How do wearable-measured sleep duration and fragmentation change before, during, and after pregnancy?

Sleep first stretches out, then contracts as pregnancy advances

The clearest pattern across wearable studies is a rise-then-fall curve. In a large cohort of Fitbit users, time in bed began increasing around week 4 of pregnancy and peaked near week 8, when participants spent about 27 more minutes in bed than their baseline; total sleep time rose from roughly 425 minutes to about 448 minutes at week 9 [2]. After that peak, time in bed declined by about 1.3 minutes per week, returning toward pre-pregnancy levels around week 30 and dropping further as the due date neared [2]. A smaller smart-ring study found the same direction of change later in pregnancy: total sleep time fell by about 0.28 hours per day in the second trimester and 0.58 hours per day in the third, compared with the first [3].

The early gain appears to come mostly from lighter sleep rather than deeper sleep. In the Fitbit cohort, the first-trimester increase was largely driven by more light sleep, while deep sleep declined across pregnancy by about 19 minutes on average by the end, and REM sleep dropped by about 9 minutes later in pregnancy [2]. That detail matters for interpretation: a longer night does not necessarily mean more restorative sleep.

The nights get longer early but also more broken and less efficient

Duration is only half the story. Sleep efficiency — the share of time in bed actually spent asleep — slipped modestly in the Fitbit cohort, from a median of 88.3% to 86.8% across pregnancy [2]. The smart-ring study similarly found lower overall sleep scores and about a 1.75-percentage-point drop in sleep efficiency by the third trimester relative to the first [3]. In plain terms, women were spending more time in bed but a slightly smaller fraction of it asleep.

One finding complicates the intuition that later pregnancy means more tossing and turning. In the smart-ring cohort, measured restlessness was actually lower in the third trimester than in the first [3]. The authors do not settle why, and it is worth treating as an observed signal rather than an explanation. What the same study did find was longer awake time in the second trimester among women who were not overweight or obese, and lower light-sleep proportions in the second and third trimesters among women with a higher pre-pregnancy body mass index [3]. Maternal age did not significantly change these trimester patterns [3].

What wearables can and cannot yet say about before and after pregnancy

The "before" side of the timeline is only partly captured. The Fitbit analysis used a pre-pregnancy baseline value as its reference point and described sleep returning toward that baseline around week 30 before falling further [2]. A large Oura-ring observational study of over 10,000 pregnancies reports trajectories spanning pre-pregnancy through pregnancy, but its available summary does not give post-pregnancy sleep results [4]. So the early-pregnancy rise is best understood as a departure from a pre-pregnancy starting point, even though few studies report that starting point in detail.

The "after" side is the weakest link. The Fitbit study consented new parents with a child under one year but its reported results focus on pregnancy [2]; the smart-ring study followed women only across the three trimesters [3]; and a pilot study of 23 pregnant women tracked sleep during pregnancy without analyzing before or after [1]. None of the available summaries provides a clear wearable-measured picture of how duration and fragmentation behave in the months after birth. That gap is worth naming plainly: the postpartum curve is largely uncharted in this evidence.

Why the shape of the curve is more than a curiosity

These patterns are not just descriptive. In the pilot study, average sleep duration fell from about 8.4 hours at week 12 to about 4.7 hours at week 40, and the authors frame early identification of poor sleep as a way to flag risk, since lack of sleep is linked to adverse pregnancy outcomes [1]. Self-reported tiredness and stress were highest in the first trimester, dipped in the second, and rose again in the third — a rhythm that does not simply mirror the sleep-duration curve [1].

In a separate cohort of 170 women at moderate to high risk for preeclampsia, poorer subjective sleep quality correlated with lower birth weight and shorter gestational age, and a combined sleep-plus-stress model predicted preeclampsia better than either alone [5]. That study used questionnaires alongside a Fitbit and was recruited during the COVID-19 period, so its context should not be read as a general statement about all pregnancies [5]. Still, it supports the practical point: wearable tracking is most useful when it catches the early-pregnancy window, when sleep is longest and habits may be easiest to shape, rather than only at the exhausted end.

Sources used in this answer

Related research questions