# 5 Years of bipolar disorder conversations on Reddit: Methods, key topics and future directions

**Authors:** Caden Poh, Cheyanne Head, John-Jose Nunez, Laura Lapadat, Emma Morton, Erin E. Michalak, Souparno Mitra, Souparno Mitra, Souparno Mitra

PMC · DOI: 10.1371/journal.pone.0338622 · 2026-03-06

## TL;DR

This paper analyzes five years of Reddit conversations about bipolar disorder to identify key topics and needs of people living with the condition.

## Contribution

The study introduces a novel application of topic modeling to extract and categorize patient priorities from online discussions about bipolar disorder.

## Key findings

- The most common topic was BD misdiagnosis and differential diagnoses.
- Other key topics included coping with daily struggles and medication use.
- Patients emphasized the need for better clinician communication on these issues.

## Abstract

Bipolar disorder (BD) is a prevalent mood disorder that can be associated with serious personal, societal, and economic costs. Growing attention is being paid to the importance of including the experiences of people living with BD in research and healthcare system advancement. However, there is still much to be learned about what people with BD are prioritizing, and where their needs are not yet being met in research and clinical settings. For the past five years, the Collaborative RESearch Team to study psychosocial issues in Bipolar Disorder (CREST.BD) has facilitated the world’s largest online BD question-and-answer event, an “Ask Me Anything” (AMA) hosted via Reddit. This event allows people internationally to submit BD-related questions to be answered by expert panelists. Altogether, 159 panelists with diverse expertise in BD have participated, and more than 2,000 questions addressed. After five years, the Reddit AMAs have become a large repository of questions that represent the needs and interests of individuals with BD and their supporters. This study used topic modeling, a natural language processing (NLP) method, to algorithmically extract key topics from AMA queries collected over five years. Queries were extracted using a python script and the Reddit API; BERTopic was used for topic modelling. Topic models of 10, 20, and an unrestricted number of topics were run; the 20 topic model was chosen as it best balanced specificity and breadth. The most common topic was BD misdiagnosis/differential diagnoses. Other topics included coping with daily struggles; understanding hypomania and suicidality; medication and use of substances such as psilocybin and ketamine; and supporting loved ones with BD. Future research in these areas would be beneficial for individuals with BD and their loved ones, and patients with BD may benefit from clinicians addressing these topics. Ethical issues are also discussed.

## Linked entities

- **Diseases:** bipolar disorder (MONDO:0004985), BD (MONDO:0007191)

## Full-text entities

- **Diseases:** Mental Illness (MESH:D001523), sleep disorders (MESH:D012893), anxiety disorder (MESH:D001008), Cyclothymia (MESH:D003527), ACADEMIC EDITOR (MESH:D007859), fatigue (MESH:D005221), elevated mood (MESH:D019964), weight gain (MESH:D015430), mental health (OMIM:603663), ADHD (MESH:D001289), Hypomania (MESH:D000087122), cognitive and memory difficulties (MESH:D003072), mental health condition (MESH:D000071069), allergy (MESH:D004342), Bipolar Disorder (MESH:D001714), Depression (MESH:D003866)
- **Chemicals:** Lithium (MESH:D008094), Lamictal (MESH:D000077213), Mitra (-), psilocybin (MESH:D011562), CBD oil (MESH:D002185), Alcohol (MESH:D000438)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Figures

4 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12965559/full.md

---
Source: https://tomesphere.com/paper/PMC12965559