Infant sleep has no single norm

The ranges published by professional bodies, on the order of twelve to fifteen hours per twenty-four before the age of one, are set by committees synthesising observational studies. They describe a span judged compatible with ordinary development, not a target to hit.

These documents almost always state their own level of evidence, generally described as low. That mention disappears in popular coverage, where the range becomes a norm and its lower bound an alarm.

Studies that publish percentiles rather than averages reveal considerable spread: at the same age, total daily duration can vary by several hours between the tenth and the ninetieth percentile, with nothing else distinguishing those children.

The same spread holds for the number of night wakings and for the age at which nights become continuous. An average computed over that distribution describes a situation many children never experience.

Most large datasets rest on parental report, by diary or questionnaire. Comparisons with actigraphy, which records wrist movement, show systematic gaps: parents do not perceive brief wakings, and overestimate total duration.

Actigraphy has limits of its own: it codes stillness as sleep. Polysomnography, the only reference method, is carried out in a laboratory on small samples and over one night, which is not an ordinary night.

Each method therefore answers a slightly different question: the questionnaire tells what parents perceive, actigraphy what the body does, polysomnography what the brain does. The figures in circulation almost always come from the first, without the source being stated.

Cross-country comparisons find substantial gaps in reported duration and bedtime, with shorter durations and later bedtimes in several East Asian countries than in English-speaking ones.

Those gaps can be read two ways. As real differences in sleep, or as differences in norms, hence in reporting: what parents count as a waking, as a nap, as a bedtime, is not defined the same way everywhere.

Bed-sharing, an ordinary practice across much of the world and a minority one in other countries, also changes what is observed: a waking followed by a feed in the same bed is not always experienced, or recorded, as a waking.

The gradual consolidation of sleep over the first months is a robust fact, found everywhere: the share of night sleep rises and episodes lengthen, independently of practices.

What is not established is the diagnostic value of a departure from the average. Studies following the same children over time show marked instability of individual trajectories, which makes reading a single measurement hazardous.