HRV, explained without the mysticism

Heart rate variability (HRV) is the beat-to-beat variation in the time between heartbeats. A heart that is being actively steered by the parasympathetic "brake" varies a lot from beat to beat; a heart running under stress, illness or fatigue beats more like a metronome. Higher HRV therefore usually means a more rested, more adaptable nervous system — but the number depends heavily on age, sex, posture and how it was measured, so there is no meaningful "normal". The only comparison that works is against your own readings, taken the same way, over time.

A steady heart is not a healthy heart

The first thing that surprised me about HRV is that it runs against intuition. I assumed a calm person's heart would beat evenly, like a clock. It is the opposite. At rest, a healthy heart is constantly being nudged — sped up a little on each inhale, slowed on each exhale, adjusted by blood pressure reflexes, hormones, temperature¹². The intervals between beats might read 980 ms, then 1,040, then 910. That jitter is the nervous system doing its job.

When you are under load — an argument, a deadline, a cold coming on, three bad nights — the body shifts toward sympathetic activity and the parasympathetic influence on the heart tends to weaken. The heart tends to settle into a faster, flatter rhythm. Variability tends to fall. Research links lower HRV with stress, illness and fatigue¹.

What the number actually is: RMSSD

There are dozens of HRV metrics, and the fact that apps rarely say which one they show is a big part of the confusion. The one iKima reports is RMSSD: the root mean square of successive differences. In words: take each pair of neighbouring beat intervals, find the difference, square it, average the squares, take the square root. The result is in milliseconds.

Two properties of RMSSD matter here. It mostly reflects vagal (parasympathetic) influence on the heart and correlates closely with high-frequency HRV², so it tracks the parasympathetic brake specifically. And it can be calculated from short recordings — how short is still debated. In research the conventional short-term recording is five minutes, which is the 1996 Task Force recommendation as Laborde and colleagues describe it². Shaffer and Ginsberg's review lists studies in which RMSSD from 10 to 30 seconds tracked the five-minute value, but judges many of them methodologically limited, and concludes that for healthy people a resting baseline as short as one minute may be sufficient if artifacts are carefully removed¹. Laborde and colleagues note that one-minute ln RMSSD has shown good reliability against the five-minute value, and that recordings of 10 to 50 seconds have been shown reliable under certain conditions². iKima is not a medical device, and its reading is a deliberately short one, so no single number is a verdict. What helps is repetition: measure several times, under the same conditions, over days and weeks, and your own baseline — the range that is usual for you — builds up from the readings.

Why there is no "normal HRV"

People ask me what a good number is. I understand the question and I refuse to answer it, because any answer would be misleading. Nunan and colleagues reviewed short-term norms in healthy adults, and Shaffer and Ginsberg report the RMSSD result: a mean of 42 ms with a standard deviation of 15 ms, from 15 studies whose results ranged from 19 to 75 ms¹. Those recordings used free or paced breathing and varied body positions, and the review as a whole included three large populations aged 40 and over, which may explain their lower values¹ — so even a "healthy-adult norm" is not one number. The spread comes from things you cannot change, and from how the reading is taken:

• Age. Time-domain HRV measures decline with age; for RMSSD in particular, one study of 24-hour recordings found a decline from 40 to 60 and a rise after 70¹. The same number therefore means different things at different ages.
• Fitness and other personal factors. Time-domain measures rise with aerobic fitness; sex, heart rate and health status also matter¹.
• Posture and breathing. Shaffer and Ginsberg list body position among the factors that change HRV measurements, from subtly to greatly, and report that values from normal and from paced breathing can differ significantly. RMSSD is less affected by breathing than respiratory sinus arrhythmia is, but the effect of breathing rate on it is uncertain¹ — so measure the same way each time. This is also why comparing an iKima reading with the HRV your Apple Watch records is not like for like: different metric, posture, window and time of day.
• Device and window. In one study of 29 healthy adults at rest, five-minute rMSSD from smartphone photoplethysmography and from a chest strap agreed closely with ECG, with trivial differences³. That is agreement over a matched window. A 20-second reading is a shorter sample than a five-minute one, and how well short windows reproduce the five-minute value is still being studied¹².

Laborde, Mosley and Thayer's guide to HRV research in psychophysiology — recommendations for experiment planning, data analysis and data reporting — spends a large part of its length on exactly this: keeping posture, time of day and recording length consistent, and controlling breathing and things like sleep, alcohol and recent exercise, so that comparisons mean something². If researchers with lab equipment have to be that careful, any app that turns one reading into a verdict about you is overreaching — iKima included, which is why its own labels are deliberately coarse and say so. More on that below.

Three rules for reading your own number

1. Same conditions, every time. Sitting, before the session, before coffee if you can. The reading is only as good as its consistency.
2. Trend, not point. A single reading is weather. Seven readings are climate. Look at the week, and look for the direction, not the value.
3. Before and after is the fair test. iKima reads you before a session and again after. That comparison holds the person, the device and the measurement method constant, and posture too if you sit the same way both times, because the two readings are ten minutes apart. If the after-reading is consistently higher, that fits the models in which slow paced breathing raises vagal tone² — though two readings cannot prove why.

The part that is not a number

I built the HRV reading because I wanted to see it in data, not just feel it. I could tell the practice was helping, but "I feel calmer" is exactly the kind of claim a tech person does not trust from himself. I wanted a number that would hold up if I was fooling myself — and, just as much, one that would confirm it if I was not. It did: the after-readings were higher, most days, from the first week. What I did not expect was that after a couple of months the number stopped mattering. I could feel the shift before I measured it. The reading had taught me what "settled" felt like from the inside, and then I did not need it any more.

That is the honest pitch for HRV in a breathing app. Not a score to optimise. A way to learn a feeling, with the numbers there to keep you from fooling yourself while you learn it.

How the phone reads it: fingertip PPG. Why the exhale is what moves it: the extended exhale.

Questions I get asked

My HRV dropped after a session. Did I do it wrong?

Probably not. Single readings are noisy — a swallow, a shift in posture, thinking about an email. Look at the week. If after-readings are lower than before-readings most days, check the technique. In my own experience the usual cause is an inhale that is too big.

Can I raise my HRV permanently?

You can influence your baseline only within the range your age and physiology allow. Time-domain HRV tends to be higher with greater aerobic fitness¹, and slow paced breathing is the technique that several models of HRV biofeedback propose for raising vagal tone². These are associations and proposed mechanisms, not guarantees, and none of it is fast. The trend is what you own.

Should I measure lying down for a higher number?

The number will probably differ — Shaffer and Ginsberg list body position among the factors that change HRV measurements¹ — and it will not be comparable to your sitting readings. Pick one posture and keep it. iKima's before-and-after design assumes you stayed in the same position.

Is low HRV dangerous?

Shaffer and Ginsberg's review lists autonomic dysfunction — which HRV is used to index — among the risk factors for cardiovascular disease and mortality¹, but a consumer reading cannot diagnose anything, and "low" has no fixed threshold. If you are worried about your heart, that is a conversation for a clinician with an ECG, not an app.