Mangiaruga Anastasia, D'Atri Aurora, Scarpelli Serena, Alfonsi Valentina, Camaioni Milena, Annarumma Ludovica, Gorgoni Maurizio, Pazzaglia Mariella, De Gennaro Luigi
Department of Psychology, Sapienza, University of Rome, Rome, Italy.
Department of Biotechnological and Applied Clinical Sciences, University of L'Aquila, Coppito, L'Aquila, Italy.
Sleep. 2022 May 12;45(5). doi: 10.1093/sleep/zsab284. Epub 2021 Dec 5.
Sleep talking (ST) has been rarely studied as an isolated phenomenon. Late investigations over the psycholinguistic features of vocal production in ST pointed to coherence with wake language formal features. Therefore, we investigated the EEG correlates of Verbal ST as the overt manifestation of sleep-related language processing, with the hypothesis of shared electrophysiological correlates with wake language production.
From a sample of 155 Highly frequent STs, we recorded 13 participants (age range 19-30 years, mean age 24.6 ± 3.3; 7F) via vPSG for at least two consecutive nights, and a total of 28 nights. We first investigated the sleep macrostructure of STs compared to 13 age and gender-matched subjects. We then compared the EEG signal before 21 Verbal STs versus 21 Nonverbal STs (moaning, laughing, crying, etc.) in six STs reporting both vocalization types in Stage 2 NREM sleep.
The 2 × 2 mixed analysis of variance Group × Night interaction showed no statistically significant effect for macrostructural variables, but significant main effects for Group with lower REM (%), total sleep time, total bedtime, sleep efficiency index, and greater NREM (%) for STs compared to controls. EEG statistical comparisons (paired-samples Student's t-test) showed a decrement in power spectra for Verbal STs versus Nonverbal STs within the theta and alpha EEG bands, strongly lateralized to the left hemisphere and localized on centro-parietal-occipitals channels. A single left parietal channel (P7) held significance after Bonferroni correction.
Our results suggest shared neural mechanisms between Verbal ST and language processing during wakefulness and a possible functional overlapping with linguistic planning in wakefulness.
言语梦呓（ST）作为一种孤立现象很少被研究。近期对ST中发声的心理语言学特征的调查表明其与清醒时语言形式特征具有一致性。因此，我们研究了言语ST的脑电图（EEG）相关性，将其作为睡眠相关语言处理的明显表现，并假设其与清醒时语言产生具有共同的电生理相关性。
从155个高频ST样本中，我们通过视频多导睡眠图（vPSG）记录了13名参与者（年龄范围19 - 30岁，平均年龄24.6±3.3；7名女性）至少连续两个晚上的情况，共记录了28个晚上。我们首先将ST者的睡眠宏观结构与13名年龄和性别匹配的受试者进行比较。然后，我们比较了6名在非快速眼动睡眠（NREM）第2阶段报告了两种发声类型（言语和非言语）的ST者中，21次言语ST之前与21次非言语ST（呻吟、大笑、哭泣等）之前的EEG信号。
两因素混合方差分析（组×夜）交互作用对宏观结构变量无统计学显著影响，但组的主效应显著，与对照组相比，ST者的快速眼动睡眠（REM）百分比更低、总睡眠时间、总卧床时间、睡眠效率指数更低，而非快速眼动睡眠（NREM）百分比更高。EEG统计比较（配对样本t检验）显示，在θ和α EEG频段内，言语ST与非言语ST相比，功率谱下降，强烈偏向左侧半球，并集中在中央 - 顶叶 - 枕叶通道。经Bonferroni校正后，单个左侧顶叶通道（P7）具有显著性。
我们的结果表明，言语ST与清醒时的语言处理之间存在共同的神经机制，并且可能与清醒时的语言规划存在功能重叠。
