(689g) Experimental Design for Estimating “Just-in-Time” States in Control-Oriented Behavioral Interventions for Physical Activity | AIChE

(689g) Experimental Design for Estimating “Just-in-Time” States in Control-Oriented Behavioral Interventions for Physical Activity

Authors 

Rivera, D., Arizona State University
Klasnja, P., University of Michigan, Ann Arbor
Park, J., University of California San Diego
Hekler, E., University of California San Diego
Kim, M., University of California San Diego
In spite of a wealth of information available regarding the benefits of physical activity (PA) to human health, insufficient levels of PA is a major issue affecting adults worldwide. Nearly 53% of adults in the US are sedentary [1]. The lack of engagement in sufficient amounts of PA leads to deterioration of body function and is a major contributor in the rise of chronic illnesses such as heart disease, cancer, obesity, and diabetes [2]. The question remains on the best methods to support individuals to engage and maintain healthy levels of PA. Control systems engineering has proven its benefits in many fields, with its adoption in behavioral medicine representing a growing area of research [3,4]. Efforts have been diverse, which include developing process engineering dynamic models of theories of health behavior using fluid analogies for the design and implementation of experimental trials (such as Just Walk) that enable the use of system identification principles to obtain dynamical models, and development of decision-making frameworks using Model Predictive Control towards the implementation and dissemination, on a large scale, of behavioral interventions that promote healthy levels of PA behavior [3,5,6].

Fluid analogies relying on chemical engineering principles such as conservation of mass enable mathematically representing theories of behavioral change as systems of ordinary differential equations (ODEs). In previous work, a fluid analogy representation of the Social Cognitive Theory (SCT) has been developed to describe dynamics at a daily level [5,6]. In the constructed computational model, interrelations between external stimuli including temperature, daily goals and goal attainment (depicted as inflows/outflows) and the main constructs of SCT such as behavior, self-efficacy, and behavioral outcomes (represented as inventories) have been proposed to model and predict engagement of an individual in healthy behavior. While earlier research findings have been instrumental, the scope of the work was focused on daily level dynamics and did not incorporate within-day dynamic processes in real-world settings that can facilitate (or hinder) engagement in PA. There is still a need for a better understanding of the multi-timescale dynamic processes that constitute an individual's behavior, which would allow for the prediction of the behavior over the timescales of interest, and by extension dynamic decision-making in behavioral interventions. The concept of the “just-in-time” adaptive interventions (JITAI) has been advanced in the field of behavioral science [7], where support is provided when the person has the need for a specific type of support, the receptivity to receive the support, and the opportunity to respond favorably to the provided support. This can facilitate the adaptation over time that yields a meaningful and sustained behavioral change. Thus, it is important to provide “just-in-time” support, which can be conceptualized as a multi-timescale problem, in which robust prediction and decision-making are needed in both the short timescale, and the longer timescale.

In this work we propose an innovative experimental design to identify “just-in-time” states and multi-timescale dynamics for a digital health intervention, Just Walk JITAI [8]. The experimental design relies on concepts from control engineering and system identification leading to input signals that elicit participant response resulting in an improved understanding of JIT behavior. The design leverages and improves on prior domain knowledge about processes that influence PA in the form of a dynamical model inspired by (SCT) [3,4,5], and are incorporated in our computational model. Two input signals are designed to address two principal components of the Just Walk JITAI intervention: a multisine signal to handle goal setting, and a combination of a pseudo-random binary signal (PRBS) with a three-level random sequence for “inspiring bouts” which define JIT support provision rules for need, receptivity, and opportunity. Each signal design ensures that there is enough excitation of states along different ends of the frequency range (lower frequency for the goal setting, and higher frequency for the inspiring bouts). The orthogonal design of the signals facilitates analysis of the data. The improved SCT computational model is used in simulations to analyze plausible responses for different kinds of participants to the proposed input signal design, in order to assess the quality of the data that will be produced out of the intervention. These simulations indicate that the proposed experimental design will result in experimental data of sufficient quality to capture the multi-timescale dynamics of behavioral change and lead to improved control-oriented interventions. These ideas are currently being evaluated in the Just Walk JITAI.

References:

[1] Centers for Disease Control and Prevention. (2022, February 17). Adult physical inactivity prevalence maps by race/ethnicity. Centers for Disease Control and Prevention. Retrieved From https://www.cdc.gov/physicalactivity/data/inactivity-prevalence-maps/index.html

[2] Saint-Maurice, P. F., Troiano, R. P., Bassett, D. R., Graubard, B. I., Carlson, S. A., Shiroma, E. J., Fulton, J. E., & Matthews, C. E. (2020). Association of Daily Step Count and step intensity with mortality among US adults. JAMA, 323(12), 1151. https://doi.org/10.1001/jama.2020.1382

[3] Rivera, D.E., Hekler, E.B., Savage, J.S., Downs, D.S. (2018). Intensively Adaptive Interventions Using Control Systems Engineering: Two Illustrative Examples. In: Collins, L., Kugler, K. (eds) Optimization of Behavioral, Biobehavioral, and Biomedical Interventions. Statistics for Social and Behavioral Sciences. Springer, Cham. https://doi.org/10.1007/978-3-319-91776-4_5

[4] Hekler, E. B., Klasnja, P., Riley, W. T., Buman, M. P., Huberty, J., Rivera, D. E., & Martin, C. A. (2016). Agile science: Creating useful products for behavior change in the real world. Translational Behavioral Medicine, 6(2), 317–328. https://doi.org/10.1007/s13142-016-0395-7

[5] Martin, C. A., Rivera, D. E., Hekler, E. B., Riley, W. T., Buman, M. P., Adams, M. A., & Magann, A. B. (2020). Development of a control-oriented model of social cognitive theory for optimized mhealth behavioral interventions. IEEE Transactions on Control Systems Technology, 28(2), 331–346. https://doi.org/10.1109/tcst.2018.2873538

[6] Freigoun, M. T., Martin, C. A., Magann, A. B., Rivera, D. E., Phatak, S. S., Korinek, E. V., & Hekler, E. B. (2017). System identification of just walk: A behavioral mhealth intervention for promoting physical activity. 2017 American Control Conference (ACC). https://doi.org/10.23919/acc.2017.7962940

[7] Perski, O., Hébert, E. T., Naughton, F., Hekler, E. B., Brown, J., & Businelle, M. S. (2021). Technology‐mediated just‐in‐time adaptive interventions (JITAIS) to reduce harmful substance use: A systematic review. Addiction, 117(5), 1220–1241. https://doi.org/10.1111/add.15687

[8] U.S. Department of Health and Human Services. (n.d.). Control Systems Engineering for Counteracting Notification Fatigue: An Examination of Health Behavior Change. National Institutes of Health. Retrieved April 10, 2022, from https://reporter.nih.gov/search/4nbzE01LtkiJMuHh9psRIw/project-details/10359062
R01LM013107: National Institute of Health, National Cancer Institute.