O018 Night-to-night variability in temperature is associated with decreased sleep regularity

Abstract Objective Global daily temperature variability is projected to increase up to 100% in some regions by 2100. Previous studies have shown that sleep duration and timing is strongly impacted by temperature. In this study we investigated associations of night-to-night temperature variability with sleep duration timing and regularity. Methods Data were collected from global users of the Withings Sleep Analyzer (n = 125 295) and ScanWatch (n = 238 550) between January 2020 and September 2023. Sleep regularity was assessed using intra-monthly standard deviation in sleep duration and timing (onset, midpoint and offset). Ambient temperature data were extracted from climate models used to calculate night-to-night temperature variability. Fixed-effects models assessed the association between night-to-night temperature variability and sleep regularity using participant-year intercept. Results Participants (n = 313 822) were middle aged (50 ± 15 years), 70% male, with an average 17 months of data. During months with high vs low night-to-night temperature variability (99th percentile; 4.3°C vs 50th percentile 1.9°C;), sleep onset variability was higher in southern (mean [95%CI]; 17.9 [16.1, 19.6] min), central (6.5° vs 1.0°C; 32.2 [29.8, 34.6] min) and northern locations (5.5° vs 2.1°C; 11.4 [11.2, 11.7] min). We observed greater variability in sleep midpoint and offset with higher night-to-night temperature variability, with similar magnitude effects. High night-to-night temperature variability was associated with <10 min difference in sleep duration irregularity. Conclusions Our findings highlight that high night-to-night temperature variability is associated with a 20 to 50% increase in sleep timing irregularity. Targeted interventions to mitigate the impact of temperature variability on sleep health are an important public health priority.