Research Paper Volume 13, Issue 6 pp 7900—7913

Deep longitudinal phenotyping of wearable sensor data reveals independent markers of longevity, stress, and resilience

class="figure-viewer-img"

Figure 4. Effect of lockdown and the future risks of COVID-19 ranked by wearable BAA and blood-based bioage models. Association of BAA with the future incidence of COVID-19: (A) BAA in the form of the negative logarithm of daily step counts, (B) GeroSense BAA, (C) CBC-based DOSI [4], (D) CBC and Blood biochemistry hazards model, and (E) Blood-based PhenoAge [2] (all data are given for UK Biobank participants aged 45−75 y.o.). (F) Longitudinal data obtained by smartwatch show that while GeroSense is more sensitive to future risks it is also more selective and does not change immediately (green line) due to merely walking less during March to May 2020 lockdown while the average activity level does (red line). The values in panel (F) were scaled to zero mean and unit variance for comparison; shaded areas show one standard deviation range at each time point.