Symptom Expression and Emotional Distress in Online Mental Health Narratives

Introduction: Online forums use as a source for mental health support has surged among individuals worldwide. Social media has become popular for sharing personal experiences and seeking information and support. Studies have analyzed posts on social media forums, focusing on frequency of engagement by users, why individuals engage, and how they engage on these web-based platforms. However, key questions about how mental illness is experienced, discussed and emotionally expressed from the user's perspective is needed to add important insight. Objective: This study aims to investigate how mental health related disorders and conditions are discussed, experienced, and emotionally framed in online discourse, specifically focusing on how mental health symptoms and distress language across mental health dialogues are expressed and examining the text-based communication beyond prevalence-based analyses and simplified sentiment analysis through symptom and experience-centered approach to uncover patterns in how symptoms are articulated and emotions expressed, and how distress is framed across multiple mental health conditions, by systematically analyzing digital textual data associated with various mental illnesses. Methodology: A retrospective observational design was conducted. The dataset used in this study was scrapped from YouTube between 11/2/2025 to 11/30/2025 by using a predefined keyword resulting in a total sample of 646 279 comments. The data was prepared and preprocessed using standard NLP procedures. Descriptive analysis of Disorder Representation, Symptom Expression Analysis, Emotional Tone and Distress Analysis, Cross-Disorder Statistical Comparisons and Crisis-Oriented Language Analysis was conducted. Kruskal–Wallis test a non-parametric analysis revealed no statistically significant differences in emotional proportion scores across mental health condition categories. Pearson’s chi-squared test indicted a robust and statistically significant association between mental health disorder type and symptom category. Result: Among the mental health conditions discussed online, content related to anxiety made up the largest count of the dataset (n = 125,001; 19.3%), followed by depression (n = 100,281; 15.5%) mental breakdown (n = 93,836; 14.5%), and PTSD (n = 110,935; 14.2%). However, Obsessive-Compulsive Disorder exhibited a robust engagement (Eng = 158.5); and panic attack–related posts showed higher levels of engagement (Eng = 189). Mental health conditions such as panic attacks (0.1050), anxiety (0.0532), depression (0.0490), and mental illness (0.0497) demonstrated intense emotions. For the category Anxiety Terms, the most negative terminology was recorded with the most negative sentiment score (−62,667). Stigmatizing revealed a net negative sentiment (−7,787) while Self-Disclosure also revealed a net negative sentiment (−2,344). Empathy showed the highest positive sentiment score (50,432), followed by Supportive (24,867) and Advocacy (4,897) categories.No statistically significant differences in emotional proportion scores across mental health disorder categories were revealed (X2 (6) = 0.118, p=1.000). However, a robust and statistically significant association between mental health disorder type and symptom category was identified (X2(18) =11, 623, p<0.001), suggesting that each mental health condition presents different symptom profiles across cognitive, emotional, and somatic dimensions. Conclusion: In conclusion, this study makes several important contributions to mental health research and practice in understanding mental illness as a lived experience rather than solely a diagnostic category. This finding also provides empirical support for conceptualizing OCD as a cognitive-based disorder, where distress is often expressed through intrusive thought patterns and not solely emotional states.