Back to 2019 MSM Agenda
Session Description: The purpose of the Intensive Longitudinal Health Behaviors initiative is to establish a cooperative agreement network of U01 projects and 1 U24 Research Coordinating Center (RCC), to collaboratively study factors that influence key health behaviors in the dynamic environment of individuals, using intensive longitudinal data collection and analytic methods. The network will also assess how study results can be leveraged to introduce innovations into longstanding behavioral theories to advance the field of theory-driven behavior change interventions. The knowledge gained will inform the development of personalized prevention strategies and best implementation strategies for communities, including health disparity populations, towards the goal of reducing disease risk and maintaining ideal health.
Behavioral science places strong emphasis on theoretical models to systematically explain and predict behaviors and events influencing health outcomes. Although these theories are useful frameworks for developing behavioral change interventions, their ability to explain and predict behavior has been only modestly successful. The research funded by this initiative will examine theoretical constructs and health behaviors from a different scientific perspective and approach than has been traditionally used and is critical for moving health behavior science towards more effective health behavior interventions for reducing disease.
Health behavior theories have developed and been evaluated primarily from a between-person perspective, attempting to explain why some people engage in health behaviors while others do not. While such questions remain important, this between-person focus has contributed to theoretical research that is predominately cross-sectional in nature and that emphasizes dispositional variables such as attitudes and normative beliefs which are relatively static over time and more trait-like in nature. In contrast, a within-person approach to health behavior theory research seeks to explain why a given individual engages in healthy or risky behaviors at one time versus another. Within-person analysis of intensive longitudinal data is likely to provide insight into the dynamic factors in the physical, social, and/or built environment that facilitate or hinder engaging in certain behaviors at specific points in time, in addition to the interaction between factors.
U01 FOA, https://grants.nih.gov/grants/guide/rfa-files/rfa-od-17-004.html
R24 FOA, https://grants.nih.gov/grants/guide/rfa-files/RFA-OD-17-005.html
2014 MSM Session on Behavioral Modeling
Speaker Bios and Abstracts:
Dr. Wolff-Hughes is a Health Scientist Administrator for the Office of Behavioral and Social Sciences Research (OBSSR). In this role, her primary responsibility is to provide technical guidance and direction to research efforts related to the development and evaluation of behavioral and social outcomes using mobile and wireless health technology. She directs the Intensive Longitudinal Health Behaviors initiative.
Professor Pavel holds a joint faculty appointment in the Northeastern University College of Computer & Information Science and Bouvé College of Health Sciences. His background comprises electrical engineering, computer science and experimental psychology, and his research is focused on multiscale computational modeling of behaviors and their control, with applications ranging from elder care to augmentation of human performance. Professor Pavel uses these model-based approaches to develop algorithms transforming unobtrusive monitoring from smart homes and mobile devices to useful and actionable knowledge for diagnosis and intervention.
Sy-Miin Chow is a Professor in the Department of Human Development and Family Studies at the Pennsylvania State University. The focus of her work has been on developing and adapting dynamic models such as differential equation models, time series models, and state-space models for use with intensive longitudinal data and other commonly encountered longitudinal data (e.g., panel data) in the social and behavioral sciences. She has developed novel methods that serve as practical alternatives for addressing data analytic problems such as variable and model selection in behavioral modeling, missing data, and nonlinearities/nonstationarities in human dynamics. She has engaged in numerous interdisciplinary projects with scientists from the physical sciences, engineering, computer science, statistics/biostatiscs, as well as scholars in emotion, aging, child development, family dynamics and prevention research.
Check out slides and posters from the ILHBN group here: https://www.imagwiki.nibib.nih.gov/content/ilhbn-theories-and-models
Interactive Discussion (please put you name before your comments):
@Regarding dynamic models that are ready to be used in apps, Billie Nahum Shani mentioned that relatively simple versions of the models we discussed have already been employed in commercial settings (see related publication below).
Bidargaddi N, Almirall D, Murphy S, Nahum-Shani I, Kovalcik M, Pituch T, Maaieh H, Strecher V. To Prompt or Not to Prompt? A Microrandomized Trial of Time-Varying Push Notifications to Increase Proximal Engagement With a Mobile Health App. JMIR mHealth and uHealth. 2018;6(11):e10123.
@Regarding study on clinicians' behavior, we are not aware of any intensive longitudinal studies focusing on clinicians' behaviors. But there is preliminary work that begins to tap into computational possibilities, such as this paper here. This is surely an interesting future direction! Sinclair, Katerina & Molenaar, Peter. (2008). Optimal control of psychological processes: A new computational paradigm. Bulletin de la Société des sciences médicales du Grand-Duché de Luxembourg. Spec No 1. 13-33.
@About ethical consideration: this is indeed a thorny issue. With regard to the studies on suicidal behavior and thoughts, mania/psychosis, and also the CoTwins study (which collect information such as song choices, conduct image analysis, etc.), these studies are aimed at prediction and explanation of behaviors, not intervention studies. Participants will have to give consent for sharing what they are willing to share with the researchers, and can withdraw anytime. Thus HIPAA policies still hold, so the researchers are not allowed to share any sensitive events or identifiable information with anyone outside of the research team without the participants' consent. Prediction is the first step. Figuring out how to best use the prediction results while keeping these ethical issues in mind is a different but also very challenging problem.
Comments From Elizabeth Ginexi, NIH - Here are some responses to questions posted below:
PAR-18-331 is not related to the Intensive Longitudinal Health Behaviors initative discussed in the session today. You can click on the links about the ILHBN U01 announcement url links above to read about it. The PAR you are asking about is just a general announcement issued by NIMHD that OBSSR is supportive of which does offer investigators an opportunity to apply computational models to study aspects related to health disparties in health. Keep in mind that it is not necessary for you to submit your computational modeling grants to a specific program announcement. You may always submit your grants to the NIH through the parent R01 and R21 FOAs. I would be happy to discuss your research ideas with you and help you locate an appropriate institute program contact and relevant study section to review your application.
For those who asked questions about the future promise of just in time behavioral interventions, this is an exciting new field. I do think there well may be some more general use for consumers in the future. I recommend a recent article on the subject if you want to learn more: https://www.ncbi.nlm.nih.gov/pubmed/27663578
To Misha, can you try and
Is there any projects to
Is there any projects to study clinician's behaviors? the goal is to improve healthcare delivery, reduce human errors.
Could the panelists discuss
Could the panelists discuss how potential ethical implications of data collection methods like monitoring teenagers' phone usage (text messages, music choices, etc.) are factored into this research? e.g., if these analysis techniques are shown to be highly effective, how can they be widely implemented in a way that is not seen as highly intrusive.
Is this program related to
Is this program related to the discussion today? https://grants.nih.gov/grants/guide/pa-files/PAR-18-331.html
To Misha, can you try and give examples of personal technological applications that can appear in the next 5 years that use models you mentioned and other similar models you are aware of. In other words, what models seem close enough for consumer use in the foreseeable future?