Discovery of the Dreaming Sleep Stage: A Recollection

A SYMPOSIUM DEVOTED TO “THE 50^TH ANNIVERSARY OF THE DISCOVERY OF PARADOXICAL SLEEP” WILL TAKE PLACE IN LYON, FRANCE, IN 2009. IT IS OF interest to recall the different steps that led to 1 of the major topics of today's sleep research.

Although older descriptions of the behavioral correlates of dreaming can be found from Lucretius¹ to Hervey de Saint-Denys² (for reference, see Gottesmann³), the forerunner of the contemporary history of rapid eye movement (REM) sleep began in 1936 with Derbyshire et al.'s⁴ result concerning cats: “When sleep was apparently less tranquil judging by twitching of the vibrissae, there were only small [electroencephalographic {EEG}] waves as in the alert waking state.”^4p582 The next significant result was the description of an EEG sleep stage in humans with associated dreaming made by Loomis et al⁵ in 1937. Showing sleep histograms for the first time, their paper noted that dreaming occurs during low-voltage EEG activity. In the same year, Blake and Gerard⁶ observed, also in humans, “feeble” EEG activity during sleep that was associated with a high auditory arousal threshold. Finally, still in 1937, a rich year for sleep research, Klaue⁷ described a “deep sleep” stage in cats with-low voltage EEG, “complete muscular relaxation,” and jerks. One year later, Jacobson, who was the first to record eye movements in humans during waking,⁸ wrote “when a person dreams… most often his eyes are active.”⁹ Soon afterward, dreaming was observed to always occur after provoked arousal following foot twitches or jerks.¹⁰ Finally, the now well-known erection during sleep, which had already been noticed by Hervey de Saint Denys,² was fully described by Ohlmeyer et al.^11,12

However, the first description of eye movements associated with low-voltage EEG, peaks of overt body activity, hyperpnea, and simultaneous dreaming was made in 1953 by Aserinsky and Kleitman,¹³ who completed their study on eye movements 2 years later¹⁴ It is of interest that Aserinsky learned the method for eye-movement recording from Jacobson, following Kleitman's request.¹⁵ As previously mentioned,³ Aserinsky once told me that, in 1953, when he was defending his thesis at the University of Chicago, the examiners told him that doctorates are awarded not for results obtained but in the hope of future findings! Soon afterward, in 1955, Dement,¹⁶ also from Kleitman's laboratory, described the association of eye movements with low-voltage EEG during dreaming in patients with schizophrenia, who recalled their dreams in a fashion similar to that of people without schizophrenia.

Next, in 1957, Dement and Kleitman published 2 full papers devoted to the dreaming stage of sleep.^17,18 The first provided a complete description of EEG, eye movements, and dream recall during and outside of the eye-movement periods and of the relationship between eye movements and dream content. The second paper detailed the body and limb movements that occur during this sleep stage and reported an increased arousal threshold by auditory stimulation. The following year, Dement and Wolpert¹⁹ first showed that episodes of this sleep stage increase in duration during the night and that there is a relationship between body movements and the changing of dream content. Moreover, “eye-movement periods with minimal ocular activity are associated with passive dreams, while frequent and large eye movements are associated with active dreams.”^19p552 A second paper of Dement and Wolpert²⁰ examined the manifest content of successive dreams during the night. Although “no single dream was ever exactly duplicated… and each dream seemed to be a self-containing drama relatively independent of the preceding or following dream…the manifest content of nearly every dream exhibited some obvious relationship to one or more dreams occurring on the same night.”^20p569 Also in 1958, Dement,²¹ working on cats, described a sleep stage with low-voltage EEG, eye movements, “complete abolition of muscle potentials,” and increased arousal threshold by auditory stimuli. This stage was first called “activated” sleep, and then “deep sleep”^22,23 because of the highly increased reticular arousal threshold²²; later, it was mainly called “rapid eye movement sleep”²⁴ thus, REM sleep.²⁵

In 1959, Jouvet and Michel²⁶ described the “paradoxical stage” or “phase” of sleep, during which neodecorticated cats showed periodic loss of tonic neck-muscle activity, the presence of nictitating membranes covering the strong myotic eyes, and small vibrissae movements. These sleep periods did not last for more than 10 minutes. In addition, “spindles” were observed in the pontine reticular formation. In partially decorticated preparations, the remaining cortex displayed rapid EEG. In mesencephalic preparations, the paradoxical phase could be observed by the pontine spindles and muscular atonia. There was no mention of eye movements. A second paper, involving the same mesencephalic sections,²⁷ included a nice figure recalling the differences in the sleep electrophysiologic activities of normal, decorticated, and mesencephalic animals, although it included no eye-movement recordings. Clearcut pontine spikes were observed in the nucleus reticularis pontine caudalis. The authors wrote that this “archeosleep” of rhombencephalic origin was “deeper” than the slow wave sleep stage, since the auditory arousal threshold was higher. The third paper²⁸ focused on intact animals. During the rhomencephalic or paradoxical phase, there was theta activity in the hippocampus, eye movements (described for the first time), jerks of the paws and tail, slowing of the heart rate, wide and rapid breathing movements accompanied by arrhythmias, and increased arousal thresholds by auditory and reticular stimulation. Finally, cortical and reticular auditory evoked potentials were considerably attenuated. The rectal temperature was decreased by between 0.2°C and 0.3°C. During the 4 hours of recording, there were 2 to 3 such periods. The authors mentioned that, although Dement²¹ and Dement and Kleitman¹⁸ thought that the depth of this sleep stage was intermediate between waking and sleep (slow wave sleep), Jouvet's team estimated that it was a deeper stage. However, in this paper, slow wave sleep is still called “deep sleep.” The authors again spoke about archeosleep to define this “paradoxical phase,” the name of which was later given up because of Pavlov's²⁹ use of the same term to characterize the decrease or loss of reactions in response to strong conditioned stimuli.^29p269 The term “paradoxical sleep” was definitively adopted in 1965.^30,31 These 1953-1959 results opened the way to the extraordinary contemporary era of sleep research.

DISCLOSURE STATEMENT

The author has indicated no financial conflicts of interest.