Because sleep is inextricably linked to virtually every health outcome and insufficient sleep quality and duration are most relevant when present on a persistent, daily basis, the longitudinal measurement of sleep (as opposed to single night, gold-standard polysomnogram [PSG]) is a crucial need in healthcare. Appropriately verified consumer facing wearable sensors, will allow, for the first time, long-term, inexpensive, objective sleep measurement alongside other health metrics and provide the opportunity for granular, real-time sleep assessments without excess burden to the patient.  Therefore, in collaboration with mathematics (Drs. Olivia Walch and Danny Forger) we have developed algorithms adapted to raw data derived from the Apple Watch to estimate sleep with excellent sensitivity (90%), specificity (67%) and accuracy (88%) compared to PSG. 

Dr. Goldstein hopes to incorporate sleep into Michigan Medicine’s precision health initiative with the ultimate goal of developing patient-specific, non-pharmacological sleep interventions to improve disease management and quality of life.  The first step in this goal is to leverage the MIPACT data set, to accurately phenotype longitudinal sleep patterns in individuals with overnight use of the Apple Watch, and determine the sleep disturbances and patterns relevant to disease management and progression.