the required journal style, including removal of unnecessary date notations such as “(April 15)” and hyperlinks. Volume, issue, page numbers, and DOIs have been added wherever applicable.
3. Consistency Check: We carefully cross-checked all in-text citations

Major comments

 

1. There are no recommendations or limitations for this study

 

Thank you for your thoughtful comment. In response, we elaborated on the discussion section to include more detailed policy recommendation, and we expanded the conclusion section to address the study’s limitations as follows:

 

[Revised Discussion Section] (page 11)
“These findings underscore the urgent need for gender-sensitive policy responses to pandemic-induced mental health inequalities. International research has consistently emphasized that women disproportionately bore the burden of unpaid caregiving during COVID-19, contributing to psychological distress and labor market detachment [15, 35]. To address these disparities, governments should prioritize expanding affordable childcare services and increasing public investment in care infrastructure [36].

Moreover, flexible work arrangements—such as telecommuting and adjustable work hours—have been identified as protective factors that mitigate mental health deterioration among working women [38-39]. Policymakers should encourage employer adoption of these practices beyond the pandemic. Finally, targeted mental health interventions, including subsidized psychological services for high-risk groups (e.g., single mothers, high-income professionals facing burnout), are essential to build long-term resilience.”

 

[Revised Conclusion Section (page 12)
“Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results”.

 

 

2. There is no conclusion

 

Thank you for your important comment. In response, we have added a dedicated Conclusion section following the Discussion to summarize the main findings, highlight policy implications, and acknowledge key limitations and directions for future research. The newly added text reads as follows

“6. Conclusion

This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].

A notable strength of this study lies in its use of objective, clinically diagnosed mental health outcomes, rather than self-reported survey data. This enhances the validity of its findings and distinguishes it from many prior studies relying on subjective measures of psychological distress [35, 40].

Nevertheless, the study has several limitations. First, due to the inherent nature of administrative health data, the analysis lacked important explanatory variables such as telecommuting practices, household division of caregiving labor, or access to mental health services—all of which are known to mediate pandemic-related mental health outcomes [15, 35]. In addition, the dataset does not include information to distinguish between voluntary and involuntary job loss, which limits interpretation of the employment-related mental health effects. Second, the use of health insurance premiums as a proxy for income may not fully capture household economic vulnerability, a limitation often cited in administrative labor studies [42]. Third, the analysis focused exclusively on depression and did not include other critical dimensions of mental health, such as anxiety, burnout, or stress, thereby narrowing the scope of interpretation [41]. Fourth, our definition of depression that only counts individuals utilizing healthcare services could introduce detection bias. Finally, the cross-sectional design limits causal inference, as it does not track mental health trajectories before and after the pandemic, unlike panel or repeated-measures studies that can capture psychological change over time [40]. In addition, as this study is based solely on South Korean data and institutional context, the generalizability of the findings to other countries with different labor market structures, cultural norms, and public health systems may be limited. Comparative studies across diverse national settings are needed to validate the broader applicability of these results.

Despite these limitations, this Korea-based study makes an important empirical contribution to understanding pandemic-induced mental health inequalities. The findings offer evidence to inform gender-sensitive and equity-oriented public health responses.”

 

 

3. There is no simple summary as all MDPI journals

 

Thank you for your helpful comment. In response, we have added in conclusion section.

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups.”

 

4. There is no ethical approval statement

Thank you for pointing this out. In response, we have added the ethical approval statement in the Back Matter section of the manuscript, as follows:

“Institutional Review Board Statement: The study protocol was reviewed and exempted by the Institutional Review Board of Incheon National University (IRB No. 7007971-202204-002), as it involves analysis of anonymized secondary data.”

 

5. There are no authors’ contributions

Thank you for pointing this out. In response, we have added authors’ contributions in the Back Matter section of the manuscript, as follows:

 

“Author Contributions: Conceptualization, LEE S.J., WEE H.S., JUNG S.H., LEE J.M.; Data curation, LEE S.J.; Formal analysis, LEE S.J.; Methodology, LEE S.J.; Project administration, JUNG S.H.; Writing—original draft preparation, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.; Writing—review and editing, LEE S.U., WEE H.S., JUNG S.H., LEE J.M.”

 

6. There are no acknowledgements

 

Thank you for pointing this out. In response, we have added Acknowledgments in the Back Matter section of the manuscript, as follows:

 

“Acknowledgments: This study was conducted using data provided by the National Health Insurance Service (NHIS) of Korea. The research management number assigned by the NHIS is NHIS-2024-1-525.”

 

7. What is /are the creativity of this study

 

Thank you for your important question. The main contribution and originality of this study lie in its use of large-scale, objective administrative health data to examine pandemic-related mental health outcomes. Furthermore, the study explores key social factors—such as job loss, childcare burden, and gender differences—providing a more nuanced understanding of mental health disparities during the COVID-19 pandemic. These points have been reflected in the revised Abstract as follows:

“Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression.”

 

Detailed comments

8. Abstract : There are no highlights –why ? There is no graphical abstract Abstract is very short—expand your data, Abstract should be divided into backgrounds/aims/methods/results and conclusion, LN/11—public health implications—describe in detail, LN/15-16—describe why ???LN/18-19—rewrite and be more concise

Thank you for the comment. We have rewritten the abstract upon your suggestion.

“This study investigates the impact of the COVID-19 pandemic on clinically diagnosed depression in South Korea, focusing on gender disparities and structural risk factors such as job loss and childcare burden. Although mental health inequalities have received growing attention during the pandemic, most existing research relies on self-reported survey data with inherent limitations. To address this gap, we utilized administrative health data from a 2% stratified random sample of the total population (N = 297,368) in the National Health Insurance Database, focusing on employed individuals without a prior history of depression. Multivariable Cox proportional hazard regression revealed that women had significantly higher risks of depression than men, particularly among those in their 20s to 40s, those who experienced job loss, had children aged 7–9, or belonged to high-income groups. These findings suggest that the intersection of employment instability and caregiving responsibilities disproportionately affected women’s mental health during the pandemic. The results underscore the urgent need for gender-sensitive public health policies that expand childcare support, institutionalize flexible work arrangements such as telecommuting, and enhance access to targeted mental health services to reduce pandemic-induced gender disparities in mental health.”

 

 

9. LN/23—add South Korea/Viral diseases/Pandemic and socio-demographic criteria to the keywords

 

Thank you for the comment. As you suggested, I have update keywords as following:

“Keywords:  Mental health; Depression; Gender disparities; Job loss; Childcare burden; South Korea; Viral diseases; Pandemic and socio-demographic criteria”

 

 

10. Introduction

LN/27/29/37/42/59---add references

LN/29—OECD—detailed then abbreviate

What is /are the differences between pandemics/parademics and epidemics

LN/60---cross-sectional surveys--/self-reported—clarify role/each

The introduction is y very long –why ? rewrite it gain and be more concise  

Aims should be more clarified

Novelty needs to be more highlighted

 

Thank you for your detailed and insightful comments regarding the Introduction. In response, we have substantially revised the introduction and added a new “Background” section to improve clarity, structure, and conciseness. The revised Introduction now focuses more directly on the research context, rationale, and aims, while the Background section provides supporting national context regarding school closures, caregiving burden, and economic impact

“1. Introduction

The COVID-19 pandemic has posed unprecedented challenges to mental health worldwide, leading to a substantial rise in related issues across the globe [1]. Data comparing the pre-pandemic period to 2020 shows a notable surge in depression prevalence across 15 major OECD (Organisation for Economic Co-operation and Development) countries, with most experiencing more than a twofold increase. South Korea emerged with the highest recorded level of depression at 36.8% and was thus the leading nation [2].

Women’s mental health has emerged as a critical concern during the pandemic, with studies showing they are disproportionately affected compared to men [1,3]. Investigations in countries including the US, Canada, the UK, Italy, China, and Chile have reported more pronounced adverse effects on women’s mental health, which emphasizes a rise in depression, anxiety, and distress rates [3-8]. In South Korea, the research on gender differences in mental health during COVID-19 has yielded varied results. Earlier studies that investigated mental health shortly after the outbreak (beyond day 55) found no significant gender differences [9], but subsequent analyses have revealed disparities between men and women and especially at certain age groups [10, 11].

Two key factors often cited as contributors to the gender gap in mental health during COVID-19 are employment disruptions and childcare burdens. First, the pandemic has severely affected women’s participation in the labor market, which led to deteriorated mental health outcomes for females [4, 10]. In contrast to earlier economic recessions, which primarily impacted male-dominated industries such as manufacturing, the COVID-19 crisis has disproportionately struck women’s employment, especially in the service sector [12]. Data from South Korea in 2020 show a marked decline in the employment rate of married women compared with married men during the pandemic [13], which highlights a change in the way economic crises influence gender-specific employment and mental health. Furthermore, the increased burden of childcare, which is intensified by the shutdown of daycare centers and schools, has placed a disproportionate strain on women, who typically assume the childcare role, creating a significant source of stress [4, 10, 14, 15].

Existing research frequently relies on self-reported, cross-sectional surveys, which can be biased by individuals' subjective interpretations of their mental health symptoms. To address these limitations, our study utilizes observed data from the National Health Insurance Database (NHID) of South Korea, encompassing healthcare utilization and socioeconomic factors such as income level, employment status, and family structure. We focus on the gender-specific mental health impacts of the COVID-19 pandemic, especially targeting a cohort of insurance subscribers with employee health insurance status in 2019, right before the pandemic. Depression was selected as the primary mental health outcome because it is one of the most prevalent and disabling conditions globally and is strongly associated with key pandemic-related stressors such as job loss and increased caregiving burden [16]. Depression can escalate to severe outcomes such as suicide, making early detection and treatment a priority for public health prevention [17]. Among various mental health conditions, depression is also more consistently recorded in administrative health data due to its clear diagnostic coding and treatment pathways, allowing for reliable identification in large-scale datasets [18]. Our retrospective cohort study aims to identify risk factors for newly diagnosed cases of depression in the employed population during the pandemic. Based on previous literature, we hypothesize that working women, particularly those who have lost jobs or have young children, have been more adversely affected [1, 3-8, 10-11, 14, 15l. This research has important policy implications because it supports the development of more effective health and economic interventions for vulnerable populations concerning mental health, which enhances preparedness for future health crises.

 

2. Background

In South Korea, the COVID-19 pandemic led to one of the longest periods of school closure among major economies. Beginning with the first confirmed case in January 2020, the government implemented a series of partial and full school shutdowns aimed at preventing in-school transmission. The start of the spring semester in 2020 was delayed multiple times, and even after resumption, in-person classes remained restricted until the first half of 2022. According to UNESCO (United Nations Educational, Scientific and Cultural Organization), South Korea experienced a total of 76 weeks of partial or full school closures from March 2020 to October 2021—surpassing durations observed in the United States (71 weeks), Germany (38), the United Kingdom (27), China (27), and Japan (11) [19].

These prolonged disruptions to childcare and education services substantially increased the caregiving burden on households, particularly for working mothers. An international survey conducted by Boston Consulting Group in 2020 across five countries—the US, UK, Germany, Italy, and France—found that parents nearly doubled their weekly hours spent on childcare and domestic work during the pandemic [20]. Similarly, a study conducted in Ireland found that the surge in homeschooling brought on by school closures was associated with increased negative emotions, highlighting the emotional strain parents—especially mothers—faced during this period [21]. In Korea, although nationally representative time-use data for this period is unavailable, a 2020 survey of parents with children under the age of nine showed that 69.7% of mothers experienced increased stress related to caregiving, compared to 54.1% of fathers. This suggests a heightened caregiving burden disproportionately affecting women [22].

The economic consequences of the pandemic further widened existing gender inequalities. According to an analysis by the Bank of Korea, women were more likely to experience employment disruptions than men—particularly among those with young children. Korean women were also more likely to work in face-to-face service sectors, limiting their ability to work remotely. Additionally, they were more frequently employed in temporary or daily-wage positions, with 21.7% of women in such jobs compared to 10.3% of men, making them more susceptible to job loss and economic insecurity [23].

These structural vulnerabilities translated into mental health challenges. A national survey conducted by the Ministry of Health and Welfare in 2022 found that individuals who experienced income loss during the pandemic were nearly twice as likely to be at risk of depression (22.1%) compared to those whose income remained stable (11.5%) [24]. When combined with increased caregiving responsibilities and limited employment flexibility, these factors may have exacerbated mental health burdens, particularly for working mothers.”

 

 

11. Material and Methods

LN/76-89—add reference

NHIS—detailed then abbreviate

The most descriptive methodologies are without references

Is there no reference for statistical analysis ??

There is no plan for the study area

M& Methods----It is very long and not well organized

Rewrite it again

 

Thank you for your helpful suggestions regarding the Materials and Methods section. In response, we have added a citation describing the National Health Information Database (NHID), which provides context on its structure, representativeness, and suitability for health services research [Kim et al., 2022]. We have also included an appropriate reference for the statistical methodology used in the study, specifically the Cox proportional hazards regression analysis, which served as the primary analytical approach [Ramaswami et al., 2021

“The National Health Information Database (NHID), developed and maintained by the National Health Insurance Service (NHIS), is a nationwide administrative health dataset that includes medical claims data, insurance eligibility, and demographic information for approximately 98% of the Korean population [25]. As enrollment in the NHIS is mandatory for all residents of South Korea, the NHID serves as the most comprehensive and representative national data source for analyzing healthcare utilization and socio-demographic factors. The database is constructed from healthcare provider claims submitted for reimbursement, allowing for detailed tracking of medical diagnoses, procedures, prescriptions, and service utilization. Its use of administrative health data enables objective measurement of clinical outcomes and minimizes recall bias common in self-reported surveys [25].”

 

“multivariable Cox proportional hazard regression analyses were performed to estimate the hazard ratios (HRs) and 95% confidence intervals [29].”

Kim, M. K., Han, K., & Lee, S. H. (2022). Current trends of big data research using the Korean National Health Information Database. Diabetes & Metabolism Journal, 46(4), 552–563. https://doi.org/10.4093/dmj.2022.0193

Ramaswami, S., Mohan, P., & Ghoshal, U. C. (2021). Survival analysis: A primer for the clinician scientists. Indian Journal of Gastroenterology, 40(5), 541–549. https://doi.org/10.1007/s12664-021-01199-8

 

 

Results

12. What about the clinical symptoms of the diseased patients  

 

Thank you for your comment. We have now specified the clinical diagnostic categories associated with the mental health conditions examined in the study

“The relevant disease codes, based on the Korean Standard Classification of Diseases (KCD-8), include F32 (Depressive episode), F33 (Recurrent depressive disorder), F34 (Persistent mood [affective] disorders), F38.1 (Other recurrent mood [affective] disorders), F40 (Phobic anxiety disorders), F41 (Other anxiety disorders), F42 (Obsessive-compulsive disorder), and F43 (Reaction to severe stress and adjustment disorders)”

13. What about the mortality percentage

 

Thank you for your comment. We have added information on mortality outcomes during the follow-up period.

“Regarding follow-up outcomes, the COVID-period mortality rate among the sample ranged from 0.11% to 0.16%, with 314 individuals dying in 2020, 385 in 2021, and 482 in 2022.”

 

14. What about the PM changes

Describe the other available methods for diagnosis and vaccination

How did the people go back to work after pandemic closure

It is very long

Rewrite it again

 

Thank you for your comments. The questions regarding changes in PM, alternative methods for diagnosis and vaccination, and the return-to-work process after pandemic closures are important public health considerations. However, these aspects were not directly analyzed in our study due to data limitations. Our analysis relied on administrative health data that did not include detailed information on vaccination types, diagnostic testing strategies, or individual-level return-to-work pathways. As a result, we did not include these topics in the main text to maintain the focus and integrity of our analysis.

 

 

15. Discussion

Authors should discuss the results obtained with those of the previous investigators’ results

Rewrite it again

Conclusion

There is no conclusion –why ?

 

Thank you for your helpful comment. We have newly added a Conclusion section to incorporate comparisons with previous literature, as follows:

 

“This Korea-based study provides empirical evidence on the gender-specific impacts of the COVID-19 pandemic on mental health among employed individuals. Women had a higher risk of depression than men specially among those who experienced job loss, had young children, or belonged to high-income professional groups. These results align with findings from other countries, including the UK and the US, which also reported intensified gender inequalities in mental health during the pandemic [15, 35, 40].

The study highlights how the intersection of employment instability and caregiving responsibilities—particularly for school-aged or younger children—exacerbated psychological distress among women. This reflects international evidence that women are disproportionately burdened with unpaid care work, leading to role conflict, time poverty, and adverse mental health outcomes [15, 35, 41]. Interestingly, high-income women exhibited a 41% higher risk of depression than their lower-income counterparts, suggesting heightened pressure from work-life imbalance and professional expectations—echoing concerns in both high- and low-income countries regarding gendered responsibilities and insufficient caregiving infrastructure [42].”

 

 

16. References

Some cited references need to be more updated

LN/290—delete (April.15) etc., apply for all

 

 

All references should be rewritten

Some cited references with missing data—recheck

As volume/issue/pages/number—all available—so no need for the link(s)—apply for all

 

 

Thank you for your thorough feedback regarding the references. In response, we have made the following revisions:

1. Updated References: We reviewed all cited sources and updated several references to include more recent and relevant literature.
2. Formatting Corrections: All references have been revised to conform to the required journal style, including removal of unnecessary date notations such as “(April 15)” and hyperlinks. Volume, issue, page numbers, and DOIs have been added wherever applicable.
3. Consistency Check: We carefully cross-checked all in-text citations