The four-month sleep change, and why regression is the wrong word for it

Everything here earned its place by changing an outcome. Nothing about the four-month sleep change is included to round the number up.

The wrong word for the right observation

Parents widely report a period somewhere around the middle of the first year when a baby who had been settling reasonably starts waking far more. The observation is real and the label is misleading, because nothing is going backwards and no ability has been lost. What is described as a regression is generally a maturation: the sleep itself is changing structure into something more adult-like.

Calling it a regression makes parents look for what they did wrong, which is both distressing and pointing in the wrong direction. Understanding it as development does not produce more sleep, but it stops the fortnight becoming an accusation.

What actually changes

Newborn sleep has a simpler structure with a large proportion of the light, active kind and relatively little differentiation between stages. Over the first months that structure becomes more organised, with clearer stages and more distinct transitions between them. More distinct transitions mean more points at which a baby can surface, and surfacing is what a parent experiences as waking.

Once feeding settles, a baby who previously drifted through those points now arrives at them properly awake and notices where it is. That is the mechanism, and it explains why the change often shows up as waking every cycle rather than once or twice.

The timing is not fixed

The four-month label is a convenient handle rather than a schedule, and reported timing spreads across a considerable range. Some babies show a clear episode, some show a gradual shift with no obvious start, and some parents notice nothing in particular. Babies born early are usually considered against their adjusted age, which shifts the whole picture by however many weeks.

A baby who was never sleeping in long blocks may not appear to change at all, since there was no consolidated pattern to disrupt. None of these versions is a better or worse outcome, and there is nothing to conclude from which one you get.

Why it often coincides with other things

The same months tend to contain rolling, more active daytime engagement and in some babies the beginnings of teeth, which makes attribution hard. Parents naturally reach for the most visible explanation, and teething in particular gets blamed for a great deal it may have nothing to do with. Because several things overlap, the honest answer is usually that you cannot tell which one is responsible, and it rarely matters.

Around the four-month mark, what does matter is that a sudden change accompanied by anything else concerning is a reason to contact a professional rather than to assume it is developmental. A baby who seems unwell is a clinical question, and no developmental explanation should be used to talk yourself out of asking.

Getting through the period

Households that treat it as a fortnight or so to be absorbed do better than those that respond by changing everything at once. Reverting temporarily to whatever settles the baby is a reasonable choice and does not permanently undo anything. Splitting the night explicitly between two adults matters more during this period than at almost any other point in the year.

It is a poor moment to start a new approach, since you would be attributing the results to the wrong thing either way. Most parents describe it settling into a new normal rather than reverting to the old one, which is consistent with development rather than regression.

What comes afterwards

The new sleep structure is the one the baby keeps, so the period is a transition rather than an interruption. Once transitions are being handled more smoothly, many babies consolidate further, though the timeline for that varies enormously.

Week by week, further unsettled periods usually turn up later in the year, often around large developmental changes, and they tend to be shorter. By then most parents recognise the shape of it, which makes the second and third episodes considerably less alarming than the first. The useful takeaway is that unsettled stretches during the first year are usually about something being built rather than something being lost.

Everything above, in order of what to do first

1. The wrong word for the right observation. Parents widely report a period somewhere around the middle of the first year when a baby who had been settling reasonably starts waking far more.
2. What actually changes. Newborn sleep has a simpler structure with a large proportion of the light, active kind and relatively little differentiation between stages.
3. The timing is not fixed. The four-month label is a convenient handle rather than a schedule, and reported timing spreads across a considerable range.
4. Why it often coincides with other things. The same months tend to contain rolling, more active daytime engagement and in some babies the beginnings of teeth, which makes attribution hard.
5. Getting through the period. Households that treat it as a fortnight or so to be absorbed do better than those that respond by changing everything at once.
6. What comes afterwards. The new sleep structure is the one the baby keeps, so the period is a transition rather than an interruption.

The takeaway

Read the harder fortnight as sleep growing up rather than falling apart, and split the nights while it happens.

The good-enough version, done calmly, is the one that helps.