Neuroscience and Yoga Online Conference Experiment Results
In this 2021 YouTube presentation, Jonathan Rosenthal walks the audience through the results of a month‑long online conference that paired live yoga classes with neuroscience experiments. He begins with acknowledgments and extensive caveats about the uncontrolled, low‑sample, real‑world nature of the data, emphasizing the need for cautious interpretation. Rosenthal then explains key physiological concepts—heart‑rate variability (HRV) as a proxy for autonomic balance, the peripheral and central nervous systems, and the sympathetic‑parasympathetic interplay—using simple neuroanatomy visuals. The talk is organized by the three thematic days of the conference: 1. Yoga & Cognition (Day 1) – Participants completed a Stroop test before and after a 1‑hour yoga session. HRV increased ~16% (non‑significant) and Stroop reaction times improved 10‑14% (significant for incongruent trials). Improvements were seen across beginner, intermediate, and advanced practitioners, though advanced yogis showed less HRV change, possibly due to ceiling effects. No correlation was found between HRV change and Stroop improvement. 2. Yoga & Mental Health (Day 2) – The State‑Trait Anxiety Inventory (STAI) and an Emotional Stroop (neutral vs. emotional words) were administered before and after yoga, with a “stress” block of scary music. Contrary to expectations, the music elicited a parasympathetic response, so HRV did not drop. Overall HRV rose modestly, and emotional Stroop reaction times improved ~10‑11% (significant). Both state and trait anxiety scores decreased markedly. Beginners/intermediates showed HRV gains during the stress block, while advanced yogis improved mainly after yoga. A positive correlation (higher HRV ↔ slower reaction) was observed, opposite the original model. 3. Yoga & Pain Processing (Day 3) – Heart‑rate variability and the Pain Catastrophizing Scale (rumination, magnification, helplessness) were measured pre‑ and post‑a 1.5‑hour yoga class. HRV rose ~17% (non‑significant). Magnification scores improved significantly; rumination trended toward improvement. Beginners/intermediates showed broader benefits, whereas advanced practitioners only improved on magnification. No HRV‑pain catastrophizing correlation was detected. Rosenthal highlights methodological constraints (small N, lack of standardization, variable home environments, limited validation of the Well2Re app) and proposes future steps: larger randomized samples, validated measurement tools, and more controlled protocols. The session ends with a Q&A covering topics such as the purpose of yoga research, measurement of sympathetic activity, and the influence of class enjoyment on physiological outcomes. He thanks collaborators (Ekaterina, Kristen, Deepika) and invites the community to stay connected. Overall, the presentation conveys tentative evidence that brief yoga sessions can modestly boost parasympathetic activity, improve cognitive control, reduce anxiety, and lessen pain catastrophizing, but the findings are limited by sample size, uncontrolled conditions, and mixed alignment with the hypothesized neuro‑physiological model.
Chapters
Okay. Hello everyone. It's so nice to see all of you. And I, I'm thrilled that we're able to go over the experiment. It took a long time. I see Ekaterina is here and she and Helen and I have been working on this a month. There were a lot of data generated by this conference. So I, I'm so glad you can all be here. I'm really excited to share these results. And our biggest goal is to share neuroscience and yoga with everyone. So I really hope you'll walk away with a better understanding of both the theory and the practical behind the neuroscience of yoga. And I also hope that you'll be inspired to support and participate in scientific research on yoga. I know that that was a big mission of the conference, and now that we're a month out, you have another opportunity, hopefully. And so obviously today is all about the experiment results from the conference, and there are some people here, I believe, and maybe some people who watch the recording of this, um, that were not able to attend the conference. And so just so everyone is on the same page, we, uh, had yoga classes during our conference, and before and after every yoga class, we had experiments. And, uh, the experiments were live hands-on neuroscience experiments that were related to each of the 3 themes of the days. So we have experiments relating to yoga and cognition, yoga and mental health, and yoga and pain processing. And we're going to be going through all of that. The flow of this— so this, you know, during the conference we took questions at the end. I'm actually going to take questions after each section. So we'll go through each day and questions as they come is perfect. So I just want to start with a huge acknowledgement to Helen and Ekaterina. They so expertly analyzed these data and they worked really hard. We had multiple late-night meetings outside of our regular day jobs working on this, so huge shout out to them. And a disclaimer about the interpretation of these data. So these studies were not conducted under controlled conditions. They were not standardized research protocol studies. Sample numbers were quite small. And so all of that is to say, proceed with caution. And what I mean by these experiments were not controlled is they were conducted in a very real-world setting across time zones. Maybe there were some glitchy internet. We weren't able to control people's home environments. Maybe someone's dog ran by while they were doing the experiments. So there were a lot of variables here. And as you should with all conclusions, you should see it with some skepticism. And these in particular just recognize the lack of control of the variables. Furthermore, in terms of sample sizes, so each of the days we had between 13 and 22 participants per experiment, which is a low number. So that also reduces the generalizability and strength and confidence of these conclusions. Another thing I think is really important to bring up is that Certain tests, particularly the cognitive tasks— tests, the— every participant did them before the yoga class and after the yoga class. And of course, when you're more familiar with a cognitive test, it's easier the second time around and you do better. And that is a very real confound. We tried to mitigate this, and what we did was we had practice rounds before people actually did the experiments. And the idea of that is, you know, you were already getting used to it, and so now that difference is lessened. But of course it still exists. And I think those are all of my disclaimers. So let's talk about some definitions. This is really where we're going to start going over the background of the experiments. One type of experiment occurred across every day. So we conducted heart rate variability experiments each day before and after yoga classes. And, um, what is heart rate variability? So this is very different than heart rate, which people often confuse it with. Heart rate is how many times does your heart beat in 1 minute, or Whereas heart rate variability is the physiologic variation in the time between 2 heartbeats. And this particularly occurs with the breath. So for example, if someone's heart were to have a beat, 800 milliseconds later have another beat, and then 800 milliseconds later have another beat, and then 800 milliseconds later another beat, that would be zero heart rate variability, right? Every beat's the same. There's no variability between them. So heart rate variability would be 800 milliseconds, 900 milliseconds, 1,000 milliseconds, 600 milliseconds, right? So that has variability. And this particularly occurs with the breath in a very notable way. During inhalation, the heart rate speeds up, and during exhalation, it slows down. And so the time between heartbeats when you're inhaling is going to get smaller and smaller and smaller. And then as you exhale, it gets bigger and bigger and bigger. And why is this so important? You know, a lot of people ask, especially in our past events, why would we have heart rate variability? Why, why do we want this? And you could speculate why. You know, when you're inhaling, maybe you want more blood flow to these oxygen-rich lungs. Maybe you Uh, you know, when you're inhaling, the pressure in the thoracic cavity goes down, so blood return increases. So maybe you need to pump more of that blood. Um, so, you know, we, we could have answers like that, but really the important thing is the inference we can make about heart rate variability, because heart rate variability is controlled by a special part of the nervous system called the autonomic nervous system. And it's particularly controlled by the balance between the sympathetic and the parasympathetic nervous system. So heart rate variability, you measure it, and now you can say something about the autonomic nervous system. That's what's so special about it. So to walk you through this, I want to start with some brief neuroanatomy. And take you through the difference between the peripheral nervous system and the central nervous system. The central nervous system includes the brain and the spinal cord, while the peripheral nervous system includes all of the nerves that come out of it. And the way you can think about it is the peripheral nervous system takes information in— sensory information, touch, sight, hearing— and it also puts information out. Motor responses, muscle contractions, heartbeat, and then the central nervous system is really in the middle of those two. So the central nervous system is taking the information in from the senses, integrating it, processing it, making decisions, and then putting that information out through those motor responses. So that is really. A 1,000-foot view of the nervous system and what it's really doing. And all of this is made up of these little cells called neurons. And I just have to say, you know, this is introductory, but these things are supercomputers. They're made like other cells. They have fatty membranes and proteins, but they take on these incredible intricate shapes such as the one you see here. And they do this in order to fulfill their function of taking in information, processing it, and having outputs. And what's so incredible about them is that they can customize their membranes with these very programmable receptors and channels. And all of this allows them to send incredibly fast electrical signals And to take in information, make decisions, and communicate those decisions with other parts of the nervous system. But let's just pause and all praise neurons. Okay, so I could go on forever. I could talk till I'm blue in the face about neuroanatomy, but I think it's much better if you learn from the experts. Who taught me. So I just want to share this video of some of the experts on neuroanatomy. And now the parts of the brain performed by the brain. Yes. Neocortex frontal lobe. Brainstem. Brainstem. Hippocampus neural node right hemisphere. Pons and cortex visual. Right stem, right stem. Sylvian fissure pineal, left hemisphere. Cerebellum left, cerebellum right. Synapse hypothalamus, striatum dentrite. Axon fibers move. Gray matter, gray. Brainstem, brainstem. Central tegmental pathway, temporal lobe. White core matter, forebrain skull. Brainstem, brainstem. Central fissure, cord spinal, parietal. Pia mater, meningeal vein, medulla oblongata and lobe limbic. Microelectrodes. Oh! The brain. That ought to keep the little squirts happy. Yeah. Okay, I, I couldn't help it. I'm sorry, I love that little film. I think it's hilarious. Um, so now we can really focus on one of the parts of the peripheral nervous system, which is the autonomic nervous system. So this is one of the ways the nervous system puts information out to the rest of the body. And the autonomic nervous system controls all of the processes that are subconscious. Um, some people argue they may be a little less, uh, you know, a little more than subconscious, so sometimes they can be conscious. Um, but these are things like how dilated are your pupils, um, how much saliva are you producing, how, uh, how open are your blood vessels to different parts of your body like muscles or the GI tract? How dilated are your lung airways? How fast is your heart beating? How quickly are you digesting food? Are you urinating or not? And of course, even sexual arousal, another autonomic function. So, um, this is just an overview, and you can see that on the left, the sympathetic nervous system has all of its nerves arriving— arising from the spinal cord. And these tend to be deploying resources for mobilization, for running, for fighting, for basically getting blood and energy to muscles in order to have a response to a threat. Whereas the parasympathetic nervous system, all of it is arising from the brainstem. I think you can see that, that little part. You can see my pointer, I believe, right? So all of this is the brainstem, and out of it is arising all of these parasympathetic fibers, which primarily come out through the vagus nerve right here. And you can see that those are then going to the same organs and having the opposite effect, and they tend to dampen down stress responses and improve digestion and Prepare the body for rest and recovery. One more thing to note is that they actually, because they go to the exact same cells, the sympathetic and the parasympathetic nervous system, they have different chemical signaling properties. Right? So if you went to the same cell and you signaled the same way, how would you tell the difference? So they use different chemicals to distinguish them, and the sympathetic. Fibers are in green here, which corresponds to NE, which is norepinephrine, and then in yellow, ACH, acetylcholine. So these are the main differences between the sympathetic and the parasympathetic nervous system. So let's zoom in on how the parasympathetic nervous system interact the whole autonomic nervous system interacts with the heart. So this is a beautiful representation of the heart. We're looking at the back of the heart, and sitting inside the back here, um, between the right atrium and the left atrium is the pacemaker of the heart. It's a group of cells that have a periodicity to them. They basically Every once in a while generate an electrical signal, electrical signal, electrical signal, and that then gets propagated throughout the whole heart in order to stimulate a contraction. And so how is that— when the snap goes off, how is that determined? Well, it's a balance of acetylcholine and norepinephrine. And so if you put more norepinephrine on it, does anyone want to put in the chat what they think would happen? Excellent. So it would fire more rapidly. You would basically be simulating a sympathetic response, and that would increase the heart rate. And there's actually great chemical examples of this that happen all the time. So The opposite of it really, which is a very common class of medications called beta blockers like propranolol. This medication is used for blood pressure, for migraines, lots of different things, and what it's doing is blocking the effects of norepinephrine. And so what would you expect that to do to the heart rate? I'll just answer that one. So it would fire more slowly, right? So if you give people beta blockers, their heart rate slows down. Now for the parasympathetic nervous system, there, uh, it's going to be using acetylcholine, and that's going to slow down the heart rate. And what's amazing about this is this really varies with the breath, and you may hear people talk about vagal tone. Which is how much the vagus nerve is firing. So it's not on or off. Is it firing at 20 times per second, 30 times per second, 40 times per second, 50 times per second, right? So the higher that number, the higher the tone. And that is what is varying. This vagal tone varies with the breath and really controls the heart rate. You know, in a variable way. So you'll see here, this is just a recording from me on my favorite app, Camera Heart Rate Variability. So as I'm breathing in, you can see heart rate goes up quite a bit, and then I exhale and the time between heartbeats lengthens a lot, heart rate goes down. Again, inhaling speeds up, and then exhaling it goes down. So very beautiful. Recording, I think. So how is all of this related to the brain, right? Because we're looking at the output of the brain, but can we make any inferences about what's going on actually inside the brain? And this is a model of how the brain controls heart rate variability. And I want to share an expression that I think is so powerful. Have you heard this? All models are wrong. But some are useful. All models are wrong, but some are useful. So this is a model. It's helpful for understanding things. And of course, there are parts missing from it, and there's a lot more to it. Systems are way more complicated than just— what is this, 8 boxes? But it is really helpful to see how these sort of higher-order parts of the brain That can be affected by yoga, then interact with some lower parts and lower parts still and lower parts still, and that ultimately can have effects on the outputs of the brain in the autonomic nervous system. So this is a really useful model, and the top of this model, the prefrontal cortex, we'll spend a good amount of time talking about. And the middle is really this network of the insula, the amygdala, and the cingulate cortex. And this network helps to process situations and stimuli to help determine how stressful they are and how to respond to them. So are you going to have fear? Are you going to freeze? Are you going to run away? Are you going to embrace them? Will you keep an open mind? And this prefrontal cortex actually can have a significant effect on how these networks behave, and it can change their their patterns of responses. So for example, one thing I noticed about this model just now that I didn't notice before is that one part of it that's wrong is there's a connection between the prefrontal cortex and the amygdala that's not shown here. I see Kristen is nodding at that one too. And as this kind of cycle of information processing goes on, it is then passed down through the hypothalamus And the brainstem, which ultimately has control over the sympathetic and parasympathetic outputs. And then we are just measuring the heart rate variability, which is the net output, right? And there's so much going on behind it. Okay, so that is the overview of heart rate variability. We'll go over heart rate variability data from every single day, and we are now about to start with day one. So yoga and cognition, day 1. So if you keep— if you— while we're going, if you put questions in the chat, I will catch up to them when we get to question parts. So what is cognition? I think we really need to start with that. And cognition really refers to all of the intellectual functions of the brain, everything between the sensory inputs and the motor outputs. It is such a huge box of nervous system functions. It includes memory, judgment, attention. Um, and it is a topic that's, um, you know, essentially a whole field of neuroscience with many journals, many papers. And we're really going to be focusing on one part of it, which is cognitive and cognitive control. Cognitive control is the ability to prioritize behaviors that meet chosen goals. And what this really kind of operatively means is that you are suppressing motor patterns that aren't achieving your goals, and you are enhancing motor patterns that achieve your goals. The cognitive control aspect that we're going to focus on is through a very famous test called the Stroop test. And this is one of the most validated, well-studied tests in all of neuroscience. And it is essentially a test that measures cognitive control by presenting stimuli and asking participants to respond to a particular feature of them. And you have to suppress other features, so they, they're often conflicting, right? Right? And the way that this is done is the word on the screen will spell out red, but the text, the color of the text, will be blue. And you need to respond with blue. And in order to do that, you have to suppress that the text says red. And this is a form of cognitive control. You're literally suppressing your ability to read and focusing just on the color information. And this is an incongruent stimulus, right, where they disagree with each other. So let's go through just a very intuitive example of what cognitive control looks like. Okay, so here's the deal. There's a marshmallow. You can either wait and I'll bring you back another one so you can have 2, or you can eat it now. So you can eat it now or you can wait and I'll bring you back 2. Okay? Okay. Okay, I'll be back. Okay, so I have 1 marshmallow for each of you. Okay, this is 1. And here's the deal. You can either eat it now or you can wait till I get back and you can have 2. Okay? Okay. So eat it now or wait till I get back and you can have 2. And I'll be back in a little bit. If we wait, we— you'll give us 2? Yep, if you Wait a bit too, or you can eat it now, whichever you want. Okay, I'm gonna wait. I'm gonna wait. I wonder what we're gonna. 2. No, you're gonna eat it. I just want to eat one bite. Well, we're still not gonna give 2. But if you wait until she gets back, she'll give you 2. She still won't give you 2 because you ate some of it. But I didn't eat a single bite more. The goat shall work it. This is a long video, but I think you get the point, right? You have to suppress the stimulus of this marshmallow in front of you because your goal is to get to— So in the Stroop test, you're given a stimulus. Here it is congruent. The text says green, the color is green. So I'm going to ask you to respond with the color of the text. You just have to say it out loud. Don't worry, you're on mute, so if you get it wrong, it's okay. But I'm going to take you through this quite fast. So together, green. It's quite hard, right, in rapid time to say just the color when you're competing with what the word says. It's very challenging. And so how does a brain achieve this? So again, we must go back to a model, and this time we're going to focus on the prefrontal cortex. And there are different components of the prefrontal cortex. We're going to focus on this so-called dorsolateral prefrontal cortex, which is very important for the Stroop test in particular. And what the dorsolateral prefrontal cortex does is it integrates with our motor program part of the brain and allows us to flexibly allow motor responses, motor programs, to link to perceptual inputs, right? So the color of the text links to a certain motor program. And that is incredible, right? That you can so flexibly reprogram the brain. And I know neuroplasticity is a huge buzzword. It garners so much attention. I actually often get frustrated. I find that the word is used as if it's this rare thing that never happens, and it is so novel. And I think that almost everything we're doing is neuroplasticity, right? So I'm in awe at this reprogramming. These, these neurons are millisecond by millisecond changing their channel response properties in order to allow us to change our way of interacting with the world. So these are very Short-term changes, right? So you're you're changing how you respond to this motor program, maybe just a few minutes while you're doing the stroop test. And so that is a very quick change that can take place. And then of course, you know there are more permanent changes. So if people do the stroop test a lot, they can literally change the structure of neurons that are doing this in order to get really fast at the stroop test. So let's go over our experimental design for the conference and the hypotheses. So over time, there were 4 measurements taken around the yoga class. The first was heart rate variability. We're going to call this HRV1. The second was a Stroop test, HRV1— sorry, Stroop 1. And then the yoga class. This was the 1-hour yoga class on Friday. And then HRV2 and Stroop test 2. And just based on this and based on the model and the theory we've talked about, we can generate some simple hypotheses that are really designed to test the model, right? If the model is right, these hypotheses should work. So first, HRV2 will be higher than HRV1 because yoga increases the activity of the parasympathetic nervous system. That's our first hypothesis. The second hypothesis is that the reaction time on Stroop Test 2 will be lower than on Stroop Test 1, and that will be because yoga improves cognitive control. Remember that all of these were in the same kind of neural network, the same kind of model of different parts of the brain interacting with each other. And then the third thing based on that is that the changes in Stroop will correlate with changes in heart rate variability. So some pretty interesting hypotheses, I think. So how did the data come? So this is for the first hypothesis. This is heart rate variability before and after 1 hour of yoga. On the y-axis here, we have a measure of heart rate variability. It's called the root mean squared of successive differences. And On the x-axis, we can see before yoga and after yoga. And what we notice is a 16% increase after 60 minutes of yoga. And, you know, I know we all obsess over statistical significance, but I think that this is still a meaningful finding. So there is a 10% chance that this was due to— sorry, there's a 10% chance that this was due to chance. There's no other way to say it. Um, typically we say if something is less than 5% that it was due to chance, then we take that as statistically significant. But this change we're going to see again and again and again, and if we pool all our samples, I'll show you what we find. So this is, um, in support of hypothesis 1. Now looking at Stroop test performance. So I break this down by the congruent condition, so that's where the word Spells the color of its text, so says red in the color red, and then the incongruent condition where they're different. And this is reaction time on the y-axis and before and after yoga. So we can see that before to after yoga in both conditions there's a decrease, but the most dramatic decrease occurred in the incongruent To test. And that's exactly what you would expect if yoga is going to affect cognitive control, because you need more cognitive control in the incongruent condition than you do in the congruent condition. So this was a 10% increase, and this was a 14% increase. And I will say, I actually went back to our data from a couple of years ago where we used the same experiment. We did it in person that time, but we found the same magnitude, about a 10% improvement in reaction time after yoga. And this one was statistically significant, so I will use asterisks whenever there is a statistically significant finding. Okay, our 3rd hypothesis, that there will be a correlation. So I will say previous studies have shown this correlation. We did not in our sample, and you know, it's hard to interpret why that is, but we, we really don't see any correlation. So this R-squared value is what percent of the reaction time change is explained by the heart rate variability change. And it's essentially less than 1% here and about 1% here. So there's really no correlation. So this raises a few questions. You know, first, is it that our model is wrong or is it that there is something wrong with our data? And I think we would really need to see repeats of this and kind of more controlled data to really say whether there was a problem with the model. Especially because previous studies have shown this correlation. So now let's go into a little more advanced data analysis. Now we're going to look by self-reported yoga level. So there was only one beginner in the entire conference, which I was shocked by. Everyone else identified as intermediate or advanced. And so we're going to look at beginner and intermediate versus advanced yoga practitioners. So this is the reaction time in this group test, and we can see that across all conditions there is improvement: blue to blue, green to green, blue to blue, green to green. Right? So whether you were a beginner, intermediate, or advanced, on average it seems that everyone saw an improvement in their reaction time on this group test. And what I find interesting is so I kind you know this hypothesis wasn't one of our original ones, but. you could hypothesize that advanced practitioners might have a higher performance or a lower reaction time at baseline because they do more yoga, and so they should have more cognitive control. This was not what we found in our data, but I think we talked about during the conference what a confound is, which is when outside data or outside variables are affecting your data. And it's important to keep track of those and look for them. And one of them that I was able to find in the demographic data that we collected was that beginner and intermediate yogis had more advanced educational backgrounds. So they were more likely to have advanced degrees, which is known to improve Stroop test reaction time. So that may be why we don't see this in our data. In other words, the advanced yogis and the beginner and intermediate yogis were not the same. Right? They came from different populations. And now for heart rate variability by level of yoga practice. This one is quite interesting. The improvement in heart rate variability was primarily seen by the beginners and the intermediate practitioners. Advanced practitioners did not have much of an improvement. So few possible explanations for this. One is that perhaps advanced practitioners really live at a higher heart rate variability, and it takes, you know, there's a ceiling effect. Heart rate variability can only go so high, and so perhaps it was just harder to get theirs up. Another possibility is that this was one of the shorter yoga classes, so maybe it just takes more time doing yoga for advanced practitioners to see a benefit. Either way, very interesting result. So just to summarize the hypotheses and our conclusions from them. So our first hypothesis, heart rate variability 2 will be higher than heart rate variability 1. We did see that. And while this did not meet statistical significance, we can, you know, say that this change was still meaningful. And we will show that later as well. And really beginners and intermediate practitioners had the most effect from this 1-hour yoga class. Our second hypothesis was that the reaction time on Stroop Test 2 would be lower than on Stroop Test 1, and we saw this for sure very strongly, and this was across all levels— beginners, intermediate, and advanced. And then the last was based on our big picture model of the whole system of the brain connecting to the parasympathetic nervous system. Well, we did not see this correlation. And again, in order to definitively say that this undermines our model, we would really need to standardize and improve our experimental design. Okay, what questions do people have? Yes, so I see Sarah asked a question about the shape of neurons, and Kristen answered it. They come in— oh, maybe we'll— Kristen, behind you. Hi. Yeah, so if, if people want to look at Kristen's background, she has a beautiful background of some of the different forms of neurons. I see some Purkinje cells there, some basket cells, pyramidal cells. Beautiful. Yeah, so why are we doing these experiments? Ooh. Interesting. Are you asking because you— maybe you can elaborate why you're asking why we're doing these experiments. We do the experiments to help understand science and to improve our understanding and to have fun with science. But maybe I'm missing the question. And then Mark is asking about the third category. So absolutely. So, Mark, going back to our model, we have that the higher parts of the brain, the prefrontal cortex, which really controls the effects of cognitive control, and that is really linked through all of these other networks through to the heart rate variability, the parts of the brain that control heart rate variability. So we might hypothesize that if prefrontal cortex activity improves and cognitive control improves, then heart rate variability should improve too. So they should be correlated, in other words. And so we essentially used our data looking at each individual. And so each point on this graph is an individual, and we're looking at their change in heart rate variability. So did their heart rate variability decrease or or increase, and their change in reaction time. So if it decreased or increased, and if they were correlated, meaning as your heart rate variability went up, your reaction time went down, then you would see a line like this. But that is not what we see. We see kind of more of a flat line. And same thing here, it would look like this. But we see a flat line. And Arshika, this is a great question. So the benefits of yoga are already known and proven. And I, I love this question because it really shows the different levels of evidence. So in order for something to be known and proven, you really want to attain a certain level of confidence in scientific terms. And science really is all about how do you prove things, how do you know things. And the lowest level of that is anecdotal, meaning that people just experience it and talk about it. And that's really kind of the level that a lot of yoga knowledge is at right now, which is great. I do think yoga has a ton of benefits, and obviously I'm a yoga practitioner myself. But I think in order to really support yoga and to know more about it, we would want to have it at a much higher level, meaning like controlling the variables and knowing that it is the yoga that is having the effect. And Eddie, yes, these are from your class, so excellent results. Is it possible that the meditative aspects of yoga With lower Stroop test scores because of lower alertness? That's a really, really good question, Jessica. I'm just going to draw a little— I have some math on the board, I'm going to erase it, but okay. That's a really good question because there's a famous, famous graph called the Yerkes-Dodson graph, which is performance versus stress level or arousal level. And, you know, some people might think that this looks like this: you get more stressed, your performance goes down, right? But that's actually not what this graph looks like. This graph looks more like this. And what this really means is that you need a certain amount of stress in order to have optimal performance. But if there's too much stress, performance drops, and if there's too little stress, performance drops. And so yes, it is in theory possible that some of the calming effects could decrease performance on certain cognitive tasks. Could you please explain congruent and incongruent again? Absolutely. So this is congruent. The text says green, the color is green. And the— this one would be incongruent. The text says blue, but the color is yellow. So they don't agree with each other, right? So it's harder to answer this one. It takes longer. So Eddie, I don't think that disappointment is in order. I think actually it's very good, especially because the sample size of this was so low. So if we look at the graph, these confidence intervals are so wide, and that's really the reason that's statistical significance was not attained. And I, you know, I think we all obsess over statistical significance, but there's, you know, something could be statistically significant and not meaningful. And I think this is meaningful but not statistically significant. And we're going to pool all of the data to really reach more of the numbers needed. So instead of 15 people in one sample, we'll be able to get to 50-something. And there you'll see More, you know, a statistically significant result across all the yoga classes. So now we move on to day 2. So day 2 is yoga and mental health. So mental health is a huge topic. In fact, this was one of the first things I really noticed yoga had effect— an effect on me when I was practicing. really mood improvement and, you know, stabilizing anxiety and depressive symptoms. So we're going to talk about 2 different measurements. One is the State-Trait Anxiety Inventory, which is a standardized survey used to reliably measure anxiety. And this measures it in 2 different ways. One, as a general character baseline, this is called trait. So how anxious are you just as a person? And then at a given moment, and this is called state, and that's why it's the State-Trait Anxiety Inventory. This one is a subjective measurement of anxiety. The second one is the Emotional Stroop Test. I just realized I spelled Stroop wrong on here. I apologize. I will fix it later. But this is similar to the regular Stroop Test, which is the cognitive one we talked about previously. But this one adds a little twist to it, and what it does is it— the words, instead of being spelling out colors, they spell out emotional or non-emotional words. So emotional or neutral words. So an emotional word would be like hate, shame. And a neutral word would be like Gate or same. You can see how these are very similar, but they totally change, you know, you have a totally different response to them. And in fact, when people are presented with emotional words, they can take a little bit longer to have their reaction time. And this is particularly the case in individuals who have a mood disturbance, so who maybe are experiencing depressive or anxiety symptoms. And it takes cognitive control in order to suppress those. And then it also is, you know, just having a good mood in general suppresses that response. So those are really— instead of competing 2 perceptual signals, we're competing with our internal state and, and whether the word is emotional or not. Does that make sense? Okay. So this is just an example of what the, the test might look like. So for a neutral block, you're faced with words like table, path, chair, street, and you're just responding with the color. So here it would be green, blue, purple, red. Whereas here you're faced with words that are very emotionally charged— death, kill, shame, murder— and you still respond with the color of the words. But for people who have more of a reaction to this, that interferes with their ability to respond quickly, right? So there's an emotional interference. And again, the dorsolateral prefrontal cortex is going to be able to flexibly switch how we respond to these perceptual inputs and to dampen that emotional interference. So this was our most complicated experimental design of the conference. So this one had heart rate variability 1, an emotional stroop test, and then repeating heart rate variability. So this one is now heart rate variability 2, and it's before yoga. And what we also did is we had participants listening to scary music during this measurement. And the idea was between the emotional Stroop test and the scary music, these will serve as stressors. And so what do you think we expected? Well, actually, it's one of the hypotheses. Never mind, I'll say it in a second. So we, we had this and then we repeated the paradigm after yoga. So again, heart rate variability, an emotional Stroop test, and then heart rate variability with stressors, scary music, and then after the emotional Stroop. So our hypotheses here: heart rate variability 2 will be lower than heart rate variability 1 because the emotional stroop and the scary music will act as stressors, meaning they would deploy resources through the sympathetic nervous system. Hypothesis 2: heart rate variability 3 will be higher than heart rate variability 2 Because yoga will restore parasympathetic nervous system activity. Hypothesis 3, heart rate variability 4 will not decrease as much as heart rate variability 2 did, and this will be because yoga has been shown to be protective against future stressors. And then finally, hypothesis 4, emotional Stroop test 2 reaction time will be faster than emotional Stroop Test 1. And this is because yoga will reduce emotional— will reduce emotional interference and improve cognitive control. Okay. So, absolutely bizarre results. We expected our stressors to reduce heart rate variability. Apparently, our sample really liked the scary music. I think some of you told us that you enjoyed the scary music, and it shows. So people did not have a sympathetic response to this. They actually had a parasympathetic response to it. And, and we see that just over the entire course of the experiment, people just had increases in heart rate variability, very little of it attributable to yoga, which would have occurred between these 2 right here. Um, okay, so I just want to show you data from a couple of years ago. This is when we used a different stressor. We used a cognitive test called the n-back. And this one, you can see that heart rate variability won. Stressor with the n-back, it drops. Yoga, it gets restored. And then there's less of a drop. All right, so that's kind of what we were expecting, but you know, using an n-back as a stressor is very different than using scary music. So I'll play the scary music for you now. You can tell me if you like it. I can't listen to it that long. It makes me feel so weird. I don't know if people Really liked it. I don't know. Okay, so now we can look at the emotional Stroop test reaction time performance. So similar to on the regular Stroop test, we see a 10 to 11% improvement. So on the left is when it was emotional words like hate and shame, and on the right is when it was neutral words like. Gait and same. Both showed an improvement, but really even more of an improvement was observed in the emotional condition. And if we look at the state-trait anxiety scores, what we see is that the— so these are going to be the trait, so like the intrinsic character traits of a person, and these are the state. So it's very surprising that yoga improved the trait qualities, right? So after, you know, 2.5 hours of yoga, people were feeling like they as a person were just less anxious in general, in addition to having their current state being less anxious. Very interesting finding. And then in terms of the statistical significance of this, it was very, very Statistically significant. Essentially everyone had an improvement. Okay. So now if we look at this by level, it gets even weirder. So blue is beginner and intermediate and green is advanced. So what we see is that the beginner and intermediate participants had primarily their improvement in heart rate variability with the stressors. While the advanced people had their improvement in heart rate variability primarily with the yoga class. I don't— I, in terms of explaining this, I would rather chalk this up to our data being noisy. I think this is a very bizarre finding. I really welcome other people if they have any comments about this and any interesting thoughts. Because I can't explain this. Interesting. So, the scary music felt like it was— people felt like it was intriguing or like a sound bath. Interesting. Wow, yeah, everyone liked the music. Okay. So this is emotional Stroop test performance reaction time by level. So in blue is beginners and intermediate, and in green is advanced. And interestingly, unlike the cognitive Stroop, we see that advanced participants were actually advanced yoga practitioners were actually lower at baseline, right? So their emotional interference was more suppressed at baseline, and this did not seem to depend on educational level. And again, all groups saw improvement. So beginner, improvement in reaction time. Advanced, improvement in reaction time, whether it was the emotional or the neutral condition. Okay, and now state trait anxiety. This one I find very funny. So at baseline, advanced yoga practitioners were more anxious. I'm sure this has to do with, you know, running around teaching yoga classes. It's very— I think it's very stressful. I remember during summers when I used to teach a lot of yoga classes, I was like, how do people do this? Eddie, I'm sorry that I'm running late. If anyone has to go, the rest of this will be recorded. And so, yes, but again, across all groups, an improvement. In terms of correlations, so this one is interesting. So if you remember, we would expect a correlation that looks like this, meaning as heart rate variability increases, reaction time should decrease, right, based on our model. And here what we see is actually a positive correlation, meaning— and this was significant, this was a very significant correlation. It explains about 20% of the data variability. And if you just look at it, it really has the shape of a line. Of course, there's some, you know, noise around it, but overall it is significant. And the— what this means is, as heart rate variability went up, reaction time went up, which is bizarre according to our model. Now, one way that you might explain this is actually there's quite a bit of research that shows that yoga practices enhance empathy and emotional responses. So you could imagine that actually the people who at baseline had more heart rate variability were maybe more sensitive to these emotional responses, but that wasn't consistent with the rest of our data. So this is also challenging to interpret and put together. And just to go over our hypotheses and their conclusions. So our most complicated day had the most challenging results. So heart rate variability 2 will be lower than heart rate variability 1 because the emotional stroop and scary music will act as stressors. Well, everyone loved the scary music. So nope. Hypothesis 2, yoga would improve heart rate variability. So heart rate variability 3 will be higher than heart rate variability 2. Well, it was slightly, and that was mainly driven by the advanced yogis having improvement with yoga. But overall, this was not the case, probably because again of the ceiling effect, right? The music made people have such a relaxed response that you couldn't go much higher. And third hypothesis, heart rate variability 4 will not decrease as much as heart rate variability 2 did because Because yoga is protective against stressors. Well, we didn't have stressors, so can't really support this hypothesis. And then the 4th hypothesis was that emotional Stroop test reaction time will be faster than emotional Stroop test 1 because we'll reduce emotional interference and improve cognitive control. And we saw this at all levels of yoga practice. Okay, questions. I see some in the chat. When you say suppressed, do you mean emotions weren't disturbed, or— that's a great question. So emotional interference refers to when you have an emotional response and it interferes with cognitive processes like choosing the right color response. And so suppressing emotional interference means that those emotional responses either don't happen as strongly or they don't have as much of an effect on cognitive tasks. All right, I'm going to speed up a little bit so that we can finish in a timely manner. So day 3 was all about yoga and pain processing, and we really looked at pain catastrophizing, which is a measure of the mental experience of pain, and it looks at domains like rumination, magnification, and perceived helplessness. We measure this in a scale called the Pain Catastrophizing Scale, and it's a standardized set of questions that really are a subjective measurement of pain catastrophizing. And experimental design. So we had heart rate variability, Pain Catastrophizing Scale, a yoga class, and then we repeated those 2. And the hypotheses were very straightforward. Heart rate variability 2 should be higher than heart rate variability 1, again because yoga increases parasympathetic nervous system activity. Pain catastrophizing scale 2 will be lower than the first because yoga will improve pain catastrophizing by improving executive function. And finally, that the changes in the pain catastrophizing scale will correlate with heart rate variability. These were our heart rate variability data. Here we see a 17% improvement after 1.5 hours of yoga. Now, this was still meaningful, but it did not meet statistical significance. Again, the sample size was, I think, in this one around 12 or 13. If we look at pain catastrophizing, we're going to break it down into the different domains. So this is rumination, magnification, helplessness, and overall. which is just an addition, adding up all these scores. And in blue is before yoga, in green is after yoga. And you can see after 1.5 hours of yoga, our sample showed significant improvements in magnification. There was also an improvement in rumination, but this was not— this did not reach statistical significance. And what does this really mean? So rumination is The process of going over and over thoughts about the pain. Could this be cancer? Could this be, you know, what is this? What am I feeling? You know, just all the thoughts that we generate about pain. Whereas magnification is like, oh my God, only this pain exists right now. This is the only thing I can pay attention to. And then helplessness is I can't do anything about this pain. There's nothing to be done. I'm stuck with this. And so those are all the components of pain catastrophizing. And we can see really magnification was helped the most. Here, if we look at heart rate variability by level, what we really notice is that beginners and intermediate have the improvement while advanced people do not. And if you remember, we saw this from the data from day 1 as well. Right, so it seems that if I were to try to put this together, the shorter yoga classes had most of the improvement for beginners and intermediate practitioners. While the longer class on Saturday with the long yoga nidra had most of the improvement in heart rate variability for the advanced people. That would be one way of trying to make it make sense. If we look at pain catastrophizing by level— so I know this is a big graph, there's a lot of data here. So now in blue are the beginners and intermediate, and in green are the advanced. And we can see rumination before yoga, After yoga, before yoga, after yoga. These are magnification. Helplessness before yoga, after yoga, and overall. And really what we see is that beginner and intermediate practitioners saw improvements across the board, while advanced practitioners really saw improvements only in magnification. And this was why magnification was the only one that achieved statistical significance, probably, but There may be a more pronounced benefit for people who are new to yoga in pain catastrophizing. And there was no correlation with heart rate variability and pain catastrophizing scale. And I'm just going to summarize, you know, why I think our heart rate variability and correlation data might have been so weird. You know, this was our first time using the Well2Re app. There are not great validation studies for it, and we do need to, in order to improve our data collection process and experimental design validate our measurement tools. So that is definitely something we have to work on in the future. Conclusions from day 3. So was heart rate variability 2 higher than 1? Absolutely. Again, it did not meet statistical significance, possibly related to the small sample size. I can say definitely related because when we combine them, the effect is there. And beginners and intermediates saw the most effect. Pain catastrophizing also improved, and we did not see the correlation. Okay, so I see there are some questions. Is this linked with enjoyment of the class? Is the general consensus that yoga classes lower stress in the moment as opposed to some perceptions that yoga increases awareness and can after time and practice increase stress threshold? Really deep question, Jess. So enjoyment of the class probably does have an effect. If people are there, they disagree with the teacher about things and are feeling a lot of resistance and are actually having a stress response to it, they will most certainly not have a parasympathetic nervous system response. And the question about increasing awareness, increasing the stress threshold, so actually paradoxically, No. And this is a really important point in the yoga science, you know, the research on yoga, which is that even though attention improves and the sensitivity to different signals increases, they actually have less of an effect on driving emotion and on driving our stress response. So it's a very interesting paradoxical finding. What was the app that I prefer? It's called Camera Heart Rate Variability. It is by Marco Altini. I'll put it in the chat. It does cost a little bit of money, but I think it's very good and it's validated. There he has published a good number of studies on it. All right, some bonus data. So this is the pooled heart rate variability data across yoga classes and You can see very clearly over all of the yoga classes from the conference, there is a significant, statistically significant 24% increase in heart rate variability. And this was a higher sample size, of course, because all of the data were pooled. And if we leave out day 2, you know, day 2 was a little wonky with our, you know, not stressful stressors. What we see is that the effect remains. So still statistically significant. There was an 18% increase in heart rate variability after a single yoga class, and these were the shorter yoga classes. Okay. So I think that is all I have. These are just some references. Oh, I just closed it. I'm sorry. I clicked too many times. But I'm sorry I ran a little bit late. I couldn't help myself. I am blown away by these results. I think they're absolutely fascinating. You know, I think we can obviously say that there were significant changes observed, and they really support this integrated model of, you know, higher parts of the brain and lower parts of the brain all interacting from the effects of yoga. So with that, I'm going to Open up the microphone. Anyone is welcome to comment. I can open up the slides, talk more, and I really look forward to hearing from all of you, all of your thoughts. Thank you for sharing. Thank you. Yeah, but I wanted to know the intention. For example, We always carry out research, of course, with certain hypotheses, but to prove something besides the fact that there is enough anecdotal evidence and you're looking for more statistical evidence, is that the main primary reason behind this or what else? Yeah. So I think, you know, yoga research is very important because in order for us to really recommend something, anecdotal evidence is not enough. You really need to know what is the effect size, who has that effect and who doesn't, and what are the risks and what are the limitations of any intervention. Yoga is obviously, we all believe very strongly in yoga, but we do need to understand it at a very deep level in order to recommend it, especially to people who might be patients and have diseases and really need the benefits that we suggest it has. So we need to prove it to them. We owe it to them to prove it to them, I think. And I think that really is what the mission of yoga research is. Now, for our experiments, they're really for fun. We do this to help, you know, our organization is a neuroscience outreach organization, and we're really trying to help people understand neuroscience. And we really want the neuroscience world and the yoga world to talk more. So that these benefits can really be discussed and studied and shared with everyone. You mean recommending yoga as a precautionary or a pre— I don't know what the terminology is, but as a treatment? Correct. Got it, got it. But I think what it discounts is in your research, the background, everybody's background, because everybody's background kind of has more contribution to their results than just the environment. Yeah, that's a really good point. And actually, that's one of the hardest things about conducting research is that there's in— there's individual differences, right? And there's actually a method in science to, um, kind of eliminate those variables, and it's called randomization. Essentially what you do is you get a very large sample of people and you randomize them to different conditions. And, uh, the idea is that because they're being sorted randomly, they will actually cancel out all of those individual differences. Then you can really just see what is the effect of yoga on the population. Yeah, but the sample size has to be really, really large. Yes, and that is for sure. Okay, cool. Thank you for your question. Jonathan, I have a question. Yes, Mark. With heart rate variability, you were talking about the parasympathetic nervous system. becoming more effective. Did you— were you able to measure— I mean, sample size too small, but measure if there was any change in the sympathetic? That is a fantastic question. Um, so the answer is, and this is really important, so it doesn't have to be that they're like kind of this seesaw, right? It doesn't have to be that one goes up the exact same amount as the other goes down. And With heart rate variability, they're, they're varying with the breath, as we talked about, the parasympathetic nervous system activity. So there is a mathematical way of analyzing heart rate variability to extract both the inputs from the sympathetic nervous system and the parasympathetic nervous system. And this gets very advanced. It's very fancy math that uses these different— one of them is called time domain analysis versus frequency domain analysis. And there are still ongoing validation studies for that. So it's, it's supported but not truly confirmed that we can extract sympathetic nervous system data just from heart rate variability. But it is very widely supported that we can extract parasympathetic nervous system data from it. We did not collect those data, and the people who do it find a decrease in sympathetic nervous system activity. Okay, thanks. Of course. Were the duration of classes long in the past experiments conducted for advanced yoga practitioners? Shilpa, can you— sorry. For past experiments, do you mean the ones from previous events? Yes. Okay. Yes. So you mentioned that, you know, it took a little more, more like a ceiling effect, right? So that's what I wanted to understand. That's a great question. You know, the answer I think is no. I have to look back at our charts, which I can pull up really quickly right now, but I do not remember seeing that effect ever before, which is why I'm so baffled by it. And so I'm going to pull up 2019's data. And, you know, we didn't, we didn't break it down by that. So I will have to do an analysis on our old data and get back to you. Sure. Thank you. Of course. Right, I'm going to hop off. I'm sorry I ran late, everyone. I can't tell you how good it is. I feel like I know you guys. So your faces I've seen so many times throughout the conference, and it's just really, really nice to feel like we have a little neuroscience and yoga online community. Everyone, I wish you all a great night. Please keep in touch. I'm going to send out an email hopefully soon. If I don't do it tomorrow, it might take me a while. I'm going to be on nights for 2 weeks, so I'm just going to be a little bit focused on work for a while, but I will definitely send it. And if I don't, you always can remind me and I will follow up. And I have to thank a few of the people here. So Ekaterina is here who helped analyze all the data. Kristen is here, who we're talking about ways to kind of do online yoga research in a more standardized way, and we're working on that right now. And Deepika is here, who helped to design the experiments and make them work on the backend. All right, everyone. Thank you so much, everyone. Bye. Bye-bye. Bye-bye.
