One important dimension of your measuring that is important to understand is how the length of time that you measure. Heart Rate Variability (HRV) is the variability of your heartbeats over a period of time. The standard readings (rMSSD, pNN50, etc) are based on some reading over a specific period.
That means if you are studying the effect of eating donuts on your HRV you need to ensure that reading 1 and reading 2 have approximately the same length (5 secs, 30 secs, 2 mins, etc). I prefer 30 seconds snapshots. Readings of different lengths are apples and oranges because you are getting an average over that time period.
If you eat a donut and read your HRV for 5 minutes the reading will an average for the 5. Eat the second donut a day later and read for 35 minutes you will get the average for 35 minutes. If your blood sugar changes more during the 35 minutes simply because your body has more time to digest it will look like a lower number on day 2 and you may draw wrong conclusions based on an error of measurement.
If all other variables are kept constant and you vary the measurement times you will get incompatible readings. Make sure you keep an eye on your timer and know that the similar timing of your readings are your most important source of useable measurements.
After you have your basic kit you can start taking readings and compare numbers. To save time and avoided taking readings that don’t compare well to other readings it is important to do a good job defining what you are looking to study. This helps keep what you are reading consistent and easier to implement.
Primarily you will be looking at the difference in your Heart Rate Variability (HRV) as you vary conditions in your environment. For example, if you decide to look at the effect of sleep on your HRV you will have to pick a time, conditions and length of time that you will measure your HRV.
For example, if you read your HRV each morning as part of your sleep study, you have to understand that reading HRV immediately after waking is different than 30 minutes after you wake up. Your SNS and PNS are in a different state during those two periods. Additionally, if you walk upstairs to retrieve you heart rate belt on one morning and you didn’t the day before, the two readings will not be comparable.
So it is best to define precisely what you are going to do when you take your readings. In your sleep study example write out “Two minutes after I wake up I take my reading with my heart rate belt reading to the app for 5 minutes as I lay on my bed.” Each morning that you meet those conditions include that reading in those you compare. If one condition is violated, unfortunately, you will have set those readings aside.
Find out more by reading about Measurement Times.
Because of the limitations of pulse oximetry I recommend you start with a Polar H7 heart rate belt. It is affordable (about $79) and available on Amazon or a local sporting goods store. If you don’t choose an H7, ensure your heart rate belt is Bluetooth Low Energy (BLE) so it syncs with the updated iPhone or Android device. Garmin uses a different standard (Ant+) and won’t work with what you want to do.
There are many, many apps that will take the readings from your Polar H7 and convert the data. The three best are iThlete, SweetwaterHRV and Marco Altini’s Heart Rate Variability Logger. iThlete is limited in that it is for a daily, short reading to determine your state of physical training. SweetwaterHRV is a good app made by very nice people and it is a bit difficult to extract data (you download and email it). It has a lot of interpretations of the data in the app.
Marco Altini’s app gives nice readings in real time and has a dead simple Dropbox or iCloud download feature. You take a reading, press a button on your device and the .csv file is in your file folder. That simple. Once you have a .csv file you can import it to excel and play with the numbers. You can also see the derivations and start looking there.
Find out more by reading about Setting Up What You Study.
You will have numerous sensor available to you and they break two types. One type uses pulse oximetry, which shoots light through your skin to sense the heartbeat by the change in color of your skin based on your blood flow. When you use a watch (Apple Watch, Mio, Basis, etc), an ear clip (Heartmath), or a finger clip (as in those common in hospitals), they all use pulse oximetry.
This type of sensor is good for measuring heart rate, but NOT as good for Heart Rate Variability (HRV). This is because the way it measures heart beats will have a lot of errors when you move your body. There are medical studies on these error rates. Nothing wrong with your Apple Watch telling you your heart rate on a run or during the day. But if you are trying to get precise HRV information from that same watch while you are moving around you will have a problem. So you can use pulse oximetry if you plan to be motionless for your studies, but if you want to look at anything where you move your body you will need to use a different method.
The way to accurately measure your individual heart beats and then your HRV is based on reading the electrical impulses that make your heart beat. These sensors are heart rate belts or more expensive adhesive patches. These give you the required accuracy because they are reading the electrical output from your heart when it beats. I have tested pulse oximetry (Heartmath, Mio wrist band) vs Polar heart rate belt and the differences in accuracy when in motion are noticeable.
Find out more by reading about the Basic Equipment Setup.
You capture the Heart Rate Variability (HRV) measurement as expressed in RR intervals by using a sensor and having it read its output to a software application with which you can see a display of the data or download it for viewing and manipulation.
To get your HRV readings from your body you can use a heart rate belt or other type of sensor and read the output on an app. To actually play with the data you will have to find an app that downloads the data in a format such as .csv or .txt that can be ported to a program like excel or R.
To learn how HRV works, I highly recommend you start by looking at the RR intervals so you can get familiar with the source data. Using apps that have already derived some measure is easier once you understand the underlying mechanics of the RR intervals. I wasted a lot of time trying to understand the apps before the underlying mechanics. You can avoid that.
Find out more by reading about Sensors.
Heart Rate Variability (HRV) is measured by looking at the time interval between successive heart beats over a specific period. That means that if you examine ten heart beats, there will be a difference in intervals between each successive beat and those differences are averaged into some number that indicates there was either a lot of variability between all of the beats or very little variability during the period the ten beats occurred.
The way HRV is captured is by a string of numbers called “RR Intervals.” The RR Intervals are measured in milliseconds. For perspective, a resting heart rate of 60 beats per minute means that there is an average of one heart beat per second. The difference between two of those beats will be a percentage of that once second, so milliseconds are the best unit of measure.
All of the complex measurements that you will encounter (Fourier transformations, High Frequency/Low Frequency, rMSSD, pNN50, etc) are all derived from this string of numbers. It can be hugely frustrating looking at academic papers and articles/blog posts without understanding that because they all dive right into these derivations and at first it looks like HRV has scores of different measurements. It is one measure with a lot of interesting interpretations. And you will develop your own favorite over time.
Find out more by reading about How To Get Your Numbers.
Your heart beats at varying intervals based on the state of your nervous system. If you are nervous, scared, or exercising your heart will beat regularly and fast. If you are relaxed your heart will beat at irregular intervals and more slowly.
This is how that works. Your Autonomous Nervous System is made up of two subsystems, the Sympathetic and Parasympathetic Nervous Systems. The Sympathetic Nervous System (SNS) is your accelerator. It raises your heart rate and pushes you into fight/flight mode. When you are in fight/flight mode your Heart Rate Variability (HRV) is low as your heart is pumping blood regularly and fast in order to get you through the immediate danger.
Your Parasympathetic Nervous System (PNS) acts as a brake on your SNS. When there is no danger, the PNS is braking your SNS, slowing the beats and the results is a lower heart rate and higher HRV. When this is happening you are resting and digesting.
The variability in your heart rate is the interplay of your body’s accelerator (SNS) and brake (PNS). As you start examining your HRV readings, it will be a bit awkward because low HRV numbers mean you are stressed and high HRV numbers mean you are relaxed. HRV does not read stress, it measures your heart’s variability and the measures are inversely correlated with stress. High variability is low stress, low variability is high stress.
Find out more by reading about RR Intervals.
Heart Rate Variability (HRV) is the most accessible measurement of stress and health for someone who is not in the medical profession, and there are a shortage of plain English explanations of what it is, how to capture it, and how to use it to improve your life.
It took me almost a year to playing with HRV measurements using numerous devices and different apps before I felt I understood it enough to explain it to others. Along that way I had some very generous help from colleagues and friends who shared key concepts with me. And I went down many unproductive paths that wasted a fair bit of time. My goal here is to help you past those unproductive paths and give you the summary benefit of the great advice I have gotten.
With the purchase of some inexpensive items and reading through this summary you should be able to have some fun taking readings that are specific to your physiology. The interesting element of measuring your own HRV is that the baselines and tolerances are uniquely yours. Aggregate data from others gives you tips on where to look for those readings that are yours and yours alone.
The best thing to do is get some simple kit and start playing with it. You’ll learn as you go and I am happy to share as many tips as possible here. Enjoy!
Find out more by reading about Variable Interval Heart Beats.
I had an opportunity to compare five meetings that had similar content and the same attendees over a one month period. My colleagues and I were preparing a big launch and we were looking at the plan in a series of review meetings. In the first meeting I had created a first draft plan and had to present to executives I was meeting for the first time. I was not sure of how they liked to consume information and was on edge. So the first meeting looked like this:
What you see here is that I was in “overdrive” 33.8% of the meeting time.Overdrive means I my parasympathetic nervous system had stepped aside and my sympathetic nervous system had me in fight/flight mode. The blue lines in the chart are those heartbeats where the difference in time between beats was under 17 milliseconds for at least 10 consecutive beats. This meeting was almost two hours long, I was answering a lot of detail and we were finding our way together so I was in overdrive for one third the time.
We returned to review the progress from the first meeting a week later. In this I had my materials memorized and I knew how the executives consumed information. The meeting went very well, and we still had a lot of work to do. Using the same definition of Overdrive here is the chart:
This second meeting was almost two hours long and because I was so prepared I was in Overdrive only 10.8% of the time. As you can see from the chart there were only periodic physiological accelerations. Big difference. In the next meeting, the executive I was supporting and I did not have a lot of time to prepare for the meeting. We went in without synchronizing. You can see the chart here:
I was in Overdrive 15.2% of the meeting. You can see that my physiological fight/flight lines are concentrated early in the meeting as the executive that I was supporting and I were synching up. We found our way pretty quickly and you can see the blue lines even out.
In the fourth meeting we had taken another week to make progress on the launch. A lot of the details were worked out and we were in pretty good shape. When we got together the same executives were in the room and my supervising executive and I had a chance to coordinate. You can see the results in this chart:
Overdrive was only 4.4% of the time and the meeting was smooth. I felt good in the meeting and the readings show things went smoothly. We had one last meeting to get final check off an approval. I would through this data out because the environmentals of the meeting completely threw things off. You can see the data is very different:
Here you see a complete physiological meltdown as I was in Overdrive 87.2% of the time. Turns out the office I was taking the meeting from was extremely hot. I was perspiring and uncomfortable. It was a distracting situation. The meeting went well. We got approval and the communications afterward were universally positive. I believe the physical discomfort overrode the comfort with the materials.
So it appeared that reviews of familiar material with the same team of people did results in less time in Overdrive. Comfort with the material and people improved my performance. The last meeting is odd and I can’t definitively explain it with the hot temperature. But the first four seem to indicate improvements.
I wanted to look at how often I entered the Parasympathetic Flatline while in a 1:1 conversation with a colleague by phone. For the discussion I read by heart beat intervals using a Polar H7 heart rate belt and the Heart Rate Variability Logger app for iOS. I also recorded my side of the meeting on a smart phone. When the meeting was complete I downloaded my heart beat intervals via csv file and pulled them into excel. Once in excel I used a formula mechanism I created that graphs segments where more than 10 consecutive interbeat intervals are less than 17 milliseconds apart.
During the 60 minutes session I measured 4,451 heart beats and the intervals between them. Of those intervals, 14% were in groups of consecutive intervals that were close together, meaning during 14% of the meeting I was in what I call Parasympathetic Flatline. This measured the periods where I was in fight/flight mode during the discussion.
Here is a vizualization of the meeting:
In the session the forty-seven stress events triggered. Of these, 22 of 47 occurred when I was talking and presenting information to my colleague. 25, or 53%, occurred when I was listening to my colleague. When listening to the recording, it is clear that the stress event, even when occurring when I am talking, begin when I was no agreeing with my colleagues response or trying to move him to a different position. The stress response was a result of not liking the direction the conversation was going.
Again physiology has shown that anticipating and trying to shape another person’t response is the source of stress in a 1:1 interaction. I once thought presenting my own opinion was a source of stress but that has turned out not to be the case. The stress, it appears, is not agreeing with someone else presenting their opinion.
Added clarification: From Twitter, fellow QS’er Gustavo (@GGlusman) asked the percentage of time I was talking versus not. Pushed by the question I went back beat by beat and looked at the session. As I reported above 736 beats were “in stress ” meaning that those beats were in a grouping with more than ten beats that occurred with a difference in beat interval less in 17 milliseconds to the adjacent beat. Of those beats in stress, I found that 238 were while I was talking and the remainder while listening. So that means 38.5% of the stress beats occurred while I was talking, 61.5% while I was listening. Impatience while listening was clearly more stress creating than flapping my gums. Thanks to Gustavo for asking the clarifying question!