헬스리터러시와 주관적 정신건강의 상관관계 및 예측: 설명가능한 인공지능(XAI)을 활용한 탐색적 연구

Health has gained attention as a critical public health concern, particularly in the aftermath of the COVID-19 pandemic. This exploratory study examines the relationship between health literacy and perceived mental health issues using Explainable Artificial Intelligence (XAI). Data from the Korean Health Panel, including 11,057 adults aged 19 and older, were analyzed. Health literacy was measured with the EU-HLS-Q16 tool, while perceived mental health issues were assessed through self-reported levels of stress, depression, anxiety, and suicidal ideation. The findings suggest that lower health literacy levels are linked to higher risks of perceived mental health problems, including depression, anxiety, and suicidal ideation. Higher risks were observed among women, younger adults, individuals with chronic illnesses, and those with disabilities. However, due to class imbalance and limited data, the AI prediction model's performance remained exploratory. The XAI model helped visualize key variable interactions but highlighted the need for further performance improvements. This study serves as an initial exploration of the relationship between health literacy and perceived mental health issues. It underscores the potential of AI-based mental health screening systems while emphasizing the need for better data collection and model refinement to support early detection and prevention strategies.