Changes in Depressive Symptoms Across the COVID-19 Pandemic and Recovery Periods Among Korean Adults: A Repeated Cross-Sectional Analysis of KNHANES 2018–2024

Background/Objectives: The COVID-19 pandemic was associated with changes in population mental health, but whether elevated depressive symptoms persisted during subsequent recovery periods remains unclear. This study examined changes in depressive symptoms before, during, and after the COVID-19 pandemic among Korean adults and explored whether temporal patterns differed by sex. Methods: Data were obtained from 21,228 adults aged ≥20 years who participated in the Korea National Health and Nutrition Examination Survey (KNHANES) in 2018, 2020, 2022, and 2024, representing the pre-pandemic, pandemic, early recovery, and later recovery periods, respectively. Clinically relevant depressive symptoms were defined as a Patient Health Questionnaire-9 (PHQ-9) score ≥ 10. Survey-weighted analyses accounting for the complex sampling design was performed, including multivariable logistic regression and period-by-sex interaction analysis. Results: The weighted prevalence of depressive symptoms was 4.3% in 2018, increased to 5.3% in 2020, decreased to 4.7% in 2022, and was 4.3% in 2024, the same point estimate as in 2018. In the fully adjusted model, the odds of depressive symptoms were higher in 2020 than in 2018 (odds ratio [OR] = 1.316, 95% confidence interval [CI]: 1.056–1.641), whereas the estimates for 2022 (OR = 1.187, 95% CI: 0.945–1.490) and 2024 (OR = 1.125, 95% CI: 0.905–1.398) were not statistically significant. Weighted mean PHQ-9 scores differed across periods (p = 0.037). Sex-stratified analyses indicated greater temporal variation among men; however, the period-by-sex interaction was not statistically significant (p = 0.069). Conclusions: Depressive symptoms showed modest worsening during the COVID-19 pandemic, but no statistically significant differences from 2018 were detected in 2022 or 2024. Although temporal variation appeared more pronounced among men, the period-by-sex interaction was not statistically significant (p = 0.069), and the observed sex-specific differences should be interpreted with caution.