(this post was updated 4/6/2023)
Apple Watch and other fitness trackers have the capability to provide us with information on cardiovascular parameters which reflect the activity of the autonomic nervous system (ANS). Measures of the activity of the ANS reflect the balance between the sympathetic nervous system (which activates fight and flight responses) and the parasympathetic nervous system (which activates “rest and digest” activities) and have been shown to be powerful predictors of mortality.
Most of the attention in this areas has been on heart rate variability (HRV) with various wearables trying to promote HRV as a surrogate marker for stress. The OURA ring people for example state without evidence that “high heart rate variability is an indication of especially cardiovascular, but also overall health as well as general fitness.”
Although unimpressed with the HRV data from Apple Watch or the OURA ring I have recently discovered that I can get a more useful parameter of ANS tone from my Apple Watch-Heart Rate Recovery.
What Is Heart Rate Recovery?
Heart Rate Recovery (HRR) is the rate of decline in heart rate after the cessation of exercise. Basically you measure heart rate right when you stop exercising and again a minute later (and/or two minutes later) and subtract one from the other. An alternative term is pulse rate drop.
Unlike HRV you don’t really need any high tech devices to make this simple but highly reproducible measurement. You can simply measure your pulse the old-fashioned way by putting a finger on your carotid or radial artery and counting the beats.
What happens to the heart rate during exercise has long been considered to be due to the combination of parasympathetic withdrawal and sympathetic activation.
The fall in heart rate immediately after exercise has been shown to be a function of the reactivation of the parasympathetic nervous system. It is accelerated in athletes and blunted in patients with heart failure.
Heart Rate Recovery As A Predictor Of Mortality
A 1999 study published in the New England Journal of Medicine found that abnormally low HRR doubled the risk of dying over 6 years.
The study examined outcomes in 2428 consecutive adults (mean age 57 years, 63 percent men) without significant prior cardiac disease who were referred to the Cleveland clinic cardiac lab for nuclear stress testing. Patients underwent symptom-limited exercise on a treadmill using a standard or modified Bruce protocol.
Heart rate was recorded at peak exercise and then patients walked upright and were walking at a speed of 1.5 miles per hour at a grade of 2.5 percent when heart rate was checked a minute later.
Median HRR was 17 beats per minute, with a range from the 25th to the 75th percentile of 12 to 23 beats per minute. Abnormally low HRR was selected as <13 beats/min and was found in 639 patients (26 percent).
In univariate analyses, a low value for the recovery of heart rate was strongly predictive of death, conferring a four-fold increased risk. After adjustment for multiple confounding factors including age and exercise capacity, patients with HRR <13 beats/min had a two-fold risk of dying.
This 20 year old study and HRR remain highly relevant. The paper has been cited 1001 times since publication and thus far in 2019 58 papers have referenced it.
In a follow up study this same Cleveland Clinic group looked at nearly 10 thousand patients undergoing treadmill ECG testing and found HRR <13 beats/min doubled the 5 year risk of death. In the figure below mortality jumps markedly as HRR drops below 13 and quite dramatically if <10 beats/min.
Subsequent studies from different investigators confirmed that HRR is associated with mortality, independent of workload and myocardial perfusion defects, treadmill risk score, and even after adjusting for left ventricular function and angiographic severity of coronary disease.
There has been a lack of consistency in these studies in stress protocols, activity post-exercise and optimal duration of heart beat measurement post exercise.
This 2001 JACC paper determined that a 2 minute HRR <22 beats/min provided a better cut-point than one minue HRR <13 beats/min in predicting mortality at 7 years in male veterans. Individuals underwent maximal treadmill followed by lying down and those with an abnormal HRR were 2.6 times more likely to die. The HRR was equivalent to age and exercise capacity for predicting death.
Apple Watch and Heart Rate Recovery
It’s not entirely obvious how to view the heart rate recovery data on your Apple Watch but it is routinely logged if you record an activity and end it precisely at the end of the activity. To see it you must leave the activity app and open the Heart Rate APP.
Scroll to the bottom of the screen and you will see HR data on your most recent activity including the peak HR and one minute recovery heart rate.
Click on that tab and the full and awe-inspiring graph of your recovery heart rate over 3 minutes is revealed. Here is mine which followed a 1.5 mile run at 6-7 MPH. I did not walk at 1.5 MPH on a 2.5% grade in recovery which would be needed if one wanted to more carefully compare a personal HRR to the numbers from the 1999 NEJM study.
The Watch only stores data on your last workout but if you go to the Activity app on your iPhone (something I had never previously done) you will find under the workouts tab a complete listing of all previous workouts.
Click on the workout of interest and all the data from the workout is wondrously revealed including cadence, pace and near the bottom heart rate changes. Swipe the heart rate changes during exercise to the left and the heart rate recovery graph is revealed. This time you will have to do the subtraction for yourself
Heart Rate Recovery-Simple, Powerful And Intuitive Measure of Autonomic Tone
So there you have it. Heart Rate Recovery (unlike HRV) is a simple parameter, easy to understand and measure.
In fact it is so simple you don’t need a fancy fitness watch or wearable device to check it. Just put your finger on your radial (wrist) pulse or carotid (neck) pulse and count the beats for 10 seconds. Alternatively, if you are on a HR measuring device at the gym (treadmill, elliptical, etc.) have the machine check it. Check immediately on stopping exercise and then one minute later. Multiply by each measurement by six and subtract one from the other.
It yields information on your vagal/parasympathetic tone and has been proven to be a powerful and independent predictor of your overall mortality.
It makes more sense to pay attention to HRR if one wants a measure of your body’s autonomic tone than HRV.
If your one minute HRR is <13 beats per minute or two minute HRR <22 beats per minute this is a bad prognostic sign. If you have not been diagnosed with significant cardiovascular disease consider seeing a physician for evaluation..
For those who have been sedentary and are deconditioned or overweight, consider an abnormal HRR as a wake-up call to modify your lifestyle and improve your mortality.
For healthy, asymptomatic individuals the HRR can serve as a marker for your overall cardiovascular fitness. Monitor it along with your exercise capacity, peak heart rate and resting heart rate to raise your awareness of how your exercise is influencing your overall autonomic nervous system balance.
Autonomously Yours,
-ACP
29 thoughts on “Heart Rate Recovery: A Simple and Powerful Predictor of Mortality”
I am a 46 year old female who recently underwent heart surgery to remove an asymptomatic atrial myxoma. Prior to surgery I was a fitness enthusiast – cycling, running, and lifting weights 6 days a week and closely monitoring my resting heart rate (low 50s) and HRV (60 – 70). I am now six weeks post op and my resting heart rate is in the high 80s and my HRV is below 10. Throughout any given day my heart rate frequently exceeds 100 bpm while resting in a chair. I am not on beta blockers due to intermittently low blood pressure. My cardiac surgeon has said that anything is considered normal 90 days post op, and during these 90 days I should avoid strenuous exercise. I would like to see my HRV improving, as it scary to see it so low. Should I be working to improve my HRV through exercise? Is it dangerous to exercise with an HRV this low? My cardiologist does not seem comfortable with recommendations related to HRV and exercise.
Elain,
I would not be so concerned about your low HRV. The higher heart rate is likely related to some deconditioning in the post-operative period.
If I had a patient with your postoperative symptoms I would advise progressively advancing cardio type exercise as symptoms allowed. If the high HR is related to deconditioning it should gradually decline as exercise capacity increases.
dr. P
After a recent round of exercise I had a 2m HRR of 67 BPM, although it was from 194 (and my HR peaked at 199, and I have at other times seen it go to 204). (Resting heart rate is 50-70.)
https://i.imgur.com/VTOY0SL.jpeg
I am 40 years old, isn’t this an abnormally high peak HR?
https://en.wikipedia.org/wiki/Heart_rate#/media/File:MaximumHeartRateByAgeVariousFormulae.png
Wondering if it’s a good thing, a ‘Go see a Doctor’ bad thing or just one of those things.
Just stumbled across your excellent article and your sharing of your personal experiment. Motivated by what you did, I went for a 5-mile run at lunch today and measured my HR 3 minutes later (no grade like the study.). My HR was down 59 bpm 3 minutes later so I guess I’m in OK shape.
By the way, I’m impressed with your low peak HR during your workout experiment.
Thanks again.
”There’s virtually not a single item to nitpick on this interval from the Apple Watch’s optical HR sensor.
Seriously. It’s about as perfect as you can get.”
-DC Rainmaker
My cooling down routine after exercise helps prevent Delayed Onset Muscle Soreness (DOMS) which frequently, for me, takes place 3-4 days after the event. Most people do not have this problem.
30 years ago I had to retire from competition squash due to a hip injury, so moved to the gentler game of competition tennis, along with my squash buddy.
Our doubles matches were often prolonged affairs, with lobs to the baseline, drop shots to the net, and long rallies – typical old man’s tennis. Very energetic. Cunning old codgers – we loved beating the youngsters.
After each match I walked until my HR dropped to below 100, then did some leg stretches. My physical therapist devised this regime, with the theory that the lactic acidosis that had built up during the match exercise was dissipated during the cool down, with reduction in DOMS. It worked in practice, as well as theory.
So I still follow this routine, and note the time taken for HR to drop below 100 – gentle walking on treadmill is the cool down exercise.
With regular exercise, my waking HR drops to 56, and the time it takes to drop HR to 100 after exercise also drops. Both good indicators of cardiovascular health. BP 120/81. At 73 these are values I am happy with, but always aiming for improvement.
Like Steve, I use waking HR as a proxy for overtraining, or overdoing things when skiing at high altitude. I find that rest days are now needed, especially when keeping up with the grandchildren.
We are all different, so it helps for each of us to track our progress, and find what works for us. My own personal HR max during workouts is 130, taking into account the beta blocker effect.
I have moved away from high intensity routines, and more towards longer slower sessions, which better suits my body, a bit fragile after numerous sporting fractures and injuries. I too follow the Soviet methods, known as Greasing the Groove. Works for me.
Besides exercise, what are ways to improve Heart Rate Recovery, i.e. to improve the re-activation of parasympathetic/vagal tone?
(for example, I’ve heard of various music therapy and meditation techniques, but wondering if this group here has any other lifehack suggestions?)
Thanks
I am an symptomatic atrial fibrillation (AFib that comes and goes) patient, I had as ablation that lasted three years but it came back. I refused another ablation because of the reliability of a cure and elected to go with the Rate Control (Dilitiazem, Verapamil, Multaq) and Antiarrhythmic drugs (Flecainide). The side effects of these drugs for me are terrible so I am looking at this as a new approach. Any information on this from others would be appreciated.
https://medicalxpress.com/news/2019-05-zap-ear-ease-a-fib.html U of Oklahoma
After six months, the researchers found that atrial fibrillation was reduced by 85 percent in the active treatment group compared to the placebo group. When the researchers looked at both the three- and six-month time points, atrial fibrillation was reduced by 75 percent compared to placebo.
https://medicalxpress.com/news/2014-08-ear-good-heart.html University of Leeds
Lead researcher Dr Jennifer Clancy, of the University of Leeds’ School of Biomedical Sciences, said: “The first positive effect we observed was increased variability in subjects’ heartbeats. A healthy heart does not beat like a metronome. It is continually interacting with its environment—getting a little bit faster or a bit slower depending on the demands on it. An unhealthy heart is more like a machine constantly banging out the same beat. We found that when you stimulate this nerve you get about a 20% increase in heart rate variability.”
Back in the day, when I pretended to be a REAL cyclist, we used to measure our resting HR and HRR. If resting HR was higher or HRR was slower we used that as a marker for overtraining. (We would also use HRR with intervals, but I usually convinced myself I didn’t need intervals, because they HURT).
Haven’t used the iPhone workout app–thanks for the tip. I’ve always ignored it when I’m prompted to say if I’m starting a workout.
Thank you Steve, detecting overtraining is very important but often ignored. I think it contributes to people giving up on exercise because they begin with too much enthusiasm. It may interest you to know that a technique used by some Soviet Block weightlifters was daily grip strength measurement.
One’s resting heart rate can change dramatically due to ablation for atrial fibrillation.
https://academic.oup.com/europace/article/7/5/415/427352
My HR was in the mid to high 50s before my 2012 paroxysmal AF ablation. It’s been in the low 70s since.
I’ve had various bouts of sustained atrial tachycardia since. And three (3!) more ablations. AF has never recurred, but I’m wondering if all that followed the first ablation was iatrogenic.
Would exercise training such as what is described here make a difference to a person with such a modified ticker??
Jeff,
I would need to see more data on this. That paper does not mention beta blockers or calcium channel blockers. It seems to me highly likely the patients were such rate slowers prior to the ablation and off after which would automatically raise HR> Also less antiarrhythmics means higher heart rate. The increase HR they saw could be due to such changes.
My anecdotal experience is that patients post ablation tend to have slower heart rates.
I think exercise training should be effective and similar post ablation. it is mostly going to be dermined by vagal effects on the sinus node and the netiher one of these are typically damaged/effected with ablation
The Cleveland Heart Clinic has extensively researched 3 heart rate variables which are highly related to all-cause morbidity and mortality. Resting heart rate alone has been found to be a better measure of CRF than HRV. Heart rate recovery and resting heart rate together just about capture all the information that is needed for determining CRF. The importance of vagal tone can not be underestimated in patients with heart disease as low vagal tone or sympathetic nervous system activation is a major marker of stress and this stress has a major effect on the heart and vascular system. You can train to improve both resting hear rate and heart rate variability by doing aerobic exercise of at least 10 METS or greater 3 to 4 times per week. Yesterday, I went and did one of my favorite exercises on the incline treadmill. Here is how it works. Set the incline treadmill at a 30% grade and walk for 3 minutes at 2.5 miles per hour then recover for 1 minute by moving the speed down to 2 mph then increase the speed to 3 mph. Repeat this for 40 minutes to 1 hour. Yesterday, I did this for 40 minutes while reaching a maximum speed of 3.6 miles per hour for the last 5 minutes. Immediately after the last 5 minutes of exercise I measured HRR. Heart recovery after 2 minutes was 167-89 which is 78 BPM. I added to this post exercise 3mg of melatonin to help reduce inflammation. After resting for 6 hours I took my vitals which were: 1. RHR 30 BPM, BP 106/52, O2 saturation was 98%. The RHR of 30 BPM indicates very high vagal tone.
Erwin,
Actually, in some studies HRR turns out to be a more powerful predictor of mortality than resting HR which surprised me.
Sounds like a great exercise routine.
I have trouble believing your 2 minute HRR, perhaps the peak HR is not accurate.
Your RHR is amazingly low.
Another interesting post, thank you.
I have a related question on recovery.
I use the somewhat fashionable HIT program for aerobic conditioning 2-3 times per week.
On an elliptical machine (arms and legs), after a 3-minute warmup, I go as fast as I can for 30 seconds, then cruise for 90 seconds, sprint 30 seconds etc.
Rather than use TIME between sprints, I’m wondering if heart rate recovery would be better, i.e., wait until pulse has dropped BY x, or maybe TO y.
I am male, 75, resting pulse 45-55, maximum rate after each sprint 155-160.
What are your thoughts?
Regards
Bob Guthrie
Bob
Short answer-I have no idea.
What is your HRR with this routine?
I have not been observing my HRR but will and respond. I have been finishing with a 90-second cool down. From other posts it would seem I should stop that – at least for HHR assessment.
My apologies for the delay Dr Anthony: my part of Australia has been covered in smoke so I took the official warning to avoid vigorous exercise.
Today I just did two 30-second sprints on an elliptical with 90-seconds between them at a reduced pace.
starting HR 56
end of 3min warmup 107
after first sprint 147
after second sprint 157
after 1 min 144
HRR1 13
after 2 min 133
HRR2 24
So my HRR rates are only just OK 🙁
I wonder if the intensity of my sprints is a factor (absolutely 100%)
Other posts seem to have much lower maximums.
I expect it is not wise clinical practice to have patients repeatedly smash themselves so maybe the recommendations come from a different protocol.
157 is about the theoretical maximum for 75-year old although it did get to 180 on the largest of the sand hills in this year’s Outback Marathon.
BTW If you wish I can in a separate post discuss the differences between chest strap and wrist HR readings.
Regards
Bob
Bob,
We have hard much of the California wildfires but little of Australia’s bush fires. I hope the smoke clears for you.
The intensity of the sprints is absolutely a facctor. Recently, I was experimenting with a HIIT (high intensity interval training) and was alternating running at 9 mph for 30 seconds with walking at 4 MPH. My HRR was much less if I quit at the end of the 9 MPH segment versus quitting at the end of the 4 MPH segment. I tried to describe in detail in my post the protocols that have been utilized in the studies showing HRR predicts mortality but they have been in general the Bruce treadmill protocol which involves increasing the speed and incline every 3 minutes. At peak for most individuals the speed does not exceed 4.2 MPH. Also the speed and incline are maintained for 3 minutes so there is equilibration. If you spring as I did for 30 seconds the body is still adjusting and HR still goes up for a while after you stop.
I would definitely be interesting in your thoughts on chest strap versus wrist HR readings.
Heart Rate Recovery depends on the workout. One of my standard workouts is 30 minutes on an elliptical with a 5 minute warm up, the rising exertion until my heart rate holds around 115 at minute 15 to 20, then reduced exertion to 90 bpm for 5 minutes and 80 bpm for 5 as a cool down. By the time I stop my heart rate is 84. It returns to normal 65. Expecting it to fall below 62 is unlikely.
I had stints 6 years ago and I’m still alive like most people.
Stephen,
Actually, I think most people who ever lived are now dead but congradulations on being among the living!
Your protocol would not be optimal for HRR.
For comparison to these studies, I would recommend doing either treadmill, outdoor run, elliptical or bike until you have reached the maximal that you can do without being uncomfortable then stopping either completely or at a very low rate, say 1.5 MPH on a treadmill.
Did that. Stopped after the 5 hardest minutes of the workout. Recovery was 22 BPM in 1 minute then plateau for next minute and additional 23 BPM by 3rd minute. 105 to 60 BPM. The actual top rate was 115BPM but I didn’t stop that instant. I guess I’ll live awhile longer than my 71 years.
Should we finalize the workout on the watch the second we stop running/cycling for better measurement of the first two minutes of recovery?
During the cooldown (walk on treadmill after jog / slow cycle) my HRR is about 10 per minute for the first two minutes. If I skip the cooldown and just stand still, it drops about 20 per minute. Which is closer to the correct way of measuring?
M,
Yes. No cooldown if you want HRR as measured in the studies. End the workout on Apple Watch at peak exercise and then walk slowly.
By the way, I’ve never been a fan of the cooldown. We use it on patients getting stress tests but personally I just hop of the device and go lift weights.
Thank you sir! I’ll definitely monitor my HRR using this method going forward. When I look at my workout history the HRR is all over the place, but I never paid much attention to when exactly I finish the workout on the watch.
One thing I noticed when I do a cooldown and only end the workout after, is that my HR will increase as I walk around the gym or toward my house in those first two minutes. My HRR in the app then shows something like 1st minute: +5, second minute: +7. But I guess my walking home might be more intense than my cooldown in this scenario so that could cause it. I was worried when I first saw that in my workout history after reading your article haha
I have the same question as adboucher and also questions regarding the intensity of exercise (e.g. walking moderately fast vs climbing stairs) and the effect of age.
I found these articles interesting:
https://www.newsgram.com/ai-tool-mortality-heart-failure-patients
and https://www.sciencedaily.com/releases/2018/09/180904140542.htm
Jane
Thank you for this very interesting column Anthony.
After my coronary bypass surgery in january 2018, and subsequent fitness recovery, I became very interested in HRV and other heart health parameters. I had a low HRV, by any type of measurement (SDNN, RMSSD, LnRMSSD… measured with Qardio Core). I looked into training aimed at improving HRV (…short, intense aerobic exercises)…
My HRV remained low… (BTW I have no extensive myocardial damage).
However, discussing this with my cardiologist, I soon realized it is a known fact that ß-blockers, such as atenolol, lower HRV, but that this interaction has no prognostic significance…
My own HRV (measured with Qardio Core) almost doubled after I switched from atenolol to nebivolol.
In any event, I do not rely on that figure anymore. I watch closely my HRR, and it looks very encouraging.
Do you know if medication, such as ß-blockers, influence HRR?
This paper suggests beta blockers do not influence HRR. https://www.nature.com/articles/jhh201686.pdf?origin=ppub
Wow! Thank you! Missed that one…