Mental illness discussion variations on X: topic modelling and sentiment analysis

• Javier Domingo-Espiñeira ^[1] ; Andrea Varaona-Santos ^[3] ; Francisco J. Lara Abelenda ^[4] ; María Montero ^[3] ; Emilio Fernandez-Egea ^[2] ; Leticia I Muñoz-Manchado ^[5] ; Miguel A. Ortega ^[6] ; Melchor Álvarez-Mon ^[7] ; Miguel Ángel Álvarez-Mon ^[8]
  1. [1] Universidad a Distancia de Madrid

     Universidad a Distancia de Madrid

     Madrid, España

     
  2. [2] University of Cambridge

     University of Cambridge

     Cambridge District, Reino Unido

     
  3. [3] Department of Medicine and Medical Specialities, Faculty of Medicine and Health Sciences, University of Alcal´a, ´Alcala de Henares, Spain
  4. [4] Department of Signal Theory and Communications, Telematics, and Computing Systems, Rey Juan Carlos University, Fuenlabrada, Spain
  5. [5] Unidad de Gesti´on Clínica de Salud Mental, ´Area de Gesti´on Sanitaria Norte de C´adiz, Hospital Universitario de Jerez de la Frontera, C´adiz, Espa˜na; Severe Mental Disorder Research Group, Department of Neuroscience, University of C´adiz, Instituto de Investigaci´on e Innovaci´on Biom´edica de C´adiz (INiBICA), C´adiz, Spain
  6. [6] Department of Medicine and Medical Specialities, Networking Research Center for Liver and Digestive Diseases (CIBEREHD), University of Alcal´a, 28801, Alcal´a de Henares, Spain ;Ram´on y Cajal Institute of Sanitary Research (IRYCIS), 28034, Madrid, Spain
  7. [7] Department of Medicine and Medical Specialities, Faculty of Medicine and Health Sciences, University of Alcal´a, ´Alcala de Henares, Spain;Department of Medicine and Medical Specialities, Networking Research Center for Liver and Digestive Diseases (CIBEREHD), University of Alcal´a, 28801, Alcal´a de Henares, Spain ;Ram´on y Cajal Institute of Sanitary Research (IRYCIS), 28034, Madrid, Spain
  8. [8] Department of Medicine and Medical Specialities, Faculty of Medicine and Health Sciences, University of Alcal´a, ´Alcala de Henares, Spain; CIBERSAM-ISCIII (Biomedical Research Networking Centre in Mental Health), Madrid, Spain

  Mostrar afiliaciones +
• Localización: European journal of psychiatry, ISSN 0213-6163, Vol. 39, Nº 2, 2025
• Idioma: inglés
• DOI: 10.1016/j.ejpsy.2025.100300
• Enlaces
  • Texto completo
• Resumen

  • Background and objectives: Social Media Listening has become essential for understanding societal attitudes about mental health. However, there is a lack of studies focusing on Spanish-language discussions and long-term an- alyses. This study aims to investigate the evolution of Spanish and English public discourse on nine psychiatric disorders over 15 years on X, identifying frequently addressed topics and associated emotions.

    Methods: We collected and analyzed tweets in English and Spanish from 2007 to 2022, referring to nine psy- chiatric disorders (autism, ADHD, schizophrenia, bipolar disorder, depressive episode, anxiety, addictions, insomnia, and suicide). Advanced topic modeling and sentiment analysis were applied.

    Results: The study revealed differences in tweet volumes across disorders and languages. Anxiety was the most prevalent term in both Spanish and English tweets. Spanish discussions emphasized insomnia and suicide at- tempts, while English tweets highlighted autism and schizophrenia. Tweets about bipolar disorder, depressive episode, and addictions were consistently low in both languages. Over time, Spanish tweets on insomnia, anxiety, and suicide attempts increased, while English tweets on autism, anxiety, suicide attempts, and schizophrenia spiked in 2018 and 2019. Topic modeling indicated that Spanish tweets focused on personal experiences with mental disorders and their impact on life, primarily associated with anger, fear, sadness, and joy. English tweets centered on social awareness and advocacy, eliciting fear, followed by treatment and prevention, associated with fear and sadness.

    Conclusions: Despite fewer discussions on X about severe mental disorders like bipolar disorder, schizophrenia, and addictions, X serves as a supportive environment for mental health discourse. Spanish-speaking users share personal experiences, fostering self-expression, while English users promote social awareness and advocacy. This highlights X’s role in normalizing mental health conversations, reducing stigma, and increasing support.