Auditory arousal thresholds after selective slow-wave sleep deprivation Michele Ferrara * , Luigi De Gennaro, Maria Casagrande, Mario Bertini Dipartimento di Psicologia, Universita Á degli Studi di Roma `La Sapienza', Via dei Marsi 78; 00185 Rome, Italy Accepted 30 June 1999 Abstract Objectives: The aim of this study is to assess Auditory Arousal Thresholds (AATs) three times during an undisturbed baseline night and to compare them to AATs during the recovery night that follows two consecutive nights of selective SWS deprivation. The presence of a time- of-night effect on AATs will also be assessed. Methods: Ten male Ss slept in the laboratory for 6 consecutive nights. The ®rst two nights were undisturbed. The 3rd night was considered as baseline. During the 4th and 5th nights, selective SWS deprivation was obtained by means of acoustic stimulation. The 6th night was a recovery. In the last 4 nights Ss were awakened three times, after 2, 5 and 7.5 h of sleep, respectively. All the awakenings were carried out from stage 2 (after at least 5 consecutive min of stage 2), by means of 1000 Hz ascending tone series. The AAT determination was based on EEG-EMG criteria: at least 10-s of clear alpha rhythm and/or a 10 s movement arousal. Results: During both deprivation nights, SWS amount was close to zero. In the ensuing recovery night a signi®cant SWS rebound was found, accompanied by a signi®cant increase of AATs with respect to the baseline. Furthermore, there was a signi®cant linear decrease of AATs during the night. Finally, the individual correlations between AATs and SWS amount were signi®cant in 4 out of 10 Ss. Conclusions: These results con®rm that AATs are a reliable index of sleep depth by showing that the SWS rebound following selective SWS deprivation is paralleled by a signi®cant AAT enhancement. The experimental paradigm also allows us to claim that AATs show a decreasing linear trend during the night, having excluded any procedural bias. Finally, AATs can be directly related to SWS amount that preceded the awakening, although the individual correlations between AATs and SWS have to be considered with caution, given the high inter-subject variability and the small number of observations. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Auditory arousal thresholds; Selective Slow wave sleep deprivation; Sleep depth; Slow wave sleep rebound 1. Introduction Auditory arousal thresholds (AATs), that re¯ect a subject's arousability from sleep, have been traditionally used to assess depth of sleep. Studies that compared sleep stages with respect to AATs have reported consistent evidence that AATs progressively increase as NREM sleep deepens from stage 2 to stage 4 (e.g. Williams et al., 1964). In general, the highest AATs are associated with slow-wave sleep (SWS), while REM and stage 2 have comparable but lower thresholds (Goodenough et al., 1965; Rechtschaffen et al., 1966). According to a homeostatic model of sleep regulation, an AAT decline in the course of the night as sleep need is progressively satis®ed, should actually be expected (Zepelin et al., 1989). However, although it is commonly believed that AATs decline overnight as sleep accumulates, outcomes regarding AAT trend during the night are not univocal. On the one hand, several reports have found that AATs during both REM (Goodenough et al., 1965; Rechtschaffen et al., 1966) and stage 2 (Rechtschaffen et al., 1966; Watson and Rechtschaffen, 1969; Zimmerman, 1970) are higher early in the sleep period than in the later part of the night. However, Zepelin et al. (1989) reported an overnight decline of AATs only with repetitive stimulation ± as compared with the up-and-down method ± pointing out the relevance of procedural factors in the assessment of AATs. On the other hand, Zimmerman (1970) did not ®nd any clear time-of-night effect for REM awakenings. In addi- tion, Bonnet et al. (1978), having excluded from the analysis all awakenings occurring within 10 min of a body move- ment, did not ®nd any time-of-night effect for stage 2 or REM AATs. On the contrary, Williams et al. (1964) reported AAT elevation in the course of the night when a behavioral response was used as an awakening criterion, and no time-of-night effect considering EEG response as an awakening criterion. The latter results could be ascribed to habituation effects based on repeated overnight exposure to the AAT measurement procedures. Clinical Neurophysiology 110 (1999) 2148±2152 CLINPH 98746 1388-2457/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S1388-2457(99)00171-6 www.elsevier.com/locate/clinph * Corresponding author. Tel.: 139-06-4991-7508; fax: 139-06-4451- 667. E-mail address: ferraram@axrma.uniroma1.it (M. Ferrara)