Support Needs and Stigma in Online Mental Health Discussions: Evidence from Large-Scale Comment Analysis

Purpose: This study aimed to characterize mental health discourse in user comments on YouTube videos about depression and bipolar disorder, focusing on how personal experience, stigma, and community building are expressed in digital interactions.Methods: A total of 100 videos were identified using the keywords “depression” and “bipolar disorder.” Using the YouTube API, video metadata and 16,241 comments were collected. Most videos were produced by individual doctors, media outlets, and general hospitals, focusing on treatment, diagnosis, and symptoms. After removing non-informative, duplicated, or commercial posts, 13,183 comments were analyzed. Using GPT-5, comments were classified into eight predefined categories.Results: “Personal Experience” was the most frequent category (9,551), followed by “Needs for Community Building: Emotional” (3,842) and “Needs for Community Building: Instrumental” (2,873). Actual “Community Building: Emotional Support” (1,345) and “Community Building: Instrumental Support” (549) appeared less frequently. Stigma-related discourse included “Social Stigma” (1,717) and “Internalized Stigma” (1,184), while “Non-evidence-based Advice” appeared in 1,146 comments. Co-occurrence analysis revealed strong associations between “Personal Experience” and “Needs for Community Building: Emotional” (.56), “Needs for Community Building: Instrumental” (.36), and “Internalized Stigma” (.33).Conclusion: Expert mental health video comment sections functioned primarily as experience- and emotion-centered spaces where users expressed substantial emotional and informational needs but received limited reciprocal support. These findings highlight the need for digital mental health environments integrating evidence-based information with structured support frameworks.