From the Departments of Psychiatry and Behavioral Sciences (C.C., C.D., Y.L., K.Y.), Neurology (K.Y.), and Epidemiology (K.Y.), University of California, San Francisco; Neuroimage Analytics Laboratory and Biggs Institute Neuroimaging Core (M.H.), Glenn Biggs Institute for Neurodegenerative Disorders, University of Texas Health Science Center at San Antonio; Center for AI and Data Science for Integrated Diagnostics and Center for Biomedical Image Computing and Analytics (M.H.), University of Pennsylvania, Philadelphia; and Department of Preventive Medicine (M.R.C.), Northwestern University Feinberg School of Medicine, Chicago, IL.
Neurology. 2024 Nov 26;103(10):e209988. doi: 10.1212/WNL.0000000000209988. Epub 2024 Oct 23.
To determine the association between early midlife sleep and advanced brain aging patterns in late midlife.
Using the CARDIA study, we analyzed sleep data at baseline and 5 years later, focusing on short sleep duration, bad sleep quality (SQ), difficulty initiating and maintaining sleep (DIS and DMS), early morning awakening (EMA), and daytime sleepiness. These were categorized into 0-1, 2-3, and >3 poor sleep characteristics (PSC). Brain MRIs obtained 15 years later were used to determine brain age through a machine learning approach based on age-related atrophy.
This cohort study included 589 participants (mean age 40.4 ± 3.4 years, 53% women). At baseline, around 70% reported 0-1 PSC, 22% reported 2%-3%, and 8% reported >3 PSC. In multivariable linear regression analyses, participants with 2-3 or >3 PSC had 1.6-year (β = 1.61, 95% CI 0.28-2.93) and 2.6-year (β = 2.64, 95% CI 0.59-4.69) older brain age, respectively, compared with those with 0-1 PSC. Of the individual characteristics, bad SQ, DIS, DMS, and EMA were associated with greater brain age, especially when persistent over the 5-year follow-up.
Poor sleep was associated with advanced brain age in midlife, highlighting the importance of investigating early sleep interventions for preserving brain health.
确定中年早期睡眠与晚年大脑老化模式之间的关联。
利用 CARDIA 研究，我们分析了基线和 5 年后的睡眠数据，重点关注睡眠持续时间短、睡眠质量差（SQ）、入睡和维持睡眠困难（DIS 和 DMS）、早醒（EMA）和白天嗜睡。这些被分为 0-1、2-3 和 >3 个不良睡眠特征（PSC）。15 年后获得的脑 MRI 用于通过基于年龄相关性萎缩的机器学习方法确定脑龄。
这项队列研究包括 589 名参与者（平均年龄 40.4 ± 3.4 岁，53%为女性）。基线时，约 70%的人报告有 0-1 个 PSC，22%的人报告有 2%-3 个 PSC，8%的人报告有 >3 个 PSC。在多变量线性回归分析中，与报告 0-1 PSC 的参与者相比，报告 2-3 个 PSC 或 >3 个 PSC 的参与者的脑龄分别大 1.6 岁（β=1.61，95%CI 0.28-2.93）和 2.6 岁（β=2.64，95%CI 0.59-4.69）。在个体特征中，睡眠质量差、DIS、DMS 和 EMA 与更大的脑龄相关，尤其是在 5 年随访期间持续存在时。
不良睡眠与中年时的大脑老化模式有关，这突显了研究早期睡眠干预以保护大脑健康的重要性。
