ELECTRIC SLEEP-PRODUCING DEVICES: AN EVALUATION USING EEG MONITORING -D., FRANCIS A. J. TYCE, M.D., A" LAWRENCE W. WOODS, M ]REGINALD G. BICKFORD, M.B. 2 Although Kerbikov(5) first reported the positive effects are of a low order of mag- use of electric sleep in the English ]item- nitude. was shown in this Because electroencephalographic moni- ture in 1955, little,interest form of treatment u2ntil the appearance in toring has allowed a more precise definition states of alertness and of sleep, it this countr -i -made in- of the y of several forl gn di- was used to determine 2 the existence of sleep struments. ilecently there have been in cations that the sleep-producing devi are before and after treatment. Furthermore, ces ormal subjects, 'in psychiatric, EEG monitoring was considered desirable 2 effective inn muscle-tension as a means of detecting an possible dam- patients and in patients with y syndromes(l, 2, 4). More extravagant age resulting from the passage of electric claims witho2ut documentation have ap- current through the head. Apart from the peared in newspaper and magazine arti-" possibility that damage-induced EEG slow- cles(7), and it. has been said that the Rus- wive activity might occitr, there was also a sian space scientists were experimentiniz question 2as to whether electric sleep would wiih the use of such devices for producing produce disturbances of fast activity such long periods of sleep in astronauts on in- as frequently accompany sleep initiated by terstellar flights. hypnotic drugs. Our own interest 2in these devices con- cemed two rather practical questions: 1) POOcEDu" Are they capable of producing sleep with The observations were carried out on two such facility that they would be useful in groups of subjects: one of normal volun- conjimction with EEG 2recordings as part teers and one of psychiatric patients. The of the clinical test? 2) Is the phenomenon normal subjects (Table 1) ranged in age of electrically induced sleep of sufbcient from 24 to 55 years, and the majority of intensity that it would be useful in the them were technicians or p2hysicians inter. treatment of various psychiatric disorders ested in the possibility of obtaining sleep associated with sleeplessness or disturbance by this method. Their attitudes toward of the sleep cycle ? these devices ranged from mildly skeptical In studying the changes 2in such a subtle to openminded. Observations on the nor- function as the onset of sleep there is neces- mal subjects were carried out in the mom- sarily a formidable problem of controlling in or afternoon. 9 factors of inte2rnal and external condition- Psychiatric patients (Table 2) who were ing. Consequently, in the experiments to be iinder treatment at the Rochester State I-los- described our interest was essentially con- pital formed the patient group. Observa- fined to validating the contention that2 sleep tions were carried out in the morning or onset could be controued and manipulated afternoon. The patients were told that this reliably by the apparatus at our disposal. instrumentation was part of their diagnosis We leave to others the formidable task of and treatment, and they were assured that 2 excluding factors of suggestion when the this was not a form of electroshock UeAt- menl They were not told that the device Tlus work was supported @m part by Public to put them 2to sleep. Health Ser%ice grants NB-02056 and NB-03225 was meant from the National Llsbtute of Neurological Dis- Conventional, bilateral, frontal parietal, eases and Blindness. tal and ear electrodes were secured From the Rochester State 2Hospital (Dr. Tyce) occipi lodion and filled with electrode jelly. and the Section of Physiology, Mayo Cunic and by col Mayo Foundation (Dr. Bicldord). Dr. Woods The appropriate eye and mastoid electrodes 2 iduate of the sleep machines were applied with was a Fellow in Psychiatry in the Mayo Cr. School of Medicine, University Of Minnesotr4 Rochester, Minn. head straps after the interpolation of sa- 2 153 N JoUft-.AL OF PSYCHIATRY, [Reprinted frt)m THE AMERICA. 19651 VoL 122,.No. 2. August, I rA ELECMIC SLEL7-PRODUCING DEVICFS r August TABLE I Normal Subjects Participating In Evaluation of Russian Electrosisep Machir4 CYCLES PER SUBJECTrilL cut AGE TIIIAL stcomo VOLTI at UFM ELECUOEKCEPMOGRAIA 2 1 24 1 Drowsy Reduced alpha amplitude 5.5 10 2 37 I 10 10 Drowsy Reduced alpha amplitude 3 2 28 1 5.5 IZ5 Drowsy No change 4 32 1 7.5 12.5 Alert No change 5 32 1 5.5 10 Very drowsy; No change 2 sleep (7) No record 6 28 1 10 11 Drowsy 7 33 1 7.5 15 lntemittent: Sleep Changes drowsiness and $loop (7) 8 33 1 7.5 9 Alert No change 2 2 25 10 Alert No change 3 14 15 Alert No change 9 33 1 10 14 Alert Reduced alpha amplitude 10 34 12 7,5 is Drowsy Reduced alpha amplitude 11 32 1 14 13 Sleep No record 12 55 1 is 5@ Alert No record line-soaked pads over the eyes and 2 mas- electroencephalograph before and immedi- toid processes. The subject then relaxed ately after the use of the sleep machine. comfortably in a quiet room with low il- Becorclb,,ig during the application of electro- lumination. 2 sleep was not possible due to artifact caused The EEG instn=ent was usually oper- by the stimulus current. After the patient ated in the same room but was not in use had become relaxed and after adjustment during the elclctrosleep period. Six channel of eye and mastoid sleep electrodes for recordings were made on a Grass type maximum comfort, the initial five- to ten- IIIC (time constant 0.18 second) standard minute EEC record was taken. After com- 2 TABLE 2 Psychiatric Patients Participating In Evaluation of Japanese Efectrosloop MaeMm Fixed Frequency 10 Cycles Per Second; Voltage 10 to 15 Volts 2 SUWECTIVE CUE PSYCHIATRIC POTOBLEM AGE TRIAL IMCT ELECTROENCEMLOGRAM 1 Reactive depression; hypochondriasis 57 1 Drowsy No effect 2 2 Drowsy No effect 2 Reactive depression; insomnia 40 1 Drowsy; Sleep Sleep 2 Alert No effect 3 Phobic reaction , insomnia 62 2 1 Drowsy Reduced alpha amplitude 2 Sleep Sleep 4 Anxiety 43 1 Alert No effect 2 Alert No2 effect 5 Anxiety -, depression 53 1 Increased anxiety No effect 2 Alert No effect 6 Reactive depression; insomnia 37 1 Alert No effect 2 No effect 2 Alert 7 Reactive depression 21 1 Drowsy; sleep No effect 2 Sleep 7 Sleep 8 Chmnic sniiety; kpcchmwtiuis 42 1 Reoess No effect 2 Drowsy No effecf BICKFOP.D 155 F. A. T. "YcE AND A. C- 19651 L. W. WOODS, Japanese instrument. The Ja2panese sleep 'Good Sleep' of a satisfactory EFG recording, pletion own as ins2trument used is kn t_he Kawasald d is manufactured by the eicctrosleel'i session was commenced with application of the current at 11 low an n nt. The Electric Company (Figure 2) This i stru- value unfelt by the subject Or patie strength ment differs from the Russian model in that 2 current was then adjusted to a t is entirely battery-operatecl from a 22-5- producing mild but undisturt)ing flicker ini en tested on a load of field of vision..Tbis was continued until volt dry cell. Wh 2 the to sleep or sug- 10,000 ohms, it produced an output voltage. the subject or patient went, gested that he did not feel drowsy and did ranging from 3 to 22 volts, square pulses milliseconds duration durin2g a repeti- not wish to continue longer with the pro- of 1.5 lation tion frequency of 1,0 cycles per second. cedure. Periods of electrosleep stimu This frequency cannot 2 be altered on the nged from 20 to 45 minutes. ra - model. After the Russian 'Good Sleep model was damaged, the 2 Japanese sleep A"ARARUS ete the study 4nstrament was used to compl Russian ins"ment.Ibe Russian sleep in2- strument (Figure 1) was imported by on the normal and paiient subjects. the Waters Corporation, Rochester, Minn. FIGURE 2 It consisted essentially of a mains-operated 2 Eye electrodes, bead bands and bilateral mastoid else- i ibmtor circuit giving vacuum-tube mult v trades of Japanese transistarind battery-operated "good 2 frequency control from 1.5 to 60 cycles Sleep" ofectrosloop device are attached to subjeel and a voltage range (with a load of 2000 ohms) of 0 to 19 volts. A square wave out- 2 range ac- put was produced. The volta-,e tually employed in these observations was between 10 and 15 volts. There were four metal electrodes attached to a rubber head- 2 band. Contact with the eyelids and with sldn over the mastoid processes was en- sured by soaking cotton wadding in normal saline for placement under the- electrode-- The negative electrodes were placed an- 2 teriorly and the p ositive, posteriorly. The Russian instrument was used on normal serio s accidental dam- subjects only since 11 age of the electronic mec2hanism prevented its use in the second part of the study. RFSULTS FIGURE I Nt)rmal subjects (Rtisvian device). Eleven Eye electrodes, bead strap and connecting lead of m(,ii raiigiiig in tge from 24 to 5:' ) year.% and Russian mains-operated vacuum-tube atectrosicep device one womaii, agic,,(l 50 years, were tested a2re In place. Bilateral mastoid electrodes are not In view. wit,L ti IC IRIIS.,iian devic(, during single trials of one-half liotir eacit. (One %til)it,ct -A-as 2 test t-d on three successive occ-,isions.) An attempt was made to use the fre(luency and voltage ratiig(! 2 %%,hich was detectable to the 'patient but produced no partictilar d'tscom- fort. Tli(,- frequency usually ranged frotii 5 2 to 15 cycles per second, and the voltage ranged from 10 to 15 volts. Sensations of -Iiiig, I ing, tapping and flickering 2 prick )UM were perceived most fre(luently, often catising some di.-icoijifort which clisapl)car(,-d 5 as the voltage was reduced. The eyes and 156 =CMIC SIrr -PRCOUCING DMCES August the jaw were the mmon sites of discom- this techni co 2 que, if present, were so minimal fort. Nine of the subjects bad EEG record- that they would be of no consequence in ings before, during (with the current helping to solve the problem of sleep re- turned off) and after each trial. cordings in electroencephalogr2aphy. Many subjects complained of discomfort Nor=l subjects and patients (Japanese due to the harness or to sensations pro- device). Three normal male subjects were duced by the electricity. Definite sleep oc- tested for 40-minute test intervals before curred in only tWo subjects. This was con-' the Japanese 'Good Sleep' instrument was firmed by EEG changes as shown in Figure used on patients (data not included in Table 3. The subject in Figure 3 showed rather I). As in the case of the Russian instrument, sparse alpha rhythm in the control record- sensation2s consisting of vibration and flicker ing seen in the upper part of the tracing were usually felt around the eyes. Of these before the stimulus had been applied. After subjects, one fell into a deep sleep within stimulation for 20 minutes at 15 cycles per 10 minutes, but the other two were res2tless second and 7.5 volts, the subject was noted and showed no evidence of sleep. to I)c sleeping. Sleep continued after re- The instniment was then tested on eight moval of the stimulus as indicated in the female patients for two consecutive trials second part of the tracing2 which shows slow of 45 minutes each (see Table 2). The pa- waves and spindle activity. tients were told to go to sleep if they felt like it but were not told that this was a FIGURE 3 sleep machine. 2 The psychiatric problems of EEG recordings on normal subject before and after an these patients consisted largely of anxiety, of Russian alsetrosleep device reveal changes In sleep phobias, hypochondriasis and depression, record which show slow waves, spindles. and occipital usually of the reac2tive type to situational Ilambdoid waves" associated with sleep presumed to problems and accompanied by insomnia. have been Induced artificially. The tests were monitored by taking an CoNr2ROL EEG recording before and after each trial and, if the subject ap peared to be asleep d2uring the trial, the electrosleep machine 6d off and EEG recordings taken. FIGURE 4 get Rifriw ew EEG of p2atient having reactive anxiety and depression zon 'M SliowlefWr I$eyct"Awc. ?$"OS - S/ofpot shows failure of 45-minute stimulation wfth Japanese $loop device to produce changes characteristic of sleep. 2 Notice widespread alpha rhythm In control and post- stimulation meardl CoNrRa In the few subjects in whic sleep oc- 2 curred, it was light and intermittent. Slight drowsiness occurred in four subjects as in- dicated by some diminution in alpha am- plitude. Most sleep or drowsiness occurred in four subjects including the subject on 2 AFAW 45 &,in. somwm loma loer-leslue - $*W whom several trials were performed. No person slept for long intervals or for the duration of the complete test, and all sub- jects were easily aroused I)y external stimu. 2 li such is noise. Some persons volunteered the information that they would have slept more easily without the apparatus. It was concluded that sleep effects from L. W. WOODS, P. A. TYC:E AND R. C. BICKFORD 157 Of thc!sc cight stibiects, only three showed ure to obtain a significantly greater inci- dtbiiiie sleep and then on only one of the dence of sle(,-I)2 or an increased rate of onset m been p t tl trials. In other instances, some slight drow than ight have ex cc ed in natun sizif--ss was indicated by reducld alpha am- unv-,isisted sleep in these S11I)jects. plitilde in the EEG. No EEG patterns attributed to cerebral An example of failure to produce sleep is in tiry were observed after stimulation with 2 shown in Figure 4. It shows a patient 'th either the Russian or Japanese device. W' an active anxiety depression whose control recording shows widespread and persist- COmmENT ent alpha activity. In a second 2 series of Originally the Russians based this method tracings taken 45 minutes after stimulation, of electrosleep on Pivlovian conditioning at 10 milliamps and 10 cycles, there was theory(l) ; that is, they invoked a phe- still widespread alpha rhyth2m but no sleep nomenon called 'parabiosis' which is de- fined as a special condition of excitable tis- had occurred. On the other hand, Figure 5 shows an sue, a condition produced by the effect of example of a pa2tient with depressive reac- stimuli unusual for the tissue and which tion who failed to sleep during the fint represents a persistent state of changed test, but who showed moderate to deep excitability. Pam osis is considered to be sleep, with spindles, on a second test car- 2 inhibitory. Pavlov postulated that three ned out two days later. The parameters types of external stimuli give rise to in- of stimulation were the same in both in- hibition; namely, the very weak, the very stances. In other cases there was moderate strong and the unusual. He b2elieved that restlessness or increased anxiety during the foreign stimuli, when repeated, led to the trial. development of inhibitory conditions Of As in the case of the Russian device the cerebral cortex. Prolonged stimulation 2 to profound inhil)ition which encom- used on normal subjects, there was a fail- lead.% FIGURE 5 Atte mpted sleep induction failed In patient having reactive depression 2an first test with Japanese Instrument but succeeded in second test 2 days later. In lower figh% note absence of alpha rhythm and pros--Once of sleep spindles Indicating EEG evidence of sleep. CONRROL ACRER 45ma STIMIAARVN 10-0,locycleslsec.-,VOSL,-Ep L Lo RO Rigm ew - 2r4rrsr 4 CV5NRROL AFRER 20@k*t STAWAARION *W IOCICIOSAWC - SLCEPING 'r, DMCZS 158 ELECMC SLEEP-PItODUCr.% August the cerebral he hl-res and by if long continued, flicker timulation may passes nusp S tion de duce a significant hypnotic effect. irradia scends to . the lower brain- pro stem centers. Hence, sleep is thought of as produced by cortical ftihibition directly summ"y with secondary2 irradiation to subcortical When tested on normal subjects and pa tients with various psychiatric diso centers. in Russia, electrosleep appears to be the electric "sleep-inducing' devices psychiatrists(3, 6, 7, 9) been found to be 'meffective from a practi- used primarily by Various Americans touring the Soviet Unioti cal standpoint, although behavioral observa- have described the use of electrosleell tions and EEG monitoring have indicated where it is given for two-hour interval,2- the onset of normal drowsiness and sleep cts and patients. No -7 on consecutive days for approximately patterns in Some subj@ 't clinicall2y m'dent EEG disturbance to 20 treatments(8). Improvement is S of a claimed for such conditions as reactive pattern attributed to injury have resulted neuroses, astlicnic states and psycboreactive from the use of the devices. There is some forms 2of schizophrenia. Hypertension and o@estion whether currents of the magnitude have also @mployed would penetrate the cranium in rheumatic encephalitis (chorea) a been bene2fited, but involution I melan- sufficient intensity to produce the complex made worse. clianges theorized in the Russi literature. cholias are believed to be an The results reported in this investigation cate that both Russian a2nd Japanese REFERENCES mdi sleep machines are relatively ineffectual in- 1. Electric Sleep: A Clinical-Physiological In- struments. However, as indicated previous- vestiption-U.S.S.R. (Translation). New ly, the design of 2 these experiments was not York: U. S. joint Publications Research such that would detect very minimal sleep- Service, 1960. (Distributed by Office of inducing properties. It is possible, for in- Technical Services, U. S. Dept. of Com- merce.) stance, that some 2of our subjects went to 2. Forster, S., Post, B. S., and Benton, J. C.: sleep more quickly than they would have done without the device. It would require Preliminary Observations of Effects of Transcerebral Application of Law Voltage sive control observations and a 2 much I exten Low Frequency Pulsating Currents, -Elee- larger series of subjects and patients t3 trosleep,' in Certain Problems of Rehabili- settle this matter. tative Medicine, Arch. Phys. Med. in press. 2 Apart from a practical failure to demon- 3. Hocb, P. H., and Pennes, H.' in Bellak, strate effective slccp-inducing properties, L., ed.: Schizophrenia: A Review of the it can indeed be doubted whether the Syndrome. New York: Logos Press, 1958, relatively small currents produced by 4264465. these instruments could have a significant 4. Hiiischnecker, A. A.: The Sciences. New York: N. Y. Acad. of Sciences, 1961, I. neurol)liysiologic effect on the briinstf--m or cortex, as has been claimed by the Russian 5. Kerbikov, 0. V.: Treatment of Mental workers. It is more probable that the Disease by Sleep, Lancet 1:744-745, 1955. most significant part of this current travels 6. Lesqe, S. - Current Clinical and Research Trends 'M Soviet Psychiatry, Amer. the skin and does not penetrate the M Psychiit. 114:101'8-1022,1958. cranium and that a more likely explanation 7. Solomon, C.: The Russians New Sleep 2 of sleel)-indticing effects, if they can be Machine, This Week, Jan. 13, 1963, M. demonstrated, is the motiototioxis flicker 8. Wayne. C. J. - Impressiont of Soviet Psy- sensation indtice(i in the visual fields. chiitry, Calif. Med. 93:351-354, 1960. Cliiiietil experience with I)Iiotic stimula- 9. Wortis J.: A Psychiatric Study Tour of tion aii(I observations ori&riiially reported by th(, U.S.S.R., J. Ment. Sci. 107:119-156, Pavlov indicate that at low intensities, and 1961.