Sleep & heart · topic 4 of 5
Resting heart rate
Measured against your own normal, not a population range
Against yourself
Against the level
The tile takes whichever of the two reads worse. Below your usual is never coloured, because a rested morning and an endurance athlete both land there.
Video walkthrough — Resting Heart Rate
In production, coming soon
Verve compares today's resting heart rate with the median of your own last 30 days. Within 5 bpm is ordinary day-to-day variation; around5 bpm above, sustained for a few days, is worth noticing; around10 above is a clear departure. The instrument shades only the high side, because a low resting rate is not a finding — a rested morning and an adapted endurance athlete both land there.
The level, and what a reading cannot say.Beside the comparison with your own normal, the level itself matters: lowest risk sits around 50–60 bpm, and above 75–80carries higher risk even when it never moves. Under 50 is common in trained people and usually adaptation, worth raising only alongside dizziness, fainting or breathlessness. The slow direction counts on its own — a baseline falling over months tracks lower risk, a rising one higher, whatever today says. Alcohol, heat, a late meal, poor sleep, illness, dehydration, thyroid and some medicines all raise it; a wearable reading is not diagnostic.
Radin et al — Wearable data and population health, Lancet Digital Health 2020
A population-based study using resting heart rate and sleep data from Fitbit wearers across several US states. Deviation from an individual's OWN baseline — not from a population range — improved real-time state-level prediction of influenza-like illness. The principle Verve borrows is the comparison, not the diagnosis: several beats above your own normal, sustained across days, is a signal that any single reading inside 60–100 bpm can never be.
doi.org →
ARIC — Heart rate change and mortality, JAMA Cardiology
The ARIC cohort showed that the DIRECTION of change matters on its own. Any increase in resting heart rate from a prior visit tracked progressively higher all-cause mortality and incident heart failure, while decreases — particularly over 12 bpm — were protective. The lowest-risk range sat at 50-60 bpm, with risk rising progressively above 60-66. This is why Verve leads with movement against your own baseline rather than a single number, and why it now also says where that number sits in absolute terms.
doi.org →
Why a personal baseline, and why the median
CommentaryThe usual healthy range spans roughly 60 to 100 bpm — wide enough that someone who normally rests at 52 can wake 11 beats up and still sit comfortably inside it. Verve compares you with the middle of your own last 30 days. It uses the median rather than the mean so that one feverish night, one late flight or one heavy evening cannot drag the very line it is being judged against, and it excludes today's reading from its own baseline, which would otherwise shrink every deviation it is meant to reveal.
doi.org →
Why the absolute number is shown as well
CommentaryDeviation alone is not enough. The association between resting heart rate and all-cause mortality, cardiovascular mortality, myocardial infarction, heart failure and stroke is graded and near-linear, independent of traditional risk factors — above 75-80 bpm carries markedly higher relative risk than below 60, with the steepest gradient for sudden cardiac death. So someone steady at 82 bpm was reading 'at your normal' in green, which was true and beside the point. Verve now reports both. The bands are a reading aid for a continuous relationship: nothing changes at 75 that did not apply at 74, and no band is a diagnosis. Under 50 is left uncoloured — it is ordinary in trained endurance athletes, and flagging it would alarm the fittest readers.
doi.org →
What Verve reads, and what it cannot settle
CommentaryVerve uses Apple Health's own daily resting heart rate, derived from heart rate during detected stillness and weighted toward sleep and early-morning quiet — not an average of raw samples, which would fold in activity, stress and digestion and sit well above true resting physiology. Apple, Fitbit, Garmin and Whoop each window differently, so absolute values are not comparable between brands; within-device trending for one person is the reliable part. An isolated elevated reading is not diagnostic. A tachyarrhythmia question needs a 12-lead ECG or ambulatory monitoring, and a seated manual reading after 5-10 minutes of rest remains the clinical reference. An elevated trend is also often not cardiac at all — infection, dehydration, anaemia, thyroid overactivity, medication effects, anxiety, pain, POTS or pregnancy, with motion artefact and device error to exclude first.
doi.org →