Making sleep behaviors interpretable: adapting the two-process model of sleep regulation to longitudinal Fitbit sleep and activity behaviors for health insights

Abstract Background As sleep data from wearable devices are increasingly available in health research, there are new opportunities to understand sleep regulation behaviors as modifiable risk factors for disease. At such a large scale (tens of thousands of people over millions of day-level observations), prioritizing and interpreting sleep behaviors is challenging while maintaining biological relevance and modifiability. In this work, we aim to address this challenge by proposing a framework to interpret Fitbit data through a well-known neurobiological framing of sleep regulation, the two-process model. Methods We use data from the All of Us Research Program, a national biobank with passively collected Fitbit data for 32,292 people across 15,754,893 total days. We map Fitbit behaviors ( b ) to either circadian (C) or homeostatic (S) processes. Using iterative exploratory factor analysis to obtain weights, the Fitbit C b and S b are then weighted at the level of each day to create C b and S b scores. Findings C b and S b scores were found to align with expected real-world relationships with age, seasonality, shift work, and napping. C b and S b scores were interpreted with relation to depression, where it was found that S b scores are highly associated with likelihood of diagnosis (OR = 1.5, p < 2e-16) while C b and S b scores are equally associated with severity (S b score β = 0.2, C b score β = 0.21, p < 2e-16). Interpretation C b and S b scores support longitudinal interpretation (e.g., changes in S b around treatment), aggregation (e.g., differences in C b between two groups), and actionable modification (e.g., reduce naps to improve poor S b ). Overall, our behavior scores allow for interpretation of wearables sleep data and can be utilized across many disease contexts to better understand how sleep influences health. Funding This work was supported by NIH training grant T32GM145734 and NIH R21HL172038.