Abstract MP40: Associations Between Meal Timing and Sleep in Older Adults

Abstract

Introduction: Poor sleep health has been associated with increased risk and worse outcomes for a number of diseases such as cognitive dysfunction, dementia and Alzheimer’s disease. An emerging area of research has highlighted the association between dietary factors and sleep. Meal timing is one aspect of diet that has unclear effects on sleep. Few studies have examined this relationship in older adults, an important distinction because sleep architecture and dietary habits change significantly across the lifespan. By identifying the involvement of meal timing in sleep health, specific behavioral interventions can be developed for older adults with sleep-related morbidity. Hypotheses: We tested the hypothesis that a longer time duration between the last meal of the day and bedtime is associated with positive changes in sleep quality Methods: Forty-six healthy older adults (72% (33 of 46) female, 13% (6 of 46) Black, mean age=70.1 years) recruited from the community in Chicago were included in the analysis. Participants underwent sleep assessment with 7 days of sleep actigraphy, one night of sleep profiler, and sleep questionnaires. Two 24-hour diet recalls within 1 week were administered with the ASA24 tool to evaluate meal timing. The data was analyzed with linear regression models adjusting for age, sex, educational attainment, and BMI. Results: For every 1-hour increase in time from last meal to bedtime, sleep onset latency (SOL) increased by 7.4 minutes (SE=3.0, p=0.019), total sleep time (measured from actigraphy) decreased by 10.6 minutes (SE=4.4, p=0.022), and sleep midpoint was delayed by 17.5 minutes (SE=0.09, p=0.002). Other sleep measures showed no significant associations with time from last meal to bedtime. Conclusions: Meals consumed closer to bedtime were associated with shorter SOL, longer total sleep time, and delayed sleep midpoint in older adults. These findings suggest that meal timing may correlate with sleep, however further investigation with larger sample sizes should be done to clarify this link.