What an Apple Watch adds to sleep tracking

Per hour of sleep, not per hour in bed

Without a watch, an app has to assume that the time between starting and stopping tracking is time asleep. That inflates the denominator: an hour spent reading in bed makes every per-hour rate look better than it was.

With sleep stages from Apple Health, the index is calculated per hour of actual sleep, which is how clinical indices are defined. The difference is not cosmetic — on a night with a long time to fall asleep it can change the number substantially.

• Sleep stages: core, deep, REM and awake, with the times for each
• When you fell asleep and how long after tracking started
• How many times you woke, and how much of the night was awake time
• The index recalculated per hour of sleep rather than per hour in bed

Blood oxygen, as a second opinion

Sound is one kind of evidence. Blood oxygen is another, and it is independent of what the microphone heard. If both point the same way on a night, that agreement is worth more than either signal alone.

Apple Watch samples oxygen at intervals during sleep rather than continuously, so it will miss individual events. What it is good for is the shape of the night: whether readings cluster comfortably in the high nineties or repeatedly drop towards the 90% line.

Position is the finding that changes behaviour

The wrist accelerometer can estimate whether you are on your back, on your side or face down, and how often you turn. In practice this is the single most actionable thing a watch contributes to snore tracking.

"You spent 73% of the night on your back" is a sentence people act on, because the fix is cheap and immediate. It also lets the app's advice be specific: positional measures are recommended when your own position data supports them and held back when it does not.

A nudge during the night, with a budget

Reporting in the morning is useful. Acting during the night is different. When snoring builds into a sustained episode, the watch can deliver a gentle haptic tap — enough to prompt a change of position without waking you fully.

Two things keep this honest. There is a nightly budget so the watch cannot buzz all night, and the outcome is measured: Loir records the sound level before and after each nudge and reports how many episodes actually resolved and by how many decibels. If it is not working for you, the numbers will say so.

• Three escalating intensity levels, with a limit per night
• Off entirely, or restricted to snoring or to breathing events
• Before and after sound level recorded for every intervention
• A resolved rate and an average drop in decibels shown in trends

Heart rate, variability and wrist temperature

Heart rate and heart rate variability during sleep give context about recovery rather than about the airway. They go into the night score as small components with small ceilings, because they matter but they are not what the app is measuring.

Wrist temperature is reported against your own baseline rather than as an absolute number, because the absolute value differs between people. A deviation from your baseline is the informative part — it moves with illness, alcohol and the menstrual cycle.

Everything still works without a watch

This is worth stating plainly: Loir is an iPhone app that records and analyses sound. Snoring, breathing pauses, cough and speech detection, the night score, trends, remedies and the doctor report all work with the phone alone.

The watch adds context and one action. Sections that need watch data simply do not appear when there is none, and nothing in the app pretends to have measured something it did not.

Automatic start is optional too: if you enable it, the watch can begin tracking when it detects that you have fallen asleep, so you do not have to remember to tap anything.