Andrew Flatt

“But wristbands can still be used to track HRV trends even if they aren’t 100% accurate”. This assumes that the error is fairly constant across the range of values (e.g. it consistently underestimates values by 10 ms whether RMSSD was 20 or 120 ms). If this were the case, then these could be used to track trends. Unfortunately, studies show that error is highly inconsistent across the range of RMSSD values. In some cases, error increases as RMSSD increases, with the error occurring in either direction. This means that the true RMSSD might be substantially higher OR lower than it appears on the app. You won’t know in which direction it is off. In other cases, there is overestimation at low values and massive underestimation at high values. This means your true low values will appear higher (e.g. 15 ms true value appears as 30 ms) and your true high values will appear lower (90 ms true value appears as 60 ms). This is unacceptable for tracking changes and quantifying daily variation (CV) with longitudinal monitoring, and for rank ordering individuals cross-sectionally (lowest to highest HRV). Note that many previous validation studies did not test for these types of bias (though it can be seen in figures), leading to overly optimistic conclusions about device accuracy. It’s concerning enough that people are using these recreationally and being misled. It’s much worse that these are increasingly being used for conducting research.

Ambulatory (24 h) blood pressure is more predictive of cardiovascular outcomes than clinic measurements, providing a more reliable assessment of one’s long-term cardiovascular health. Cardio is the most effective type of exercise for reducing ambulatory blood pressure.