Back

Charting infant sleep cycle development using actigraphy: Longitudinal evidence for cycle lengthening within the first year of life from 35,000 hours of sleep

Hammad, G.; Schoch, S. F.; Engelmann, M.; Spock, Z.; Kurth, S.; Winnebeck, E.

2025-07-16 physiology
10.1101/2025.07.10.664225 bioRxiv
Show abstract

Sleep is marked by ultradian cycles, coinciding with the alternation of rapid eye movement (REM) and non-rapid eye movement sleep (NREM). These sleep cycles are a striking feature exhibited from infancy through adulthood, yet their underlying mechanisms and functional relevance remain elusive, calling for large-scale longitudinal studies. Here we leverage Locomotor Inactivity During Sleep (LIDS), previously established in adults as an accessible means to track sleep cycles, to chart ultradian sleep cycle dynamics within the first year of life at scale. Specifically, we analyzed > 35, 000 hours of sleep from a longitudinal dataset of 152 infants with 10 days of actigraphy at 3, 6, and 12 months of age. Using advanced signal processing techniques, we demonstrate the existence of rhythmic patterns in LIDS in infants with a cycle length of[~] 60 minutes. Cycles were not only shorter in infants than parents (- 11.8 min, 95% C.I. [- 13.8, - 9.8]) but also increased by[~] 10 min from 3 to 12 months of age, mirroring previous results in smaller samples for NREM-REM sleep cycles. This increase was found to be partially mediated by an increase in sleep bout duration with age (+1.0 min/h, [+0.9, +1.2]). Longer cycles were also found in infants still breastfed at 12 months (+2.5 min [+0.4, +4.5]) and their breastfeeding mothers (+6.7 min [+0.5, +12.9]). We also explored variations in non-rhythmic LIDS patterns in relation to endogenous and exogenous factors. Overall, our results provide evidence of a link between rhythmic patterns in limb inactivity and sleep cycles in infancy, underscoring the potential of studying these in large-scale population cohorts for developmental and health outcomes.

Published in SLEEP · not in our set (fewer than 10 published preprints to learn from) · training set

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.