Pediatrics · Part Four · What Families Notice and How Care Works
Lesson 44 / 57
Sleep in the Early Years: How Much a Normal Child Actually Varies
Across 5,107 children, the most striking feature of the sleep charts was the sheer variation at every age in an entirely normal population.
Infant sleep consolidates gradually as the circadian system and arousal networks mature. In a longitudinal cohort of 5,107 children, 24-hour sleep fell from a mean peak of 14 hours at 4 to 6 months to 10 hours at 9 years. The authors describe huge variation at every age in a normal population. A systematic review of 74 papers found reference values disagreeing widely while the direction of change was consistent. The Unified Model of Tone reads sleep as tone on a schedule.
Mean 24-hour sleep at 4 to 6 months
14 hours
Mean 24-hour sleep at 9 years
10 hours
Children in the cohort
5,107
Papers in the reference review
74
Sleep consolidation
Consolidation is the gradual gathering of sleep into longer blocks, mostly at night. A newborn sleeps in short stretches distributed around the clock. Over months the longest single stretch lengthens and shifts toward darkness. That shift, rather than total hours, is what parents experience as improvement.
The circadian system
The circadian clock is a hypothalamic pacemaker that anchors daily rhythms to light. It is present at birth but not yet entrained, which is why a newborn has no day-night preference. Entrainment develops over the first months through light, feeding and the rhythms of daily handling.
01How much children actually sleep
The variation is larger than the average
Population data give a clearer picture than any single recommendation. Price and colleagues drew on the Longitudinal Study of Australian Children, with 5,107 children recruited at 0 to 1 years, using prospective 24-hour time-use diaries. Mean 24-hour sleep fell from a peak of 14 hours at 4 to 6 months to 10 hours at 9 years (Price 2014).
The composition of that decline matters. Mean sleep onset moved later, from 20:00 to 21:00, and day sleep fell from 3.0 hours to almost nothing. Most of the reduction in total sleep is naps disappearing rather than nights shortening.
The finding the authors themselves emphasized
They describe the most striking feature of their centile charts as the huge variation at all ages in sleep duration, sleep onset time and especially wake time, in a normal population. That is the sentence worth carrying into any conversation about whether a child sleeps enough.
02Findings
What the research shows
The figures below come from population sleep data and from a systematic review of reference values.
03Why quoted norms disagree
The direction of change is more reliable than any number
Published sleep norms vary widely, and knowing why prevents a lot of unnecessary worry. A systematic review of infant sleep-wake behavior across the first 12 months included 74 papers and found high discrepancy between studies in reference values (Dias 2018).
The same review found more congruence regarding changes than values, particularly in the first six months. Sleep duration decreased across the year, night wakings decreased, bedtime moved earlier, and the longest single sleep period at night lengthened.
The same lesson the milestone data teach
This is the pattern that runs through the whole section. The WHO milestone windows are months wide while the sequence holds across populations (WHO 2006). Infant crying varies enormously between countries while the decline after 8 to 9 weeks stays consistent (Wolke 2017). Direction is dependable. Absolute values are not. The developmental sequence makes the same case for movement.
04What is being built
The clock, the arousal networks and the brain they serve
A baby practices sleeping before anyone holds her. Fetal recordings late in pregnancy show alternating cycles of movement and stillness, the first crude version of an active state and a quiet state taking turns. The brainstem and hypothalamic structures that will later run arousal are already laying down their wiring.
Sleep then consolidates because those systems keep maturing, and the scale of what is being built over those months is easy to underestimate. Total brain volume rises 101 percent across the first year and cerebellar volume 240 percent (Knickmeyer 2008). Synaptic density in auditory cortex peaks near three months while the middle frontal gyrus waits until after fifteen (Huttenlocher 1997).
The clock and its two signals
The suprachiasmatic nucleus is the small hypothalamic pacemaker that anchors the daily round. It is present at birth but not yet entrained, which is why a newborn distributes sleep around the clock without preference. Entrainment arrives as that nucleus becomes responsive to light through the retinohypothalamic tract and the pineal gland takes up its own nightly release of melatonin.
Cortisol carries the opposite signal, rising toward morning to drive waking, and its rhythm matures alongside melatonin. The axis producing it is highly responsive in early life. That is one concrete reason a calm, predictable, softly lit evening does real work, since bright light and irregular routines reach the same clock.
Feeding carries timing information
The entraining signals are not only light. Breast milk composition follows a daily rhythm, with higher cortisol and activity-promoting amino acids during the day and higher melatonin and tryptophan at night (Akanalci 2024). A feed delivers timing information alongside nutrition, which is one reason feeding pattern and sleep pattern are hard to separate in a young infant.
05The shape of a night
A newborn night has its own architecture, and the brainstem runs it
Newborn sleep is not adult sleep in miniature. It alternates between active sleep, marked by fluttering eyelids, irregular breathing and small twitches, and quiet sleep, which is still and regular. Those two states plus wakefulness are the whole repertoire at the start.
The circuitry producing them sits low in the brain. The infant brainstem alone contains a fundamental sleep-wake circuit sufficient to produce transitions among wakefulness, quiet sleep and active sleep, and that circuit supports rapid transitions between states (Blumberg 2014). A drowsy newborn can surface fast because the switch she is using is built to move fast.
What consolidation actually is
Consolidation is not the brainstem getting better at its job. Later in development the interactions between brainstem and forebrain strengthen, and they run in both directions. They contribute to the consolidation of sleep and wake bouts, to sleep homeostatic processes, and to the emergence of diurnal or nocturnal rhythms (Blumberg 2014). Longer blocks are the visible result of more traffic between two levels.
The population data show the same shift from the outside. The systematic review above found night wakings decreasing across the first year and the longest night sleep period lengthening, particularly across the first six months (Dias 2018). Those two changes together are what families experience as a baby starting to sleep through.
Why the phrase is misleading
Sleeping through the night is not a single achievement a child either has or lacks. It is a gradual lengthening of the longest block, sitting inside a population whose variation at every age is described by researchers as huge (Price 2014). A child at the shorter end of that distribution is not failing at anything.
Waking between cycles is a normal feature of sleep at every age, including adult sleep. What changes with maturity is the ability to return to sleep without full arousal, and that ability is built rather than issued.
06In the office
An evaluation reads posture, head turning and how a baby comes back down
Families bring an unsettled infant for a chiropractic evaluation with one question underneath the others. Why does this baby surface so hard, and stay up so long, when the waking itself is ordinary? The assessment begins there rather than with a technique.
What gets read is the body a nervous system is arousing and settling inside. Resting posture. The symmetry of the primitive reflexes against age. The resting curve of the neck, how far the head turns to each side, and which way it drifts at rest. Then palpation over the occiput, the upper cervical tissues and the cranial base.
The contact that follows is a sustained, light hold, no more than the pressure a person could comfortably rest on a closed eyelid. Force is graded to the age and size of the child. There is no twist, no thrust and no audible release, and a settled baby very often stays asleep through the visit.
Why the model expects that input to matter
A newborn is soothed from outside long before she can soothe herself, and the channels doing the soothing are physical ones. Rocking, holding, feeding and steady contact deliver vestibular and proprioceptive input to the same brainstem networks that set arousal. The vestibular system follows that route in full.
The brainstem circuit that produces those state transitions sits low in the head, directly beneath the upper neck (Blumberg 2014). That region carries some of the densest position information in the body. The model holds that crowded or one-sided information from it is read as a state rather than as a local complaint. A system spending effort to hold a pattern has less to spend on coming back down.
The timetable is what makes this window interesting. Settling and sleeping consolidate over the same weeks, with mean fussing and crying falling to 68 minutes a day by 10 to 12 weeks (Wolke 2017). The model expects sleep to be among the first readings to move when the organization underneath it changes.
07Safe sleep
Where a baby sleeps is settled guidance
Sleep position is not an open question. Babies sleep on the back, on a firm flat surface, without loose bedding. Nothing in the developmental data qualifies that.
The developmental data are sometimes misread on this point, so it is worth stating clearly. A study of 351 healthy term infants did find that prone sleepers reached several motor milestones earlier than supine sleepers (Davis 1998). Every infant reached every milestone within the normal range, walking age did not differ, and the authors state directly that the finding is not a reason to abandon safe sleep recommendations.
Where that experience belongs
Supervised awake time on the stomach, which the same literature notes is the most commonly overlooked half of the advice families receive. Tummy time and the curve covers what that position builds and why symmetry matters more than minutes.
08The model’s claim
What the Unified Model of Tone predicts about sleep
The sleep figures above come from population data and systematic review. What follows is this model’s reading of them, stated as ours rather than drawn from the papers cited.
Tone is carried through oscillation, coupled across scales. A living body is a nested set of rhythms, from heartbeat through breath to the day-long circadian round. A healthy body holds them in phase, each rhythm supported by the ones above and below it. The model reads sleep consolidation as the slowest rhythm in that stack finding its phase: not a behavior a child acquires but the clearest visible expression of how well the whole organization is timed. The model also takes the night shift literally. Sleep is when the glymphatic system flushes the brain’s metabolic debris most actively, so the rhythm being consolidated is also the rhythm the housekeeping runs on. A system that can shift cleanly between arousal and rest, on a rhythm matched to the day, is displaying the flexibility the model calls health.
The prediction
From that follows a claim the sleep literature does not make. The model predicts that sleep consolidation will rarely be independent of the other readings of the same organization, though compensation can hold any single reading steady. A child whose sleep is poorly consolidated should also differ on settling, feeding coordination, autonomic measures and postural symmetry, because these are readings of one variable rather than four. In canonical terms, the return to sleep after a night waking is recovery time, autonomic variability is variability structure, and the phase relation between feed rhythm and sleep rhythm is cross-frequency coupling read directly.
That prediction is consistent with what the developmental data already show, since crying, feeding and sleeping all consolidate over the same weeks. Sleep also changes more across childhood than anything else here. Mean 24-hour sleep falls from 14 hours at 4 to 6 months to 10 hours at 9 years (Price 2014). A range that wide makes it a sensitive readout of the organization underneath.
The model adds one more expectation about where that organization can be reached. It treats the upper neck as a dense source of position information feeding the brainstem circuit that switches these states, so it expects head rotation range and settling to travel together in the same infant.
This is a claim about organization rather than about treatment, and safe sleep guidance is not negotiable. What it claims is that better-organized tone yields greater adaptive capacity, whichever appropriate input delivered the useful information. If sleep consolidation, settling, feeding coordination, autonomic variability and postural symmetry are shown to move together, the unification claim is confirmed.
09The tone reading
How sleep expresses tone
Every topic in this library expresses all of tone. In sleep three aspects carry the signature, because a system moving from 14 hours at 6 months to 10 at 9 years is retiming itself continuously.
Oscillation
Sleep is the clearest rhythm a body runs. Consolidation is the daily cycle gathering into one long block matched to darkness.
Time course
Night wakings fall and the longest sleep period lengthens across the first six months. Consolidation is a schedule rather than a skill.
Set point
Arousal returns toward a baseline after each waking. How readily it does that is what parents experience as settling.
The remaining foundations run through sleep as well. Coupling: feeding and sleeping entrain each other, and milk itself carries a daily rhythm. Input quality: light, feeding and predictable handling are the signals a clock is set by. Gain: how strongly an ordinary noise rouses a child is set centrally rather than at the ear. Prediction: a predictable evening lets a nervous system anticipate the shift rather than resist it. Load: building a brain at this rate is metabolically expensive, and sleep is when much of it happens. Constraint: a baby who is uncomfortable in most positions has fewer ways to settle. These are readings of one organization rather than separate systems, which is the core claim of the Unified Model of Tone.
10Across the library
How this page relates to the rest of the library
Sleep consolidates alongside settling and feeding, over the same weeks.
Crying data from 8,690 infants, and the same lesson about direction being more reliable than any number.
The feeding side of the same weeks, with prevalence figures that follow a similar curve.
Where prone experience belongs, and why safe sleep and tummy time are separate decisions.
The same principle in movement: order holds across populations while dates vary by months.
The regulatory machinery behind settling, being built over exactly this period.
Sleep as a measurable state across the lifespan, with the instruments used to read it.
The foundation this page leans on hardest. Why rhythm is how a regulated system keeps time.
11Frequently asked
Questions families ask about infant and child sleep
How much sleep does my child actually need?
Less precisely than most charts imply. In a cohort of 5,107 Australian children, mean 24-hour sleep fell from a peak of 14 hours at 4 to 6 months to 10 hours at 9 years. The researchers describe the most striking feature of their centile charts as the huge variation at all ages in sleep duration, onset time and especially wake time in a normal population. A single target number does not describe that.
Why do published sleep norms disagree so much?
Because they are measured differently. A systematic review of infant sleep-wake behavior across the first 12 months included 74 papers and found high discrepancy between studies in reference values. The differences came from different measures, ages, self-report versus direct observation, and countries. The same review found much more agreement about the direction of change than about the numbers themselves, which is the part worth trusting.
When will my baby sleep through the night?
It is a gradual lengthening rather than a night it happens. The systematic review above found night wakings decreasing across the first year and the longest single night sleep period lengthening, especially over the first six months. Brief arousals between sleep cycles are normal at every age including adulthood. What develops is the ability to return to sleep without fully waking, and that develops rather than being present from the start.
Does feeding affect my baby’s body clock?
It appears to carry real timing information. Breast milk composition follows a daily rhythm, with higher cortisol and activity-promoting amino acids in daytime milk and higher melatonin and tryptophan at night. A feed therefore delivers a time-of-day signal alongside nutrition. That is one reason feeding pattern and sleep pattern are so difficult to separate in a young infant, and why feeding on demand tends to settle both together.
Can my baby sleep on the stomach if she sleeps better that way?
No. Babies sleep on the back, on a firm flat surface, and nothing qualifies that. The study often cited on this point did find prone sleepers reaching several motor milestones earlier. Every infant still reached every milestone within the normal range and walking age did not differ. The authors state directly it is not a reason to abandon safe sleep recommendations. Prone belongs in supervised awake time only, never in sleep.
Is my child a bad sleeper, and what would an evaluation involve?
The variation in normal populations is wide enough that most children described as bad sleepers sit inside it. Snoring or pauses in breathing, daytime sleepiness in an older child, a sudden change in an established pattern, or disrupted sleep alongside poor growth are the specific signals worth raising with a pediatrician. An evaluation here reads resting posture, head turning to both sides and the symmetry of the primitive reflexes. The contact that follows is a sustained light hold at the upper neck, graded to the child, with no twist and no thrust.
What does the Unified Model of Tone say about sleep?
That sleep is the clearest visible expression of how well the whole organization is timed, rather than a behavior a child acquires. From that the model predicts sleep consolidation will rarely be independent of settling, feeding coordination, autonomic measures and postural symmetry, because it treats them as readings of one variable, with compensation deciding which of them shows the change. Crying, feeding and sleeping consolidating over the same weeks is the shape of evidence it expects.
12The sources
References
9 primary sources, each linked to its PubMed record. Figures quoted on this page were checked against the published abstract.
Related evidence