Light exposure and sleep architecture in real-world settings

Abstract Light exposure can influence sleep timing, quality, and architecture, yet evidence from real-world settings remains limited. We investigated personal light exposure, quantified as melanopic equivalent daylight illuminance, and sleep parameters using consumer-grade wearables in a convenience sample of UK adults ( n = 89). Participants simultaneously wore light sensors and sleep trackers for 7 days, yielding over 500 days of data, complemented by daily sleep diaries. Earlier sleep/wake timing was associated with longer-duration daytime light exposure, as well as more regular, less fragmented light patterns across the week. Higher interdaily stability and lower intradaily variability of light exposure were further linked to enhanced deep sleep intensity during the initial portion of the night. Subjective and wearable-derived measures were generally correlated, although disrupted sleep architecture was associated with greater discrepancies between subjective and wearable sleep estimates. These findings demonstrate the feasibility of naturalistic protocols and highlight the potential importance of brighter days and more stable, less fragmented light exposure patterns for sleep.