Scholarlink

<p>Sleep deprivation has a significant impact on health and quality of life, with approximately 30% of adults worldwide suffering from insomnia. Physical activity has been recognized as a promising non-pharmacological intervention method, but the daily association between sleep and physical activity has not been clarified because key factors such as circadian rhythms and activity times have not been considered. Therefore, this study analyzed physical activity by time of day and intensity, and divided participants into circadian rhythm groups to examine the impact of physical activity on sleep on the same day. The results showed that physical activity had a significant influence on sleep efficiency, achieving a high accuracy of 0.8 or higher across all groups. Additionally, through explainable artificial intelligence, the study identified differences in the effects of physical activity by time of day, revealing that low-intensity activity in the evening, 12–15 hours after waking, had the greatest impact across all groups. These findings could serve as an important foundation for developing personalized, non-pharmacological intervention strategies to improve sleep quality. This study demonstrates the potential of explainable machine learning approaches to elucidate the complex relationship between physical activity and sleep, and suggests practical applications for promoting sleep health.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Associations between daily physical activity timing and sleep efficiency revealed by explainable machine learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月22日

查看全文
  • Jihee Choe,
  • Yujin Oh,
  • Chanwoo Park,
  • Young Kim,
  • Changwon Wang,
  • Se Dong Min

摘要

Sleep deprivation has a significant impact on health and quality of life, with approximately 30% of adults worldwide suffering from insomnia. Physical activity has been recognized as a promising non-pharmacological intervention method, but the daily association between sleep and physical activity has not been clarified because key factors such as circadian rhythms and activity times have not been considered. Therefore, this study analyzed physical activity by time of day and intensity, and divided participants into circadian rhythm groups to examine the impact of physical activity on sleep on the same day. The results showed that physical activity had a significant influence on sleep efficiency, achieving a high accuracy of 0.8 or higher across all groups. Additionally, through explainable artificial intelligence, the study identified differences in the effects of physical activity by time of day, revealing that low-intensity activity in the evening, 12–15 hours after waking, had the greatest impact across all groups. These findings could serve as an important foundation for developing personalized, non-pharmacological intervention strategies to improve sleep quality. This study demonstrates the potential of explainable machine learning approaches to elucidate the complex relationship between physical activity and sleep, and suggests practical applications for promoting sleep health.