Heart Rate Variability

“Small is the number of them that see with their own eyes and feel with their own hearts.”
― Albert Einstein


In my previous post I discussed a paradox regarding the evolution of complex living organisms and entropy. Entropy in its classical definition is always directed from away from order and towards disorder. Living systems are able for a time to resist this entropic force maintaining internal integrity by flexibly adapting to an outside environment.

This interplay between internal stability and external adaptability is made possible through the  communication of many nested interacting physiological systems. It is known that autonomic nervous system modulation plays an integral role in many aspects of this complementary interplay by maintaining stability through homeostasis, and by changing in response to the environment through allostasis. The autonomic nervous system has a prominent role in this process and has two complementary sub-components corresponding to the sympathetic and parasympathetic branches. The sympathetic nervous system is predominate during stress while the parasympathetic nervous system is predominate when relaxed.

So What is Heart Rate Variability

Recently there has been a great deal of research related to heart rate variability (HRV) as a non-invasive bio marker of stress. Heart rate variability is related to the regulation of the sinoatrial node, the natural pacemaker of the heart.

The rhythm of the heart is primarily under the control of the vagus nerve, which inhibits heart rate and the force of contraction. When you inhale, you take your foot off the parasympathetic brake and your heart rate accelerates. When you exhale, you press down on the parasympathetic brake and your heart rate slows. This change in your heart rate from beat to beat is called heart rate variability (HRV), and HRV is a widely used method for studying cardiac autonomic modulation.

HRV can also be considered a measure of vagal tone, and is elevated in those with active lifestyles and low in those who are sedentary. Paradoxically HRV is higher when relaxed and lower under stress. A high resting HRV indicates that the heart is manifesting complex patterns subtly responding to its environment in an adaptive manner. This is a good sign and tends to correspond with a sense of well-being.

Low HRV is thought to reflect excessive sympathetic and/or inadequate parasympathetic activity and is a strong predictor of mortality in patients with Congestive heart disease. Reduced HRV is a powerful and independent predictor of an adverse prognosis in patients with cardiac disease. It has a potential to become a non-invasive diagnostic and prognostic index in clinical practice. HRV is also lowered in psychological disease states, such as anxiety, depression, and PTSD.

Julian Thayer has been one of the leading researchers relating neuroimaging studies and HRV indices to stress. Here is a review of much of his work. Thayer and his colleagues propose a model they refer to the "Neural Visceral Integration Model"

"We further propose that the default response to uncertainty is the threat response and may be related to the well known negativity bias. Heart rate variability may provide an index of how strongly ‘top–down’ appraisals, mediated by cortical-subcortical pathways, shape brainstem activity and autonomic responses in the body. If the default response to uncertainty is the threat response, as we propose here, contextual information represented in ‘appraisal’ systems may be necessary to overcome this bias during daily life. Thus, HRV may serve as a proxy for ‘vertical integration’ of the brain mechanisms that guide flexible control over behavior with peripheral physiology, and as such provides an important window into understanding stress and health."

This model specifies a central autonomic network (CAN) brain network including prefrontal and sub cortical regions that function to support adaptability and health. The primary output of the CAN are sympathetic and parasympathetic neurons that innervate the heart. Top (prefrontal) down (subcortical) inhibition is associated with increased vagal input (^ HRV). 

A Neurovisceral Integration model (NIM)

A neurovisceral integration model (NIM) proposes that the autonomic nervous system is a final common pathway that links psychological and physiological states and that HRV can be a useful index of NIM and organism self-regulation. This can be thought of as a western model that describes the heart-mind concept that is familiar to eastern philosophy, traditional Chinese medicine and tai chi practice. Or more simply as Nelson Mandela said:

"A good head and a good heart are always a formidable combination."

Heart rate variability is also gaining a great of interest in the field of endurance exercise.  Endurance training is one of my interests and I own a watch and heart rate monitor that allow me to capture HRV data.  Oscar Wilde was onto something when he said  “Hearts Live By Being Wounded”,  yet there is a fine line between the stress that causes a healthy progressive adaptation, and the stress that can lead to imbalance and disharmony. Attention to HRV levels over time may help those undergoing strenuous training  attempting to attain peak condition while avoiding over training.

There are many measures used to quantify heart-rate variability some of which come from complexity, chaos and information theory. Included among these HRV measures are measures of entropy. In future posts if time avails I plan to keep something of a diary capturing my HRV in various activities, rest, biking, running, cognitive stress states of composing a blog or writing up statistical results at work. This should allow me to see how sensitive and reliable these indices are to various activities over time. Here are some examples:

First this is my HRV data yesterday at rest before heading out on my bike to work:

Beats Per Minute = 50.36
Shannon Entropy =2.84
Correlation Dimension (D2)=3.44

Beats per minute is simply a measure heart rate. Low beats per minute in a resting heart rate is associated with high HRV, thus 50 beats per minute resting HR suggests a likely high HRV. For the measure labled Shannon Entropy, a low score indicates more complexity and high HRV. Compared to this study of engineering students at rest in which their average Shannon Entropy was 3.17 my score of 2.84 is good sign. HRV reduces with age (I am 50 yrs old), but perhaps my endurance training is paying off (or maybe engineering students have trouble resting ) . Correlation Dimension (D2) is another measure of complexity. In this case a high score means high HRV. Again my results here look good as the average for the engineering students was 2.83.

Now here is my HRV data on the 1st 30 minutes of my bike to work (the software only allows a 30 minute max period).

Beats Per Minute = 124.8
Shannon Entropy =5.32
Correlation Dimension (D2)=0.547

This all makes sense. My HR speeds up due to the stress of the physical activity. Both the entropy and D2 measures indicate less complexity (thus lower variability) in my heart rate.

Now here is my HRV data on a mildly stressful writing task sitting at work:

Beats Per Minute = 60.67
Shannon Entropy =3.39
Correlation Dimension (D2)=3.24

Again this all makes sense. My HR is 10 beats higher than at rest in the morning. The entropy and D2 measures both indicate less complexity in the HRV thus somewhat more stress. Interestingly the paper I linked above with the engineering students was a study of HRV and stress on exams. According to table 5 in the paper I wouldn't have qualified as being under stress with the entropy or D2 measure (although D2 was very close). The data does suggest more stress than my earlier resting test. These measures seem to be doing a good job so far of quantifying stress loads.