Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA.
Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, USA.
Sleep. 2023 Oct 11;46(10). doi: 10.1093/sleep/zsad206.
Healthy aging and many disorders show reduced sleep-dependent memory consolidation and corresponding alterations in non-rapid eye movement sleep oscillations. Yet sleep physiology remains a relatively neglected target for improving memory. We evaluated the effects of closed-loop auditory stimulation during sleep (CLASS) on slow oscillations (SOs), sleep spindles, and their coupling, all in relation to motor procedural memory consolidation.
Twenty healthy young adults had two afternoon naps: one with auditory stimulation during SO upstates and another with no stimulation. Twelve returned for a third nap with stimulation at variable times in relation to SO upstates. In all sessions, participants trained on the motor sequence task prior to napping and were tested afterward.
Relative to epochs with no stimulation, upstate stimuli disrupted sleep and evoked SOs, spindles, and SO-coupled spindles. Stimuli that successfully evoked oscillations were delivered closer to the peak of the SO upstate and when spindle power was lower than stimuli that failed to evoke oscillations. Across conditions, participants showed similar significant post-nap performance improvement that correlated with the density of SO-coupled spindles.
Despite its strong effects on sleep physiology, CLASS failed to enhance motor procedural memory. Our findings suggest methods to overcome this failure, including better sound calibration to preserve sleep continuity and the use of real-time predictive algorithms to more precisely target SO upstates and to avoid disrupting endogenous SO-coupled spindles and their mnemonic function. They motivate continued development of CLASS as an intervention to manipulate sleep oscillatory dynamics and improve memory.
健康衰老和许多疾病都表现出睡眠依赖性记忆巩固减少，以及非快速眼动睡眠（NREM）脑电波振荡相应改变。然而，睡眠生理学仍然是一个相对被忽视的改善记忆的靶点。我们评估了睡眠期间闭环听觉刺激（CLASS）对慢波（SOs）、睡眠纺锤波及其耦合的影响，所有这些都与运动程序性记忆巩固有关。
20 名健康的年轻成年人进行了两次下午小睡：一次是在 SO 状态上升期间进行听觉刺激，另一次是没有刺激。其中 12 人返回进行第三次小睡，刺激时间与 SO 状态上升有关。在所有的小睡期间，参与者在小睡前进行运动序列任务训练，并在小睡后进行测试。
与没有刺激的时期相比，上状态刺激会干扰睡眠并引发 SO、纺锤波和 SO 耦合的纺锤波。成功引发振荡的刺激比未能引发振荡的刺激更接近 SO 上状态的峰值，并且当纺锤波功率较低时。在所有条件下，参与者在小睡后都表现出相似的显著的记忆改善，这与 SO 耦合的纺锤波密度相关。
尽管 CLASS 对睡眠生理学有很强的影响，但它未能增强运动程序性记忆。我们的研究结果表明，有几种方法可以克服这一失败，包括更好的声音校准以保持睡眠连续性，以及使用实时预测算法更精确地靶向 SO 上状态，并避免干扰内源性 SO 耦合的纺锤波及其记忆功能。这些发现为进一步开发 CLASS 作为一种干预手段，以操纵睡眠振荡动力学并改善记忆提供了动力。
