Sleep — the operating system of development

Sleep isn't "downtime". During sleep, the baby's brain consolidates memory, prunes unused synapses, releases growth hormone, and calibrates circadian rhythms. Chronic sleep deprivation in adults is equivalent to drunkenness; in babies, it sabotages the very engine building the brain. On the other hand, baby sleep doesn't look like adult sleep — and trying to make it look like that is one of the main sources of unnecessary family suffering.

This pillar gathers what evidence says about how much sleep is normal by age, how to sleep safely, why awakenings are expected, and what the literature really shows about sleep methods.

1. Why sleep matters so much

Baby sleep has two peculiarities:

• More REM sleep than adults. About 50% of sleep time in the first months is REM (active sleep, with rapid eye movements), a proportion that drops to ~25% in adulthood. REM is associated with procedural memory consolidation and emotional processing. That's why babies wake more — REM is light by design.
• Growth hormone is released in pulses during deep sleep (NREM3). Chronic poor sleep affects somatic growth.

There's also declarative memory consolidation (words, faces, sequences) during sleep, cortisol regulation, and immune system maturation. It's no exaggeration to call sleep the operating system of development — when it fails, several other things fail with it.

2. How much sleep is normal — and the huge variability

The Iglowstein cohort at the University Hospital of Zurich followed children from birth to adolescence and established widely-cited reference values^{Iglowstein et al. 2003}:

The range is enormous. Healthy babies can sleep 11h or 19h and be within normal. Comparing your baby's sleep to a friend's almost never helps.

By 6 months, most can manage 5-6h overnight stretches, but a significant minority continue waking 2-3 times a night past the first year — and that's normal variation, not a clinical problem.

3. Safe sleep — the AAP 2022 consensus

The American Academy of Pediatrics 2022 guideline synthesizes what has the most evidence for reducing sudden infant death syndrome (SIDS) and accidental sleep death^{AAP 2022}:

• Back to sleep for every sleep (nap and night), through 12 months. The "back to sleep" campaign cut SIDS mortality by more than 50% in countries that adopted it.
• Firm, flat surface in the crib — rigid mattress, well-tucked sheet.
• Empty crib — no pillows, loose blankets, bumpers, toys, positioners. Nothing besides the baby.
• Same room, separate beds, ideally for the first 6 months (at least the first 4).
• No smoking, before or after birth. Tobacco is one of the largest modifiable SIDS risk factors.
• Breastfeeding reduces SIDS risk — the longer, the greater the effect.
• Pacifier at sleep onset (after breastfeeding is established) reduces SIDS risk by ~50-60%^{Hauck et al. 2005} — see the dedicated pacifier article.
• Up-to-date vaccination is associated with lower SIDS risk.

4. Awakenings are normal — not "failure"

Popular culture sells the idea that babies "should" sleep through the night. Neurobiology says otherwise: baby sleep cycles last 50-60 minutes (vs. 90 in adults), and at the end of each cycle there's a micro-awakening — a moment when the baby checks whether the environment is safe.

Babies who are fed (night feed), warm, and whose parent is nearby go back to sleep on their own. Babies in changed environments, hungry, or in the middle of a developmental leap may call — and that's biological design, not regression.

Waking 2-3 times a night through 6-12 months is within normal variation. The right question isn't "how do I make her sleep through the night?", but "how do I organize our sleep sustainably given her biology?".

5. Co-sleeping — an honest reading of the evidence

Co-sleeping (bed-sharing) is controversial because there are two seemingly contradictory bodies of evidence:

• Epidemiological studies show increased SIDS risk with bed-sharing, especially under 4 months.
• But Blair et al. (2014) re-analyzed with control for specific risk factors (smoking, alcohol, drugs, sofa, extreme tiredness): in absence of hazards, the risk in non-smoking parents who breastfeed, without alcohol/drugs, on adequate surface, is statistically small after 3 months^{Blair et al. 2014}.

AAP maintains the recommendation of separate beds in the same room — the safest position. But parents who choose bed-sharing after assessing risk factors and following safe protocols (La Leche League's Safe Sleep 7) are making informed, not negligent, decisions.

The truly unsafe scenario is the exhausted parent who falls asleep on the sofa with the baby — multiple times higher risk than a planned bed. If you're exhausted, the safer solution is to lie down with the baby on adequate surface, not try to resist on the sofa.

6. "Sleep training" — what the literature really shows

The American Academy of Sleep Medicine review by Mindell et al. (2006) consolidated 52 intervention studies and concluded that behavioral methods (extinction, graduated extinction, fading) improve sleep short-term in ~80% of families studied, with no documented adverse effect on development or attachment^{Mindell et al. 2006}.

But there are important nuances the literature also shows:

• Effectiveness ≠ obligation. Working doesn't mean every family needs to do it. Many babies "settle" without formal intervention.
• Age matters. Before 4-6 months, extinction methods are considered inappropriate by most researchers. The baby still lacks neurological maturation for self-regulation.
• Pure "cry it out" vs. graduated methods. Graduated methods (chair method, fading) have the same effectiveness in meta-analyses and generate less stress for parents and babies.
• Culture and context matter. Studies are predominantly in Western families with access to separate rooms. There's no evidence that babies in different arrangements (intentional co-sleeping, sleeping together longer) have worse outcomes.

The balanced reading: if sleep is sustainable for the family, there's no reason to "train". If it's unsustainable, behavioral methods are a legitimate tool — chosen without guilt, ideally after 6 months, with method compatible with the baby's temperament.

7. Tools with a narrow useful window

Two evidence-rich tools deserve their own articles — read alongside this pillar:

• Pacifier — strong SIDS protection (~50-60% reduction), with timing and weaning rules.
• White noise — real induction effect, but with strict caution on volume and duration.

Both follow the same principle: tool with a narrow useful window, not a panacea.

8. Screens and sleep

WHO is categorical: zero screens before age 2, maximum 1h/day between 2-5 years, always co-viewed^{WHO 2019}. An important reason is sleep: blue light from screens suppresses melatonin, and rapid visual stimulation raises cortisol and delays sleep transition. No screens at bedtime, at any age.

9. Practical synthesis

If you take 5 things from this pillar:

1. Sleep is active neural construction. Protect it like you protect breastfeeding or feeding.
2. Huge variation is normal. Your baby isn't a "bad sleeper" if she sleeps less than the neighbor's.
3. Safe sleep has a clear manual — back to sleep, empty crib, same room, no smoke, with pacifier after breastfeeding is established.
4. Waking is design, not bug. Welcome it without anxiety.
5. Training methods exist and work — but are only needed if current sleep is unsustainable. There's no prize for a baby who sleeps too early or ignores parents.