Varied Interpretations of Heart Rate Variability

If you use any type of personal sensor device like a smart watch or a fitness tracker, you are probably already familiar with the concept of heart rate variability or HRV. You might have seen this statistic interpreted in many different ways -- as a measure of cardiovascular fitness, of stress, of mindfulness or resilience in the face of stress, or even of risk for negative cardiovascular events like heart attack and stroke. What is HRV, and what does it mean?

HRV is a measure based on the duration between one heartbeat and the next. This is usually measured in milliseconds per beat -- so, for instance, if your heart beats 60 times per minute, then the average interval between two heartbeats is 100 milliseconds (60 seconds / minute x 1 minute / 60 beats = 1 second per beat or 100 ms per beat). The heart has an off-beat as the chambers expand and contract in rhythm, so the technical definition is based on the type of ticker-tape record of heartbeats that's produced by an electrocardiogram. Notice that each beat actually has several high and low points, which are called a "qrs complex." The q step is the low point before the uptick, the r step is the peak (sometimes also called n), and the s step is the trough after it. You can analyze the interval between the peaks, the interval between the troughs, or other metrics. The most common way to look at this is based on the peaks, so it's called an "RR interval" or "NN interval" -- i.e., the time it takes the heart to get from one R/N peak to the next.

To get the measure of heart rate variability, we need to know not only the average value of those intervals over a period of time, but also the range of different intervals. An easy way to calculate this is by subtracting each time point from the one before it and then averaging those differences -- you have to square each difference and then back out of that by taking a square root after you average them all together, which is just to get rid of the negative signs created when some intervals are higher than the average and others are lower, but the "average difference" is the idea. This method is called the Root Mean Square of Successive Differences, or RMSSD. Another well-accepted HRV measures is the standard deviation of n-to-n intervals, or SDNN, which you get by subtracting each of those interval lengths from the average value of a difference over some time period (that's the "deviation" part), and then again averaging the deviation scores with squaring and square roots to get rid of negative signs. These common metrics are both measured in milliseconds.

Other HRV metrics are not as easily interpretable as the length of a beat-to-beat interval, however. Some metrics used in research include the percentage of successive beat-to-beat intervals that differ by more than 50 milliseconds (pNN50); the difference between the highest and lowest heart rates observed in a particular time period (HR max - HR min), which is measured in beats per minute; and a variety of power measures that are calculated based on the frequency of beats over time and are measured in squared milliseconds or Hertz, like the frequency of a sound wave. Power measures are calculated using a method called "spectral analysis" that is again like the way sound or light waves are studied. Some of these measures convey more reliable information than others, with SDNN often being among the best-performing metrics and the difference between minimum and maximum heart rates being among the worst. In my own research, HRV based on SDNN was more useful as a predictor of people's daily fatigue than any of the other heart rate metrics we examined, including average, minimum, maximum, or resting heart rate.

One major problem with consumer-grade sensors that provide HRV metrics is that they often don't tell you which of these methods was used! The numbers produced may therefore be comparable to other data produced using the same sensor device, but are often not interpretable on their own. It is therefore hard to tell someone whether they have a "good" or "poor" HRV metric. Instead, the best we can do is to say that a particular person's HRV is higher in some circumstances and lower in others, or that some people have higher or lower HRV in general. This is not a strong basis on which to make any type of decision about one's health. It's also true that the type of task being performed during an assessment of HRV will affect the readings one gets -- HRV during exercise, for instance, will be different from HRV collected during a deep-breathing relaxation task, and both may differ from a laboratory-based stress task like trying to compute math problems under time pressure.

Assuming that we can compute an accurate measure of HRV from raw heart rate data, what does that HRV number actually mean? I'm going to look at four possible explanations, each one a little more theoretical than the one before it. "More theoretical" means either more important and true, or more divorced from reality, and I'll leave it for you to decide which explanation you prefer.

• Cardiovascular fitness - Most heart rate metrics, like peak heart rate or resting heart rate, simply measure whether someone has a healthy heart. The simplest interpretation of HRV is that when it is high (more ups and downs during the course of a day) then the heart is responding appropriately to a wide variety of situations. It also probably means that you are getting enough exercise, because your heart rate is higher at those times and lower when at rest. Low heart rate variability might simply mean that your heart rate stays in a steady range all the time, which means you aren't ever giving it much of a workout. Age is also a strong predictor of HRV, with younger people having the greatest variability and older people the least. High HRV then means your heart is in good working order, and low HRV means that it may not be. And even a small sample of heart rate data can predict HRV metrics over a much longer time period. In people with chronic heart conditions, a HRV score that's down in the range below 70 milliseconds per beat can actually predict the chance of dying from a cardiovascular event. This is probably because the person has a chronically low heart rate, which also means low variability in the rate over time -- their heart doesn't beat as reliably and it can't handle much of a strain.

• Cardiovascular stress - The reason for an increased heart rate is usually some form of stress. For example, your heart might beat faster because of exercise (a good kind of stress), anxiety because you have to speak in front of a group (a more negative kind of stress for most people), or just because you drank several cups of coffee (also a stressor from the heart's point of view, even though it is one we often seek out). Stress isn't bad in and of itself; it's just a normal part of life, and it tends to increase HRV which is a sign of health. But chronic stress can be bad for you, resulting in a heart that is always on "high alert" and never has a chance to come back down from a stressed-out state. A heart rate that is always high will have low variability -- it won't change much over the course of the day -- which is a different way in which low HRV might be a sign of poor health. People who experience daily stressors as more severe tend to have lower HRV. And people who take time to relax or meditate tend to have higher HRV. In my own research I have used low HRV as a marker for chronic stress, and it shows a reasonable level of agreement (but not total overlap) with people's level of self-reported stress based on a questionnaire.

• Cardiovascular recovery - As I said above, stress itself isn't bad; if you don't exercise you're likely to have low HRV, and even psychological stress can we a positive experience sometimes -- think of the thrill of winning a game or competition. But some people have a harder time than others in "coming down" after a stressful experience is over. Some psychological research has shown in particular that people with high levels of hostility -- a worldview in which they feel perpetually under attack -- have lower HRV and higher levels of mortality. [Earlier research that identified this same pattern blamed it on a "Type A personality" in which people showed competitiveness and time urgency, but later studies identified hostility as the specific personality aspect with negative effects on the heart -- good-natured competitors don't have the same risk]. In people with high levels of hostility, stressful events produce the normal cardiovascular response. However, the person's blood pressure tends to stay elevated longer after the event is over, and people with more negative feelings about a stressor and more negative emotions overall also have lower HRV after exposure to stress. In one study subjecting the person to actual verbal harassment during a stressful task had no effect beyond the internal trait of hostility -- the type of person experiencing the stress mattered more than the stressful event itself. Finally, there's evidence that mindfulness-based stress reduction can ameliorate some of the negative effects of stress and improve heart rate variability.

• Sympathetic vs. parasympathetic nervous system activity - heart rate increases occur because of the sympathetic nervous system (SNS: epinephrine [adrenaline] and norepinephrine from the adrenal glands, plus adrenocorticotropic hormone from the pituitary gland) activating what is usually called the "fight or flight" response. Heart rate decreases are caused by a different hormone, acetylcholine, which is part of the parasympathetic nervous system (PNS). Together, these two systems work to maintain a state of balance in the body, with each one compensating for potential excesses of the other. Polyvagal theory is a framework for understanding the relationship between SNS and PNS activity, with SNS activation indicating a more primitive stress response and PNS activation suggesting a more evolved or sophisticated stress response that includes the potential for social engagement and self-soothing. Low HRV means that the SNS is overly dominant in a person's pattern of stress response, while greater PNS activation will result in greater HRV. In this view, HRV serves not only as a marker of stress response, but as a way of measuring prosocial tendencies and psychological health. The HeartMath approach to HRV isn't based directly on polyvagal theory but it takes a similar perspective in interpreting HRV as an indicator of SNS versus PNS activity. Here, more positive emotions like joy and appreciation are expected to produce a more regular (or "coherent") pattern of heartbeats, which also tends to have more ups and downs and therefore a higher level of HRV. Negative emotions are expected to produce jagged wave-forms that stay high or low for longer intervals, and therefore show up as lower HRV. One small-scale study showed that compassion-focused imagery produced an increase in HRV and a decrease in another stress hormone, cortisol. These changes were even greater than those shown with a standard relaxation task, but (in a finding that seems more closely related to the third bullet above) they were weaker for people who were more anxious or self-critical overall. The idea of coherence isn't identical with HRV, but heart rate variability is one way (along with respiration rate, blood pressure, or self-reported emotion) that researchers who use HeartMath principles have proposed to measure what is essentially a psychological or spiritual experience of connection and peace.

Which of these interpretations of HRV is correct? Can your fitness tracker really measure your spiritual health, your personality type, or your level of connection to the universe? Or does it just show that your heart beats faster on days when you get more exercise? It's hard to know at this point. The association between higher HRV and good cardiovascular health is indisputable, but the links to stress are also quite convincing. And in both published research and some ad hoc experimentation, I have found HRV to explain more than simpler measures like maximum or average heart rate. Whatever it means, HRV is an intriguing physiological metric that may give us insight into the Intuitive Mind -- experiences and behavioral tendencies that are usually outside of our conscious awareness.