Wearables and Health

Wearables that track some biometric are popular among the exercise crowd, but also in more extensive use for health management purposes and of course, all kinds of fantastic health benefits are claimed for these wearables.  Alas, as usual the hype is not matched by the actual research evidence to date.  A meta-analysis published by Nature Digital Medicine finds no real clinical effect from these devices.   (Nature Article)   The authors reviewed randomized clinical trials from 2000 to 2016 on the use of wearable biosensors.  They found 27 applicable studies, 16 of which were considered high quality, which evaluated a clinical outcome.  These outcomes included body-mass index, weight, waist size, body fat percentage, and systolic and diastolic blood pressure.  The devices used included blood pressure monitors, spirometers, cardiac activity trackers, electronic scales and physical activity trackers, among others.  The primary disease indications for which the tracking was done were cardiovascular in 11 trials, pulmonary in 6, weight in 6 and a mix of miscellaneous conditions.  Most of the studies included education, some included incentives for use and some had a coaching dimension.  Almost all of the trials involved a clinician in using the data to interact with the patient and/or modify treatment regimens.

The meta-analysis found no significant difference between intervention and control groups in regard to body mass index, weight loss, waist size or body fat percentage.  There was similarly no significant difference in blood pressure outcomes.  Discouraging.  The notion of being able to continuously, in real-life settings, monitor and record various biometrics, including heart rate, heart activity, blood pressure, blood sugar, etc., is intuitively valuable.  But it has to be used by the individuals and clinicians in a way that modifies behavior or adjusts treatments in order for all that great data to have a real impact on health.  That is where the rubber always meets the road.  So although the trial evidence to date is not good, I view that as more a function of how the data is used than its underlying value.  And the results of the meta-analysis support this, as the authors note that there were certain trials where aspects of feedback and clinician interaction suggest some benefit to the remote monitoring and data analysis.  This technology will ultimately be useful, but has to be coupled with behavioral interventions.