What HRV actually is
Your heart does not beat like a metronome. Even at complete rest the gap between consecutive beats varies by tens of milliseconds, and heart rate variability is the measurement of that variation. It exists because two branches of the autonomic nervous system are constantly negotiating: the parasympathetic branch (mostly the vagus nerve) slows the heart beat-to-beat, and the sympathetic branch speeds it up. A flexible, well-recovered system shows plenty of beat-to-beat variation. A stressed, inflamed or exhausted one shows less.
So HRV is not a fitness score. It is a read on autonomic state — the balance between recovery and stress right now. That is why it responds within a single night to alcohol, illness, a hard interval session or a bad argument, and why it is genuinely useful for deciding whether today is a day to push.
RMSSD versus SDNN — the measurement matters
| Metric | What it captures | Who reports it |
|---|---|---|
| RMSSD | Root mean square of successive beat-to-beat differences — mostly parasympathetic (vagal) tone | Whoop, Oura, Garmin, Elite HRV, HRV4Training |
| SDNN | Standard deviation of all intervals — both autonomic branches, plus breathing and circadian effects | Apple Health, clinical Holter reports |
| lnRMSSD | Natural log of RMSSD, sometimes ×20 — compresses the scale so changes are easier to read | Some training apps and research papers |
This is the single most common source of confusion in consumer HRV. If your Apple Watch says 65 and a friend’s Whoop says 42, you have learned nothing — those are different quantities over different windows. Pick one metric, one device and one measurement window, and only ever compare it to itself.
Why comparing to other people is nearly useless
Look again at the chart above. The gap between the 10th and 90th percentile at any single age is larger than the entire decline from 20 to 70. Genetics, resting heart rate, body size, breathing rate and posture all shift the number substantially in healthy people. Two equally fit, equally recovered 40-year-olds can sit at 25 ms and 70 ms — and both are completely normal.
This is Helix’s whole philosophy, so we will say it plainly: baselines beat norms. A population table can only tell you whether your number is unusual. Your own 30-day baseline tells you whether you are recovered — and that is the question you actually wanted answered. A 38 ms reading is meaningless out of context; a 38 ms reading against your own 55 ms average is a clear signal that something is taxing your system.
Practically, that means: ignore the percentile after the first look, establish two to four weeks of readings under identical conditions, and then watch for deviations from your own rolling average. Helix does this automatically from your Apple Watch data — it never scores you against a table.
What moves HRV day to day
- Alcohol — the most reliable suppressor. Two drinks commonly cut overnight HRV by 20–40%, and the effect can still be visible the second night. If you want to prove HRV is measuring something real, this is the experiment.
- Sleep. Short or fragmented sleep lowers HRV and raises resting heart rate the same night. Accumulated debt keeps it suppressed — our sleep debt calculator quantifies the shortfall, and the sleep calculator helps you fix the timing.
- Training load. A hard session suppresses HRV for 24–48 hours; that is the adaptation signal working. Suppression that lasts more than a few days, especially with a rising resting heart rate, is the classic non-functional overreaching pattern.
- Illness. HRV often drops a day or two before you feel symptoms — one of the more genuinely useful early warnings a wearable provides.
- Psychological stress. Deadlines and conflict suppress it about as effectively as physical load, which surprises people who expect a training metric.
- Late meals, heat and dehydration. All measurable, all modest, all confounders if your routine changes.
Measurement conditions, in order of importance
- Same window every time. Either your wearable’s overnight average, or a short morning reading — never a mix. Overnight values run higher because deep sleep has the highest vagal tone of the day.
- Same time of day. HRV has a clear circadian rhythm. First thing after waking, before you get up, is the standard.
- Same posture. Lying, seated and standing readings differ substantially. Pick one.
- Before caffeine, food and your phone. All three shift the number.
- Breathe normally. Slow deliberate breathing inflates RMSSD dramatically — great as a practice, useless as a measurement.
Raising the trend
Nothing raises HRV overnight except removing whatever suppressed it. Over months, the interventions with the best support are the boring ones: a large base of easy aerobic training, consistent and sufficient sleep, less alcohol, and a training plan where hard days are genuinely followed by easy ones. Our heart rate zone calculator gives you the zone-2 window that builds that base, and the max heart rate calculator helps you set it from a realistic ceiling rather than 220 minus your age.
Slow-paced breathing at around six breaths per minute reliably raises HRV during the session and appears to improve resting values over a few months. Aerobic fitness itself correlates with higher HRV, which is why the VO₂max calculator is a useful companion to this one — but the correlation is loose enough that you should never infer one from the other in an individual.
Finally, the honest boundary: HRV is a state signal, not a diagnosis. A single low reading means you slept badly or drank wine. A sustained, unexplained collapse in your baseline, particularly alongside symptoms, is a reason to see a doctor rather than to adjust your training plan.