Poster Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2026

Digital traces of psychological distress in an online brain tumour support community: a computational linguistic analysis (145875)

Julia L Robertson 1 , Louis Sanzogni 1 , Nagesh Shukla 1 , Khondker M Zobair 1
  1. Griffith University, Nathan, QLD, Australia

Background:

People affected by brain tumours experience high levels of psychological distress, yet distress is often assessed only during clinical encounters. Online peer support communities may offer an opportunity to identify emerging psychosocial concerns through natural language. This study examined linguistic markers of distress, trajectories preceding suicidal ideation disclosure, and patterns of community engagement within a large online brain tumour support community.

Methods:

A total of 111,315 comments posted between 2013 and 2026 by 10,119 users in a Reddit-based brain tumour support community were analysed using computational linguistic methods. Eight distress-related language categories were assessed and combined into a composite distress score. Analyses included validation of established linguistic markers, longitudinal modelling of distress trajectories, user clustering, and assessment of COVID-19-related changes in community language.

Results:

Suicidal ideation or self-harm language was identified in 154 comments (0.14%) from 121 users. Comments containing suicidal ideation language showed significantly higher rates of first-person pronoun use than non-flagged comments (7.37 vs 5.15 per 100 words, p<0.0001), replicating established psycholinguistic findings. Among users with sufficient posting history before the first disclosure, 56.5% demonstrated rising distress trajectories before suicidal ideation language appeared. Three distinct user profiles emerged: Long-term Community Regulars (82.8%), Brief/Newcomer Burst Posters (16.6%), and High-Distress Disclosers (0.6%). COVID-19-related discussions were associated with increased isolation language and higher distress scores.

Conclusions:

Linguistic analysis of online brain tumour support communities may identify early signals of psychological distress and potential escalation before explicit disclosure. These findings support further investigation of AI-enabled approaches to complement psycho-oncology care, while highlighting the need for prospective validation and ethical implementation.

  1. Rude S, Gortner EM, Pennebaker JW. Language use of depressed and depression-vulnerable college students. Cognition and Emotion. 2004;18(8):1121-1133.
  2. Rai S, Stade EC, Giorgi S, Francisco A, Ungar LH, Curtis B, Guntuku SC. Key language markers of depression on social media depend on race. Proceedings of the National Academy of Sciences. 2024;121:e2319837121.
  3. Al-Mosaiwi M, Johnstone T. In an absolute state: Elevated use of absolutist words is a marker specific to anxiety, depression, and suicidal ideation. Clinical Psychological Science. 2018;6(4):529-542
  4. Yeskuatov E, Chua SL, Foo LK. Leveraging Reddit for suicidal ideation detection: A review of machine learning and natural language processing techniques. International Journal of Environmental Research and Public Health. 2022;19:10347.
  5. Nunez JJ, Leung B, Ho C, Ng RT, Bates AT. Predicting which patients with cancer will see a psychiatrist or counsellor from their initial oncology consultation document using natural language processing. Communications Medicine. 2024;4:69.