INTRODUCTION
  
The human brain has 2 fundamental behavioral states: sleep and wakefulness[1], and these are separated by wake-sleep transition[2]. The sleep state is defined by electroencephalogram results[3-6] or behavior[7]; in previous behavioral studies, sleep was defined by the failure to respond to external stimuli, while wakefulness was con-sidered a cognitive response to external stimuli[7]. Studies of human brain sensation have shown that sensor areas remain fully activated during wake-sleep transition[8-21]. However, these studies utilized only event-related potential (ERP) or functional MRI (fMRI). Subsequently, the behavioral mechanisms of how sensitivity influences transition from wakeful to sleep state remain poorly understood. Based on the signal detection theory, the present study compared the sensitivity index (d’) between wakefulness and wake-sleep transition[22-24] to measure sensitivity differences.

RESULTS

Quantitative analysis of participants
Each subject recorded a good night’s sleep prior to testing. During wake-sleep transition testing, the starting time of wake-sleep transition testing and the initial time of losing behavioral response were recorded for each subject (Table 1). All subjects slept well, except two. One subject reported not having fallen asleep for one session and the other one for two sessions. In the remaining sessions, sleep onset latency > 30 minutes was considered abnormal (insomnia)[25]. Subject 1 exhibited sleep onset latency longer than 30 minutes in 1 session, and subject 5 with 2 sessions (Table 1). All sessions were valid for later analysis except the two situations mentioned above.
Morphological changes in reaction time
The time series of wakefulness and wake-sleep transition are shown in Figure 1. During wake-sleep transition testing, 2 typical phases were measured. When stage A, characteristic morphological features included temporal fluctuation of behavior response. Reaction times varied between 17.88 ms and 4 716.9 ms, and almost all subjects displayed a high accuracy rate ranging from 72% to 99.5%. While in stage B, subject’s behaviors responded stability and behavioral responses were undetectable.

Sensitivity differences between wakefulness and wake–sleep transition
Because the magnitude of reaction time significantly altered with time, individual differences in reaction time between wakefulness and wake-sleep transition (stage A) were tested. Sensitivities of wakefulness and wake-sleep transitions were calculated based on the signal detection theory (Table 2). In addition, paired-sample t-tests showed significant differences at the 0.01 level (t = 5.083, P < 0.01; Table 3).

DISCUSSION

The wake-sleep transition is a complex process; a previous study[26] reported that subjects detect and respond to external stimuli during the EEG-defined stage 1, which suggests that behavioral measurements can be used to detect wake-sleep transition.



Another study confirmed that sleep is a state in which subjects fail to respond to external stimuli[7]. The fin-ger-tapping task also provides comparable information on wake-sleep transition and function as a behavioral measure of sleep onset[27]. In the present study, subjects responded to serial, external stimuli during stage A, but exhibited no behavioral responses after a short period, although target stimuli were maintained. Based on sleep behavioral criteria, the initial time that the subject lost behavioral response to target stimuli is considered as the sleep state[7]. Therefore, morphological characteristics of stage B were consistent with sleep behavioral characteristics, which suggested that the subject fell asleep during stage B. These results suggested that the interval between initial time of sleep and beginning of wake-sleep transition represents sleep onset latency.
In general, sensory inputs are consciously perceived only during wakefulness. However, electrophysiological studies have shown that sensory processing is preserved during non-rapid eye movement (NREM) sleep. For example, early auditory-evoked potentials generated during sleep suggests that early stage sensory processing is not affected by sleep[28]. One study utilized fMRI and EEG confirmed bilateral activation in the auditory cortex, thalamus, and caudate during NREM sleep[8]. These above-mentioned results suggest that the human brain processes auditory stimuli and detects meaningful events during sleep. At the behavioral level, simple selective signal detection extracts the sensitivity index d’. In the present study, normal subjects’ sensitivity decreased during the wake-sleep transition, demonstrating the possibility of significantly decreased sensitivity. Sensitivity represents the transformation of sensory processing of physical stimulus energy into an internal representation, which suggests a significantly decreased sensation capacity for acquiring outside information.

SUBJECTS AND METHODS

Design
Self-controlled experiment.
Time and setting
This experiment was performed at the Key Laboratory of Human Being Development and Mental Health, Huazhong Normal University, China from April to June, 2009.
Subjects
According to the Pittsburgh Sleep Quality Index[30] approval and informed consent, 14 healthy, undergraduate volunteers from Huazhong Normal University, who were free of sleep aid and drugs, participated in the present study. The subjects comprised 11 females and 3 males, aged 18–22 years. All subjects reported good sleep before the study.
Methods
Stimuli and task
Two kinds of short audio stimuli were randomly presented. The target stimulus was clear, and the non-target stimulus was relaxing. Intervals between two adjacent stimuli were adjusted to 2-5 seconds. During testing, subjects were asked to make response when they heard the target stimulus. To minimize response disturbances on sleep, a previously developed behavioral detection method was utilized[7, 30]. A digital sensor (Hong Kong Smilebuyer Technology, Hong Kong, China) was attached to the subject’s finger, behavioral responses would be recorded once the subject pressed the sensor slightly.
Signal detection theory
A parameter estimation of the signal detection theory was applied to compare sensation performances between wakefulness and wake-sleep transition. In this theory, responses are classified into 4 categories: (1) hit, (2) false alarm, (3) miss, and (4) correct rejection. The sensory process is assumed to continuously output based on random Gaussian noise. When a signal is present, the signal combines with the noise. When the signal intensity is fixed, sensitivity measurements of the sensory process are based on differences between mean output with and without signal conditions. Large differences represent high sensitivity and vice versa. The sensory process was analyzed by employing response proportions in the index calculation of sensitivity. When standard deviations of the 2 distributions were equal   (μn＝μs＝1), sensitivity was represented by d’[31]:

μs represents mean of signal-plus-noise distribution, μn represents mean of noise distribution, δn represents standard deviation of noise distribution, and δs represents distribution of noises and signals.
Procedure
The sleep room was soundproof and lightproof. Each subject was tested for 4 consecutive nights, and the first night was considered as adaptation period[32]. The entire experiment was performed over 35 days.
Each subject was asked to arrive at the laboratory 1 hour before his or her typical bedtime. Testing for sensation of wakefulness was performed after 10 minutes of rest. The testing session lasted 20 minutes, followed by another 10-minute rest. If the subject felt sleepy, testing for sensation of wake-sleep transition began. If the subject did not respond within 30 minutes, testing was terminated. Following testing, a subjective assessment of sleep quality was performed[33].
Statistical analysis
Data collected included previous night sleep quality, time of study, initial time of falling asleep, reaction time, subjective perception of sleep, demographic data. To obtain the sensitivity index d’ of wakefulness and wake-sleep transition, the time course of testing was analyzed based on signal detection theory. Sensitivity for wakefulness and wake-sleep transition were compared using the paired-sample t-test. P < 0.01 was considered statistically significant. Statistical analyses were performed with SPSS 17.0 (SPSS, Chicago, IL, USA).

Author contributions: Chuang Gao and Wei Wei conceived and designed the study. Yang Zhang, Fengjuan Zhang, and Shanyu Ke performed the experiments. Chuang Gao, Yang Zhang, and Fengjuan Zhang analyzed the data. Chuang Gao and Fengjuan Zhang wrote the manuscript.
Conflicts of interest: None declared.
Ethical approval: This study was approved by the Institutional Review Boards of Huazhong Normal University, China.
Acknowledgments: This work was supported by grants from the Research Council of Huazhong Normal University. We thank the volunteers, who participated in this study.

REFERENCES

[1] Kahn D, Pace-Schott EF, Hobson JA. Consciousness in waking and dreaming: the roles of neuronal oscillation and neuromodulation in determining similarities and differences. Neuroscience. 1997;78(1):13-38.
[2] Merica H, Fortune RD. State transitions between wake and sleep, and within the ultradian cycle, with focus on the link to neuronal activity. Sleep Med Rev. 2004;8(6):473-485.
[3] Nieuwenhuijs DJ. Processed EEG in natural sleep. Best Pract Res Clin Anaesthesiol. 2006;20(1):49-56.
[4] Benini F, Trapanotto M, Sartori S, et al. Analysis of the bispectral index during natural sleep in children. Anesth Analg. 2005;101(3): 641-644.
[5] Ogilvie RD. The process of falling asleep. Sleep Med Rev. 2001; 5(3):247-270.
[6] Nittono H, Momose D, Hori T. The vanishing point of the mismatch negativity at sleep onset. Clin Neurophysiol. 2001;112(5): 732-739.
[7] Ogilvie RD, Wilkinson RT, Allison S. The detection of sleep onset: behavioral, physiological, and subjective convergence. Sleep. 1989;12(5):458-474.
[8] Portas CM, Krakow K, Allen P, et al. Auditory processing across the sleep-wake cycle: simultaneous EEG and fMRI monitoring in humans. Neuron. 2000;28(3):991-999.
[9] De Gennaro L, Ferrara M, Ferlazzo F, et al. Slow eye movements and EEG power spectra during wake-sleep transition. Clin Neurophysiol. 2000;111(12):2107-2115.
[10] Bódizs R, Sverteczki M, Mészáros E. Wakefulness-sleep transition: emerging electroencephalographic similarities with the rapid eye movement phase. Brain Res Bull. 2008;76(1-2):85-89.
[11] Badia P, Wright KP, Wauquier A. Fluctuations in single-hertz EEG activity during the transition to sleep. In: Ogilvie RD, Harsh JR, eds. Sleep Onset: Normal and Abnormal Processes. Washington: American Psychological Association (APA). 1994: 201-218.
[12] Harsh J, Voss U, Hull J, et al. ERP and behavioral changes during the wake/sleep transition. Psychophysiology. 1994;31(3):244-252.
[13] De Gennaro L, Ferrara M, Curcio G, et al. Antero-posterior EEG changes during the wakefulness-sleep transition. Clin Neurophysiol. 2001;112(10):1901-1911.
[14] Casagrande M, Bertini M. Laterality of the sleep onset process: which hemisphere goes to sleep first? Biol Psychol. 2008;77(1): 76-80.
[15] Casagrande M, Bertini M. Night-time right hemisphere superiority and daytime left hemisphere superiority: a repatterning of laterality across wake-sleep-wake states. Biol Psychol. 2008; 77(3):337-342.
[16] Bódizs R, Sverteczki M, Lázár AS, et al. Human parahippocampal activity: non-REM and REM elements in wake-sleep transition. Brain Res Bull. 2005;65(2):169-176.
[17] de Lugt DR, Loewy DH, Campbell KB. The effect of sleep onset on event related potentials with rapid rates of stimulus presentation. Electroencephalogr Clin Neurophysiol. 1996;98(6): 484-492.
[18] Baker FC, Maloney S, Driver HS. A comparison of subjective estimates of sleep with objective polysomnographic data in healthy men and women. J Psychosom Res. 1999;47(4):335-341.
[19] Shinar Z, Baharav A, Akselrod S. Changes in autonomic nervous system activity and in electro-cortical activity during sleep onset. Comput Cardiol. 2003,2003:303-306.
[20] Tanaka H, Hayashi M, Hori T. Topographical characteristics of slow wave activities during the transition from wakefulness to sleep. Clin Neurophysiol. 2000;111(3):417-427.
[21] Foulkes D, Vogel G. Mental activity at sleep onset. J Abnorm Psychol. 1965;70:231-243.
[22] Swets JA. Is there a sensory threshold? Science. 1961;134: 168-177.
[23] Lewis O, Harvey Jr. The critical operating characteristic and the evaluation of expert judgment. Organ Behav Hum Decis Process. 1992;53(2):229-251.
[24] Swets J, Tanner WP Jr, Birdsall TG. Decision processes in perception. Psychol Rev. 1961;68:301-340.
[25] Alexander ZG, Howard MK, Louis GK. Sleep Psychiatry. London: Taylor & Francis Group. 2004.
[26] Johnson LC, Hanson K, Bickford RG. Effect of flurazepam on sleep spindles and K-complexes. Electroencephalogr Clin Neurophysiol. 1976;40(1):67-77.
[27] Casagrande M, De Gennaro L, Violani C, et al. A finger-tapping task and a reaction time task as behavioral measures of the transition from wakefulness to sleep: which task interferes less with the sleep onset process. Sleep. 1997;20(4):301-312.
[28] Burton SA, Harsh JR, Badia P. Cognitive activity in sleep and responsiveness to external stimuli. Sleep. 1988;11(1):61-68.
[29] Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193-213.
[30] Gao C, Chen B, Wei W. Dynamic detection of wake-sleep transition with reaction time-magnitude. Neural Regen Res. 2009; 4(7):552-560.
[31] Krantz DH. Threshold theories of signal detection. Psychol Rev. 1969;76(3):308-324.
[32] Tamaki M, Nittono H, Hayashi M, et al. Examination of the first-night effect during the sleep-onset period. Sleep. 2005;28(2): 195-202.
[33] Yang CM, Han HY, Yang MH, et al. What subjective experiences determine the perception of falling asleep during sleep onset period? Conscious Cogn. 2010;19(4):1084-1092.
 (Edited by Pan JY, Qin YL/Su LL/Song LP)