What if a wearable could measure your blood pressure all day without repeatedly squeezing your arm? Cuffless blood pressure monitoring provides a promising and convenient alternative to study blood pressure changes in daily life. However, the cuffless method – which estimates blood pressure based on the speed of blood flow – is not yet accurate enough for clinical use.

In this study, Myrte Schoenmakers and colleagues examined the cuffless method extensively using eleven conditions varying from relaxation to mental and physical stress tasks. Three modeling methods to estimate the blood pressure values were applied to demonstrate how using the ‘double-dipping’ method (most commonly used in literature) inflates your accuracy results. Next, both personalised and person-invariant models were created. Although reliability was high, construct validity (can we detect changes in blood pressure between conditions?) and convergent validity (do the estimates correspond with the gold standard?) were moderate with mean absolute errors from 10.1-12.9 mmHg and only 59.2–68.5% of esti¬mates differed less than 10 mmHg with the reference values. Therefore concluding that cuffless blood pressure monitoring has insufficient validity for clinical or stress research applications, even with person-specific calibration.