Episode Transcript
[00:00:00] Speaker A: Welcome to the Heart Rate Variability Podcast. Each week we talk about heart rate variability and how it can be used to improve your overall health and wellness.
Please consider the information in this podcast for your informational use and not medical advice. Please see your medical provider to apply any of the strategies outlined in this episode. Heart Rate Variability Podcast is a production of Optimal LLC and Optimal HRV. Check us out at optimalhrv.com Please enjoy the show.
Welcome friends to Heart Rate Variability Podcast. I am Matt Bennett and I am excited to explore a topic that I'll be honest, I don't know a ton about. But I'm so excited. Dr. Judith Andersen said you got to talk to Dr. Chan about her work on heart rate fragmentation.
So I've been nerding out a little bit of this, preparing for this episode. But Dr. Chan, welcome to the show. I would love just to start out, just a quick introduction of yourself and your work.
[00:01:05] Speaker B: Yeah, thank you. My name is Dr. Jennifer Chan. I am currently a defense scientist with DRDC Canada, so also known as the Department of Defense Development and Research of Defense Canada.
And I am a psychophysiologist. So I primarily focus on research looking at the intersections of both psychology and physiological measures. A majority of my work has been in cardio cardiovascular measures in the subject of health psychology. So what components of health, both physical and mental, how do all these things intersect for good health?
And a lot of my research comes from my graduate and dissertational studies at the University of Toronto, as well as focusing a lot on police research. So my background is actually quite a bit on police performance, health and behavior.
[00:01:58] Speaker A: Awesome. So let's, let's just dig right into it. As someone who has been like obsessed with heart rate variability, the heart autonomic nervous system, I was sort of surprised as I was talking to Dr. Andersen that I never heard the term of, you know, heart rate fragmentation. So let's just go for the basics.
What is it and how did you where. When did it make your radar?
[00:02:26] Speaker B: Yeah. So heart rate fragmentation is a relatively new concept. I cannot take credit for it. That is the good work of over from Harvard Medical Research, Drs. Madeleine Acosta, Roger Davis, as well as Mary Goldberger.
It initially came across my radar during my dissertational studies.
I, like many individuals, was very affected during the COVID pandemic through my research. A lot of my research, while I've done some cardiovascular research before, I actually departed away from it during my graduate studies because of the different issues we would have with measuring cardiac activity, especially in the field with police officers. So we're often going in the field with their training, they're running around, we're dealing with movement and sweat and just even Bluetooth connectivity, trying to keep up with all of those things. And so I was actually moving away from cardiovascular measures for a little bit and focusing on saliva measures, looking at more endocrine research until the pandemic hit. And as you can imagine, it's very difficult to collect saliva and posit any research to collect saliva when you're in the midst of a pandemic.
[00:03:47] Speaker A: Nor would that be the number one thing on my list to do on a typical Tuesday during a pandemic.
[00:03:52] Speaker B: Exactly it actually I kind of came back to cardiovascular research as a bit of a pivot to try to be able to continue my research and my graduate research while still remaining in the realms of physiology. But I was hesitant to dive straight back into HRV again because of the issues we had with it. So we were discussing different measures and different approaches to research and you know, around this time I was becoming more knowledgeable and the understanding with my colleagues about other forms of nonlinear based heart rate variability and the general movement of cardiovascular research and science towards nonlinear measures like things like sample entropy, things like even frequency measures of low and high frequency. And so going through that and trying different things, we were still having the issues of how do we measure this in an applied field sample, how can we interpret this given the fact that while I'm a physiologist, I do not have the same amount of years of experience as a cardiologist when we're trying to look at things like point hair plots and sample entropy numbers and individually trying to assess what is healthy for a person given that is so personal and it can vary so much. And mostly what we're working with are people who are considered physically healthy, if anything more towards a peak health of just very active and physical. So it was one of those trying to find those interpretations. And so through the readings as well as, you know, collaborations with my research with Dr. Anderson, who was my supervisor at the time, as well as another one of our colleagues, Dr. Jo we came across this research from the Harvard laboratory, looking at it and it, it was just, it was so interesting and not, not necessarily the, a new, brand new concept, but definitely encourage your perspective and in a way that was just so easy to grasp. You know, it's so intuitive I find in understanding it and applying it especially for applied when you're trying to create these interpretations for non scientists and non cardiovascular researchers. So that was really the enticing aspect of it at the beginning. And when these original publications were starting to come out at the time relatively new, they were only out for a few years. I think the earliest publication was 2017. Maybe at the time. Yeah.
[00:06:40] Speaker A: It doesn't go back too far.
[00:06:41] Speaker B: Yeah. So most of the research and still is primarily focus in the cardiovascular field and cardiovascular science. And I really wanted to try to take that into health, psychology and physiology and try to see how it applies. Can it be transferred or are we going to see the same issues in which, you know, we see sometimes with applied cardiovascular research, people tend to run away with the concept of HRV without necessarily applying it in the most appropriate scenarios.
[00:07:12] Speaker A: Yes, absolutely. So. So let's just dive into the. The definition for those that aren't like. Like me initially not familiar with the term because it seemed like, you know, at least my basic understanding, they could help us with some of the issues that if you kind of touched on that heart rate variability as we were talking with. We've had a Judith and another woman working with police forces. It's like you stand up and your data skewed right there with heart rate variability. So while it's a powerful metric, I found it's also incredibly sensitive and artifacts can come in there and just wreak havoc, especially if you're studying in real time. So I would love just to dig into the definition of around fragmentation and how it might differ from other metrics when we're talking about more related to heart rate variability.
[00:08:07] Speaker B: Yeah. So heart rate fragmentation is referring to the concept of changes in acceleration and heartbeat. So we sometimes slow up and we slow. Sorry, slow down and speed up when we're looking at heart rate. So we're looking at how often do we change from going from an acceleration to a deceleration state. If we kind of think about it as when you're driving in a car, when you are on the highway and you don't want to put sudden hitting from the gas to the brake, to the gas to the brake. That's really bad for your engine, bad for your fuel economy is just not good for your car. Your. Your heart is the same way. You don't want to constantly be switching from acceleration to deceleration. So sometimes if someone is very fragmented, they can be changing from speeding up to slowing down almost every single heartbeat. And that can be physically extremely taxing for the heart and can lead to a lot of issues.
So when we're talking about fragmentation, it's that sudden change. And the more that happens, the more taxing that is the More physiologically problematic that can be versus something that's more gradual, where we're having gradual changes in acceleration and deceleration. That is what we would call more fluent or a healthier pattern. So you can still have a certain amount of variability, but also maintain fluency. And you could also have high variability and have a very fragmented, a very fragmented pattern. So it allows us to add an additional layer on top of heart rate variability because it is a variability measure. We are looking at the changes in dynamics, but we're adding this extra aspect of it, another dimension of it with these patterns and directionality of these patterns.
[00:10:00] Speaker A: Very cool. So I, I know our listeners are, or most of them at least are familiar with things like residence frequency breathing. Where we, we see on the inhale, the vagal break releases, we get that sympathetic. Then we put the vagal break on it, we see a slowing in heart rate. Heart rate variability is using that to, to measure it. So with the frequency, are we talking about a longer measures than on the inhale, exhale. Are we, are we still utilizing some of those that same systems? Like where does this kind of come from?
Kind of building upon what many of our listeners will know with resonance frequency breathing and the heart rate and the breath.
[00:10:44] Speaker B: Yeah. So a lot of resonance frequency breathing is, as you've probably discussed, very dependent on vagal mediation or this vaguely driven activity. So when we have the coupling doing heart rate and respiration and it is very apparent we can have that driving force of vagal tone modulation. However, that's not necessarily the case when we have more short term changes. Heart rate variability is not only inputted by vagal input. It's not the only thing depicting its activity. So when we're kind of looking at these signatures and these more frequent changes, it's taking a look at this, the idea of sinus arrhythmia or these erratic sinus rhythms that may not be completely driven by vagal mediation.
[00:11:37] Speaker A: Great. So, so is it. What if it's not the that? What, what is it? What are we looking at that. I mean I could imagine like, like you said, driving on the interstate, somebody cuts you off in traffic, that that would probably, we know that, you know, puts your system into hopefully the flight response and gets you to safety. But, but what is, what is sort of behind that? The fragmentation?
I'm sure it may not be just one thing, but what are we seeing when we see, I would assume, if I'm using the term right, increase fragmentation.
[00:12:10] Speaker B: Yeah. So as far as the exact biological mechanisms are still research Being understood. Trying to understand what exactly? If we, if I had a pinpoint, I said, okay, you know, is it a valve? Is it, you know, is it a specific node in the heart that we can use and measure? It's not necessary. We haven't necessarily determined that aspect of it yet. But what we do understand it as is that when you have this higher level of fragmentation, it's representing an overall output of dysfunction in your cardiovascular network. So we're seeing that this coupling of, you know, physiological activities both with your heart and your breathing, they're breaking down. We're not having that resonance, we're not having that matchup. So we're having this literal dysfunction and physiological breakdown of some or many components of this network that is starting to emerge that exceeds your physiological capability to adapt or respond appropriately.
[00:13:14] Speaker A: Gotcha. So maybe a sign of going out of homeostasis at that point and struggling to adjust in some ways.
[00:13:24] Speaker B: Yeah, absolutely. I mean, our body works on, you know, waves of patterns. We love resonance. We love those predictable waves of pattern. And heart rate is another one of those frequencies. Right? We have that continuous. So if you are in a fluent state, you are getting those gradual change in, you know, sinusoidal patterns. But a fragmented pattern breaks out of that frequency, it breaks out of that rhythm. So we, that, that's a really good way to match it.
[00:13:57] Speaker A: Excellent. So do we see, you know, correlations between like whether it's RMSSD or a frequency domain. Do we see a relationship there between the, you know, fragmentations and I would assume oftentimes lower heart rate variability. But I don't want to assume too much because so, so is there any relationship there? Are we talking about sort of separate entities that may be telling us a little bit different information about our current state?
[00:14:35] Speaker B: I think that it's a little bit more nuanced. They are inherently tied together because they're both time domain based variability metrics. So inherently they are tied together. You have more fragmentations and, you know, more fragmentation and lower HRV is less healthy, but higher HRV and lower fragmentation is more healthy. I think the ways that it can separate and so far in the research has been showing, is then we're trying to look at healthy activity and we can see a bit of a departure of how well it can discriminate healthy versus unhealthy individuals. So for example, RMSD in healthy versus cardiovascular, you know, individuals with cardiovascular disease or issues.
We know that it can discriminate quite well between the two sometimes in these, you know, distinct groups. However, it can be more difficult for things like frequency domain measures. It can be more difficult for things like the PNN 50.
[00:15:48] Speaker A: Yeah.
[00:15:49] Speaker B: What we're seeing with fragmentation is that it's actually much better at discriminating healthy versus non healthy groups. So we're seeing that direct translation from unhealthy activity to unhealthy physiological outcome that may not be representative from other HRV measures. I mean, we have that paradoxical relationship where as individuals get older, their HRV also gets healthier, it gets higher and better. And so when we have these paradoxical relationships, it kind of suggests to us that maybe it's not working in the best way as intended, or maybe it's also not working in the most efficient way for our interpretive and applicative uses. So that's kind of where it starts differentiating from each other. But they are both based on the same idea that there are certain patterns or certain activities that trend towards health.
[00:16:47] Speaker A: So with the fragmentation, you know, one of the things I love and is sometimes frustrating with heart rate variability is such a, a generalized measure. I we have a joke on the, the podcast of a drinking game, but a healthy drink, of course, Google HRV and anything and see if you get a, a peer reviewed journal article, if not a meta study that that's coming out, you know. So on one hand it's an amazing generalized measure of health and wellness. On the other hand, it's hard to say, okay, is this because I took a long run yesterday or that I'm totally, I'm dealing with mental health issues? Like, you know, that there's, there's those two sides of this equation that are hard to distinguish unless you have been taking your heart rate variability every morning like I have for like the last 10 years. So you start to learn that, but it takes that time to do it. So are we looking with the fragmentation piece, are we looking at, does it do like HRV does and go beyond just cardiovascular health? Right. It sounds like we're also potentially measuring autonomic nervous system health, mental health, is it, is it sort of universal in the same way as hrv?
[00:17:58] Speaker B: So that's actually a lot of what my dissertational research and we actually have an article in the works under review right now looking into its application outside of cardiovascular health. So what we were looking at in our research was could this predict or could this just be sensitive enough to discriminate mental health differences across individuals, especially not just mental health differences in people who are clinically differentiated? Because we know that people who are, once they hit a clinical level of any form of health issue, they typically are physiologically quite different. So I was really interested again in this gray area. These are individual particularly healthy. And so what we did for my dissertational research is that we looked at undergraduate students and had them self report to different mental health characteristics. So we put them through. We had them self report on symptoms of anxiety, depression, ptsd, as well as general stress.
And based on that, we were able to come up with a group that we referred as our typical or healthy group group quotes on the round, the healthy who didn't surpass the thresholds for any of the mental health symptoms, or individuals who we called our probable mental health group who surpassed the thresholds for one of either ptsd, depression or anxiety. And what we did is we measured their responses to baseline relaxation versus slightly cognitively stressful tasks, not necessarily stressful emotionally, but just made them work and think a little bit. In our case, we used just a short mathematical cognitive test. And as well, I think we did almost like a stroop paradigm of connecting emotional faces to expressions.
It was emotional fake stroop test.
What we found, which was really interesting, was that if we used fragmentation as a way of identifying if someone's being stressed or aroused in the same way as hrv, we found that fragmentation actually did predict when they were going through these more stressful situations.
That we saw an increase of fragmentation, an increase of physiological stress, when they're going through the physiological stressor. And what was really interesting was when we tried to compare the groups of the students who were had probable mental health symptoms versus the healthy group was that the individuals who had probable mental health symptoms actually were less reactive to the stressor in comparison to the healthy group. So we're seeing actually less changes in fragmentation, which kind of aligns with what we know about stress and in other physiological aspects. So we often see that individuals who have mental health symptoms have dysregulated stress responses. So they can either be a greater response or what we call a blunted response. So in this case, we were actually seeing individuals who had probable mental health, so not clinically diagnosed. It was sensitive enough to differentiate that they had a more blunted response. So not adapting or responding to the cognitive stressor in the same way as the healthy individuals were fascinating.
[00:21:37] Speaker A: So let's bring that then that response like, so the fragmentation data, what were you. What I'm just endlessly fascinated with what were you? So was the group that had, you know, was identified with mental health, you know, maybe not clinical, but in that group, were you seeing if they weren't as reactive, were you seeing less fragmentation in that group?
[00:22:06] Speaker B: We didn't actually. So when we compared just baseline fragmentation between the healthy and the healthy students and students who had probable mental health symptoms, they didn't actually differentiate on their baseline levels of fragmentation. And for the purposes of this study, we defined fragmentation as the most extreme pattern so that switching every beat. So fragmentation is, you know, it's a continuum. You can have more or less fragmentation. But the specific pattern I wanted to focus on was the most extreme one because as far as for argumentative sake, you can't argue that that's not a healthy response. That's the worst pattern that you could have in which almost every beat. So I was looking solely at the proportion of time they were in that state. And so when we compare the healthy and the probable mental health individuals, we didn't see a difference in just those baseline amounts. So when they're at rest there's very similar amounts saying to when there's stressors, we're looking at sheer magnitude that didn't necessarily differ. And that's not an unusual thing to find, especially when we're looking at physiological measures like for example, it could be something similar like cortisol where you have a general trend and response throughout the day. But an individual who has a specific magnitude of cortisol and is unhealthy may not be the exact same number for someone else. And same thing for HRV like RMSSD and sdnn, we don't have these magnitude. This is the exact number. We have ranges. So our healthy and you know, are not so healthy potentially individuals, they were following within the same ranges. We wouldn't call them clinically or statistically significant from each other. And those baseline measures at any given state of like baseline stress or recover recovery, but it was really in that reactivity aspect that they start differentiating from each other. And that's really where a lot of stress research is moving towards as well. We're not looking at single baseline states because it, when your stress itself is not a baseline or a stagnant activity, it's a reactive activity. So a lot of times there's a mood and more. So now there's even more of a movement towards acknowledging stress as a multi stage essence. It's not, you know, a dynamic aspect. We can't look at a single state and do a comparison really. We have to look at this reactive change with this overall larger pattern.
[00:24:44] Speaker A: Fascinating. So do you then like what are, and I'm interested to bring this into your sort of work with maybe through the defense, like peak performers or, you know, others, like what, what is the utility of heart rate fragmentation? I know you, again, you've done work like that with Dr. Anderson, with police, you know, officers, like, what is some of the potential? And if, if it's like actual, and I'm using potential wrong, you can correct me there. But where do you see some of the potential benefits of this metric that may be beneficial over, you know, some of the traditional heart rate variability metrics?
[00:25:31] Speaker B: I think from a theoretical perspective, it really brings a quantifiable aspect to nonlinear measures that makes it much more objective. So that's the first thing. So we want to move towards nonlinear measures because they are more representative of cardiovascular dynamics. However, so many of them are visual based. If you look at a frequency spectrum plot or a point care plot, I cannot say the number of hours I spent trying to train myself to read and understand point care plots. And even then, if you give me two case profiles, I would not necessarily be able definitively unless they were very different shaped, you know, the nuances of, you know, one ellipse from another ellipse. Like, it's really difficult. And that is a majority of physiological scientists and researchers. We're, you know, not, we're not all cardiovascular specialists. So that's one aspect that I really like, that it brings a level of objectivity and quantifiability, especially for reproductive sake in science, for comparisons in research. It makes it. It's a lot harder to replicate a study if you're basing it on subjective analysis of a graph than, you know, a number of saying, this is 70% in this variable and this, you know, we have something much easier to work with.
So that's from a research theoretical perspective of why I think it's really great and applicable from a client or even a clinic perspective. I'm not a clinician myself, but for individuals I've worked with, Dr. Joe Arpaia in particular, he's found a higher level of applicability in the clinic that he's mentioned to us that he needs a measure when he does biofeedback with his patients that can be seen on the screen and makes intuitive sense to the patient that can show this rapid change in state within a few seconds. Again, that idea of transferability, that translation aspect, it is very easy to see someone who is in a very fragmented state than someone who is in a fluid state. And that's a lot easier than a spectrograph. I teach fragmentation very briefly to my undergraduate students. When I was Teaching at U of T. And these are students who necessarily may not have a full biology background or especially a cardiovascular background. They're coming into this with, I know what HRV is because I wear an apple watch and it's a stat that gives me, and I will throw up onto the slide this great example from Acosta of a patient with heart disease versus a patient who is healthy and it's an ecg. And I say, okay, tell me which patient is the unhealthy one? And they look at it and we're like, the bottom one, because that one is in the red slide.
That's. That's the only differentiation. And then I will throw up their graph of their NN charts or their, you know, end to end values. And you can see this fragmentation. You can see the actual movement and acceleration. And then I say again, which one looks healthier? And they say, oh, well, it's the one that looks way more erratic and looks kind of all over the place. And just from that, you know, that these are people who have been exposed to it for less than five minutes and they can immediately say, that's the one that's not the healthy one. And this is how I can identify someone who's not doing well.
[00:29:10] Speaker A: So with those, you know, and this is where. This is kind of what fascinated me about this metric, potentially. So are we at the point with fragmentation, which I don't know if we'll ever get, really, with heart rate variability like that, you can measure someone in real time.
Can we have a police officer run through a active shooting scenario drill and measure fragmentation in any meaningful way? Is it allowing us to kind of get to that level or are we still a little bit out from that?
[00:29:48] Speaker B: I think we're not necessarily completely there quite yet. I mean, there's still the same issues of movement that's, you know, measures are limited by the technology that can be used. And so with something like an applied police scenario, our major issue again is movement. And, you know, you can get fragmentation simply because a signal is connecting or not connecting and creates a different interpretation of what those patterns are. Because we do require patterns, meaning, you know, you need a few heartbeats in a row to be able to come up with a subsection of pattern and to see how that repeats and moves on. So I wouldn't say necessarily it's quite, you know, let's throw this in the field and we can just grab this immediate aspect. But it is to the point where I would say it at. Is at a similar level of, I would say HRV measures that exist already in feedback and speed in which we can say very immediately we can put someone in a chest ban, have them go through their scenarios or whatever tests they're doing and just feed that through our calculations to be able to quickly identify, okay, this is the percentage that you were in this state. We can identify at what point you were in the state in, during which sections were you in it more than less, you know, so we can definitely, from a comparative sake, I think do it just as efficiently as, you know, RMSSD or SDNN nowadays.
[00:31:21] Speaker A: So it's kind of right there with the researches. There's, it's not ideal, but given us, we, we can gather some information there and hopefully.
[00:31:30] Speaker B: Absolutely. I think, yeah, I think from like, just like a literal usage phase it very much can be the means I would, the means to be able to calculate it is not exactly widely spread. So there isn't for example, a, an app that you can have right now that can just calculate fragmentation for you. It's a really straightforward calculation, but it's also just not widely available at this.
[00:31:57] Speaker A: Yeah, well that, that's what's been in my head like you know, with the, the optimal HRV app, you know, at the end of the three minute morning reading we do where we try to get people to sit still for all the reasons we're talking about, you know, you have RMSSD sdn, you know, we, we give the max min, just in case you're interested. I don't know how many people know what that, that that means. But so if, if you think about, you know, let's just take like a three to five minute, you know, HRV reading, would heart rate fragmentation sit well within those other metrics? In other words, like, if we were as would you suggest, this very energetic person who's really excited about HRV has got this little app thing going. Would it. Do you see in the future that this would be a metric that sits alongside of those metrics and gives us a little bit, I think the answer is yes to this because of what you already said, but gives us a little bit more different information than the other metrics do.
[00:33:04] Speaker B: Yeah, absolutely. I'm not in a state to be saying that, you know, oh, fragmentation is be all, end all. We're going to replace everything we know with this one measure. I think if anything, the existence of this measure and the emerging, continuously emerging research is further proving that we need to be a little bit more nuanced in the, you know, what, how we interpret hrv. We need more than One measure.
It's not a good idea to have one measure in general. For hrv, we're still at the state that we do need to have multiple measures to create a more.
It's a complex interpretation, but it's also a complex activity. Yeah, you can't have something that's very complex be measured by a single measure and a singular way of calculation, especially something as complex as hell. So I encourage the use of this to be added to toolkits of HRV to add that extra dimension, to add that extra level of interpretation, especially when we can see conflicting, conflicting stories, for example. So I like to always use case examples in my classes. And we can throw up different scenarios. So we had a. We have a scenario of a patient who. She's a veteran and she has severe PTSD from military sexual trauma as well as from childhood. And we would throw up what her activity looked like when she had her eyes closed and not instructed to think about anything in particular versus have her eyes closed and focus on being safe. And from a comparative state, we would see that according to our base levels of rmsd, LFHF ratio, as, you know, conflicting. That maybe as a measure itself, but, you know, with those measures, it almost made it seem like she was healthier or had better activity when she was unfocused versus when she was actually focusing on being safe. And with the fragmentation, we would actually see an opposite effect of comparison. So we would see that when she was unfocused and she would let her mind wander and, you know, perhaps lead to a more stressful state. We saw a higher level of fragmentation comparatively when she was focusing on being in a safe state. So it adds that dynamic aspect of, you know, we can't just take a singular measure and say, this is it. Because if we do that, we do risk of running with these interpretations or stories that may not be the most representative. We can see different things just based on the measure you choose. So if you're only choosing one metric.
[00:36:00] Speaker A: Right.
[00:36:01] Speaker B: You are limiting yourself to only a singular potential interpretation and potentially one that may not necessarily make sense or align with what should be happening or what is happening for the individual. Right. It's. You're getting a completely different interpretation. So as someone who's an applied researcher, especially for someone who works with, you know, police, and now I work with military members, and a lot of what we do informs. Informs policy.
The danger of having an incorrect message or interpretation becomes a lot greater when you only focus on one perspective.
[00:36:41] Speaker A: Yeah.
[00:36:42] Speaker B: And so it's. We're all about, you know, multiple perspectives, making the most informed decisions we can. So taking multiple types of measures can help improve how informed you are when you're trying to make those interpretations or those takeaways, especially in a clinic or an applied scenario for a case individual.
[00:36:59] Speaker A: Yeah, what I love about your work and how you look at this and in those multiple metrics, and what I'm really obsessed about is how maybe AI can help us. Like, okay, your sdn. So just, you know, before you're teaching me about this new metrics, like, okay, your SDN and your RMSSD and your Maxman, we've got those metrics.
What are they telling us together? Like, that's where like, fragmentation could be a really.
It just sounds like it is a really another piece of information maybe differentiated enough to actually tell us something a little bit more. Because I still, if you ask me, here's my sdnn, here's my rmssd. What is these two scores telling me versus RMSSD in isolation?
You know, I might be able to take a guess, but, you know, I don't think it tells us a whole lot more. It may tell us a little, but, you know, I think you need almost something like AI to go in and really do that analysis behind the scenes. And that's why your work is so exciting, is that it might give something a little bit on, a little bit removed, but that, that is actually a strength because it gives us more information than something just a little bit different like SDNN and rmsd.
[00:38:27] Speaker B: Yeah, it's really interesting, and I mean, I'm not by any means an expert on AI or machine learning, but it's really interesting considering the limitless of potential of what AI could theoretically do to really pull and disentangle the different complexities and nuances of it. We kind of talked in theory in the lab about the idea of using heart rate fragmentation as a way of almost creating a cardiovascular fingerprint, as we like to call it, of activity in individuals, because we're looking at literal patterns. You know, I can draw the pattern that someone actually is making based on their acceleration and deceleration. That can goes into the whole aspect of symbolic dynamics, which is a whole separate thing to talk about. But as far as like interpretive sake, well enough that's interesting in which, you know, we're decreasing the emphasis on the sheer magnitude of change and more focused on the change itself and the change of direction. So rather than say, for example, okay, we're increasing acceleration, you know, it's 5 milliseconds, 5 millisecond change in acceleration versus a 3 millisecond change in deceleration. We're looking more specifically on did it go up or did it go down? And by what thresholds we apply to it, we can create these visual patterns of actually seeing that fluency and fragmentation. So I think that potentially given that AI is limitless in its application and potential use, it would be really interesting to see if there was a way combined with what we understand about the different metrics to create that idea of a digital like a cardiovascular fingerprint. I think that would be really cool. Yeah. And you know, if I let my like sci fi brain go into it, maybe a little scary, but also very cool.
[00:40:17] Speaker A: Yeah, yeah. It'll either destroy us or tell us more about our heart rate variability and cardiovascular health. Well, one or the other there. So. So I do want to follow up on something you said because it struck me as like the second aspect of heart rate variability that really fascinates me and I know a lot of our readers is heart rate variability biofeedback. And it sounds like this, we could be using fragmentation potentially for that as well. Did I hear you? You correct. Because right now like mostly focusing on low frequency like metrics during HRV biofeedback, but it sounds like, let me just throw it out here and you can correct me and take me to the right path. Is that maybe a. But I'll let you tell me because I'm thinking like, oh, if you're breathing at residence frequency, are we seeing different fragmentation scores coming out of that as well? But I would love to see how you, how this might be used or is being used by biofeedback practitioners.
[00:41:25] Speaker B: Yeah. So in general, we actually, it wasn't the primary focus of the study because it was, it wasn't designed in that way.
But when we were looking at the students, we actually did incorporate a portion of kind of teasing into residence frequency a little bit. So after, you know, before they got stressed out, we had them go through different breathing rates. And to say which one does feels physically most comfortable to you because at least from what we've been able to tell, and we're doing ongoing research, people are relatively good at identifying what's a comfortable breathing rate. And we, we, we posit that a self identified comfortable breathing rate is oftentimes pretty close to your actual resonance frequency.
And so what we actually found was that individuals who after they went through the stressful scenario were then recovering and we gave them a breathing triangle which they were following that breathing rate. We actually saw a significant Decrease in fragmentation not only from experiencing stress, but also significantly large decrease in fragmentation from baseline. So when we just told them, breathe, breathe however you want, we're not going to tell you to breathe, just sit here for, you know, 10 minutes, do whatever. We actually did see with resonant frequency breathing or this idea of paste breathing, paced recovery breathing, this decrease in fragmentation. And this was observed for both the individuals who are healthy and the individuals who had mental health threshold or surpassing threshold level symptoms.
[00:43:04] Speaker A: Awesome. So, so potential. If I, I'd say there's potential implementation for that biofeedback arena as well, which is really cool.
[00:43:14] Speaker B: Oh, absolutely. Yeah. Honestly, a lot of what we've done comes from the biofeedback aspect and its practicality and uses in biofeedback. And that's why it's so useful, especially in clinics for patients. Just again, that visual aspect, that of showing that to patients, because it's, it's more difficult, I think, to show resonance frequency sometimes to patients or participants who are not versed in it. Especially if, you know, if you've seen some of the live biofeedback screens, they're either oversimplified in which you get a little like green light if you're doing it well, even though they're not necessarily search certain of what they're doing well, or they get something that's overly complex where they're seeing those sinusoidal waves and they're like, oh yeah, it's matching up. And you know, you don't necessarily aren't necessarily mentally ready to look at something like that. Right. Especially for someone who's not familiar with looking at that every single day.
[00:44:17] Speaker A: Yeah. Yeah, that's awesome. So I, one of the things that I just love to get, I'll just throw this out here, you don't have to even respond to it. But, but what I was really excited about with some of this work is a little bit of frustration of mine. Even though the shoulders of the researchers that you and I both stand upon, I think we would both celebrate the heck out of them tremendously. But I've been a little frustrated that, you know, as I got into hrv, probably about seven or eight years ago now, there was this like European paper that was put out that defined, it's kind of a bible inverse of heart rate variability. It defined the different metrics. And I was really kind of surprised that maybe the low frequency high frequency metric is going out, but maybe a few things in that paper are going out of style. But we haven't really seemed to Add a whole lot of new stuff to this. And I'm like with our technology it was sort of surprising as somebody relatively new to the field that okay, what, what happened in you know, 2010? Did we like, like 13 years later we're still on these same metrics. And so I, I wonder like as you learned about this and where do you, is this just something you think that 10 years from now we will have that, that new metric that I've been waiting for, at least for a while. Do you think this will, will come in to be that for us?
[00:45:50] Speaker B: I think so. I mean it already is starting to and we're seeing signs of it. So yeah, I'm familiar with that European paper. It is that workshop paper of this is. This is the bible for hrv.
[00:46:02] Speaker A: Yes.
[00:46:02] Speaker B: And you know, in many ways it still is.
It, it really identified HRV and excuses for discriminating cardiovascular health. And that can't be disputed. It absolutely is. Great.
[00:46:16] Speaker A: Yes.
[00:46:17] Speaker B: And it also though and key thing I think about that paper is it does get into its own caveats of we are measuring people who are laying down, not moving, they are awake, they have either they're fine in health or they have a clinical level of cardiovascular disease that is, you know, impeding with their life and potential mortality. So we can't necessarily work with this, you know, alone. And they mentioned in their own measures, right. This alone is not enough. You need oxygen, you know, VO2 levels, you need respiratory rates, you need all these specific conditions in order to make this work. So they self identified the limitations of it in that I think and that.
[00:47:05] Speaker A: You and I can talk about a paper from the 90s and both know exactly what we're talking about. It says tremendous amounts about the work that that group group did. So it's, it's a sign of respect.
[00:47:16] Speaker B: It's one of those. Yeah. And it's absolutely use it. And I think why maybe it's not necessarily we're seeing a lot of movement from that is because of those limitations and of those caveats. It, it's. We're still I think in the phase of research as it goes in which you get something that's considered gold standard and then everyone starts and then it gets moved into the different fields and being used as the gold standard and not necessarily being disputed. So you're not necessarily having non cardiovascular scientists come back and say okay, you know, this is not the best thing we can do about it, but we're using it anyways versus say a cardiovascular scientist who would be the most likely Individual who would be really targeting and updating the research and science on it for their uses. HRV is great as it is. It is great for discriminating cardiovascular health. That's what they need it for and that's so when you're building on it, it can necessarily, you know, change and updates like fragmentation come from an as needed basis. Right. We invent things based on need and change and fragmentation is addressing the aspect of adding objectivity to a nonlinear measure. That's what it's addressing, that's what it's adding to. So we are seeing though it getting pulled in and I would imagine like any of those gold standard papers, they usually update themselves once every, you know, 10 to 20 years. Usually is usually the cycle for those gold star review papers. And we are seeing though fragmentation being pulled in other fields of aspects of identifying this exists. This is something that we're looking at more closely but we just haven't gone there and research yet. There's always a catch up time because you know, even from the state of publication, which is, you know, the earliest time that you can encounter it to go from. Okay, let me read this, let me understand it now. Let me develop a study, collect a study, analyze results and publish it. And that can be, you know, that can be a year's turnaround. I think when I, I initially conceived the study was, would have been in 2020. So it's already been, you know, four years.
[00:49:36] Speaker A: Yeah.
[00:49:36] Speaker B: And we're still in the, you know, review and feedback phase of publication. So that the turnaround can be really slow.
[00:49:45] Speaker A: Yes.
[00:49:45] Speaker B: And, and so dissemination, you know, and all those aspects I think is really can be what holds things back. But it's, it is adapting and it is changing, I do believe. Yeah.
[00:49:56] Speaker A: Awesome. Well, let me start to wrap up with, with this question because I'm fascinated about your answer for this is if you were to look ahead 5 to 10 years being on really, you know, there's a handful of people kind of working on it. You're spreading the awareness of it. Where do you see fragmentation? Where do you think we'll be? Like if I have you back on the show in five years, what are we going to be talking about? Where, where we're. This is by the way for our listeners, all speculation. We hold nobody accountable for their answers on, on this question. But just kind of if you look out, what are some of maybe your hopes, your you know, where you get excited about what this, where this metric might go in the future.
[00:50:45] Speaker B: Yeah. I think optimistically I would really like to See the research into the biomechanisms of heart rate fragmentation and to see exactly are there specific portions of, you know, the autonomic nervous system or of the heart itself that we could use to help better understand what causes this output rather than just, you know, the broad stroke of well, it's dysregulating, it's dysfunctional, which, you know, that, that specificity is so lovely. Of course, though, we don't always get what we want when it comes to science. But I think that's one aspect I would really like. I think in five to 10 years from now, I think there will likely be a lot more papers that will be attaching heart rate fragmentation, I can see, to stress physiology and to psychology and health in general.
I think the part of me that is a little cautious is the concern of it running away in the same thing that, you know, in the same way HRV has in which, you know, we, we can see the immediate consumer aspect and use of it and we grab this measure and then we pull it out of context and next thing you know, we have, you know, a risk based device that claims, you know, that has, you know, the cardiovascular society backing it as like medically grade safe, which I have, you know, you know, a whole opinion about as well as, you know, this metric that is so individual and only and is great in very specific scenarios, but you're not going to be using it in any of these specific scenarios. So we're, that's, I think my like slight concern or like word of caution or something like that. My withered years of like probably 35 at the time. And so, and so that I think that I'm really interested to kind of see how this will change and add to what we know about HRV and how we look at health in general in a way that's usable for the average person in a way of that idea of health maintenance. I mean like cardiovascular diseases, I think the top, if not in the top five causes of mortality according to World Health Organization. So it's very much something that is needed and requires ongoing regulation, ongoing maintenance or ongoing, you know, vigilance that just cannot be supported by the current healthcare system. So anything that can allow us on, you know, on the other side of that of like, that's the great thing about home devices and wristwatches, you know, we, it allows an individual to at least have some form of agency over their health and some form of being able to really track and better understand themselves in that way.
[00:53:48] Speaker A: I love it. And yeah, somebody who like has dedicated themselves to staying in the very narrow part of. It would be so nice to make up like a readiness score or something that I didn't have to just justify how I made it up. And I think that there's been a. I was getting frustrated with that, with heart rate variability is. And, but the boring thing, then you have to give an RMSSD score, which, like, you know, maybe 0.01% of the population know what that means. So it's that tricky balance where we're at with this. But I, I know that what you're doing. I'm a big fan of. I think that this is so exciting that, that we, we are just, we're on the edge of something new in here and there's, there's this whole. I believe what my hope and aspiration is that it gives us that nice little chunk that, that, you know, if you take a reading a couple times a day or in the morning or maybe integrate into your biofeedback practice, it's. It's giving you a different piece of information that, that helps to fill out the hole. And that's where I've been a little bit frustrated with, like, just trying to figure out, okay, if we had an SDNN and an RMSSD score and they're. What do those two scores tell us if you put them together and you analyze them together maybe a little bit on parasympathetic. I don't. I, you know, I've studied it and I probably couldn't tell you. An doesn't really mean anything. And this is where I love bringing in your work. I think there might be some good meaning there that can really, really help people. So you've got a fan here. And please know that as your work continues in this area, you always have a place to come to because I'm very fascinated with what you're doing. So I just want to thank you and thank you for the work.
[00:55:44] Speaker B: Thank you so much.
[00:55:45] Speaker A: Awesome. And as always, you can find show notes. We'll put some information about Dr. Chang in the show notes of this article as well, and you can find that at optimalhrv. Com. So thank you everybody, and as always, I'll see you next week.
