Muse 2 forehead photoplethysmography agrees with Polar H10 for heart rate but overestimates ultra-short RMSSD

The human heart does not beat like a metronome. Between one thump and the next, there is a tiny, natural variation in time. This fluctuation, known as heart-rate variability, is a sign of a healthy, responsive nervous system. When the body is relaxed, the heart slows down and speeds up in rhythm with breathing, creating a complex, shifting pattern. Scientists have long used this pattern to understand how the body handles stress and rest. Traditionally, measuring this required a medical-grade electrocardiogram, a machine that reads the heart's electrical signals directly from the chest. However, as wearable technology has exploded in popularity, people now want to track this same data using simple devices that sit on the wrist or forehead. These devices do not read electricity; instead, they use light to watch the pulse of blood flowing through the skin. The question for researchers is whether this optical method, which measures the arrival of a pulse wave rather than the electrical spark that starts it, can tell the same story as the medical standard.

In a recent study published in Physiological Reports, a team of researchers in Bali, Indonesia, put this question to the test using two popular devices: the Polar H10, a chest strap that records heartbeats with high precision, and the Muse 2, a headband designed for meditation that also measures the pulse on the forehead. The researchers asked a simple but critical question: if you wear both devices at the same time while sitting quietly, do they agree on the details of your heart's rhythm? They focused on two specific measures. The first was the average heart rate, which is simply how many times the heart beats in a minute. The second was a more subtle measure of variability, which looks at how much the time between beats changes from one moment to the next. This second measure is often used to gauge how relaxed a person is, but it is also much more sensitive to small errors in timing.

The study involved twenty-eight adults who sat quietly for two minutes while wearing both devices. For the first minute, they kept their eyes open, and for the second minute, they closed them. The researchers then compared the data streams from the headband against the chest strap. The results were clear and split into two very different stories. When it came to the average heart rate, the two devices were virtually identical. The headband calculated the number of beats per minute with such accuracy that the difference was less than one beat per minute. In this regard, the forehead sensor worked perfectly, confirming that it can reliably tell you how fast your heart is beating.

However, the story changed completely when the researchers looked at the subtle variations in the rhythm. The headband consistently reported that the heart's rhythm was more variable than the chest strap did. In nearly every single session, the headband showed a much larger difference between beats than the chest strap recorded. On average, the headband overestimated this variability by about nineteen milliseconds, which is a significant difference in the world of heart rhythms. To put this in perspective, if the chest strap measured a certain level of relaxation, the headband suggested the person was even more relaxed than they actually were. This overestimation was so consistent that the researchers found the headband's reading was nearly double the value of the chest strap in some cases. The data suggested that the headband was picking up a small amount of timing noise with every single beat, and because the calculation for variability looks at the difference between one beat and the next, these tiny errors added up and made the heart look much more erratic than it truly was.

The researchers also investigated why this happened. They considered whether the device was simply struggling to find the exact moment of the pulse due to the movement of the forehead or the way light travels through skin. They found that the error was consistent with a timing uncertainty of about sixteen milliseconds per beat. This could be a mix of the device's own limitations and the natural fact that blood takes a fraction of a second to travel from the heart to the forehead, a time that changes slightly with every heartbeat. Because the study did not record the actual electrical signals of the heart, they could not separate the device's error from this natural biological delay. What they did know for certain was that the two devices were not interchangeable for this specific measurement. The headband could not simply replace the chest strap when trying to measure the fine details of heart-rate variability.

The study also looked at whether the devices could track changes in the body. When the participants closed their eyes, a common relaxation technique, the chest strap showed a clear increase in heart-rate variability, indicating the body was settling down. The headband, however, did not provide sufficient evidence to demonstrate that it reliably tracked this same change. The data were compatible with anything from no tracking to moderately good tracking, meaning there was not enough proof to confirm that the headband could detect the shift, but also not enough to prove that it was impossible. This means that while the headband is excellent for telling you your heart rate, it cannot be confirmed to track shifts in relaxation based on the variability of your pulse with the same reliability as the chest strap.

The researchers concluded that the Muse 2 headband is a reliable tool for measuring average heart rate, but its measurements of heart-rate variability should be treated with caution. The device does not measure the same thing as the chest strap; it measures pulse-rate variability, which includes extra noise and biological delays that the chest strap does not see. For scientists and clinicians who need precise data on how the nervous system is functioning, the headband's numbers are not a direct substitute. The study suggests that if these devices are used to measure variability, the results should be reported as a measure of pulse rhythm rather than heart rhythm, and users should understand that the numbers will likely be higher than what a medical-grade device would show. The technology is useful, but it has a specific limit: it can count the beats, but it cannot yet perfectly capture the subtle, shifting dance of the heart's timing.