Can wearables predict our fitness status?

Cardiorespiratory fitness (CRF) is the term used to reflect the physical activity status of an individual. It is used to assess the relationship between physical activity and health status. There is an inverse relationship between CRF and health outcomes. It has a central role in evaluating the risk of cardiovascular diseases (CVD) and is a strong predictor of CVD outcomes. Maximal oxygen consumption (VO2max) is the gold-standard measure of CRF. Various models have been used in the past to indirectly calculate VO2max based on variables such as sex, age, resting heartrate, and others. Based on these models, there are recent efforts trying to estimate VO2max in free-living conditions from wearable devices. In a recently published article in NPJ Digital Medicine, Spathis et al., demonstrate via a very elegant methodological approach and the use of Artificial Intelligence the ability to predict CRF in free-living conditions through the design of dedicated algorithms and models and the use of raw wearable sensor data. In this study, large cohort data has been used. It has been demonstrated that the use of wearable data combined with other biomarkers as inputs in Neural Networks can reliably predict VO2max and its change over time measured via traditional methods.
You can access the full text here.