ETD

• actigraphy
• real-world
• sex differences
• sleep

Background - Men and women sleep differently, and these differences carry real clinical weight. But they are not uniform across life: to characterize them properly, they must be quantified along the entire lifespan.
Methods - We studied 822 participants aged 5 to 94 years (434 males, 388 females), monitored at home across 6,086 nights with wrist-worn actigraphy and AI-based sleep scoring. Eight parameters were examined: total sleep time, sleep efficiency, wake after sleep onset, the number and duration of awakenings, sleep onset and offset, and the Sleep Regularity Index. Sex-specific, non-linear age trends were modelled with generalized additive mixed models (GAMMs), with a standard generalized additive model (GAM) used for the Sleep Regularity Index, which was available as a single score per participant rather than as a nightly measure
Results - Females slept longer (+23.4 min, p < .001) and fell asleep earlier (−12.5 min, p = .003). These overall effects were not constant with age. The duration advantage was narrowest in early-to-mid adulthood and widened in older age, where the male trajectory kept declining while the female one remained stable. The onset difference was largest in the third and fourth decades and faded beyond age 60. Sleep efficiency was higher and wake after sleep onset lower in females during adolescence and early adulthood, with both differences attenuating through the adult years. Sleep offset and regularity showed smaller and more age-dependent differences.
Conclusions - This study modelled sex and age together, on objective rather than self-reported data, to show how sex differences in sleep shift in magnitude and direction across the lifespan. As accelerometry enters routine care, normative values should follow these trajectories, moving toward a more tailored sleep assessment.