Prediction model development of women's daily asthma control using fitness tracker sleep disruption

Background: Night-time wakening with asthma symptoms is an important indicator of disease control and severity, with no gold-standard objective measurement. Objective: The study objective was to use fitness tracker sleep data to develop predictive models of daily disease control-related asthma-specific wakening and FEV₁ in working-aged women with poorly controlled asthma. Methods: A repeated measures panel design included data from 43 women with poorly controlled asthma. Two components of asthma control were the primary outcomes, measured daily as (1) self-reported asthma-specific wakening and (2) self-administered spirometry to measure FEV₁. Data were analyzed using generalized linear mixed models. Results: Our models demonstrated predictive value (AUC=0.77) for asthma-specific night-time wakening and good predictive value (AUC=0.83) for daily FEV_1. Conclusions: Fitness tracker sleep efficiency and wake counts demonstrate clinical utility as predictive of asthma-specific night-time wakening and daily FEV_1. Fitness tracker sleep data demonstrated predictive capability for daily asthma outcomes.