Understanding Infants' Sleep Pattern

When I was a new parent, I told all my friends, "I totally got this. I know how to think about babies, this is gonna be really easy." Little did I know. There we were, new parents. And then all hell broke loose, and it was only because I couldn't figure out how to get my daughter to sleep through the night. [Andre Fenton, Director, Cognitive Neurobiology Lab, New York University, USA]

As a neuroscientist, it had not occurred to me to actually study sleep. But when I was a newborn parent, there's no knowledge about what is normal and what one should do. And so I decided to study sleep in infants. Surely, we could find out how to think about baby sleep in a scientifically grounded, biologically-based way.


Everyone knows that baby sleep seems erratic early on, and eventually it becomes regular. So we wanted to find out when do babies learn to sleep like we do? Does it develop gradually? Does it suddenly occur in the ninth month, for example? Or does everyone have their own rhythm?

The first thing to do was to figure out how to get the data. We discovered that there's a whole set of apps that people are using to log events about their babies. Most importantly, sleeping. That for us is a treasure trove.

So we talked to some app developers, and this collaboration between the app developers and the scientists became the NYU Baby Sleep Study.

We're going to visit some of the families who've signed up for the study and find out how it is that they're actually using this application, which will give us a sense of how the data are being collected.

The way we're used to collecting data is in a small number of subjects, but with absolutely precise and accurate measurements.

In this case, though, we have no control over anything. We're not even making any of the measurements. And so we don't know whether the times are accurate to the minute, or to the hour.


The thing that struck me the most when I got to meet some of the study participants is that they were accurate to the minute. It was clear that the error was, by assertion, no more than five minutes. To have minute-precise data, and to have millions of those data points, makes you extremely confident as a scientist in coming to conclusions.

Please meet my longtime colleague, Pascal Wallisch. Pascal has a long history of working with lots of data before they were called "big data."

What you're looking at here is a distribution of babies in the study around the world. So every dot here represents a baby. We have logged data from about 1,000 babies, and they have produced a total of six million events made up out of about 1.5 million sleep events, 2.5 million eating events, and 1.5 million diaper change events. Collectively.

If you have a large number of babies and they each have their own rhythm, but there's a common, underlying pattern to that rhythm, when you sample enough babies, the pattern will emerge. And we're beginning to answer the first simple question, when does the sleep pattern emerge?

The question that we're trying to answer is, "Will my baby sleep through the night?" That's really what they want to know, when that will happen.


We have started to analyze the data, and we see trends. But to be very clear, this is very early days. We are not comfortable to publish this yet.

But what you see here is that at one month, there's no clear pattern of when the baby's eating or sleeping. So, basically sleeping and eating at random, which is very taxing on the parents. This was in fact the problem.

But then, within about four months, an initial cycle is emerging. And by the time you get to eight months, there's a hint of a pattern. And then within a year, this clear pattern emerges.

Babies are sleeping through the night. They sleep soon after feeding in the morning, and about five hours later in the early afternoon, that's the third time in which they very reliably sleep.

If we had a big enough data set, this would be very simple to turn into a website where an individual person could come and say, "Oh, my kid is within the normal limits. Everything's fine." Or, "Everything is likely to be fine." Or, "My kid is outside of the normal limits."

That would be an early sign that you could recognize maybe there's a reason for concern, or some kind of intervention, or, at least, you know, close scrutiny.

The really amazing thing is to recognize that the transition from no sleep pattern to a sleep pattern is something that almost all babies do. And it'll take about a year.

Source: Netflix.

Post a Comment

0 Comments