Cardiac activation during arousal in humans: further evidence for hierarchy in the arousal response
Introduction
Current models of alertness regulation suggest that daytime sleepiness depends on the duration of prior sleep and on the presence of sleep fragmentation. Although sleep reduction produces an increase in sleepiness (Carskadon and Dement, 1982), several studies have shown that it is not so much the amount of sleep but the frequency of arousal that is important in the recovery functions of sleep (Williams et al., 1964, Stepanski et al., 1984, Bonnet, 1985, Bonnet, 1987, Philip et al., 1994, Roehrs et al., 1994). Evidence in favor of this hypothesis also comes from studies conducted in patients with sleep disorders such as obstructive sleep apnea (OSAS) (Roehrs et al., 1989) or periodic leg movements (PLMS) (App et al., 1990), in whom arousal density is the best predictor of daytime somnolence.
When we consider the processing of arousal response in clinical studies, definition of what constitutes arousal is critical, and criteria of detection and scoring are still controversial. Many studies have focused on microarousals (MA) (American Sleep Disorders Association, 1992, Boselli et al., 1998) and phases of transitory activation (PAT) (Schieber et al., 1971, Collard et al., 1996) corresponding to the largest periodical component of arousal response in humans. They are characterized by a combination of EEG desynchronization, appearance of alpha and low voltage EEG fast rhythms, and tachycardia, and thus they translate a ‘cortical arousal response’ induced by endogenous or exogenous stimuli.
With the inclusion of more sophisticated methods of arousal detection, recent studies (Halasz and Ujszaszi, 1991, Halasz, 1993, Halasz, 1998) opened the discussion of whether synchronized EEG sleep patterns might represent a form of arousal response in humans. By application of auditory stimuli, the authors found that stimuli-induced arousals consisted of transient EEG patterns, i.e. K-complex or delta bursts, without subsequent EEG desynchronization and associated with autonomic activation. These events, called ‘subcortical or autonomic arousal’ (Pitson et al., 1994, Martin et al., 1996), are intrinsic components of human sleep, appearing spontaneously as phases A1 of the cyclic alternating pattern (CAP) (Terzano et al., 1985, Parrino et al., 1998) and expression of levels of greater or lesser arousal.
To understand the exact influence of ‘subcortical arousals’, one must answer the question ‘Do K- or delta bursts directly cause sleep disruption; and are K- and delta bursts primary forms of an arousal response?’ There is now evidence supporting the concept of a common component in both MA and subcortical arousals in the ability to protect sleep against exogenous and endogenous stimuli (Terzano et al., 1985, Halasz and Ujszaszi, 1991, Halasz, 1993). Responses to auditory stimulation in humans induce vasoconstriction (Williams et al., 1964), blood pressure variations (Rees et al., 1994), and increase in ventilation (Carley et al., 1996), concomitant with bursts of K-complexes or delta waves. Moreover, delta bursts occur in patients with upper airway resistance syndrome (Lofaso et al., 1998) and OSAS (Berry and Gleeson, 1997) as an arousal response to airflow limitation.
Despite these findings, there is still controversy as to whether subcortical arousals truly reflect an arousal response. The most straightforward empirical method to determine an arousal response is to determine if concomitant phenomena occur in motor or autonomic systems. Measuring heart rate variation during a transitory event is one such approach. If the hypothesis that subcortical events are an arousal response from sleep is correct, the HR variation found during these arousals should be the same as that occurring during MAs and PATs. This study was undertaken to characterize the phenomenon of heart rate variations as indicators of the type of arousal response, i.e. cortical and subcortical arousals, in healthy subjects. A second goal was to obtain a more precise characterization of the EEG activity by means of spectral analysis to see whether EEG changes undetected by visual scoring could be seen before arousal onset and affect the cardiac response.
Section snippets
Subjects
The subject sample was comprised of 27 healthy subjects, 10 men and 17 women, aged 28.8±9.7 years (range 19–55 years). Subjects underwent a medical evaluation, including medical, psychiatric and sleep history, and a physical examination. They had no life history of cardiac disorder, and all were in good medical health. None was using medication that could affect heart rate, blood pressure or sleep structure. All volunteers participated in a 3 night sleep protocol, with the first night used for
Polygraphic data and EEG arousal scoring
Details of sleep parameters and visual arousal scoring are given in Table 1. All subjects had polysomnographic parameters within normal limits for adults, and the 4 types of arousals were recorded in all patients. The total number of arousals, however, varied between subjects, with an average number of scored arousals of 269.5±48.0. MA represented 36.1% of the total events, and PAT 21.1%; 16.2% of arousals were defined as D-bursts, and 26.6% as K-bursts, occurring mostly in stage 2 and slow
Discussion
The main goal of the present study was to determine whether cardiac changes occur during subcortical arousal, that is, during K-complex and delta bursts, and whether the pattern of variation was similar to that found during cortical arousals, i.e. MA and PAT. Using HR variation over time we found a significant change in HR during K- and D-bursts consisting of a tachycardia followed by a bradycardia, reflecting the changes seen during MA and PAT but to a lesser degree. Moreover, during MA and
Acknowledgements
The authors thank Dr Helli Merica for stimulating discussion and Dr Robert Blois for providing the nocturnal recordings.
References (46)
- et al.
The value of sleep recording in evaluating somnambulism in young adults
Electroenceph clin Neurophysiol
(1991) - et al.
Heart rate variability: sleep stage, time of night, and arousal influence
Electroenceph clin Neurophysiol
(1997) - et al.
The impact of cyclic alternating pattern on heart variability in healthy young adults
Clin Neurophysiol
(2000) Arousal without awakening. Dynamic aspect of sleep
Physiol Behav
(1993)Hierarchy of micro-arousals and the microstructure of sleep
Neurophysiol Clin
(1998)- et al.
The problem of identification of autonomic neurons in the lower brainstem
Brain Res
(1975) - et al.
Simultaneous changes of rhythmic organization in brainstem neurons, respiration, cardiovascular system and EEG between 0.05 and 0.5 Hz
J Autonom Nerv Syst
(1998) - et al.
Facultative coupling of reticular neuronal activity with peripheral cardiovascular and central cortical rhythms
Brain Res
(1975) - et al.
Arterial blood pressure response to transient arousals from NREM sleep in nonapneic snorers with sleep fragmentation
Chest
(1998) - et al.
Slow rhythmic rate fluctuations of cat midbrain reticular neurons in synchronized sleep and waking
Brain Res
(1982)