Digital health literacy among people with bipolar disorder in Germany – a cross-sectional survey

Linda Kokwaro^1,2*Helena Krüger¹Dennis Stratmann^1,3Daniel Schulze⁴Daniel Fürstenau^5,6Surjo R Soekadar¹Sonia Lech¹Stefanie Schreiter¹

• ¹Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität zu Berlin, Department of Psychiatry and Neurosciences, and Berlin Institute of Health, Berlin, Germany
• ²Einstein Center for Neuroscience, Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität Berlin, and Berlin Institute of Health, Berlin, Germany
• ³International Psychoanalytic University, Berlin, Germany
• ⁴Charité - Universitätsmedizin Berlin, Institute of Biometry and Clinical Epidemiology, Berlin, Germany
• ⁵Freie Universität Berlin, School of Business & Economics, 14195 Berlin, Berlin, Germany
• ⁶Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität Berlin, Institute of Medical Informatics 10117, Berlin, Germany

Background: This study assesses digital health literacy (DHL) among individuals with bipolar disorder (BD) in Germany, employing Sørensen's Integrated Model of Health Literacy (SIMHL). According to SIMHL, health literacy is a dynamic construct influenced by demographic, structural, and situational factors. With increased adoption of digital health resources, often overlooked are the skills required to use these tools. This study aims to bridge this gap by examining overall patterns of DHL among individuals with BD and employing SIMHL to investigate DHL the role of associated structural, situational/clinical, and technical access factors. Methods: A cross-sectional online survey was conducted comprising 212 individuals with self-reported BD. DHL was assessed using the HL-DIGI instrument in addition to SIMHL factors including: structural (gender, employment status), situational (manic and depressive symptomatology), and technology use factors (use of health-related websites). Structural Equation Modeling (SEM) was used to model DHL as a latent variable. Results: DHL followed a bimodal distribution, with participants clustering at either low (33.5%) or high (31.6%) levels. Evaluating reliability of online health information and using it to solve health-related tasks proved to be the most challenging. SEM demonstrated overall adequate model fit with higher DHL significantly associated with male gender, being employed and use of health-related websites. Discussion: In line with SIMHL, DHL in individuals with BD is shaped by structural and access-related factors rather than illness severity. Gender, employment status, and engagement with health-related websites emerged as key correlates. Conclusion: DHL in BD is not uniformly lower compared to the general population, underlining the feasibility of digital interventions as a promising pillar of care. At the same time, substantial heterogeneity highlights the need for routine DHL screening and tailored support for subgroups with lower skills. Future interventions and studies should systematically stratify by DHL level and target higher-order skills such as evaluating reliability and decision-making, in order to ensure equitable access and maximize the benefits of digital mental health care.