Everyone wakes at night, adults included. Sleep runs in cycles of 45 to 90 minutes, and a brief arousal at the end of each cycle is built into the design. The difference between a child who "sleeps through" and one who doesn't is rarely the waking itself: it's whether the child can slide into the next cycle without help. That's why so much of night sleep traces back to how sleep starts at bedtime.

A sleep association is whatever conditions your child falls asleep with: feeding, rocking, a pacifier, your presence, motion, or their own bed and thumb. Associations aren't good or bad, they're simply what the brain checks for at every cycle-end arousal. When the conditions changed since falling asleep (the milk is gone, the arms are gone), the brain sounds the alarm. A child whose bedtime conditions are still there at 2 AM resettles without anyone knowing they woke.

As a general shape: newborns feed around the clock, babies of 4 to 6 months commonly still need 1 to 3 night feeds, many babies of 7 to 9 months are down to 0 or 1, and most healthy babies past 9 to 12 months can get their full nutrition in daytime hours. These are patterns, not rules: growth, size, and feeding method all shift the picture, and your doctor has the final word on whether your child still needs night calories.

A split night is a long waking of an hour or more in the middle of the night where your child is calm, often cheerful, and simply awake. It happens when sleep pressure runs out: too much daytime sleep or a too-early bedtime means the tank empties partway through the night, the body treats it as a long nap, and refills enough pressure to finish the night later. The fix lives in the schedule, not in the night itself.

The night has a structure, and where the wakings fall is a clue. The first half of the night is dominated by deep sleep, so distress there points toward overtiredness, which fragments deep sleep. The early morning hours run on light sleep and almost no sleep pressure, so wakings there point toward schedule and light exposure. Wakings at tidy 1-to-2-hour intervals all night trace the sleep cycles themselves, the signature of an association check failing at each arousal.

No. There are many roads to independent sleep, from very gradual (shrinking the help step by tiny step over weeks) to faster approaches, and research supports a range of methods when applied consistently. What matters most isn't which method a family picks, it's picking one fitting their values and staying consistent with it, because switching approaches mid-week teaches a child nothing except persistence. The right method is the one you can follow through on.

Real nights are often a blend: an association plus a schedule drift, or a regression on top of a habit. Untangling which thread to pull first is the actual skill of this work, and it's hard to do from inside a sleep-deprived household. Hit reply to the email where you received this link and tell me what your nights look like. Helping families sleep is what I do.