Nightly HRV vs the morning spot check
Journal · 2026-09-15 · R. Klug · All notes
Read in: Deutsch · Français · Español · Italiano
Two measurement traditions coexist in HRV tracking. The older one: wake up, lie or sit still, record two to five minutes, log the value. The newer one: wear a sensor through the night and let the whole sleep period be the measurement. Both have serious advocates. Here is what the published comparisons actually found — numbers included, because the two protocols are not as interchangeable as they look.
Reliability: the night is steadier
Nuuttila and colleagues (2022) reported intraclass correlations of 0.92–0.97 for nocturnal lnRMSSD, against roughly 0.82 for a morning supine measurement. Day-to-day variation tells the same story: around 8% for a supine morning reading — and about 16% standing, which is why orthostatic protocols produce the noisiest series of all. A full night simply averages over far more data than a three-minute window, so random noise shrinks.
The most direct comparison so far
The most direct comparison is Nuuttila et al. (2024, Sports Medicine – Open): 24 recreational runners, two weeks of deliberate training overload. Morning measurements — supine and standing — showed no significant change through the overload. The nocturnal recordings did respond, and only the nocturnal metrics correlated with how performance changed. One study, a modest sample, one protocol — those limits are real. It is nonetheless the most direct test published so far, and within them it favours the night.
Notably, in the companion analysis it was nocturnal heart rate — the plainest number in the file — that separated overloaded athletes best, alongside a simple subjective readiness question. A useful reminder that HRV is one signal in a record, not the record itself.
The morning case, stated fairly
Advocates of the morning protocol — Marco Altini most prominently — raise points the night cannot fully answer: an evening session depresses the following night, so a nightly value partly describes yesterday’s training rather than today’s state; nighttime artifacts cannot be corrected by repositioning a strap; and a seated measurement adds a mild stressor that avoids parasympathetic saturation. These are legitimate mechanisms, and a disciplined morning routine remains a defensible protocol. What the morning has not yet shown is an outcome study where it detected a training state change that the night missed.
What both protocols demand
- Trends, not days. A single day correlates barely at all with training state; rolling baselines over several days are where signal appears (Plews et al.). ampulse compares each night against your own multi-week history, never against population norms.
- Artifact control. Bourdillon et al. (2022) showed a single uncorrected artifact can inflate RMSSD by several hundred percent. This is why every ampulse HRV window carries a quality gate, and why the percentage of valid beats is shown next to the number.
- Same method, always. Values from different protocols, devices or statistics are not comparable in absolute terms. Pick one measurement and stay in it.
What a nightly value is — and is not
A quality-gated nightly average, tracked against your own baseline, is a description of how your autonomic state moved. It is not a verdict: a higher value does not certify recovery (RMSSD rises in positive adaptation and in early overreaching — Bellenger et al. 2016), and a low morning is a reason to look at the whole record, not an instruction. ampulse takes a side only on the measurement question: record the full night, window by window, gate on quality, keep the intervals — and leave the judgment where it belongs, with you.