WHOOP Podcast - HRV-CV: WHOOP Research Study Reveals The Longevity Metric Everyone Needs To Be Tracking

Episode Date: February 25, 2026

Welcome to the WHOOP Research Series, where we breakdown into the extensive, scientific research conducted by the WHOOP Performance Science team. On this week’s episode, WHOOP Global Head of Human Performance, Principal Scientist, Dr. Kristen Holmes sits down with WHOOP Senior Research Scientist Dr. Greg Grosicki to unpack the latest WHOOP Research Study on Heart Rate Variability Coefficient of Variation or HRV-CV. Using data from 21,000+ WHOOP members and 2 million nights of sleep, Dr. Holmes and Dr. Grosicki reveal why day-to-day stability in your HRV may matter more than the number itself. HRV-CV proves to outperform traditional metrics in detecting the real impact of alcohol, sleep consistency, and metabolic health, making it a great measure of healthspan. By looking deeper into the study, Dr. Holmes and Dr. Grosicki explains how HRV-CV can help you program your training smarter, improve your healthspan and build autonomic resilience. This episode unpacks how measuring HRV-CV isn’t just about performance, but risk stratification and understanding your physiology over-time, not just day-to-day. (01:11) Intro To The Study: Why Is HRV-CV Important? (04:09) HRV-CV and Risk Stratification(08:28) Definition of HRV-CV(11:40) What HRV-CV Indicates in Athlete Recovery(14:26) Breaking Down The WHOOP Study(17:13) HRV-CV Trends in Shift Workers(21:21) Why Is HRV-CV an Important Biomarker?(25:18) HRV-CV Trends in Individuals on GLP-1 Medications(26:29) The Weather vs. The Climate: HRV vs. HRV-CV(33:41) The Results: Breaking Down the Data(36:37) How HRV-CV Responds to Certain Behaviors(39:38) Trends Between HRV-CV Between Age, Biological Sex, and Body Mass Index(49:54) How To Improve HRV-CV(56:39) New Research Roles and Opportunities at WHOOPReferences:American Journal of Physiology PublicationHRV-CV: The Key Metric for Lifestyle Consistency and StabilitySupport the showFollow WHOOP: Sign up for WHOOP Advanced Labs Trial WHOOP for Free www.whoop.com Instagram TikTok YouTube X Facebook LinkedIn Follow Will Ahmed: Instagram X LinkedIn Follow Kristen Holmes: Instagram LinkedIn Follow Emily Capodilupo: LinkedIn

We have stumbled upon a biomarker that's reflecting health. People who want to live longer and healthier lives should be tracking heart rate variability, coefficient of variation, HRV, CV. The paper that was recently published that looks at this metric consisted of more than 21,000 WOOP members in roughly 2 million person days of data. Looking at kind of this seven-day perspective and looking at the day-to-day variation in HRV was way more sensible. So from a programming standpoint, this became a really important way to think about adaptation.

And I love the word adaptation or resilience being what exactly HRV-CV-C-V-E is telling. It's a resilience score. After this paper, they'll be integrating HRV-CV as a way to really risk-stratify individuals by their health status. They're using these data to give insight into health span. Giving us trajectories as to how quickly we're aging are metabolic health, insulin insensitivity, and there's no better metric to pick that up than the HRV-CV-CV. There's a lot of compelling data here from the sports science literature to tell us that this is a metric worth chasing, but many people fall into this trap of...

This is becoming a regular occurrence, Dr. Gregorsicki. Always a pleasure, Dr. Kristen Holmes. So I feel like I, you know, I don't want to, like, overinflate our research, but we've got another groundbreaking one to talk about today. Always fun, when you have the data that we do. I know, I know. So we're going to talk about heart rate variability coefficient of variation. We have published a really cool paper that looks at this metric.

So we're going to talk about that. Before we do, though, I'd love for you to talk about the team that we collaborated with on this paper. Because it's really a bunch of greatest of all time as it relates to all things, heart rate variability and certainly HRVCV. Yeah. And when you look at heart rate variability being integrated into sports. science being used by both researchers and practitioners in the field who are working with some of the best and some of the most elite athletes in the field. This study really gave us the opportunity to collaborate with almost all of them. And so specifically due to many

of them, your wonderful connections, Dr. Dan Ploos and Paul Larson, who are just legends in the field and both have been publishing. And the good news is both have reached out to me in the past week on Instagram asking when we are going to do more podcasts. So. Oh, I love that. More opportunities to come. Perfect. Perfect. And I think another one worth calling out, certainly Dr. Andy Galpin, you know, who has loads

of experience implementing hurry variability and the most demanding sports. I mean, UFC, I think is one that comes to mind immediately. NFL players. NFL players, yeah. And Dr. Marco Altini, who's one of the greatest thought leaders in terms of understanding trends and how to interpret those trends in ways that are valid, accurate, It's meaningful. So we were really fortunate to be able to collaborate with these individuals.

And then I think just, you know, obviously I think our experience, you know, just having insight into these data daily at population level, certainly. And then obviously I'll work with teams over the years, high performance teams, collegiate teams, professional teams all across the globe, really understanding, you know, what is it like to incorporate this metric, hearty variability CV specifically and what it actually means. in terms of actioning. Yeah, yeah, no, I totally agree. And another one of the co-authors on the paper that I think is worth bringing up is Jason

Carter, who's the dean at Baylor, really just an exceptional sleep scientist and has deep expertise in cardiac autonomic function and heart rate variability in particular and is actually a big thought leader in using HRV. And after this paper, now they'll be integrating HRV-CV as a way to really risk stratify individuals by their health status, which I think is a lot of where we at Woop are going, in addition to using it from a performance metric to tell when someone's ready to perform, but also using these data to give insight into health span. Well, talk a little bit more about risk stratification because I think that's one of

the greatest utilities of this metric. So maybe just dive into that a little bit more. Yeah. So I think all of our users know that resting heart rate is big in health span. That's very intentional. and it's based on a deep and rich body of literature showing associations between resting heart rate and morbidity and mortality. People might be kind of surprised to see, though, when they look at their health span, that heart rate variability currently is not in there just yet. And I had the

opportunity to work with a number of our data scientists when we were deciding what goes in and what doesn't get included in health span. And I think important to say, and of no surprise, to anyone is that this was a very intensive, hotly debated process. We spoke about this for, we debated it for years. Yeah, I remember it started with that survey when I originally came to boop and I was like, oh, good, it's decided. And then like maybe two years later, we were like, nope, still no decision. But there was a reason for that.

And when you look at heart rate variability, it's really insightful if I look at my heart rate variability today compared to my heart rate variability yesterday or last week or a year ago. but many people fall into this trap of comparing their heart rate variability to that of someone else's. And I use the word trap very intentionally there because that's not how it should be done. And if we're using heart rate variability to risk stratify, there can be unintended consequences there. Just for instance, if we look at individuals who have eating disorders, that actually increases heart rate variability. And that's certainly not something that anyone would encourage. So there's, yeah, some really wonky things happening with heart rate variability.

Yeah, I mean, look at, even in this paper, one of the things we looked at was trends in heart rate variability with advancing age. And our colleague, Dr. Bill Von Hipple, was just doing an analysis, and he restricted it to people who were between 18 to about, I think it was about 60 years of age, and showed an impressive 20% reduction in HRV per decade. So when you age 10 years, your heart rate variability is being reduced by 20%. So if your heart rate variability is lower this year than it was last year, that's something that perhaps wouldn't be surprising. And in this paper, fortunately, we actually looked at people who were over 60 years of age. And we saw what to some people would be surprising. And this is just cross-sectional. It's not longitudinal. Meaning we just compared individuals of different

decades. We didn't compare people as they got older, right? But in the older age group, the 70 and 80-year-olds, we actually saw higher heart rate variabilities than in the 60-year-olds. And this is a trend that when you look in the literature has been shown a number of times, and there's a number of possible reasons for this survival bias, selection bias in our cross-sectional study. But one of the potential causes of this that's most commonly discussed is the fact that older individuals have these cardiac arrhythmias, meaning the heart is beating abnormally irregularly, as opposed to heart rate variability, which is generally thought to be a good thing. They're having these ectopic beats or these undesirable heartbeats that will artificially

inflate heart rate variability. And that's not something we would want to be chasing, right? And so if we're going to cardiovascularly risk stratify based on heart rate variability, we could create a serious error there. We also know there's a very interesting paradigm known as the African American paradox of heart rate variability. And that specifically we know that African Americans tend to have much higher heart rate variability than white Americans. Yet African Americans have a higher incidence of cardiovascular disease and cardiovascular morbidity. And so how is that explain? And so how is that explain? And again, that's between person, but these are a number of different situations where we can apply and see that comparing my heart rate variability to that of someone else's could be problematic.

And so that begs a question, are there any sort of within person or normalizing our heart rate variability metric as a recovery metric to ourselves that could perhaps provide complementary insight when we look at heart rate variability? And that's where the heart rate variability coefficient of variation, also known as the HRV-CV, and you say that five times fast, Yeah. You'll probably make a mistake comes into play. Define HRVCV. The HRVCV or heart rate variability coefficient of variation. And I was talking to my roommate about this last night. I like gave him a bunch of different examples. And he's like, he's like, man, just keep it simple. It's really the stability or lack thereof of your heart rate variability over multiple days. And so if I'm looking back

at my heart rate variability for, let's say the past. seven days. In the paper, we looked at the seven-day heart rate variability metrics to calculate HRV-CV. And for those who are more mathematically inclined, this is the seven-day standard deviation divided by the seven-day mean. And we did that for good reason when we look at the original sports science studies by Marco and by Dan and by Paul and all the wonderful scientific advisors we worked with on this paper. That's what's been done in the literature. And that's not to say that the 30-day or 60-day isn't any more meaningful. And in fact, Dr. Galpin was asking questions about this when he was reviewing the manuscript.

So we acknowledged in the limitations that, yeah, we didn't look at those periods. And that needs to be done, but that this is what's most commonly done. And if you think about it, it provides a nice perspective into what's been going on in the past week, where if you look at it over a longer or shorter period of time, it's providing different context. Anyway, the seven-day heart rate variability coefficient of variation looks at how stable or unstable your heart rate variability has been for the past week. It's really calculating the day-to-day variation. The day-to-day variation, yes.

It was two days ago now. I was looped into, it was a bit of a question or debatably a complaint from one of our members that their heart rate variability was the same four days in a row. And they were upset by that and thought there might be measurement error. And so our wonderful data science team quickly looked into the raw data and saw that this was a rounding thing, that the 10th place decimal was different on each one of those days. And so what they were complaining about that their heart rate variability was the same. You know, I'm getting a raw deal here.

What we want to tell them is like, okay, well, good drink like three beers tonight. Yeah, you'll see some variability. Yeah. And so, again, just to set the table from a real high level, we want to see that stability of HRVCV. So a lower heart rate variability coefficient of variation generally is what's going to be desirable or better or preferable indicating that you're waking up in a consistent or a stable recovered state where if that heart rate variability coefficient of variation is more unstable or it's bouncing around more from day to day.

That suggests that the body is dealing with some sort of either internal or external load that's causing this autonomic fluctuation, if you will. And so when you look at this, again, until this paper, studies on HRV-CV had almost been entirely focused on athletes, which is a great way to learn about a metric, but we were able to apply it to performance populations. And this analysis in the paper that was recently published consisted of more than 21,000 WOOP members in roughly 2 million person days of data. So this is a really, really big data set. So before we kind of wrap up the athlete side of things, I mean, this is a metric that,

you know, prior to arriving at Woop, I was using a megawave and first speed. I was using some other hearty variability measurement devices. And I became really frustrated as a practitioner trying to interpret these data that the data day to day was pretty noisy. And when I arrived to Woop, and I started interacting with some of the brightest sports scientists in the world who were managing and programming for individuals and teams, athletes and professional and collegiate teams, we definitely came to the conclusion that looking at kind of this seven-day perspective

and looking at the day-to-day variation in HRV was way more sensible because we had the opportunity then to reduce some of the noise associated with kind of looking at it acutely, you know, HRV day to day. So from a programming standpoint, this became a really important way to think about adaptation. Yeah. And I love the word adaptation or resilience being what exactly HRV CV is telling. It's a resilience score. It is. It is. It's fantastic like that. And just as an example of that, good buddy and colleague of mine, Andrew Flat, during his PhD. Another HRV goat. HRV wizard. It's nice to have him in the pocket. But one of his most. famous HRV-CV-CV studies was actually when he was doing his PhD at Alabama, working with the

Alabama swimming team. And they're all doing the same rough training program. And this is where Andrew's mind really gets, it's creative and it really produces some nice insights. He looked at over the course of the season. They're all doing the same training program. They're all getting the same coach. The HRV-CV-V as a possible way to distinguish between those who make made to the NCAA championships to those who were more collegiate level swimmers. So they're still very high-performing swimmers, but we're really comparing the best of the best, the national level to those who are just competitive, just competitive collegiate swimmers. And what he saw is that over the course of the season, those who are succeeding and making

it to the NCAA championships, the highest level, theoretically the fittest individuals, their HRVCVCV was nice and low. They had high-heartedly variability, they were recovering well. They had these nice, low HRV CVs, meaning that they're recovering well, recovering consistently, day in and day out. There's other studies, some done by our collaborators, right? Dan and Paul, looking at HRVCVCV as a metric of adaptation, which is a term you used and I love. And they showed that over the course of, I believe it was a nine-week training program, those whose HRVCV or heart rate variability recovery was the most consistent. and had the lowest HRV-CV-CV, they adapted best. They also had the fastest maximal aerobic speed they called it was the outcome measure.

So there's a lot of compelling data here from the sports science literature to tell us that this is a metric worth chasing. And as generally happens when we learn from the best and the fastest and apply it in the general population, the insights sometimes are even greater. Yeah. Talk a little bit about how we, in the world of sports, you know, we express HRV-CV as a percentage. Maybe just talk through kind of what are the bands, you know, if we're talking, you know, 1% to 10%, you know, 11 to just talk through that real quick. Yeah. And so we did publish in the paper. And again, 21,000 individuals stratified by biological sex. So an approximately equal distribution of males and females, something that we care deeply about to make sure that our product is working equally as well. And we also stratified it by age group by decades. So not only considering biological sex.

is a relevant biological variable, but also age because we know that these variables change as we get older, at least we presumed so based on heart rate variability. And so we actually provided in the in the paper a nice supplementary table giving just HRV-CV-CV kind of benchmarks in terms of percentiles by decade. And what was really nice is it kind of lined up with exactly what Andrew and we would just be eating dinner would tell me or talking with Dan. And that's that very hard to have a HRV CV of under 10%. And by the way, for our members, you can pull up Woop Coach right now

and ask Woop Coach for your seven-day HRVCV. It'll tell you what the heart rate variability have been. It'll give you their standard deviation and their mean and calculate your HRVCB. A new thing to optimize for folks. New thing to optimize. And if you're under 10%, just know you are in the elite of the elite.

You're like a Tour de France cyclist. Yeah, these are like the fittest of most elite. Yes. Yeah. And so in general, when we're looking at the heart rate variability, coefficient of variations in this population, it ranged from somewhere between 7 or 8 percent and probably the youngest and the most elite,

all the way up to 30, 35 or 40 percent in those who might be less fit or might be participating in what some of us internally will call naughty behaviors. Hard bad behavior weekends as I presented during one meeting. William Bon Hippel, it calls them naughty behaviors. This is one of the things that we see in our shift work data. Oh, yeah. I mean, our acute care surgeons and, you know, people who are just really, I think, have the most demanding jobs on the planet and the demands of their body are just exceptional.

You know, they have actually quite high high HRVCB. So, you know, they're in the, they're in the high 20s, you know, low 30s, which, you know, is one of those things that, you know, we need to figure out how to mitigate, you know, that damage because obviously over time. that's going to be quite a reduction in lifespan. And it aligns with our earlier conversation about using heart rate variability as a measure for risk stratification. In this case, we're showing superiority with heart rate variability coefficient of variation.

And here's why I was actually listening to podcast. I think it was maybe last week. And they were comparing cardiovascular morbidity and mortality amongst people with MBs. And you see that in the general doctor who's working in primary care, it's lower than the general population and then these acute care surgeons, it's substantially higher, right? And it's because of their shift work and it's reflected by those heart rate variability coefficient of variations that are very high in these individuals. I mean, you can imagine taking, you know, the entire ecosystem of Mayo Clinic and being able to kind of look at, you know, which jobs,

you know, inside, you know, you've got a whole kind of, you know, you've got a cohort of medical doctors to your point, right? Like, you know, the physician who's able to sleep in their bed every single night is going to be very different than the person who's on call, right? Yep. So understanding the cost, right, the physiological cost to these different types of jobs gives us an opportunity to give these individuals resources and potentially program their call just as we would an athlete in programming their days of rest. We have the opportunity to think about it differently, right?

There's just like this incredible opportunity before us to use. these data for good. No, for sure. And I think the very unique thing about HRV, CV, as we see in the paper, when we compare it to heart rate variability, is it provides us with insight that we might miss if we're just looking at absolute heart rate variability trends. Just mean HRV. Yeah. That's not to say that HRV isn't useful. And as a matter of fact, it's extremely useful. Extremely.

But when used in compliment, they tell a deeper story. And so, you know, I like, to give the analogy of, like, when is a perfect example of when heart rate variability coefficient of variation matters most. And so what if we have two individuals who, when we look back at their seven-day heart rate variability means or their averages are absolutely identical? Maybe they both have means of 80 percent or 80 milliseconds. I'm sorry, 80. That would indicate just based on that, that they're at comparable physiological states. But we see that one of the individuals had literally, perhaps, I mean, good luck doing this, 80 milliseconds every single day over those seven days. Conversely, the other individual, one day he was at 60, and then the other

day he was at 100. And so to maybe kind of put this in a different perspective, say you have doctors going through medical school. and one of them on every single course, he's scoring 80%. The other one, he's making some classes. He's not making others. And so he's getting 60% on some tests and 100 is on the other. When they graduate medical school, they both have the same GPA. Now, which one of those doctors would you rather see?

The 80% consistently. Yeah. Yeah. Otherwise, you're taking a gamble, right? Yeah, yeah. He has some knowledge there. Maybe he got a 60 in surgery. And so it's the same thing with our bodies.

If we're thinking about how prepared we are for the day ahead or the week ahead, I think we can take confidence in knowing that if we've been recovering consistently, that our bodies are- And training appropriately. And training appropriately, that our bodies are prepared for that. Yeah. Why is HRV-CV-V-V-CV a novel biomarker? I think it's, frankly, just not very well known yet. And this is one of the things I love about being here and particularly working with you. I know when you called me and said, Greg, like, I want you to come work for me.

I said, I'm going to do it. And the primary reason why, Kristen, is because you can take the science and you can get it out to the public. And that's something that scientists sometimes have a really hard time doing. And I think that Woop now has the obligation to make sure that the community of both fitness professionals, but also just health and wellness individuals, older individuals, people who want to live longer and healthier lives understand and can appropriately leverage the utility of this metric. And we can dive into it a little bit more later, but the evidence definitively shows from the paper that tracking heart rate variability coefficient of variation can be extremely

advantageous from telling us how our body is responding to the behaviors and giving us what our body is responding well to and poorly to, but also giving us trajectories as to how quickly we're aging and how our metabolic health is, as may be indicated by body mass index. So there's a lot of insights, I think, that can be coming from HRV-CV. I think with advanced labs, having those data, we're going to be able to draw some really interesting associations between heart variability CV and things like metabolic health. Certainly. Yeah.

I mean, you think insulin sensitivity being the biggest one, right? And one that is often surprising to members. And if overnight you're having massive fluctuations in blood sugar due to insulin insensitivity, then that's going to cause traumatic swings in your heart rate variability. And there's no better metric to pick that up than the HRVCV. And we see this very clearly when people report having a late meal, their heart rate variability is suppressed overnight. And you can imagine, you know, one night, not a big deal. But if that's a common habit, you know, eating food late night within an hour or so before you intend to sleep, that is going to have an impact on your metabolic function.

And, you know, that is an HRV-CV. HRV is a really great metric that that helps us understand how your body is coping with that specific behavior. Yeah, that's the perfect external stimulus example of what would be going on internally if you're a normal healthy person. versus if you're someone who may have insulin sensitivity and you might not even be eating the late night meal, but your body just can't tolerate the glucose load. It's good. Your HRV-CV-CV is going to respond to that. Perhaps provide people with preliminary or insight that they wouldn't have gotten any other way,

even with the blood test that could get it from their metrics. And you can imagine an individual in that scenario who has a really volatile HRV-CV-V-CV, and, you know, we know that they're having a big meal within an hour of bedtime. They change that habit. And then, gosh, we layer on some exercise, you know, during the day. And all of a sudden, we start to see that HRV CV stabilize. And in relationship to that, we probably see markers of metabolic functioning start to improve as well. And what's great about, I think, some of the physiological data is that we probably see

that move first. So there's some satisfaction there, whereas it might take a little bit longer to see, for example, your Homo IRR or insulin sensitivity, marker of insulin sensitivity move, right? So there could be a lag in the blood data, but we see it in the physiological data first, which is kind of nice. It's one I'm super excited to look at in this GLP1 study that we'll be kicking off pretty soon. Talk for a second about that and how you imagine HRVCVCV kind of showing up in that research. Yeah, so last year we published this 12-week GLP-1 study and a relatively modest sample of our members, 66.

And GLP-1 use as tracked by Woot members on the platform has just risen exponentially. We know that in the United States, about 10 to 12 percent of the population is probably currently taking a GLP-1. I can think of five or 10 members in my family who are taking them. It's probably a popular topic of conversation at the holiday time would be my guess. No, it's eating. Well, right? Yeah, exactly. I'm going to just skip the gLP ones this week.

But yeah, we know that in addition to weight loss, there's systemic benefits of these medications. One of them being that if it's taken, a drastic improvement in insulin sensitivity and metabolic health. And so in theory, pretty robust and detectable reductions in HRV-CV-CV would be expected to be observed. So time will tell. I know we can't wait to get into that data set. Talk about today's weather versus the climate. So let's do like some kind of side-to-side compare. So we've kind of talked around it a little bit.

But, you know, really, I think for our listeners who have been so focused on HRV, this is just a little bit of a different way of thinking. you said both matter. So just kind of lay out the how our listeners should be thinking about their data in the context of HRVCV and HRV. Yeah, I think, you know, in the analogy of today's weather versus the climate members, at least like me, wake up in the morning, maybe go to the couch, pour a cup of coffee, and then the first thing you do is open the- So see some morning light, hopefully. Exactly, right. Open the loop app and check your recovery score and it gives you this morning's heart rate variability, which is fantastic. And that would be akin to today's weather. Now, if we want to know how today's weather sits in the weather as of

recent, that's where the heart rate variability coefficient of variation comes in. How does today's weather relate to yesterday's weather and the day before that and the day before that, right? And has it been in Boston actually relatively warm recently? And there's been a pretty high volatility in weather here. It's been the 50s the past couple days and be down in 20s tomorrow. Some pretty high HRV CV in Boston right now going on. So the heart rate variability coefficient of variation would be more like the climate, giving us insight into weather patterns over a certain duration of time. And when we think about how our listener, you know, when might they want to enter in into whoop coach?

You know, what is my HRVCVCV? I know for me I use HRVCVCV to understand. I look at it on Sunday and, you know, kind of in conjunction with that. my health span. You know, I've had a lot of travel early here in the new year. So my pace of aging is not great because my sleep consistency has been a little wonky. So just some things that, you know, it's just part of life. But also my HRVCVCVCV was certainly higher than it was the month of December when I wasn't traveling. So, you know, it's interesting to kind of see, okay, you know, what happened this week? And then I use those data to help me understand my programming

for that week. And that's kind of how in the world of sports science, that's how we use HRVCV and that's how I use it personally. And it's been a really great way for me to ensure that on a weekly basis, I'm adapting to how much load my body can take on. So I don't put myself in a situation where I'm taking on more load than I can handle. And as a result, I mean, I think that's one of the core reasons why I just don't get injured, you know, knock on, knock on wood. You know, I mean, I don't get injured. Obviously, I have a whole, I've been in this my whole entire life, right? So I've just been doing this for a very long time. So it's maybe unfair. But I think anyone can leverage these data to pay down the risk of injury. Because when you have these data and

you stay within these bands of functional adaptation, you know, it's just, it's just a lot harder to get injured. Yeah. It's all about looking at the load you're taking on this past week versus the load you've taken on previously and then how your body's responding to that. So if you have a really high HRVCV, let's say, on a Sunday. How might you think about your load that week? Yeah. So you asked about when we'd look at it. I would say now. Because it's not just training. It's also psychological load, right, is going to impact your HRVCB. And this was very clear. We kind of had a psychological, we were using this with Penn State football and Florida State soccer and Arsenal and all these various teams who are on the platform over the course of the last decade. We also had kind of a psychological intake, you know, just to measure various aspects, various measures of psychological.

functional functioning so we can understand, oh, is it training load? Is it sleep? Or is it just, you know, dealing with a relationship or just to set the stage? Yeah. I think other than looking now at your HRV-CV, looking at it from a weekly perspective does kind of provide us with a really nice perspective. The triathletes and the runners I coach, I always look like on Saturday or Sunday when I'm writing their training plans for the next week. And it is like you said, some of it does reflect what they've done in training. But I coach a lady who's in her 30s and a dentist in the weeks she has on call. Despite the fact that her training load is reduced 20 or 30 percent, her HRV CV is always higher. And that's reflecting the fact that she's waking up very early in the

morning, probably to get training sessions than I didn't tell her to do. Or, yeah, or, you know, on her feet all day. Yeah, dealing with patience. Exactly. Yeah, there's stress at work, right? And so that is like a great way in my opinion to look at it. If we look at it more than that, it can kind of drive us a little bit crazy, I think. So I think looking at it kind of like at the weekly level provides a real nice perspective. And I do think it's worth mentioning and hopefully not to the point of confusion that HRV CV, like heart rate variability, we said there are cases where higher, not always better. And again, those were pretty abnormal. There are there are certain unique cases in which a low HRV CV might not be super desirable. One thing I think we want to do

when we're looking at our metrics, and this is where heart rate variability and heart rate variability coefficient of variation marry very nicely together. It's just look at our heart rate variability trend. Has it been going up? If so, then that's great. And a little extra noise, a little less instability in heart rate variability is probably acceptable there. Because overall, the data are showing that our autonomic nervous system is improving and it's very imbalance. So if your HRV is going up and you see a little uptick in your HRV CV, I wouldn't fret that. But you won't as much by design mathematically because the mean is in the calculation. But not to say, okay, but for most of us, our heart rate variability mean is generally relatively stable. And since we're all aging,

actually is probably going down a little bit. And so that's when, in my opinion, HRV, CV, CV, is the most useful. If we compare consecutive weeks where a heart rate variability has been stable and we see oscillations in our HRV-CV-CV, that's when it's most practical. Now, there are circumstances. People can think about this if heart rate variability drops off a cliff and it's maybe your average is generally 60 and you're just 20, 20, 20, 20 every day. Well, your heart rate variability coefficient of variation would be quite low. But we should know intuitively that's not a good thing if our heart rate variability has dropped from 60 to 20. So not to provide confusion, but I do think that it's worth acknowledging that.

Yeah, yeah, it's a great point. And I think really important call out. What's up, folks, if you are enjoying this podcast or if you care about health, performance, fitness, you may really enjoy getting a whoop. That's right. You can check out whoop at whoop.com. It measures everything around sleep, recovery, strain. and you can now sign up for free for 30 days.

So you'll literally get the high performance wearable in the mail for free. You get to try it for 30 days to see whether you want to be a member. And that is just at whoop.com. Back to the guests. I want to talk specifically about the paper and to some of the findings. So hearty variability coefficient of variation during sleep as a digital biomarker that reflects behavior and varies by age and sex. published in American Journal of Physiology, Heart and Circulatory Physiology.

And that was published in... It would have been late 2025. So you can find that online. You can read it yourself. Greg, why do you give us a rundown of the results? Yeah. And what your takeaway is in terms of like what should the listener really focus in on? Yeah.

And I'm going to start super high level. Great. And I'll ask you a dive in. Perfect. We already kind of dissected the population, massive sample, 21,000 moot members, stratified or approximately equally distributed by age and biological sex, approximately two million nights of data. So massive data set. And there were three questions that we wanted to answer. The first, and I think this is very important, is if we're going to calculate HRV-CV-CV, which is generally

done looking at the past seven days of data this morning and then the previous six days or seven days of data. What if we miss a day, right? We obviously tell you to wear whoop every day, but we know there are occasions when we don't. So what if we miss a day? And we only have six. And we found that the six days is still fine. Well, what if we missed two days in that seven day period? And we found that five days is fine. And the really reassuring thing there is a paper published by us at the beginning of last year showed that approximately 95% of WOOP members are wearing WOOP six to seven days a week. So we know that for almost all of our members, this metric is going to be super useful. Right. We can reliably calculate heart rate variability coefficient and variation and know

that it's going to say something meaningful about your adaptation. Now, if in the past seven days you've only worn whoop four times, then the prognostic value of this metric is probably going to be decreased. So if trying to calculate the seven-day heart rate variability coefficient of variation, you need at least five days of data, finding one. Super valuable for anyone who's trying to action this metric, right, to understand, especially for a practitioner who's running a team and responsible for the programming of the athletes and interpreting these data. This is just a great contribution to the sports science.

And the funny thing is it actually is in complete agreement with my colleague Anders' paper, which was done, and it was something like 12 or 15, maybe 20. I don't want to like undercut him because I know he'll bust my chops about it. Female soccer players and like two or three months of data. That's exactly what he recommended. And he's generally right about things and he was right about this too. But it's nice to have some affirmation. Yeah.

And he would want that replicated. You know, he's, I mean, that's what we do in science. Certainly. We try to validate each other's. So if you're going to calculate, if you want to estimate the seven day, you need at least five. Ideally you have seven. So that was finding one. Finding two, we looked at how.

heart rate variability coefficient of variation responds to behaviors. And so when we quantify behaviors, we have what we now refer to as the big four, which are what we on the performance science team, and I think as Woop, think about the biggest movers of recovery, rusting heart rate, heart rate variability, and now we know the heart rate variability coefficient of variation. These are drinking alcohol, which anybody on Woop knows. It's crushing. Absolutely crushing. Yeah, two times the cost of any other metric on recovery.

Average recovery impacts something like negative 12 or negative 13. So alcohol, physical activity, sleep duration, and sleep consistency. And so we asked when members drink more or less, when members exercise more or less, when they sleep more consistently. I will say altitude fever are also, you know, there's some other, obviously things that move around. But these are in terms of like looking at not the outliers, which is things that are happening, you know, more consistently. Yeah, less and episodically. Yes. Yes. Yes. Exactly. And as anticipated, heart rate variability,

coefficient of variation responded very well to these and just the directions we would think. Keep in mind before what I tell you, high HRVCVCV, not as good, low, better. When people drank more, HRVCVC went up. When they exercised less, HRVCVCV went up. When they slept less, HRVCVCV went up. when they slept less consistently, like over the holidays. My health span, one of the things that promotes it best, sleep consistency. What happens when you reduce your sleep consistency? HRV-CV-C-V-C-V goes up, and so we probably all saw that around the holiday times too. And so what was even more interesting is we actually compared how HRV-CV-CV responds to the big four

with heart rate variability and then thanks to a very keen reviewer resting heart rate. not something I necessarily wanted to do, but something I'm glad I did. They really push us. Those were really worse. And it was reviewer two, of course. Inside joke for any of our listeners in academia. Viewer two is always just the...

Yep, the chopbuster. But what was real interesting, and it's good to know, is that of those four, HRV-C-V-C-V was more responsive than heart rate variability, more responsive than resting heart rate variability, resting heart rate to three of them. Alcohol, sleep duration, and sleep consistency. HRV, CV was more responsive to those behaviors than HRV and resting heart rate.

Correct. That's a big statement. It is. And so I think... I just wanted to pause there. Everyone, everyone can hear it. It's important. It's important to get across. And I think it provides pretty compelling rationale for it being something that we should start tracking. Yeah, pay attention to. I love it. And then finding three is we looked at associations between HRV-CV and age, biological sex, and then body mass index. Let's start with biological sex because it's simple.

We only had two categories of that, males and females. I will say there are folks on our platform who identify outside of those two categories, but we do not have enough of those folks to include them in the analysis. Yeah, that's a good point. When you look at a lot of the variables, resting heart rate, lower in males than females, for the most part, V-O-2 Max, higher in males than females. H-RV-C-V-V, on the other hand, consistently across the lifespan, lower in females, lower being generally better than it is in males. And so here's a metric that we have reflecting health,

and particularly maybe cardiovascular health, that is lower in females than males, pretty much at any age. And when we looked at that, when we showed it to our collaborators, both in academia and an industry, a lot of people were kind of surprised by that. It wasn't what we anticipated going in either. No. But then we peel the onion back a little bit, and we started thinking about it. Okay, if we have indeed stumbled upon a biomarker that's reflecting health, and particularly cardiovascular health, we have younger men between 30 to 50 years of age, far more likely to be diagnosed with hypertension or high blood pressure than females.

We have them far more likely to die from cardiovascular disease than females. And so perhaps HRV-CV is indeed reflecting risk stratification by the fact that it is higher across the lifespan and males and females. And some of our very astute listeners may say, well, what about behaviors, right? And when we look at that literature, we know that males on average are more likely to participate in riskier behaviors and that their lives tend to be less consistent. And we also know, and this is new data, that we have not published yet, so I'd like to caveat that. But it does seem that men are not getting the sleep that they need. They're undersleeping on average relative to women. They drink more. We know that,

for sure, from the paper we have in review at Plus Digital. So you would expect based on that for males to have a higher HRV-CV. The thing is, when you control for all of that, the same pattern persists, which gives us pretty darn good confidence that there is indeed a sex difference here in heart rate variability coefficient of variation, and that it's higher in males across the lifespan than females. And this is why it is a novel digital biomarker. Correct. Yeah, very interesting.

Number two, we looked at age. So I think it's best when we think about it from the context of age that we looked at it separately in males and females. Males, we see that from 20 to about 35 to 40 years of age, it's relatively flat. Kind of a funny finding is when you look at it just kind of without controlling for behaviors and are 18 to 22 year old or college age students, it's very high. Yeah. And then when you control for things like alcohol, all of a sudden it drops a lot.

Yeah, that sure does. But when you control for alcohol from 18 to about 35, it's pretty flat. Yeah. And we know that band, the younger individuals, kind of in that 20 to maybe even up to 30 bracket, are more sensitive to alcohol. Yeah. Both men and women. Yep.

Younger, more sensitive. Yep. Yep. For sure. And then at about 35 to 40 years of age, we see HRVCV just start rising. And it rises basically continuously until all the way about 70 to 80 years of age, which tracks with cardiovascular disease burden. Sure, people in their 20s who are male get incident cardiovascular disease and unfortunately die of cardiovascular disease sometimes.

But around 35 to 40 years of age, that's when cardiovascular disease burden starts to come. We see more people being diagnosed with hypertension. And the same trend follows for HRV-CV-CV. And it's also worth pointing out that it's approximately, and I'm not saying that this is causal, but it would be associated. And we did not measure testosterone, but it is when testosterone starts to drop. And so could there possibly be- We're going to have these data soon. Right, advanced labs, yes.

But could there possibly be some endocrine mechanisms at play here? Yeah. And we're not sure. So again, males pretty stable from 18 to 35 or 40, and then it starts to rise. Females, on the other hand, lower than males across the lifespan, and actually drops until about 50 years of age. And then it starts to go up. Which is so interesting because we have kind of this tax we deal with on a monthly basis,

where there's so much variability, you know, from a hormonal perspective. Yeah. Right? Which, you know, there's variability in our respiratory rate, in our variability in a resting heart rate. I think, again, that's why this is such a valuable perspective. For sure, for sure. Because when you look at just those as individual kind of metrics,

it just appears quite noisy. But when you look at HRVCV, it does reduce a lot of that noise. And it is interesting. We had to do a lot of deep thinking in reconciling the fact that HRVCV is lower in females and males when females are having this natural variability in resting heart rate and heart rate variability over the month. And our great colleagues, Summer on the data science team, published wonderful work last year in NPJ digital medicine showing that as a matter of fact, there's associations between females who are healthier and having greater variability in resting

heart rate across the menstrual cycle. And so... And you can imagine, you know, for a woman who who is experiencing this variability, that is because things are working as they should, you know? And that's when you're getting these kind of bigger fluctuations, like, that's actually like a normal response to those hormones. And so how do you rectify that, right? We have females who have lower, more stable heart rate variability profiles than males, and yet they do have this inherent natural variability. And thank God we were working with Dan on this because I didn't know what to say when

summer and the data science team started asking, well, like, well, how do you marry these two? And he referenced a paper about the vagal tank theory. And this is excellent. Thank you, Dan. Yes, thank you, Dan. It basically says that when you're looking at cardiac autonomic control, there are a couple important factors to look at. One of them is just tonic or at rest.

And that's what our members are getting when they wake up and see their overnight heart rate variability. We're at rest. This is how our heart rate variability is how our autonomic nervous system is functioning when we're at rest. Another one to consider is what happens when there is an external stimulus. So, for instance, if someone wants to go exercise, it is in fact advantageous for the vagal nerve to withdraw if I'm exercising, meaning when I start exercising, to put that in kind of more lay terms, I want my heart rate to go up when I start exercising. It's bad if it does not. And there are

conditions, aging being one of them that are associated with the inability to raise our heart rate. And we see that in the alcohol data, right? That actually, in fact, younger individuals are more responsive to alcohol than older adults. Right. We almost see like a muted, like there's no real cardiac response to the alcohol as you age because there's less potential for variability. You have a ceiling almost. Right.

And that's exactly the context, I think, that is needed when we interpret the fact that low HRV, CV is good, but that also having this variability across menstrual cycle is also good, that within a given day in a month, females who are younger and healthier, lower BMI, have more variability in their heart rate, you know, in relation to their cycle. That's their body is responding to the stimulus, right, just as our body should respond to if we go outside and go exercise. Our heart rate should go up and our HRV should go to. down. So with the menstrual cycle, we're seeing the same thing. And I think that provides nice

context. And keep in mind, these are very slight shifts. We're talking about over the course of 30 days, shifts of maybe three or four milliseconds of heart rate variability, which isn't much, right, compared to like if we have even two or three drinks and our HRV crashes 20 milliseconds in a single day, three or four milliseconds over 30 days is relatively minor. And the last one was body mass index. Again, for those who may not be familiar, in general, to a point, lower body mass index is better between 18.5 and 24.9. We're kind of classified as being in this normal weight category, 25 and 30, this overweight category, and then if anything 30 or over, would be in this obese category. And I just want to be very clear that body mass index is crude, that people who are carrying extra lean mass, as a lot of our members probably are. from working out may very well fall into the overweight category when in fact more specific tests

like Dexa or underwater weighing would show that they actually have very low fat mass and a lot of fat-free mass. So I just want to acknowledge that. It's a limitation. It is for certain. But one of the things we saw here is that those who had a higher BMI had less stable, less desirable HRV-CVs. And so putting that all together, all the things we know, that are associated with risk for cardiovascular disease, unfortunately being male, getting older, and having a higher body mass index associated with high heart rate variability coefficient of variation, also associated with cardiovascular disease risk. So for anyone listening to this conversation, I think the big question is, and I think we

sprinkled it in throughout, but I think it's worth, you know, at the end of this conversation, kind of summarizing, if we are interested in a lower heart rate variability coefficient of variation, we're looking at, you know, how do I build this autonomic nervous system resilience and robustness? What would be the advice that we want to give to a person listening? What is the path to a lower HRVCVCB? The truth is a lot of it are the same things that are going to raise your heart rate variability and lower your resting heart rate. The biggest one, and this is straight from the data that we saw that shouldn't be surprised is to stop drinking as much. You know, there's a decrease in alcohol consumption globally, it seems.

So, yeah, stop drinking as much exercise, but exercise within yourself. And we actually have... Appropriate training. Yeah. We have a really interesting case study that I can't wait till it gets public of an individual who lost an incredible amount of weight. 30% of his body weight, no natural, no medication. And that's not to put shame on the use of medications. In fact, highly would encourage it if people are considering.

But he did. His heart rate variability, coefficient of variation dropped a lot. But what was interesting is on his weight loss journey, he trained for and completed last summer. And maybe we'll have a separate pod about this. I'd love to. Half Iron Man Triathlon. And you can see when he was really in the depths of training. And he was straining pretty hard.

Like we're talking a lot of 15, 16 average strain weeks. His HRV CVCV went up quite a lot. So I think after the Half Iron Man, he stopped doing that, obviously, training at that level and it dropped back down. So that's not to say that a couple weeks of high HRVCVCV if you're trying to do an overload are probably what you want to say. Right. So those strategic functional overloads are part of the strategy of getting fitter.

There's no question, right? So we want to see what is that percentage of increase that it's acceptable, tolerable, that gives you that overload, that strategic overload, that functional overload, without actually dipping into reserves that make it hard for you to rebound next week. I mean, that's where these data, I think, become really valuable as you start to see what is tolerable at an individual level in terms of overload. For me personally, like I, you know, for the teams that I work with, you know, we don't land on it after, you know, one week, right?

Like it takes months, really, to figure out an individual level how people respond and adapt. Yeah. And a great way to do that is give the overload and then look how quickly the body of course. Yeah. And then in controlling for some of these factors that we know are going to impact your ability to capitalize on the overload that you just went through. Like age. Age is a factor certainly, right? Sleepway consistency. I remember with Duke basketball years ago, you know, we just went through this like functional overload period. I was like, sure. Don't fudge it up. You know, because you kind of have this. This is I think one of the downsides of being an athlete. You kind of like binge, you know, these bad behaviors after. like a hard trading block as a reward, right? But I was like, no, no, no, this is this is where you gain the competitive advantage, right? Because you don't want to undo all that work that you just did by engaging in all these naughty behaviors. It's easy. It's easy to do. Yeah. If you can, if you can go through this functional overload period and then you can live a week of just engaging in kind of tapering type of behaviors, prioritize your sleepway consistency,

eat as healthy as possible, maintain a time or strict eating kind of schedule. I mean, mean, that is how you get fitter. You know, I've seen my HRV, you know, get better over the course of the last, like, eight years. Like, I think that you can kind of bypass age a little bit, right? Modifiable for sure. It's modifiable, right? Like, by thinking strategically about these overload periods and mapping that with a really strategic taper, you can make huge strides in your HRV. Yeah. And I think going back to questions about modifying and positively modifying or lowering HRV CV, So it's important to recognize that in the paper we studied the big four, right? Drinking, exercising more, sleeping more and sleeping more consistently.

But as was brought up by Andy when he was reviewing the paper, there's a lot of factors that we didn't or can't evaluate that are going to have a massive impact on HRVCV. And just today on our call with our partner Ferrari, some Formula One racing, I saw your eyes light up when they were talking about some of these modifiable factors and things that we aren't necessarily thinking about that don't fall within those big four. Things like travel and sleep and rest when we're traveling, right? When are we napping? When are we in taking caffeine? You know, when are our opportunities for sleep? You know, yeah, I mean, there's just an ocean of opportunity to help these folks deal with the demands of that job. And psychological stress being one that we can maybe physiologically pick up on, but not necessarily identify the etiology of.

Right. Like anxiety and stress before a big performance is something that can be detected by stress monitor, which mine's probably in the mid ones to twos right now. But it's challenging for the whoop to identify or for us to identify. Is that coming from the fact that I'm going for a walk? or that. And do you feel threatened or do you feel challenged? Exactly. That's another big piece of stress monitor that we know.

How we perceive stress. It's huge. It's impacting how our biometric respond. So, yeah, there's a lot of factors at play that we can and also necessarily can't evaluate yet as contributors to how HRVCV and our body response to things. Is there anything that you want to add about HRVCVVCV that you think? is important that we didn't touch on. We need to extend a big thank you and gratitude to all of our wonderful colleagues, both externally and internally, who helped with this for their support,

you know, to the American Journal of Physiology, Heart and Circulatory Physiology for being a wonderful partner. Yeah, their reviewers were, I thought, sensational. Outstanding reviewers. I think it was 15 pages of single space reviewer comments that I submitted back and I read. did the entire analysis to show them that it replicates. You like persisted that in ways that are really admirable. Well, thank you. But really we were lucky to one work with one of the coolest products in the world. Grateful for the feedback from the scientific community that supports the rigor of what we're doing. Yeah. I want to have a quick shout out and hopefully we keep this in because an executive letter we're talking about, okay, how do we recruit? You know, we've got to hire a lot

of folks this year, which is super exciting. And on the performance science team, we will be hiring probably three new scientists over the course of of this year. So hopefully people can listen to these conversations that Greg and I have about the data and about the research we do. And if it excites you, consider applying for some of the open roles that come across because we really are on the very edge of health and performance and have an opportunity to investigate. The questions that we can ask is just pretty much endless.

So having some more firepower team is going to be unreal. Yeah, it's a global experiment of human performance, right? And I don't think there are bigger datasets that exist. And more exciting. Right. The academic collaborators who we work with attests to that. So when we have people from Mayo Clinic and Baylor and IU and UCLA and the biggest and brightest of minds who are hungry to work with us. University of Iowa, Go Hawks. Yes, some super cool research coming out of. Stanford. Consistency, of course. We can't help ourselves. Stanford, Monash. I mean, gosh.

Yeah. So come work with us as the short of it. All right. And yeah, we look forward to that. Thank you so much, Greg. Thanks, Kristen. If you enjoyed this episode of the WOOP podcast, please leave a rating or review. Check us out on social at Woop at Will Ahmed.

If you have a question to us answered on the podcast, email us, podcast to Woop.com. Call us 508-4434349552. If you think about joining Woop, you can visit Woop.com. Sign up for a free 30-day trial membership. Newmever can use the code Will WI-L to get a $60 credit on Woop accessories when you enter the code at checkout. That's a wrap, folks. Thank you all for listening. We'll catch you next week on the WOOP podcast.

As always, stay healthy and stay in the green.