Heart rate variability

Heart rate variability, as the name implies, is variability in the duration between consecutive beats.  In the ECG, it is variability in the time interval between consecutive R waves in the QRS complex that represents the electrical events of ventricular contraction.  In general, the heart beats faster during inspiration and more slowly during expiration.  These variations are governed by the autonomic nervous system, which is responsible for the regulation of many of the body’s internal organs and in particular, coordinates the fight-or-flight reactions that prepare our body to deal with challenging situations.  There are two distinct divisions of the autonomic systems: the sympathetic system which tends to accelerate the heart and the parasympathetic system which produces deceleration. 

Almost certainly HRV is a crucial importance to the athlete, though the details are still a subject of debate.  There are three main issues:   1) loss of HRV  is potentially a useful indicator of the stress associated with training, and there is evidence suggesting that adjusting training schedules according to changes on HRV can increase the quality of the training and diminish the risk of over-training; 2) loss of HRV is a fairly reliable indicator of sudden cardiac death in individuals with heart disease and some evidence indicates that it has similar implications even when there is no other evidence of heart disease; 3) a resilient heart with high HRV might in fact function more efficiently and hence improved HRV might itself contribute to improved performance.   In the next few posts I will attempt to examine all three of these issues, and also to address the question of how HRV might best be increased by training.

Before examining the evidence that training might have either beneficial or deleterious effects on HRV, I decided  to look at HRV in my own heart.  Both Polar and Suunto manufacture hear rate monitoring equipment that provides an estimate of HRV, but I am still undecided about the value of investing in such equipment.   Furthermore, I thought it would be interesting to examine not only heart rate variability, but also the shape of the various peaks and troughs that make up the ECG.  I rigged up a system with a lead attached to each forearm, and fed the signal to some electrically isolated amplifiers, before digitizing it so that I could subject it to some detailed analyses. (Isolation of the amplifiers is crucial for safety).  The trace recorded with this ad hoc system in not a clinical ECG, but nonetheless corresponds quite closely to lead 1 in a clinical 12 lead ECG.  Lead 1 represents the difference in voltage between left and right arms.



The ECG reflects the flow of the electrical currents that produces contraction of the heart muscle.  The current flow produces currents in the surrounding body tissues and provides a signal detectable on the body surface.  The magnitude and average direction of the current flow determines the size and shape of the various peaks in the ECG.  For our present purposes, the feature of greatest important is the rhythm, and in particular, the variability of the interval between beats.  However, a substantial number of athletes have minor abnormalities of the shape of the waveform.  Follow-up studies indicate that these abnormities are generally benign and do not create a high risk of heart attack, though the issue of distinguishing clearly between normal variation seen in athletes and pathological variation is still a somewhat vexed question.  Therefore, I was interested to examine the waveform to see whether or not I have significant abnormalities.  I am an amateur in the interpretation of ECG’s, so if a cardiologist happens to read this, I would be delighted to hear whether or not I have interpreted things correctly. 

ECG and inter-beat interval

ECG and inter-beat interval

The figure on the left shows the ECG during two successive heart beats, while the figure on the right shows the variation in inter-beat interval over a period of about 80 seconds.  The waveform looks fairly normal to me (as an amateur). 

Atrial contraction

The P wave reflects spread of the electrical signal from its generation at the sinoatrial node, though the walls of the right and left atria, and in my recording is about 80 milliseconds wide and has a height of 0.08 mV.  After an interval of around 195 millisec, the P wave is followed by a QRS complex.  The PR interval is the time between sino-atrial node firing and the firing of the AV-node which initiates ventricular contraction.  It is an important indicator of transmission of the electrical signal from atria to ventricles. The normal value is around 200ms.  Very short values might indicate a potentially serious conduction abnormality known as the Wolfe-Parkinson-White syndrome, while long values might indicate blockage of transmission, so it is reassuring that the duration of my PR interval seem fairly normal.   Between P wave and QRS complex the trace should be fairly flat, though in my case there is a slight downwards slope.  I do not know if this has any significance. 

Ventricular contraction

The QRS complex reflects the spread of the electrical signal that causes depolarization of the muscle in the walls of the ventricles.  The depolarization of the muscle cell membrane produces the muscular contraction that ejects the blood from the ventricles.  The negative Q wave represents the spread of electrical signal downwards and left-wards to the apex of the heart and the large Q wave reflects the spread around the lateral walls upwards and predominantly rightwards from the apex. In some leads of the ECG, R is followed by a small negative deflection known as the S wave, though due to the direction of current flow at this time, the S wave often small in lead 1, and in my ECG, it is scarcely discernible at all.    In a healthy heart the width of the QRS complex should be less than 120 ms.  In my case it is only about 60 ms.

Ventricular repolarization

The final peak of interest is the T wave, a hump occurring about 350 ms after the end of the QRS complex in my trace.  This denotes re-polarization of the heart muscle, making it ready to contract again.  Prolongation of the QT interval occurs in a number of pathological conditions and can also be an unintended side effect of various medications. Because it tends to be shorter at high heart rates, it is usual to estimate a corrected value known as QTc which allows for variation in the heart rate.  In my recording QTc is about 300 ms.  The usual value is around 420 ms, so I do not have any evidence of a pathologically long QTc.

One feature that is of great clinical importance is the ST segment.  This is depressed when the heart muscle is deprived of oxygen and can be elevated once the heart muscle has been damaged.  The ST segment should be flat, but in my case it rises steadily for about 100 ms preceding T wave.  As far as I am aware, this upward slope of the ST segment is quite common in athletes, and possibly denotes thickening of the ventricle walls.  The point at which ST elevation of depression is usually measured is 60 millisec after the end of the QRS complex.  At this point, my trace shows no evidence of either depression of elevation, so I am inclined to interpret this picture as indicating a healthy though perhaps slightly hypertrophic heart.


Now to the important issue for our present discussion.  My average heart rate during this recording was 46 beats per minute, which I think  is fairly typical of a fit athlete when sitting down, trying the relax but nonetheless, with my mind focused on making sure the equipment is continuing to function.  However even more striking is the variation of the heart rate.  As can be seen from the right hand figure, over the course of about 80 seconds the inter-beat interval varied from 1.1 sec (HR 54 bpm) to 1.5 sec (HR 40 bpm).  The normal variability in beat-to-beat interval is less than 10%.  So I definitely have substantially increased heart rate variability.

Even  more interesting is the time-scale of the variation. It can be seen from the right hand figure, that the dominant fluctuation are rapid (ie high frequency) fluctuation occurring on a time is around 2-4 beats (i.e over a period of 4 seconds corresponding to a frequency of  0.25 cycles/sec.)  These high frequency fluctuations are characteristic of the parasympathetic nervous system, which in general is associated with relaxation and recovery. 

However, closer inspection also reveals underlying variation on a time scale of around 30-40 beats (40-50 seconds) corresponding to a frequency of 0.02 cycles per second.  Slow fluctuations on this time scale are characteristic the action of the sympathetic nervous system – the part of the autonomic nervous system that promotes fight or flight.  So I am encouraged to observe evidence of a balance between parasympathetic and sympathetic activity, but with a particularly strong parasympathetic component.


Sympathetic-parasympathetic balance: the Poincare plot

There is an intriguing though somewhat complex way of quantifying the relative contributions of parasympathetic and sympathetic nervous system to hear rate variability, known as the Poincare plot.  The Poincare plot is a scatter plot in which each heart beat is represented by a point in a two dimensional plane, with the value of the  inter-beat interval preceding that beat is plotted on the horizontal axis and the value of the following  inter-beat interval on the vertical axis.   This scatter plot is shown on the figure below, for the same data as is shown in the figure above.  The points look a bit like a swarm of bees scattered around a line that runs diagonally upwards at 45 degrees. 

Sympathetic - parasympathetic balance: the Poincare plot

Sympathetic - parasympathetic balance: the Poincare plot


The interpretation requires some concentrated thinking but is worth the effort.  Imagine that the heart rate variability is produced entirely by sympathetic nervous system activity.  Such fluctuations occur on a slow time scale – typically varying substantially over a period  40 beats or more. Therefore for consecutive beats, the inter-beat interval is almost unchanged.  The points representing those near identically spaced consecutive heart beats must lie near the 45 degree line.  However, if there is a slow drift in values of inter-beat interval over time, due to fluctuation in the input from the sympathetic nervous system, then the points will wander up and down the 45 degree line.

On the other hand, if the fluctuations are driven by the parasympathetic nervous system, they occur on a time scale of a few beats, and the beat by beat variation will be large.  Therefore many heart beats will be represented by points that lie far away from the 45 degree line

If we draw an ellipse that is just large enough to include most of the heart beats (allowing that there will usually be few outliers that buck the general trend) then this ellipse will have one  axis pointing along the 45 degree line and the other axis at right angles to this.  The length of the axis along the 45 degree line represents the amount of sympathetic drive and the length axis at right angles to this represents parasympathetic drive. 

In individuals a high risk of heart attack the ellipse is a long thin cigar shape lying along the 45 degree line.  For individuals with a good balance between parasympathetic and sympathetic input, the ellipse is almost round. I was delighted to find that for me, the ellipse is fat and almost round.



So in summary, I am quite pleased with what my little experiment showed.  As far as I can tell, the shape of my ECG trace is near normal, apart from the upward slope of the ST segment that I understand is relatively common in athletes.  My heart rate is slow, but even more importantly, there is a large variability, driven by a good balance between parasympathetic and sympathetic activity.  I am re-assured that I appear to be at low risk of a heart attack.


I suspect that the substantial HRV is the product of my training.  In future posts I will examine the evidence regarding the types of training that are most likely to improve HRV and also explore the question of whether measurement of HRV does in fact provide not only an estimate of likely risk of a heart attack, but also might be a useful indicator of the over-training syndrome, and hence provide a useful way to adjust one’s training schedule.

7 Responses to “Heart rate variability”

  1. Ewen Says:

    Canute, I’d be happy to have you as my cardiologist! That was most interesting. Your heart looks to be in great shape.

    I now know more about HRV than I ever did. I take it the variation when at rest is a good indicator? For instance, when I was standing still before my run today, the HR varied between 70 and 64 (and would go up and down by a few beats over a short time).

    Why does a lack of HRV indicate over-training? Or did I miss that?

    Also, I’m interested in the training methods to increase HRV in your coming post.

  2. canute1 Says:

    Ewen, Thanks for your comment. With regard to whether resting HRV or HRV during exercise is most useful as a guide to overtraining, there is evidence that HRV measured during relatively non-demanding activities (eg standing-up as opposed to lying down) is a useful guide to levels of over-training. As far as I know, Polar recommend using standing HRV in the morning as the measurement to guide adjustment of training intensity. HRV decreases as exercise intensity increases, and I am intrigued by the idea that the ability to maintain HRV during aerobic exercise might also be informative regarding current training status, but I have not yet assembled much evidence regarding this. Another future blog topic.
    With regard to why HRV indicates over-training, this is complex. Many different mechanisms regulate the parasympathetic and sympathetic input to the heart, and therefore control HRV. In the short term, circumstances that demand fight or flight lead to increased sympathetic input and decreased parasympathetic input. The information that governs this regulation comes from higher command centres in the brain (eg HR increases due to parasympathetic withdrawal when a sprinter is on the starting blocks, and this effect is greater the shorter the anticipated race), and also from many receptors sensitive to either mechanical stretch or chemical environment within the body. However, perhaps more important with regard to reduced HRV associated with cumulative stress, corticosteroids (which a released in response to many types of stress) can act to decrease parasympathetic output. This topic merits a more extensive discussion – yet another future blog topic!
    I don’t want to bore readers with my own pre-occupations, though in fact I write my blog not only in the hope of inviting comments from readers, but also for the purpose of clarifying my own ideas, so I will probably continue with heart physiology at least from time to time in the near future.

  3. Mystery Coach Says:

    Canute, Do you have an eMail where I can sent you information.



  4. Part P Training Courses Says:

    This is a cool blog and I have added you to my favourites!

  5. canute1 Says:

    PP, thanks for your comment.

  6. Talisha Bellar Says:

    Howdy! Someone in my Myspace group shared this website with us so I came to look it over. I’m definitely loving the information. I’m book-marking and will be tweeting this to my followers! Superb blog and fantastic design.

  7. digestive biscuit Says:

    I was curious if you ever considered changing the page
    layout of your blog? Its very well written; I love what youve got to
    say. But maybe you could a little more in the way of content so people could connect with it better.
    Youve got an awful lot of text for only having 1 or two pictures.
    Maybe you could space it out better?

Leave a Reply to Part P Training Courses Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: