Heart Rate Variability in Wearables: Promise and Uncertainty

The report's author, a medical student and digital enthusiast, compared Apple Watch Series 10 and Oura Ring 4 after switching devices. Both track HRV, but Apple Watch uses the SDNN algorithm, while Oura uses RMSSD. HRV measures the tiny fluctuations in the time between consecutive heartbeats, reflecting the balance between the sympathetic nervous system, which governs fight-or-flight responses, and the parasympathetic nervous system, which promotes rest and digestion. A healthy autonomic nervous system is flexible, adjusting heart rate quickly to meet challenges and then relaxing.

For much of the 20th century, medicine considered a steady, metronome-like heartbeat a sign of health, the report says. But research starting in the 1960s challenged this view. A 1965 Yale University study monitoring fetal heartbeats found that fetuses in distress had unusually regular heartbeats with reduced variability. A 1980s study of 808 myocardial infarction patients found that those with lower HRV had significantly higher mortality rates. These findings led to the recognition that HRV is a marker of the heart's adaptability, and that low variability can indicate poor health.

The clinical gold standard for measuring HRV is electrocardiography (ECG), which records precise intervals between the R waves of each heartbeat. Consumer devices use photoplethysmography (PPG), in which green LEDs detect changes in light absorption caused by blood flow. PPG is vulnerable to interference from arm movement, poor sensor contact, skin tone, tattoos, and temperature changes, which can distort frequency-domain analysis. As a result, consumer devices typically rely on time-domain methods such as SDNN and RMSSD, which are simpler and more robust.

SDNN calculates the standard deviation of all normal heartbeat intervals, reflecting overall variability, but it is influenced by changes in heart rate itself. For example, an increase in heart rate from 60 to 100 beats per minute during exercise can raise SDNN even if the intervals remain uniform. RMSSD, by contrast, focuses on the root mean square of successive differences between adjacent intervals, primarily reflecting parasympathetic activity and less affected by heart rate trends. Oura uses RMSSD to generate nightly recovery scores, while Apple Watch's SDNN is presented as a long-term trend. The report notes that Apple Watch Series 12 and Ultra 4, equipped with improved optical and electrical heart sensors, now include a "recovery HRV" metric using RMSSD. Third-party apps such as Athlytic and PeakWatch also compute RMSSD from Apple Watch data to provide recovery scores similar to Oura's.

The report cites a 2022 meta-analysis showing that people in the lowest quartile of HRV had a 1.56 times higher risk of all-cause mortality compared with others. Another meta-analysis of individuals without cardiovascular disease found that lower HRV was associated with a 32% to 45% increased risk of a first cardiovascular event. However, HRV is a sensitive but non-specific biomarker: it can signal that something is wrong but cannot identify the specific problem.

The author notes that the devices were self-purchased and that the article is not an advertisement. Oura Ring is not sold in China and requires a monthly subscription fee of $5.99. The report raises questions about the accuracy of consumer wearables compared with medical-grade equipment but does not offer a definitive conclusion.