The physical, mental, and economic costs of chronic diseases (e.g., obesity, heart disease, cancer, and diabetes) are astounding, with 7 out of 10 deaths and an estimated 75% of healthcare costs attributable to chronic diseases. Development of these chronic diseases is often driven by a lifetime of small decisions related to health behaviors such as physical activity, healthful eating, or tobacco use. Reducing the impact of chronic disease requires strategies for fostering life-long healthy behaviors.

Many believe that health behavior-change technologies (e.g., smartphone apps or wearable physical activity sensors that foster healthy behavior change) can play an important role in facilitating lifelong healthy behavior. Well there is great promise, at present, these technologies have not delivered, particularly for supporting long-term behavior change. How might we build health behavior-change technologies that can support life-long healthy behaviors?

The goal for this Dagstuhl Seminar is first to examine this question from a transdisciplinary perspective and then establish a transdisciplinary research agenda for addressing it. Sub-questions that will be explored include: what is the best way to design systems that are meant to be used for a very long time? Assuming personalization and adaptation over time will be important for supporting long-term use, what is the best way to design adaptations within the system? What are the most appropriate behavioral theories and how do we generate better specified dynamic mathematical models of behavior that can drive these adaptations? How does the ratio of the value a person gets from a system (e.g., information on how to live healthier) compared to the amount of burden placed on them for using it (e.g., needing to check-in regularly) shift over time? What are the implications of this ratio for the design of lifelong behavior-change technologies? Do individuals need just one system that will support them for a lifetime or, instead, will different apps and systems be more appropriate at different life stages? If the latter, what is the best strategy for curating digital health tools so that target users can find the right tool at the right time for them? Also, assuming past data from previous systems will be helpful for future systems, what strategies can we use to enable better sharing of data across systems? How do we evaluate the usability and impact of these life-long adapting systems?

This Dagstuhl Seminar will bring together researchers and practitioners from interdisciplinary fields such as computer science, particularly human-computer interaction (HCI), behavioral sciences such as psychology or sociology, and engineers to further articulate the grand challenges for developing effective technologies for long-term health behavior change, discuss the state of the art, and develop a roadmap for future research.