Together with other researchers, Van der Mee analyzes the technology of widely-used smartwatches under highly diverse conditions. They examine how well the signal of the smartwatches aligns with devices used in scientific research and whether the calculated stress scores are dependable.
All smartwatches, whether from Fitbit, Garmin, or Apple, use more or less the same algorithm, she explains. They measure stress levels based on bodily signals, combined with personal data such as age, gender, and movement patterns. However, the differences between individuals are "enormous" and cannot be measured solely using those indicators. For instance, one person might have a higher resting heart rate than another. "And if the heart rate does indeed increase, a distinction must be made between positive stress, like excitement, and negative stress. These watches don't do that."
Van der Mee illustrates this with an example: "My husband is gaming right now. Such a game triggers arousal: his heart rate increases. But that doesn't mean he's experiencing stress. He's feeling excitement. The watch doesn't discern that difference." However, the effects on the body are quite distinct, according to Van der Mee. "Brief excitement generally doesn't have a negative effect on the body. Prolonged negative stress does."
Thus, a smartwatch can wrongly conclude that someone is experiencing prolonged stress, when in reality, excitement and stress alternate. And this is not always without consequences. "When a smartwatch says you're stressed, many people take that as truth. This can lead them to worry or worsen their existing stress."
A smartwatch is not medical equipment, Van der Mee emphasizes, and manufacturers have too much freedom in the claims they can make. In her opinion, this is a task for legislators. "The term 'stress' should be replaced with a more neutral term. However, such a term doesn't exist yet in Dutch."
Whether smartwatches will become more reliable when additional indicators, like sweat levels, are measured alongside heart rate, remains uncertain. "At this moment, we unfortunately know too little about the relationship between these signals and stress on an individual level."