Monthly Archives: July 2014

Heart Rate Variabilty While Giving a Public Speech

I was delighted to get an invite to speak at the Bay Area Quantified Self Meetup tonight. It was a great meeting and a very welcoming crowd. Here is what it looked like from an outside view as I was on stage giving the talk (Photo courtesy of Mariana Quiroga @marianaquiroga):

Qs Speech

The words and slides that I shared will come in another post. Lets instead go from the outside view to the very inside view – what was happening with my heart rate variability as I spoke. While I gave the speech I was wearing my Polar H7 which was connected to the SweetBeat Life app. After I took the readings I uploaded the data to Kubios. And to be fair to all participants, before the speech I did a session on Breathe Sync to top up my calmness.

In this graphic I show a baseline of what my RR interval, AR Spectrum and Low Frequency(LF)/High Frequency(HF) ratio looks like when I am completely relaxed. Beneath that I show progressive parts of the speech from the start point to the one-third mark:

Slide1

You can see at the top on the data labelled “Smooth Sailing – baseline” the nice three peak wave in the left graphic, one clear peak in the AR Spectrum in the graphic outlined in blue and an LF of 4382 versus an HF of 454 and a ratio of LF/HF of 9.6. That is a picture of a relaxed and coherent state. Notice the difference at the beginning of giving my speech. Two peaks for RR Interval and a very pronounced HF spike (the yellow bit) and values of LF (22.4K) and HF (30.7K) that are off the charts. A huge surge of energy as I begin the speech.

As I progress into the speech my power (LF &HF) drops dramatically. I have no variability at the 17% but some variability by the 30% point. But by that point my frequency levels are as low as they are when running a 10K race. So it looks like from a frequency level that 30% into the speech I am expending as much energy as running a foot race.

Here is the graphic of the speech from the halfway point to the end:

Slide2

At the 50% mark I have returned to a normal level of energy (HF is 2304, LF is 141) though my variability is low (RR interval on left). My subjective experience was that while giving the speech I felt on top of the material. At the 75% mark I have another drop in HF and LF. I recall thinking these slides where complicated so I was on my guard. As the speech reached the end point I see some return to LF levels at 1897 with one peak in the RR interval. And when I have gone through an interesting Q&A there is a huge surge of LF energy. A “rush” of completion perhaps?

I have always considered myself a capable public speaker and enjoyed giving my talks. Whenever faced with a speech I dig into it knowing I can do it. Looking at the data it could be that I have a regular cycle of stress in the beginning that ends with a huge surge of “completion” energy. When I remember the speech I remember the surge, not the stress. So I seek to do it again, ignoring the multiple less enjoyable steps that led to the surge.

Its only a hypothesis. Now you have a snapshot of a public speech at the heart rate variability level.

Sign up for the QuantXLaFont Newsletter
Get our lifestyle tips and studies delivered to your inbox.
Thank you! We don't spam :)

Why I Like Breathe Sync

Breathe Sync is one of a host of paced breathing apps on the iOS App Store. I have tried five different apps for the iPhone and Android whose aim is to bring the breath to an evenly paced level. The formats vary. A few just offer the pacing clock and you breathe in accordance with a visual representation of the in breath and the out breath. Others use the camera phone to measure some level of cardiac coherence and give feedback, either in real time or at the end of the session.

I like Breathe Sync because it has given me the fastest and most effective path to get into a high level of coherence. I use it now to “power up” when I need a boost to coherence. The interface is very straightforward:

 BreatheSync

The circle in the middle expands and contracts based on in the in and out breath respectively. The heart in the upper right pulses to let you know that the app is picking up your heart rate. And the timer on the bottom lets you know how much time until the end of the session. You have your finger on the camera so it is reading your heart rate. And here is where Breathe Sync is different – it changes your breath pattern based on the state of your heart rate. Coherence occurs when your heart rhythm and your breath rhythm are moving on the same cycle and Breathe Sync gets you there faster as it moves your breath rhythm to your heart rhythm.

How much so? I measured my coherence using Heartmath Pro and for 25 sessions would fire up Breathe Sync for a 1 minute session when I was in low coherence according to the Heartmath coherence score. The average Heartmath coherence score increased was 2.7 points per Breathe Sync one minute session. From experience, that is a large increase. When I used Heartmath’s own breath pacer in comparison 1 minute sessions after hitting a low coherence score the improvement was .75 points per session.

The difference in approach is that Heartmath does not change the breathing based on your heart rhythm. You breath steadily and eventually the heart catches up. It works but it is much slower.

As with all disciplines there is a mixture of tools that gives good results. Heartmath gives a coherence score that lets you know how you are doing. Breathe Sync gets you to coherence faster than any other approach. And SweetBeat Life allows you to take detailed and accurate telemetry while on the road. Currently I use all three to get the best results.

Using Stress App While Getting Tooth Drilled At Dentist

That title does not lie. I wore a Polar H7 Heart Rate Sensor paired with SweetwaterHRV’s SweetBeat Life to the dentist where I was to have a cavity filled. I measured the entire session and have pulled the data for the specific five minutes where the drill was going and that burning smell filled the room. To that measurement in a moment.

I also visited my stockbroker today to review the state of affairs. I wore the same heart rate sensor and had SweetBeat Life reading my vitals. At one point in the discussion he asked me to make a long term decision that could effect the fortunes of my children and their children’s children. I have the data from that five minute discussion as well. To that in a moment.

As a baseline I have uploaded five minutes of a coherent session that I did some time back. It is the most perfectly relaxed and coherent session I have experienced. The data was exported and put into Kubios HRV freeware. Here is what that report looks like:

Slide1

You can see at the green arrow the waveform of my heart rate. It looks nearly like a perfect sine wave, smooth and relatively even. The blue arrow shows a nice distribution of heart frequencies which indicates good variability. And the orange arrow shows the Low Frequency to High Frequency ratio (LF/HF) of my system as 9.5. This is a smooth, relaxed state.

For comparison I pulled five minutes of a difficult work session. I don’t know what was happening that day, I was just out of sorts while doing email and tweeting. Here are those readings:

Slide2

You can see the waveform (green arrow) is much flatter, there is a much smaller distribution of frequencies (blue arrow) and LF/HF ratio (orange arrow) is 1.056. This is a stress state and the metrics are very clear that there is very little variability in the heart rate.

To our friend the stockbroker. Though I spent an hour in the appointment there was one five minute period where I was “on the spot” with respect to a decision. Here is the picture:

Slide3

The waveform is chaotic at points (green arrow), but there are clear sine waves for a portion of the period. The distribution of frequencies (blue arrow) is quite broad and the LF/HF ratio is 8.046. This is nearly the same ratio as in the relaxed state. So though I felt somewhat out of sorts at moments mostly because I never was any good at picking stocks, I was generally in a Poised state during that meeting.

The Dentist came as quite a surprise. There were periods where I was very concerned that I would feel intense pain. Fortunately I have an excellent dentist and things turned out to be quick and painless. The data shows I was not nearly as upset on a physiological level as I thought I would be:

Slide4

To remind you, this five minutes was when the drill was going. The waveform (green arrow) is a more jagged than a relaxed waveform, but there is some variability. See the difficult session above to see a waveform with little variability. The frequency distribution (blue arrow) was much broader than I would have expected. And the LF/HF ratio was 4.350, not as low as the difficult session at all. In fact, the LF power which indicates a mixture of the Sympathetic Nervous System (Fight/Flight) and the Parasympathetic Nervous System (Rest/Digest) was equal at the dentist and the stockbroker.

So the conclusion here could be that having a bad day doing email is far worse than visiting the stockbroker or the dentist.  If you are banging away at email , feeling out of sorts and find yourself in San Francisco, head over to Dr. Shek and have him the get the drill out!

The Power of Exporting Data

Here is why the ability to export data from the device is so critical if you are going to do cost effective quantification of your data. I collect heart rate variability (HRV) data with the Heartmath Pro which retails for $299. I think that is very expensive. The standard screen shows a points system and an HRV waveform as shown here:

Slide1

 

Helpful to train yourself to breathe and reframe your mental state to put you into “coherency” which the Heartmath Institute says as a combination of different physiological and mental coherencies. Coherence for Heartmath is an abstraction from the heart rate intervals the ear clip reads. 

The reason I paid the extra money for the Pro version of Heartmath is that it exports data and that exported data can be uploaded into freeware like Kubios. I exported just over 9 hours of data spread over 25 sessions where I was sitting quietly using paced breathing to ensure I was in coherence. Here is my FFT spectrum from those 9 hours:

Slide1

By moving the data from Heartmath to Kubios I can now cut the data by timeframe, aggregate it as I have done here and get a more precise reading on the amount of signal from Low Frequency and High Frequency both of which give me a ratio telling me how balanced I was in the period. I am freed up to actually discover things my changing my viewpoint and the data used. 

Not only are Heartmath interfaces fixed, the data collection tools are very visible. Fine if I am in the comfort of my own home, but I can’t conduct a negotiation with a colleague while wearing ear clip and my computer bonging away on the table. I can wear a heart rate monitoring belt beneath my shirt and capture the data on the Sweetwaterhrv.com app which also exports to Kubios. Here is a test session from Sweetwaterhrv displayed in Kubios:

Slide2The ability to interpret the date comes from the power of Kubios. And the way Sweetwater allows me to collect is unobtrusive and more useful. And the best thing of all? The app costs $9.99. By combining the power of one software with the usability and affordability of the other technology I can achieve my ends at a fraction of the cost. And I learn along the way which is really what QS is all about. 

Improvement Results In Upset Recovery

I am pleased to be presenting the results of this study at the QuantifiedSelf Meetup in San Francisco on 28 July.  As I have been crunching the numbers I thought I would give a sneak peek at the summary results.

Starting on 24 June and I sat for 71 sessions using the Heartmath Pro and a digitial recorder. I would do work on a computer and record verbally what I was thinking when the Heartmath indicated I was dropping out of cardiac coherence (Upset). I was measuring interruptions to coherency which roughly correspond to being distracted or Upset while working.

As the sessions progressed I developed recovery techniques to try and move back to coherence (Poise) as quickly as possible. I wanted to see if I could reduce the amount of time I was Upset during the sessions. The data shows that my techniques did reduce the amount of time I spent Upset in each session over the course of the study:

1PercentSession

What this graph means is that the % of time I spent Upset per session went from close to 35% at the baseline to 12% at completion. In time terms, that means in each 60 minutes of work I was able to add back 13 minutes and 48 seconds of productive time. When in Poise my thoughts are more focussed, clearer and my output higher.

As an example of how Poise made for more a more productive mental state, while working on multiple administrative issues I needed to recall a mailing address. In every session when I was in Poise I could do this with ease. During one particular session I was in Upset and I could not remember the address at all. I remember thinking “I know this information” but I had to look it up on my computer. When I returned to Poise I could again easily remember the address. This is one anecdote that reflects my qualitative experience of how clear my mind was when in either Upset or Poise.

My second question was if the number of Upsets were decreasing per session. The techniques slightly improved the average time between upsets, but not enough to account for the overall improvement show in the first graph:

2MinsbetweenUpsets

The average time between Upsets in each session went from 2 minutes to 2.5 minutes over the course of the study. My qualitative experience was that the sessions had a steady diet of distractions and upsets that would occur before I was even aware of it and the machine had to alert me that I had triggered. My pre-conscious awareness was knocking me into an Upset state at a pretty steady rate.

What accounted for the improvement? The techniques I developed to effectively react to the Upset look to be the driver. The length of time I spent in the Upset state once I detected it dropped over the course of the sessions:

3Minsperupset

Awareness and consistent response in this study was particularly effective in improving how fast I could reverse an Upset. The methods were based on the two navigational impulses I discussed in my earlier post on Feeling Lost or in the Wrong Place. I have now moved to measuring two sessions daily as training to embed the ability to reverse Upsets with a guide of keeping the % of session time in Upset at around 10%. So far the results have been positive.

Look forward to seeing you at the QS SF Meetup!

 

Design Issues Example

I don’t want to get into the product review business because I think there are so many products out there and each of us use them in different ways that it is difficult to review them and have the feedback be relevant. And I have huge respect for all these startups who are working hard to create great new products so I want to support them rather than be a Mr. Grumpus throwing in comments from the peanut gallery.

I thought to illustrate a very basic design issue to show how even the coolest of apps can miss some very basic points. Cool app in question is the Sweetwaterhrv.com’s heart rate variability (HRV) app. I downloaded both of their paid apps, the latest being Sweetbeatlife which I use during exercise. To be fair to them, most of the other apps I have downloaded are now filtered out and unused. I continue to use this one so it is the top of the pile.

What I like about the app is that is provides HRV, stress levels based on autonomic nervous system frequency readings and cool features like food tolerance and heart rate recovery after exercising. It pulls data from a bluetooth chest strap. The results go into the cloud and you have a nice historical record of your readings.

What I find challenging about the app is that I think its creators have missed some very basic points of how it would be used in the field. Case in point – the heart rate recovery mode where the user would run, cycle or do some other physically challenging task to determine fitness. Workout in it purest form.

Picture yourself on a run, or doing something strenuous outside. Sun is shining and you have sunglasses on. The FUNDAMENTAL metric you track while humping it during these sessions is time duration. How long have I been running? How much longer is this session? How much longer does this suffering continue? Look at how the duration is displayed on an iPhone (red circle):

Slide1

 

Grey 4 point font timer on a white background. Massive black area above that shows a heartbeat trace that has no relevance to the exercise session because the heartbeat average is in white in the upper right. UX 101 here. When running in a sunny place with shades on you have to have big, bold numbers to show the time. Garmin and Polar know this instinctively. With this interface it is simply impossible to see how long you have been exercising UNLESS you stop running, pull up your shades and shield the screen from the sun. Which defeats the purpose of the exercise.

And then, when and if your nervous system frequencies are out of alignment and your stress level goes to “red”, as they are and will do when you need to really engage your sympathetic nervous system to push that run up a hill, you get this irrelevant screen:

Screen breath pacerHey relax. Move to a 6 breaths per minute pace. Even though your heartbeat is 165 BPM, you are huffing for air and jamming up a hill, here is a breath pacer that is completely irrelevant to gum up the screen. And to get back to your 4 point grey timer back you have to tap the tiny button on the bottom left to get the data screen or the breath pacer stays there indefinitely. And you have to do it while you are running. In the sun. Which does not work so you have to stop again.

What you see here is an app that is designed around sitting quietly and taking a reading. Then the designers add in concepts like heart rate recovery without really doing to work of looking at how people would use the app differently when doing activities that would merit taking such readings. So you get breath pacers that pop up when the use case cannot call for such a reaction from the app.

These basic flaws are rampant in every device and app I pay for. Heart rate readers that do not read correctly. EEG monitors that drop out when you move your head. EKG readers that you have to sit absolutely still for them to read, and then there is a 60% successful session rate. The list goes on.

If we are to have wearables and quantification work, the basic design needs to work. Otherwise trust is eroded and all your energy Mr or Ms Designer will be wasted. So put on a pair of shorts, lace up your running shoes and take your app out there. That bright shiny thing above you is the sun. Look at your app. You may find that 4 point font on a white background is an obvious oversight. If you don’t see it, we certainly do. We are all in this together and in the end it benefits us all for the apps to work well. Let’s make it so.

 

 

Anatomy of an Upset & Return to Poise

I introduced the concept of the heart’s EM signature and thought I would unpack a few Upsets and returns to Poise so we could see both the context and the movement at the EM spectrum level. Both of these Upsets were during a work session yesterday. I was using Heartmath ERMPro to measure my EM signature and a simple digital recorder to hear what I was thinking at the point of Upset. The environment was quiet and without distractions while I was returning emails to colleagues to set up meetings.

Upset #1: The Undefined Meeting Date

I sent a note to a colleague suggesting a date and time to meet and immediately pinged red on the meter. I felt that I was in the “Wrong Place”, a navigational term I introduced in an earlier post. The feeling originated from not knowing if I had created a conflict with other meetings and a vague sense that this colleague had already rejected this date. All of this was at the feeling level and added up to a signal of being in the Wrong Place. The signature looked like this:

Slide1

The EM signature showed my heart in the very low band, usually associated with Sympathetic Nervous System, or “fight/flight.” This Wrong Place feeling had put me on alert at a fundamental level.

Having practiced the most effective response to this is reversing the belief, I started hunting for the underlying pre-conscious picture that was creating the trigger. After some tries I hit on it – the date and time I had proposed really did not work for me either. Here is the profile:

Slide2

Though this reversal process s aimed at creating a new picture that sends a signal to the heart all is OK, the process is rapid and verbal. It sounds like this “I’m not actually available Tuesday, he is not available Tuesday, he didn’t want Tuesday, I don’t want Tuesday, I want a different date and time.” Pop. The meter flashes green and the EM signature is as shows. Still with a lot of low frequency noise, but the mid-range pops online to balance it and coherence emerges. And I feel better in the process.

Upset #2: I need to reschedule

Another Wrong Place upset, this time when I had to email a colleague early in the morning and let him know I had to push a meeting off a couple of hours due to an error on my part. I felt out of place on a vaguely guilty level of doing my colleague wrong. The steps and frequency picture looked like this:

Slide3

Again a big spike on the lower frequency band, so much so that is overrode the pretty active frequencies near .1Hz which is associate with coherence. As a note though this graph looks like the graph from the previous example where I am in coherence the ratios are very much mathematically different. The area under the curve for low frequency is much higher here.

Using a reversal of belief approach I begin hunting for a return to Poise. It looked like this:

Slide4

The hunting for the reversal started with my commitment and feeling out of place for suggesting a last minute change, and I ranged to my colleagues displeasure, and I routed back to the fact the meeting had been rescheduled multiple times already and that we had been flexible with each other. This popped the Upset and I was back in coherence after 18 seconds. This time the lower frequencies damped down completely and the feeling was quite positive. It helped when my colleague emailed back a bit later saying “No problem.”

The reversal process is fast and iterative. The average of 18 seconds that I see on these comes from first identifying I am in Upset, moving through a Wrong Place feeling to the process of pictures and reversals. There are a lot of wrong trials so getting it below 15 seconds will be difficult.

And in the spirit of transparency, sometimes I can languish in an Upset for up to a minute due to not following the reversal of belief or misidentifying the source repeatedly.

Next I will pull together snapshots of a “Lost” upset.

Heart EM Spectrum Working & Uninterrupted Focus

Did two sessions today, one to focus for 20 minutes without any activity, the other to measure how effectively I could return to Poise after detecting an upset using the Hearmath Pro. The device also took readings of the frequency spectrum of my system and the average of my system EM power emissions over the length of the session looked like this:

Slide1

 

By Working I mean working in a quiet environment with no immediate physical danger while editing a document. While doing the uninterrupted session I was sitting in the exact same place. Instead of editing a document I was simply focussing on paced breathing. Both sessions were of similar length, 20 and 25 minutes respectively.

Each session had a spike at the .1Hz range, which is associated with the entrainment frequency of heart rate coherence. That means when your heart and breath are entrained you see the system emit a pretty steady .1Hz.

Of interest is during the Work session, the lower frequencies associated with the sympathetic (fight-flight) nervous system also showed a spike. During that session I had 2 navigational “Lost” feelings and 13 “Wrong Place” feelings. The average recovery for both was around 18 seconds.

I can not draw any conclusion from this but it certainly gives an other dimension to the training in that I can see the shape of sessions on an EM level and connect that to the shape of the emotions as they occur while I navigate through my mental territory of the work session.