Smart wearable devices: Useful tools to monitor our health

Exercise monitoring devices, or physical activity monitors, have been used for decades in the physical activity research community. Early versions of these devices simply recorded the duration and intensity of physical activity. Modern versions of physical activity monitors can additionally give direct feedback to the user and thus have the potential to promote behavioural changes, i.e., to encourage physical activity and possibly detect health problems. Review articles on the applications of these wearable devices for cardiovascular disease care have been published here and here.

Smartwatches
There are several types of devices for sale on the market. So-called smartwatches, such as Apple Watch, Fitbit, Garmin, Omron, and TomTom watches, contain sensors that measure physical activity (the number of steps, the energy spent) using the accelerometer and gyroscope, distance travelled (GPS), heart rate, as well as detect falls, monitor sleep, detect cardiac arrhythmias (electrocardiogram), and estimate blood oxygen saturation and cardiorespiratory capacity (VO₂max). These devices are increasingly popular for tracking our health and fitness, especially during physical activity sessions, including walking, running, cycling and many other sports.

Other smart wearable devices
There are also devices that are worn around the thorax (e.g., BioHarness, Polar straps), around the wrist (bracelets), or that are inserted into “patches” that are attached to the chest (e.g., Zio Patch, Nuvant MCT, S-Patch cardio), clothing (e.g., AIO Smart Sleeve), or shoes. These smart wearable devices mostly measure heart rate and record electrocardiograms. These devices are less popular because they are bulkier and their functions are more limited.

An encouragement to exercise more?
A systematic review and meta-analysis of studies on the effect of interventions based on physical activity monitors was carried out by Danish researchers in 2022. As a first step, the researchers did a systematic review where they analyzed 876 articles published on the subject, of which 755 articles were excluded because they did not meet the established quality criteria (poor study design, protocol, intervention, comparison, population, etc.). They then contacted the authors of 105 studies to obtain detailed protocols and relevant data that would not have been included in their studies. This illustrates how difficult it is to carry out this type of synthesis study and that the studies published are of variable quality.

The systematic review retained 121 randomized controlled studies with 16,743 participants, where the average duration of the interventions was 12 weeks. At the very beginning of the intervention, the number of steps taken daily was 6994 on average, and the average body mass index was 27.8 (overweight). The three main aims of the study were to assess the impact of the intervention (wearing a monitor) on: 1) physical activity in general; 2) moderate to high intensity physical activity; 3) sedentary lifestyle.

• Physical activity in general. For this meta-analysis, the results of 103 studies with 12,840 participants were included. Participants who wore a physical activity tracker did on average more physical activity than those who did not wear it. The increase was significant, albeit modest: 1235 steps more on average daily by wearers of a physical activity monitor.
• Moderate to vigorous intensity physical activity. The meta-analysis on the effect of a monitor intervention on moderate-to-vigorous intensity physical activity, which included 63 studies with 8250 participants, also showed that wearing a physical activity monitor had a positive impact on the amount of moderate-to-vigorous intensity physical activity, about 48.5 minutes more per week.
• Sedentary lifestyle. Interventions with a monitor indicated a slight favourable effect on sedentary time, a reduction of 9.9 minutes on average per day.

The authors conclude that the level of evidence for the effects of interventions based on the wearing of physical activity monitors is low for physical activity in general and moderate for moderate-to-vigorous intensity physical activity and for sedentary lifestyle. The effects on physical activity and moderate-to-vigorous physical activity of monitor-based interventions are well established, but may be overestimated due to publication bias, i.e., that researchers sometimes tend to publish their results only if their study is positive.

Detection of atrial fibrillation with a smartwatch
Atrial fibrillation is the most commonly diagnosed type of cardiac arrhythmia and affects more than 37 million people worldwide. This type of cardiac arrhythmia is associated with an increased risk of cardiovascular disease (5 times higher for stroke) and premature mortality.

Smartwatches, such as the Apple Watch for example, contain optical sensors that measure users’ heart rate by emitting green light through the skin of the wrist. An Apple Watch app uses algorithms that help identify abnormalities that suggest atrial fibrillation.

In a study funded by Apple, researchers wanted to test whether the Apple Watch could really be useful for detecting episodes of atrial fibrillation (AF). During the 8 months of the study with 419,297 participants, 2161 (0.52%) of them received irregular pulse notifications. Of these, those whose symptoms were not urgent were mailed an electrocardiogram (ECG) recorder (ePatch), which they wore for 7 days. Of the 450 participants who returned the ECG recorder for data analysis, 34% had confirmed episodes of atrial fibrillation, while 66% of them could not be confirmed. The main criticisms (see here and here) of this study are that the detection of AF by the Apple Watch is not very specific and not very sensitive, and that this type of use could do more harm than good by creating concern, promoting overdiagnosis and overtreatment and therefore causing a waste of health system resources. In addition, the high cost of this type of smartwatch means that it could not be deployed on a large scale, if the technology were to improve one day.

Smartwatches are continually being improved, so the performance gap with medical-grade devices keeps getting smaller. Physicians should be open to reviewing the data generated by smartwatches as it could yield personalized information useful for patient treatment. However, there is a lack of a regulatory framework to standardize these data and incorporate them into regular clinical practice. These devices have the potential to generate huge amounts of biometric data that could lead to unnecessary and costly testing, with consequences for patients and the healthcare system.

Other uses

Blood pressure measurement
Hypertension is a very common cause of several cardiovascular diseases and premature death. The presence of blood pressure sensors in consumer wearable devices could potentially improve the detection of hypertension, particularly nocturnal hypertension which is associated with the worst consequences. There are devices with a cuff, but devices without a cuff have recently been developed, making it possible to consider their use for the measurement of ambulatory blood pressure. The results of comparisons of these new devices with existing medical devices are encouraging, but this new technology is still in its infancy and will need to be further refined and studied.

Other sensors
There are minimally invasive biochemical sensors that measure molecules of interest in physiological fluids. For example, continuous glucose monitors measure glucose in the interstitial fluid (under the skin, not in the blood). These monitors have been clinically validated, but it remains difficult to integrate them into consumer portable devices. Sensors using sweat or saliva could be more easily integrated into wearable devices, but this remains to be developed.

Biomechanical sensors incorporated into clothing or shoes, such as the ballistocardiogram and seismocardiogram or dielectric sensors, have been developed with the aim of passively and continuously measuring cardiac output, as well as the volume and weight of liquids in the lungs. These devices could be useful for monitoring the condition of people with heart failure.

Smart wearables are constantly being perfected and they have a bright future ahead of them. However, it will be necessary to ensure that these devices are accessible to all and that they do not negatively impact the health system by generating too many biometric data and unnecessary clinical tests. In addition, smart wearables will need to be evaluated and subjected to strict standards to ensure their quality and effectiveness. The medical profession will probably have to open up more to the use of these devices, which allow remote monitoring, especially in these times of the COVID-19 pandemic when there has been a significant decrease in patient visits to their doctor’s office.