How Infant Sleep Is Measured, And Why Numbers Disagree

Sleep figures quoted for babies vary substantially between sources, and parents comparing their own child to them find no consistent target. The variation comes largely from how sleep was measured.

Parent report is the most common source

Most large studies rely on parents recording or estimating sleep, usually through diaries or questionnaires. This is cheap, scales well and covers the home environment.

Its weakness is that parents cannot observe brief awakenings, and a baby who wakes and resettles independently is recorded as having slept through. Reported night waking therefore measures waking that disturbs the household.

Estimates given from memory also compress and round, and tired people are not precise. Diaries kept in real time perform better than recall, though they still depend on somebody noticing that a waking occurred.

Movement-based measurement sees different things

Wearable devices infer sleep from the absence of movement, which captures brief awakenings that parents miss. Studies using them consistently report more night waking than parent report does.

The trade-off is that stillness is not the same as sleep. A baby lying quietly awake is often scored as asleep, and the algorithms differ between devices.

This is one reason consumer sleep trackers and research-grade devices produce different numbers from the same night. The threshold each one uses to call stillness sleep is a design choice rather than a fact.

Laboratory measurement is precise and unrepresentative

The reference method records brain activity, eye movement and muscle tone, and is the only way to identify sleep stages properly. It defines what the other methods are approximating.

It also requires equipment attached to a baby in an unfamiliar setting, which changes the sleep being measured. Sample sizes are small for the same reason.

Findings about sleep cycles and stages come from here, while findings about how much babies sleep at home mostly do not.

What follows for a parent

Ranges quoted in guidance are wide because the underlying variation between babies is wide, and the measurement disagreement widens them further. A single average is close to meaningless.

Comparing a child against a number from an unspecified source is therefore unhelpful, and comparing against another family's account is worse, since that account is parent report at its least precise.

Sleep that is causing genuine difficulty for a family is worth discussing with a health visitor or GP regardless of how it compares to any published figure.