The six-month sleep regression

Around the twenty-fifth week, our daughter forgot how to sleep. She had been doing six-hour stretches. Suddenly she was up every ninety minutes, cheerful at 2am, furious at 4am, and refusing the bottle that had worked the night before. I sat on the bedroom floor in my underpants holding a sleeping bag and tried to remember what year it was.

This is the six-month regression. It is real. It is well-named in the sense that it does happen around six months. It is badly-understood because the word "regression" makes it sound like the baby has gone backwards, when in fact they have leapt forward and their sleep is the casualty.

The good news is it ends. The bad news is "ends" means three to four weeks. The thing nobody tells you in advance is that this is the first time you and your partner will have to make a real, joint decision about sleep, and the decision is more about your relationship than it is about the baby.

What is actually happening

Three things converge between five and seven months and the regression is the collision of all three.

Sleep cycle maturation. Until around four months, babies sleep in two crude states: deep and light. At four to six months their sleep cycles mature into the adult pattern of stages, with brief partial wakings between cycles. An adult does this and rolls over. A baby does it and notices, with great alarm, that conditions have changed since they fell asleep. If they fell asleep being rocked and they wake on a mattress, that is a problem requiring management.

Developmental leap. Around six months babies are doing big motor work. Sitting unaided, beginning to commando crawl, working out object permanence. The brain is firing in new ways and the baby is, frankly, too excited to sleep. They wake up and want to practise sitting. At 3am.

Separation anxiety begins. This one most parents miss. Object permanence kicks in at around six to seven months and the baby works out, for the first time, that you continue to exist when they cannot see you. This sounds like progress, and it is, but the immediate side effect is that being alone in a cot becomes a distinct event with emotional weight. They cry not because they are uncomfortable but because they have noticed your absence.

You can address each of these. You cannot make them happen faster.

What does not fix it

Every well-meaning relative will tell you to give the baby more solids before bed. This is one of the most durable myths in parenting and the data does not back it up. Studies on infant sleep and solids consistently find no relationship between calorie intake at bedtime and night waking after six months. The baby is not waking because they are hungry. They are waking because their sleep architecture changed and they need to learn to bridge the wake-ups.

A heavier solids meal at 5pm sometimes makes things worse, because reflux is more likely and a baby with a full belly lying down is not a recipe for calm. Stuffing them with food before bed will not buy you sleep. It might cost you some.

What also does not fix it: a darker room (yours is probably already dark), a different white noise machine (it is fine), buying a new sleep sack (it is also fine), or moving them to their own room if you have not already (this can help marginally but it does not solve the underlying issue).

What actually works, and the marriage discussion

Here is where it gets uncomfortable. By month six you and your partner have to talk about sleep training and you probably disagree.

The three broad approaches:

• Cry-it-out (full extinction). Put the baby down awake. Do not return until morning unless something is genuinely wrong. The evidence base on this is large and the outcomes are good (fast resolution, no documented harm in the medical literature), but the experience for the parents is brutal and many couples cannot do it without one of them crumbling.
• Modified extinction (Ferber, controlled comforting). Put the baby down awake. Return at increasing intervals (3, 5, 10 minutes) for brief check-ins. Do not pick up. Comfort with voice, leave again. Slower than full extinction by a few days, easier on the parents, similar end state.
• Fading / gradual / no-cry. Slowly reduce intervention over weeks. Sit by the cot, then move further, then leave the room. Two to four weeks of work. Gentlest in the moment. Hardest to stay consistent on.

There is no objectively correct choice. There is a choice that fits your nervous system and your partner's, and a process for landing on it.

The conversation that has to happen, and most couples skip:

• Which of these can each of you actually do without resentment?
• What is the threshold at which one of you would override the plan? (Define it before you start, not at 2am.)
• Who is on duty for which nights? (Solo nights are easier than tag-team for consistency.)
• What is the time horizon? (Two weeks. If you are not seeing improvement in two weeks, pause and reassess. Do not run any approach for six months on hope.)

Do not start sleep training and bail out three nights in. Inconsistency is worse than any of the three approaches done with commitment. The baby learns "if I cry long enough, the rules change" and the next attempt is harder.

What it looks like when you are in it

Week one: the baby cries more than the day before, you both feel sick, you question every decision you have ever made. Sleep is somehow worse. This is normal. You are in the steepest part of the curve.

Week two: cries get shorter. The 11pm wake disappears first. The 4am wake is the last to go. You start to recognise the cry that means "settle yourself" versus the cry that means "I genuinely need you", and you stop responding to the first.

Week three: most nights they sleep through, with maybe one wake. You and your partner have one or two date conversations that do not involve the baby and you remember that you are friends.

Week four: settled. Mostly. There will be other regressions (eight months, twelve months, eighteen months, two years), but the baseline is restored.

If you are at week three and things are not better, talk to your maternal and child health nurse. There are physical causes (silent reflux, sleep apnoea, ear infection) that can mimic regression and they need to be ruled out by someone with a stethoscope.

The body of it, the body metaphor

I think of sleep training the way I think of physiotherapy after an injury. It hurts in the moment. It works through repetition. The pain is information, not damage. You stop being able to tell whether you are in week two or week three, but the body keeps adjusting in the background. One morning you wake up and realise you have not been counting hours.

The bullet list every household needs taped to the fridge

• ONE plan, agreed by both parents, in writing, before night one.
• A two-week minimum commitment, no bail-outs after three nights.
• A pre-set escape clause: what events trigger a stop and reassess.
• A division of nights so consistency does not become resentment.
• A no-judgement debrief in the morning, ten minutes, decaf coffee.
• A back-up plan for nights when one of you is sick or away.
• A written log (a phone note will do) of wake times and durations so you have data, not vibes.

Closing

The regression is real. The fix is not more food. It is sleep architecture meeting parental consistency, with separation anxiety as the third party in the negotiation. Pick an approach. Pick a partner-aligned threshold. Hold the line for two weeks.

Settle the baby. Settle the marriage. Sleep arrives.