How kids affect your sleep

Linda Meijer, May 2021
Ever wondered how kids impact your sleep pattern? I analyzed my Fitbit sleep data to see the differences in the various life phases of the last 3 years.

Preparing for maternity leave

Currently, I am on maternity leave, waiting for my second child to be born. It made me realize how fast the last few years have flown by and how much I forgot about the first year with a baby. So to refresh my memory and mentally prepare for the upcoming months, I thought it would be nice to dive into my recorded sleep data.

Regularity vs. Chaos

In the last 3 years, a lot of things have changed in my life. Starting a family has been definitely the most impactful thing. First of all, there is a huge amount of time that goes into caring. But besides that, there is also the well-known lack of sleep. Newborns need to be fed every 3 to 4 hours in the first weeks and don't come with a day and night schedule. So lots of interrupted nights! In most cases, the intervals become a bit longer after a few weeks or months. But it can take up months or even years before a baby will sleep through the night. If you haven't experienced this yourself (yet), this is how it looks like in data:

Look at all these short time intervals in the middle chart 😰

Yes, I definitely forgot about how bad it was and how much I longed for a sleep that lasted longer than 4 hours. The good news is that after a year most of the interruptions are gone. The bad news, during that time also a new pattern has been introduced: no more sleeping in on the weekend! After the very irregular start comes the period of very regular early wakings...

Technical sidenotes

  • Fitbit data format

    Fitbit analyzes your sleep and stores it in different formats. More information can be found at the Fitbit Web API documentation. I took the 'levels' data from the Sleep Web API which can be either in a format called sleep phases ('wake', 'light', 'deep', 'rem') or sleep patterns ('awake', 'restless', 'asleep').

    The recorded format depends on how well Fitbit was able to perform the analysis. E.g. when the Fitbit has a low battery or is positioned in a way that prevents consistent heart-rate reading it will record the simplified sleep pattern format.

    For simplicity, I filtered out all records of type 'wake' and 'awake'. I ignored all gaps that last less than 10 minutes and created new sleep records by consolidating multiple consecutive sleep records into a single record.

  • Missing data

    I regularly forget to charge my Fitbit. And subsequently, forget to put it back on again. So like in almost every data project: I had to deal with missing data. To make my visual look cleaner, I filled the missing dates (<10%) with a random sleep record of a window of +/- 2 weeks and additionally shifted it by 30 minutes.

  • N=1

    Of course, this is just one example. Lots of other factors than your family situation will play a role too. For example, the bottom chart was captured during the first (partial) Covid lockdown. Probably this accounts for part of the regularity as well. And maybe you have noticed the two weeks in the upper graph with more irregularity than the other weeks? That was the Christmas and New Years' holiday period 😉

    Sleep patterns of other parents during the newborn phase will also vary a lot. For one thing, it will be heavily influenced by how you divide care as parents. For example, breastfeeding is simply not so easy to split with your partner as formula-feeding. Next to that, your baby will have its own unique rhythm. When it comes to the sleep patterns of newborns, they are kind of all over the place. So if these irregular sleep patterns scare you: don't worry, your future baby might as well be one of those babies who would go for long stretches at night!

‹ Prev
View all articles
Next ›