What Is a Good Heart Rate Variability? What Your Watch Is Measuring

Six forty on a January morning in western Maine. The porch thermometer says four degrees, the truck needs ten minutes, and your ring has already filed its verdict on the day: HRV 34 milliseconds, down eleven from yesterday, recovery yellow. So what is a good heart rate variability? There isn’t a single number, and there’s no threshold that applies to everyone. The largest pooled review of healthy adults (44 studies, 21,438 people) found that five-minute RMSSD averaged 42 ms with a standard deviation of 15 ms, but the averages of the individual studies ran from 19 ms up to 75 ms, and that same review reported interindividual variation of up to 260,000 percent on some HRV measures. The number also drops steeply with age and moves with cold, sleep, training load and how the reading was taken. A single morning figure means something only against your own rolling baseline, measured the same way at the same time of day.

What Is a Good Heart Rate Variability? The Published Range Is Enormous

Heart rate variability is the variation in the gap between consecutive heartbeats, measured in milliseconds. Two names show up constantly. RMSSD is the one most consumer devices report, sometimes as a natural log. SDNN is the one most of the clinical literature is built on.

The reference numbers people quote at each other usually trace back to one place. A quantitative systematic review pooled 44 studies of short-term (five-minute) HRV in healthy adults, 21,438 participants in total, published between 1997 and 2008. Pooled SDNN came out at 50 ms with a standard deviation of 16, and a range across studies of 32 to 93 ms. Pooled RMSSD was 42 (15) ms, with a range of 19 to 75 ms. The authors noted that their values were lower than the older reference norms, and flagged “large interindividual variations (up to 260,000%), particularly for spectral measures.”

The frequency-domain numbers in the same consolidated table are stranger still. High-frequency power pooled at 657 ms2 with a standard deviation of 777, across a study range of 83 to 3,630 ms2. That is roughly a 44-fold gap between the lowest and highest study averages, in healthy people.

So a person at 19 ms and a person at 75 ms can both be sitting comfortably inside the normal literature. Neither one learns anything by comparing themselves to the other.

The Cutpoints You’ve Seen Came From a 24-Hour ECG

Somewhere in most HRV explainers there’s a tidy tier list. Under 50 ms is bad, 50 to 100 is borderline, over 100 is healthy. Those figures are real. Under the framework from the 1996 Task Force report, “patients with SDNN values below 50 ms are classified as unhealthy, 50-100 ms have compromised health, and above 100 ms are healthy.”

They are also 24-hour values, taken from continuous ECG and used for clinical risk stratification. The standards document that produced them (the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, published in the European Heart Journal in 1996) predates wrist and ring optical sensors entirely.

The scale difference is not subtle. In a 24-hour study of 2,079 adults aged 25 to 41, SDNN was 160 (40) ms in men and 147 (36) ms in women. The pooled five-minute SDNN above was 50 ms. Same metric, same units, roughly a three-fold difference, purely because of the length of the recording. Holding a five-minute or overnight reading up against those 24-hour cutpoints is the most common error in consumer HRV writing, and it leaves healthy people convinced they’ve failed a test they never actually took.

Age Bends the Curve, and That Lands Hard Here

HRV falls with age, and it does not fall evenly across the metrics. A study of 260 healthy subjects aged 10 to 99 tracked 24-hour time-domain measures across nine decades: “SDNN and SDANN reached 60% of baseline by the tenth decade; SDNN index reached 46%; pNN50 and rMSSD declined most rapidly to 24% and 47% by the sixth decade.” Gender differences appeared in the younger subjects and disappeared after age 50.

Read that twice for the shape of it. The parasympathetic measures, the ones your device is most likely reporting, do most of their falling before you turn 60. The decline isn’t a clean slope either. A separate study of 1,743 people aged 40 to 100 found that RMSSD and pNN50 “showed a U-shaped pattern with aging, decreasing from 40 to 60 and then increasing from 70.”

Which is where the local part comes in. Maine’s Office of the State Economist reports that “Maine’s median age remains the highest in the nation” and that in 2022 Maine had the largest percentage of its population aged 65 and over of any state in the nation. The same report projects the state’s 65-and-over population rising from 378,679 in 2022 to 421,663 by 2032, an increase of 11.4 percent over the projection period.

The average reader here is therefore older than the reference cohort sitting behind any app’s normal range, and the measures that age fastest are exactly the ones on the screen.

A Northern Winter Is a Different Measurement Problem

Cold moves the reading. In a controlled study, twelve men went through a 30-minute neutral preconditioning phase followed by 30 minutes of cold exposure at minus 5, minus 10, minus 15 and minus 20 degrees Celsius. “All HRV indexes of four cold exposures were significant,” and the falling temperature came with progressive parasympathetic activation and sympathetic retraction. SDNN was the most sensitive index, with good linear relationships to blood pressure, pulse and hand temperature.

NOAA’s Maine State Climate Summary puts winter average temperatures at 25 degrees F in the far south and below 15 degrees F in the northern and interior parts of the state, with a long-term average of 29.6 very cold nights a year, defined as a minimum of 0 degrees F or lower. (The count has run below that long-term average since the mid-1990s.) Zero Fahrenheit is about minus 18 Celsius, sitting inside the exact band that cold study tested.

A January reading taken with cold hands after a walk out to the barn and a July reading taken in bed are not the same measurement, and no population chart knows which one you just took. The hours around those readings differ by season too, since winter here reshapes when and how well people sleep in the first place, which is its own long conversation about Maine winters.

Accuracy claims deserve the same care. One wrist optical sensor validated against ECG did well inside its own protocol, with heart rate bias at or under 0.39 (0.38) percent and natural-log RMSSD bias of 1.66 (1.80) percent, limits of agreement plus or minus 5.93 percent. That protocol was during sleep, against ECG, with a 200 ms artefact filter applied. There is no comparable published figure for a cold wrist in motion at 10 degrees, which is the condition a lot of people in Maine and New Hampshire train in for four months of the year.

What the Number Cannot Tell You

Many apps convert HRV into a stress-versus-recovery balance, usually built on the ratio between low-frequency and high-frequency power. That framing has a problem at the root. A review of the underlying physiology found that “direct recording of sympathetic nerve activity failed to correlate with LF power in either healthy subjects or patients with heart failure,” and that the “LF peak of the heart rate power spectrum is reduced by at least 50% by either cholinergic antagonists or selective parasympathectomy.” If blocking the parasympathetic system halves the supposedly sympathetic signal, the ratio is not measuring the balance it is named after.

Direction is less informative than it looks, too. In elite endurance athletes, researchers have documented “both increases and decreases in HRV” accompanying negative adaptation, along with “increases in cardiorespiratory fitness … with atypical concomitant decreases in HRV.” A number going up is not automatically good news. A number going down is not automatically a warning.

Then there is the sleep-stage layer that recovery scores are usually wrapped around. In 53 adults studied for one night against clinical polysomnography, two-state agreement (asleep versus awake) was 88 percent for the Apple Watch and 89 percent for the Oura Ring (kappa 0.30 and 0.51). Multi-state sleep-stage agreement fell to 53 percent and 61 percent (kappa 0.20 and 0.43). The accuracy question, and the anxiety that tends to trail it, gets a fuller treatment in this look at what sleep trackers actually get right.

How the Research Teams Actually Use It

People who use HRV seriously use it as a series, under fixed conditions, and they are blunt about how much data that takes. Sport scientists monitoring training adaptation write that practitioners “should use a minimum of 3 (randomly selected) valid data points per week,” with the added value flattening out after three to four days a week in trained triathletes. One reading is not a data point in that sense. It’s an anecdote with a decimal place.

Even tightly controlled measurement wobbles. Forty elite rugby union players took two-minute recordings on four separate days: intraday reliability came in at ICC 0.96 with a coefficient of variation of 3.99 percent, interday reliability at ICC 0.90 with a coefficient of variation of 7.65 percent. Trained athletes, controlled protocol, and the number still drifts about 7.65 percent from one day to the next. Treat that as a floor for how much movement is meaningless, not a ceiling.

Length matters as well. Across 30 subjects measured at rest, during exercise and in post-exercise recovery using windows from 10 to 240 seconds, “at least 120 s was required in the post-exercise recovery or exercise conditions,” and the authors concluded that criteria established for static conditions cannot be applied to the non-static conditions of daily life. Very short snapshots and the five-minute standard stop being interchangeable the moment you are not sitting still.

Half the finding is the protocol. A season-long study of 22 professional soccer players, 504 recordings in all, took every reading sitting, early morning, fasted, for ten minutes. SDNN, rMSSD, pNN50, SD1 and SD2 all “showed an identical behaviour throughout the season, with lower values in the pre-season and the end of the season.” The season showed up in the data because everything else was held still.

What Actually Shifts It Over Months

Training does change HRV, and the clearest evidence points at steady aerobic work. A meta-analysis of 21 studies (9 randomised) in 523 people with type 2 diabetes found that exercise training increased SDNN (effect size 0.59, 95% CI 0.26 to 0.93), RMSSD (0.62, 0.28 to 0.95) and pNN50 (0.62, 0.23 to 1.00), while decreasing LF power and the LF/HF ratio. The authors state that “the level of proof is the highest for endurance training.” The limit is worth naming plainly: that population was people with type 2 diabetes, not healthy adults, so the size of the effect in everyone else is not settled.

The underlying behaviour is unglamorous. CDC’s activity guidance for adults describes 150 minutes of moderate-intensity physical activity a week, or 75 minutes of vigorous-intensity activity, plus muscle-strengthening activity on two days. Around here that often looks like a hill road, a pack and an hour, and the evidence behind walking under load is worth reading before anyone starts adding weight to it.

One last piece of context about what these features are. FDA treats recovery and readiness scores as general wellness products, defined in guidance issued January 6, 2026 as products that “(1) are intended for only general wellness use, as defined in this guidance, and (2) present a low risk to the safety of users and other persons.” The listed categories cover physical fitness, relaxation or stress management and sleep management, and monitoring heart rate during recreation appears as an acceptable example claim. That is a policy of not enforcing device requirements on low-risk products. It is not clearance, and a readiness score is not a diagnostic.

Which leaves a fairly modest job for the thing on your wrist. Your own baseline, built from readings taken the same way at the same time of day and watched by the week rather than the morning, is the only comparison that carries information. Anything that feels like a medical question, a resting heart rate that has changed and stayed changed, or a rhythm that feels wrong, belongs with a clinician and a real ECG rather than with a trend line and a color.

Sources

• Nunan D, Sandercock GR, Brodie DA. A quantitative systematic review of normal values for short-term heart rate variability in healthy adults. Pacing Clin Electrophysiol. 2010;33(11):1407-17. https://pubmed.ncbi.nlm.nih.gov/20663071/
• Shaffer F, Ginsberg JP. An Overview of Heart Rate Variability Metrics and Norms. Front Public Health. 2017;5:258. https://pmc.ncbi.nlm.nih.gov/articles/PMC5624990/
• Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Eur Heart J. 1996;17:354-81. https://pubmed.ncbi.nlm.nih.gov/8737210/
• Umetani K, Singer DH, McCraty R, Atkinson M. Twenty-four hour time domain heart rate variability and heart rate: relations to age and gender over nine decades. J Am Coll Cardiol. 1998;31(3):593-601. https://pubmed.ncbi.nlm.nih.gov/9502641/
• Sun B, Wu J, Li C, Li C, Hu Z, Wang R. Effects of different extreme cold exposure on heart rate variability. Ergonomics. 2024;67(8):1147-1163. https://pubmed.ncbi.nlm.nih.gov/37988319/
• NOAA National Centers for Environmental Information. Maine State Climate Summary 2022. NOAA Technical Report NESDIS 150-ME. https://statesummaries.ncics.org/chapter/me/
• Office of the State Economist, Maine Department of Administrative and Financial Services. Maine Population Outlook 2022 to 2032 (April 2025). https://www.maine.gov/dafs/economist/sites/maine.gov.dafs.economist/files/inline-files/Maine%20Population%20Outlook%20to%202032.pdf
• Bellenger CR, Miller DJ, Halson SL, Roach GD, Sargent C. Wrist-Based Photoplethysmography Assessment of Heart Rate and Heart Rate Variability: Validation of WHOOP. Sensors (Basel). 2021;21(10):3571. https://pubmed.ncbi.nlm.nih.gov/34065516/
• Billman GE. The LF/HF ratio does not accurately measure cardiac sympatho-vagal balance. Front Physiol. 2013;4:26. https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2013.00026/full
• Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Med. 2013;43(9):773-81. https://pubmed.ncbi.nlm.nih.gov/23852425/
• Miller DJ, Sargent C, Roach GD. A Validation of Six Wearable Devices for Estimating Sleep, Heart Rate and Heart Rate Variability in Healthy Adults. Sensors (Basel). 2022;22(16):6317. https://pubmed.ncbi.nlm.nih.gov/36016077/
• Plews DJ, Laursen PB, Le Meur Y, Hausswirth C, Kilding AE, Buchheit M. Monitoring training with heart rate-variability: how much compliance is needed for valid assessment? Int J Sports Physiol Perform. 2014;9(5):783-90. https://pubmed.ncbi.nlm.nih.gov/24334285/
• Nakamura FY, Pereira LA, Esco MR, Flatt AA, Moraes JE, Cal Abad CC, Loturco I. Intraday and Interday Reliability of Ultra-Short-Term Heart Rate Variability in Rugby Union Players. J Strength Cond Res. 2017;31(2):548-551. https://pubmed.ncbi.nlm.nih.gov/27243917/
• Kim JW, Seok HS, Shin H. Is Ultra-Short-Term Heart Rate Variability Valid in Non-static Conditions? Front Physiol. 2021;12:596060. https://pubmed.ncbi.nlm.nih.gov/33859568/
• Naranjo J, De la Cruz B, Sarabia E, De Hoyo M, Dominguez-Cobo S. Heart Rate Variability: a Follow-up in Elite Soccer Players Throughout the Season. Int J Sports Med. 2015;36(11):881-886. https://pubmed.ncbi.nlm.nih.gov/26140687/
• Picard M, Tauveron I, Magdasy S, Benichou T, Bagheri R, Ugbolue UC, Navel V, Dutheil F. Effect of exercise training on heart rate variability in type 2 diabetes mellitus patients: A systematic review and meta-analysis. PLoS One. 2021;16(5):e0251863. https://pubmed.ncbi.nlm.nih.gov/33999947/
• U.S. Food and Drug Administration, Center for Devices and Radiological Health. General Wellness: Policy for Low Risk Devices. Guidance for Industry and FDA Staff, issued January 6, 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/general-wellness-policy-low-risk-devices
• U.S. Centers for Disease Control and Prevention. Physical Activity Basics: Adult Activity. https://www.cdc.gov/physical-activity-basics/guidelines/adults.html

This article is for informational purposes only and does not constitute medical advice. Consult a qualified healthcare provider before making any health decisions.