REM SLEEP LATENCY AND WAKEFULNESS IN THE FIRST SLEEP CYCLE AS MARKERS OF MAJOR DEPRESSION A CONTROLLED STUDY VS. SCHIZOPHRENIA AND NORMAL CONTROLSVadim S. Rotenberg vadir@post.tau.ac.il, Eyal Shamir, Yoram Barak, Peter Indursky, Leonid Kayumov Progress in Neuro-Psychopharmacology & Biological Psychiatry 26 (2002) 1211-1215 Abstract Introduction: REM sleep latency is a clinically significant sleep variable that is found to be decreased in several psychiatric disorders. However, it is not known whether alteration of REM sleep latency is similar across disorders. In order to test whether incorporation of wakefulness in the first sleep cycle has a different outcome on REM sleep latency in different clinical groups, the authors have investigated correlation between sleep variables in the first sleep cycle in 25 patients with major depression, 24 patients with chronic schizophrenia, and in 10 healthy subjects. Results: REM sleep latency correlates with the duration of wakefulness in the first cycle in patients suffering from chronic schizophrenia and in healthy subjects. This correlation does not hold true in patients suffering from major depression. Conclusion: Wakefulness incorporated in the first cycle influences REM sleep latency in healthy subjects and in patients suffering from chronic schizophrenia but not in patients suffering from major depression. This finding further supports the evidence that reduced REM sleep latency is a nonflexible marker of depression. Key words: Depression; First cycle; PANSS; REM latency; Schizophrenia; Wakefulness 1. Introduction REM sleep latency (duration of sleep from sleep onset to the onset of the first REM sleep period) is one of the main objective sleep variables. In many publications, the reduction of REM sleep latency is considered as a marker of major depression (Kupfer, 1976; Reynolds and Kupfer, 1988). According to recent studies, it is not specific for affective disorders and is present in different psychiatric disorders including schizophrenia (Jus et al., 1973; Zarcone et al., 1987), borderline personality disorder (Benson et al., 1990), narcolepsy (Moscoo et al., 1984), eating disorders (Katz et al., 1984), and obsessive-compulsive disorder (Insel et al., 1982). Reduced REM sleep latency is considered to be a clinically significant pathological feature in all these disorders. Therefore, the change of REM sleep latency in the course of the disorder and in the process of treatment can presumably be used as a prognostic factor for spontaneous alteration of the disorder or as a marker of treatment efficacy. However, it is not known whether change of REM sleep latency is quantitatively and qualitatively similar in different psychiatric disorders. It is also not known whether REM sleep latency in these disorders has similar relationships with other sleep variables, such as sleep latency and wakefulness in the first sleep cycle. REM sleep latency is measured differently in different studies (Knowles et al., 1982). First of all, there are two possible definitions of sleep onset: (1) first epoch of any sleep stage including Stage 1 and (2) first epoch of any stage of sleep other than Stage 1. According to both of these definitions, the duration of wakefulness in the subsequent 10 min of sleep must be less than 2 min in order not to mix with a real sleep onset, a short sleep episode occasionally incorporated into wakefulness. According to Benca et al. (1992), the first epoch of Stage 1 for sleep onset definition was used in 51 investigations, while the first epoch of any other stage (most oftenStage 2) was used in 206 investigations. The preponderance of the second definition was present also in the subsequent investigations (Benson and Zarcone, 1993; Goldman et al., 1996; Riemann et al., 1994; Saletu et al., 1996; Thase et al., 1994). It means that in most investigations the initial Stage 1 period is ignored, although it is well known that in patients suffering from mental disorder Stage 1 may be significantly increased, including also the first cycle (Benca et al., 1992). A second problem in the definition of REM sleep latency is related to the inclusion or subtraction of the wakefulness period incorporated in the first cycle. According to Benca et al. (1992), wakefulness was subtracted in four studies (8%), when Stage 1 for sleep onset definition was used, and in 77 investigations (37%), when Stage 2 was used for the same purpose. Wakefulness was subtracted from REM sleep latency by some authors (Benson and Zarcone, 1993; Goldman et al., 1996; Thase et al., 1994, 1997) and was not subtracted by others (Riemann et al., 1994; Saletu et al., 1996). Obviously, it is difficult to compare the results of the investigations with different approach to wakefulness subtraction. In addition, while in healthy subjects awakenings in the first cycle are relatively rare, they appear more often in patients suffering from mental disorder (Rotenberg, 1980, 1982). Its subtraction may cause an artificial similarity between REM sleep latency in healthy subjects and in those suffering from mental disorders (Ganguli et al., 1987). If low REM latency in spite of increased wakefulness characterizes depression, but not healthy subjects, it would mean that REM sleep requirement is higher in major depression. In any case, it is not reasonable to perform such subtraction until we find that wakefulness in the first cycle has a similar influence on REM sleep latency in different clinical groups. To the best of our knowledge, no investigations focusing on this issue exist. Thus, the main aim of the present study was to investigate correlation between REM sleep latency and the duration of first cycle wakefulness in patients suffering from major depression compared with chronic schizophrenia and in healthy subjects. Our hypothesis is that REM sleep latency does not correlate with wakefulness in the first cycle in major depression. The effect of wakefulness incorporated in the first cycle on REM sleep latency was the topic of our study. In this study, the authors were not interested in sleep structure. 2. Methods 2.1. Subjects Our study included 25 patients suffering from major depression (12 men and 13 women, mean age 49.1 7.2 years, meanS.D.), 24 patients suffering from chronic schizophrenia (14 men and 10 women, mean age 43.212.7 years), and 10 healthy subjects (4 men and 6 women, mean age 43.310.0 years) without any subjective sleep complaints. We decided to include a group of patients suffering from chronic schizophrenia because chronic schizophrenia is often also characterized by decreased REM sleep latency and our task was to check whether there is difference between two groups in some features of REM sleep latency. The study was approved by the Institutional Review Board and the Israeli Ministry of Health Committee for Studies in Human Subjects. Written informed consent was obtained from all participants following a detailed explanation of the nature of the study. 2.2. Diagnosis Diagnosis was established according to the DSM-IV criteria. Hamilton Depression Rating Scale 21 items (1960) was used to score the severity of depression symptoms in patients suffering from major depression. The Positive and Negative Syndrome Scale (PANSS) rating scale (Kay et al., 1987) was used in order to evaluate severity of positive and negative symptoms in patients suffering from schizophrenia. All patients had no other medical problems. Patients suffering from major depression received no psychotropic medication for at least 10 days prior to investigation. They had not been treated previously with any neuroleptic agent or other medications with long half-life. The mean score for Hamilton Rating Scale for depression was 38.15.3 (meanS.D.). The mean number of previous episodes of major depression was 4. The group of patients suffering from schizophrenia was divided into 13 patients suffering from paranoid schizophrenia, 6 patients suffering from residual schizophrenia, 4 patients suffering from disorganized schizophrenia, and 1 patient suffering from schizoaffective disorder. The duration of the disorder ranged from 11 to 39 years (mean duration was 20.44.9 years). Negative symptoms dominated the clinical picture of most of the patients suffering from schizophrenia: 19 patients showed positive/negative ratio, according to PANSS, less than 1. The mean level of positive symptoms was 18.29.3 (meanS.D.). The mean level of negative symptoms was 28.26.1 (meanS.D.). All patients were on neuroleptic treatment during the study. Their mean daily dose was equivalent to chlorpromazine 300 mg/day (Kaplan and Sadock, 1995). Treatment regimen remained unchanged during 60 days prior to polysomnography. 2.3. Sleep data Sleep data were collected from all patients on two or three consecutive nights, including the first night in the laboratory, and from control subjects on two consecutive nights. The nights-to-patient ratio in the group of patients suffering from major depression was 2.7, in the group of patients suffering from chronic schizophrenia 2.5, and in the control group 2. According to this ratio, the differences between groups were not significant. In the clinical groups themselves, it was almost equal. Therefore, the ratio of the first nights to the total numbers of nights was also almost equal. Although first-night effect (the increase of REM sleep latency on the first night in the sleep laboratory) is usually present in control subjects and in many patients (Saletu et al., 1996; Rotenberg et al., 1997), we decided not to exclude the first night from analysis. The main point of this study was to find out whether any alterations of REM sleep latency were related to the intrusion of wakefulness in the first sleep cycle. Thus, it would be a mistake to exclude the extreme deviations of REM sleep latency. Table 1. Sleep variables in different clinical groups (meanS.D.)
The significant differences are mentioned in the text. 2.4. Lights Lights were usually turned off at 10:00 p.m. and turned on between 6:00 and 6:30 a.m., according to the typical schedule in the ward. Electroencephalogram (EEG), sub-mental electromyogram, and electrooculograms (EOG) for eye movement registration were recorded using an Electro-encephalograph Neurofax EEG-4400. The EEG and EOG were recorded at a low filter setting of 0.3 Hz. 2.5. Polysomnograms Polysomnograms were analyzed according to international criteria (Rechtschaffen and Kales, 1968). Sleep onset was defined as the first 30-s epoch of Stage 1 or any other sleep stage that appeared after wakefulness and was followed by no more than 1 min of wakefulness in the subsequent 10 min of sleep. Stage 2 latency was defined as the time that elapsed from Stage 1 onset to the first 30-s epoch of Stage 2. Slow wave sleep (SWS) latency was defined as the time that elapsed from the sleep onset to the first 30-s epoch of Stage 3. REM sleep latency was defined as the time that elapsed from the sleep onset to the appearance of the REM sleep episode. To be considered as an REM episode, the REM sleep duration has to exceed one 30-s epoch. The awakenings in the first cycle have been estimated according to the standard criterion of wakefulness (Rechtschaffen and Kales, 1968). The REM episode was defined as singular if there were no more than 10 min of intervening wakefulness or non-REM sleep incorporated into the REM episode. 2.6. Statistical analysis The mean data and mean standard deviations of sleep variables were calculated for each clinical group separately. The significance of sleep variables differences was determined using the t criterion and ANOVA. Pearson's Product-Moment Correlation was used for the estimation of relationships among REM sleep latency, wakefulness in the first sleep cycle, percentage of REM, and sleep latency in each clinical group. Correlations were calculated using each night for each subject without averaging data for all nights of each subject. 3. Results The main results of the present study are shown in Tables 1 and 2. Only significant (P<.05) correlations are included in Table 2. As it could be expected according to previous studies (Benca et al., 1992; Rotenberg, 1982), sleep latency, Stage 2, SWS latency, and wakefulness in the first cycle in both clinical groups are higher than in healthy subjects (P<.05). REM sleep latency, on the other hand, has a nonsignificant tendency to be lower in major depression. Wakefulness in the first cycle in major depression is lower than in schizophrenia (F=19, P<.0384). The range of the standard deviations of sleep variables in healthy subjects is less than in the clinical groups (Table 1). Only in major depression is REM sleep latency not correlated with the duration of wakefulness in the first cycle. This correlation is present in patients suffering from chronic schizophrenia and in healthy subjects (Table 2). In both clinical groups, REM sleep latency correlates negatively with the percentage of REM, while in healthy subjects the correlation between these variables is absent. It is worthwhile to stress that sleep latency does not correlate with REM sleep latency in healthy subjects or in mentally ill patients. Table 2. Significant (P<.05) correlation between REM sleep latency and other sleep variables in different clinical groups
4. Discussion The main result of the present investigation is that only in major depression is REM sleep latency not correlated with the duration of wakefulness in the first cycle. It is worthwhile to stress that night sleep of healthy subjects and of patients suffering from chronic schizophrenia displays a significant positive correlation between these two variables. The duration of wakefulness and its standard deviation is lower in healthy subjects and is higher in patients suffering from schizophrenia. In patients suffering from major depression, it is intermediate. Thus, it is not very likely that the absence of correlation between these variables in major depression is caused by either very low or very high amount of wakefulness in the first cycle. Our suggestion is that wakefulness in the first cycle regularly increases REM sleep latency. This effect does not exist in major depression where a high REM sleep pressure (increased REM sleep requirement) may cover the effect of awakenings. Increased REM sleep requirement is confirmed by the redistribution of REM sleep in sleep cycles (its increase in the first cycle) and by increased eye movement density in the first cycle (Reynolds and Kupfer, 1988; Benca et al., 1992). We have confirmed these data in our previous investigation performed on the same group of patients (Rotenberg et al., 1997). The increased REM sleep requirement in depression can be explained by the adaptive role of REM sleep in mood regulation (Kramer, 1993), in the restructuring of defense mechanisms (Greenberg and Pearlman, 1993), and in the restoration of search activity (Rotenberg, 1993). In the present study, in contrast to some other investigations (Kupfer, 1976; Reynolds and Kupfer, 1988; Jus et al., 1973; Zarcone et al., 1987), REM sleep latency does not differ significantly in mentally ill patients and in the control group of healthy subjects. We have found only a nonsignificant tendency of REM sleep latency to be shorter in patients suffering from major depression. There are two possible reasons for this result. First of all, in contrast to most other investigations, in our study, wakefulness was not subtracted from REM sleep latency and is longer in the first cycle in mentally ill patients than in the control group. Second, also in contrast to most other studies (see Introduction), we have used for sleep onset definition the first epoch of Stage 1, while in mentally ill patients this stage in the first cycle is usually increased. In our study, Stage 2 latency, which contains Stage 1, was also increased in mentally ill patients in comparison to the control group. It means that by calculating REM sleep latency it is necessary to take into consideration the difference in Stage 1 duration in the patients group and in healthy subjects. Another reason why REM sleep latency in clinical groups does not differ significantly from REM sleep latency in normal subjects is a large standard deviation of this variable in both clinical groups. It is very possible that this deviation is determined by the inclusion of the first night of investigation (in all groups of subjects) in the analysis. The group of normal subjects was monolithicthe first-night effect was present in all participants. At the same time, both clinical groups contained patients with and without first-night effect (Rotenberg et al., 1997, 1998). If present, first-night effect was often accompanied by the extremely short REM sleep latency in the subsequent nights, thus increasing the deviation of data. This "compensatory" decrease of REM sleep latency in the subsequent nights following the first-night effect may explain a common finding of the shortened REM latency in depression if the first night is excluded from analysis, as often happened. However, in the present study, excluding the extreme deviation of REM sleep latency by ignoring the first night would be irrelevant according to the main goal of this investigation. One of the limitations of the present study is that our clinical groups were investigated in different conditions. Patients suffering from major depression were investigated in treatment-free conditions while patients suffering from schizophrenia were investigated on a stable neuroleptic regimen. Typical antidepressant agents usually increase REM sleep latency and suppress REM sleep in comparison to drug-free conditions (Sandor and Shapiro, 1994). Neuroleptics increase REM sleep latency in comparison to neuroleptic withdrawal, while at the same time they decrease the duration of wakefulness incorporated in sleep (Neylan et al., 1992). Thus, it would be reasonable to expect that neuroleptic treatment could abolish or decrease correlation between REM sleep latency and wakefulness in sleep, but in any case, there are no reasons to expect neuroleptic treatment to elicit such correlation. It means that the difference between our two clinical groups is presumably not dependent on drug continuation or drug withdrawal. In our study, REM sleep latency is neither related to sleep latency in healthy subjects nor in mentally ill patients, although the deviation of the duration of wakefulness before sleep is very broad in mentally ill patients. According to the concept of stable REM sleep rhythm determined by the process (Borbely, 1982), it was reasonable to predict a negative correlation between sleep latency and REM sleep latency by taking into consideration the fixed time for lights-off in all our subjects. The absence of such correlation stands in contradiction to previous data (Knowles et al., 1987). They, however, investigated healthy subjects with naps and prominently shifted sleep. In their subjects, the range of wakefulness prior to sleep was much broader than in our control group. A very small duration and range of wakefulness before sleep in our control group may be a reason for the absent correlation between sleep latency and REM sleep latency in this group. At the same time, the absence of such correlation in psychiatric patients may be an additional sign of the increased REM sleep requirement (Cartwright et al., 1991). 5. Conclusion In contrast to healthy subjects and patients suffering from chronic schizophrenia, the intrusion of wakefulness into the first cycle does not increase REM sleep latency in a regular way in patients suffering from major depression. Thus, REM sleep latency in depression is less flexible and is more resistant to alternative factors like awakenings in the first cycle. Our findings further support the assumption that "REM sleep pressure" is an electrophysiological marker probably unique to patients suffering from major depression. Acknowledgements The investigation was supported by grants from theMinistry of Health, Israel, and of the Israel Psycho-biological Society. References
Benca, R.M., Obermeyer, W.H., Thisted, R.A., Gillin, J.C., 1992. Sleep and psychiatric disorders: a meta-analysis. Arch. Gen. Psychiatry 49,651-668. Benson, K.L., Zarcone, V.P., 1993. Rapid eye movement sleep and eye movements in schizophrenia and depression. Arch. Gen. Psychiatry 50, 474-482. Benson, K.L., King, R., Gordon, D., Silva, J.A., Zarcone, V.P., 1990.Sleep patterns in borderline personality disorder. J. Affect. Disord. 18,267--273. Borbely, A.A., 1982. A two process model of sleep regulation. Hum. Neurobiol. 1, 195-204. Cartwright, R.D., Kravitz, H.M., Eastman, C.J., Wood, E., 1991. REM latency and the recovery from depression: getting over divorce. Am.J. Psychiatry 148, 1530-1535. Ganguli, R., Reynolds III, C.F., Kupfer, D.J., 1987. Electroencephalographic sleep in young never-medicated schizophrenics: a comparison with delusional and non-delusional depressives and healthy controls. Arch. Gen. Psychiatry 44, 36-44. Goldman, M, Tandon, R., Taylor, S.F., DeQuardo, J.R., Shipley, J.E., Patel,., Reddig, S., Jibson, M., 1996. Dexamethasone nonsuppression and short rapid eye movement latency in schizophrenia: markers of an affective diathesis? Biol. Psychiatry 40, 927-929. Greenberg, R., Pearlman, C., 1993. An integrated approach to dream theory: contribution from sleep research and clinical practice. In: Moffitt. A., Kramer, M., Hoffmann, R. (Eds.), The Functions of Dreaming. State University of New York Press, New York, USA, pp. 363-380. Hamilton, M.A., 1960. A rating scale for depression. J. Neurol., Neurosurg.Psychiatry 23, 56-62. Insel, T.R., Gillin, J.C., Moore, A., Mendelson, W.B., Loewenstein, R.J.,Murphy, D.L., 1982. The sleep of patients with obsessive-compulsive disorders. Arch. Gen. Psychiatry 39, 1372-1377. Jus, K., Bouchard, M., Jus, A.K., Villeneuve, A., Lachance, R., 1973. Sleep EEG studies in untreated long-term schizophrenic patients. Arch. Gen.Psychiatry 29, 386-390. Kaplan, H.I., Sadock, B.Y., 1995. Comprehensive Textbook of Psychiatry, sixth ed. Williams & Wilkins, Baltimore, MD. Katz, J.L., Kuperberg, A., Pollack, C.P., Walsh, B.T., Lumoff, B., Weiner,H., 1984. Is there a relationship between eating disorder and affective disorder? New evidence from sleep recording. Am. J. Psychiatry 141, 753-759. Kay, S.R., Fiszbein, A., Opler, J.A., 1987. The Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261-276.Knowles, J.B., MacLean, A.W., Cairns, J., 1982. Definitions of REM latency: some comparisons with particular reference to depression. Biol. Psychiatry 17, 993-1002. Knowles, J.B., MacLean, A.W., Vetere, C., Young, P., Salem, L., Surridge-David, M., Coulter, M., 1987. The influence of prior wakefulness on REM sleep. J. Biol. Rhythms 2, 81-93. Kramer, M., 1993. The selective mood regulatory function of dreaming: an update and revision. In: Moffitt, A., Kramer, M., Hoffmann, R. (Eds.), The Functions of Dreaming. State University of New York Press, New York, USA, pp. 139-196. Kupfer, D.J., 1976. REM latency: a psychobiological marker for primary depressive disease. Biol. Psychiatry 11, 159-174. Moscoo, S.S., Shampain, D.S., Sassin, J.F., 1984. Nocturnal REM latency and sleep disturbance in narcolepsy. Sleep 7, 115-125. Neylan, .., Van Kammen, D.P., Kelley, M.S., Peters, J.L., 1992. Sleep in schizophrenic patients on and off haloperidol therapy. Arch. Gen. Psychiatry 49, 643-649. Rechtschaffen, A., Kales, A., 1968. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects National Institute of Health, US Government Printing Office, Washington, DC (Publication 204). Reynolds, C.F., Kupfer, D.J., 1988. Sleep in depression. In: Williams, R.Z., Karacan, I., Moore, C.A. (Eds.), Sleep Disorders, Diagnosis and Treatment. Wiley, New York, pp. 147-164. Riemann, D., Hohagen, F., Krieger, S., Gann, H., Muller, W., Olbrich, R., Wark, H.J., Bohus, M., Low, H., Berger, M., 1994. Cholinergic REM induction test: muscarinic supersensitivity underlies polysomnographic finding in both depression and schizophrenia. J. Psychiatr. Res. 28, 195-210. Rotenberg, V.S., 1980. Sensitivity, neuroticism and sleep disturbances: some controversial problems. Waking Sleeping 4, 271-279. Rotenberg, VS., 1982. The Adaptive Function of Sleep. Nauka Publisher, Moscow. Rotenberg, VS., 1993. REM sleep and dreams as mechanisms of the recovery of search activity. In: Moffitt, A., Kramer, M., Hoffmann, R. (Eds.), The Functions of Dreaming. State University of New York Press, New York, USA, pp. 261-292. Rotenberg, VS., Hadjez, J., Kimhi, R., Indursky, P., Sirota, P., Mosheva, T., Benatov, R., Elizur, A., 1997. First night effect in depression: new data and a new approach. Biol. Psychiatry 42, 267-274. Rotenberg, VS., Hadjes, J., Martin, ., Indursky, P., Michailov, Y, Barak, Y, Weiss, R., Gutman, E., Shamir, E., Elizur, A., 1998. First night effect in different forms of schizophrenia (pilot investigation). Dynamic Psychiatry 5/6, 421-430. Saletu, B., Klosch, G., Gruber, G., Anderer, P., Undowratn, P., Frey, R., 1996. First-night effects on generalized anxiety disorder (GAD)-based insomnia: laboratory versus home sleep recordings. Sleep 19, 691 -697. Sandor, P., Shapiro, C.M., 1994. Sleep patterns in depression and anxiety: theory and pharmacological effects. J. Psychosom. Res 38 (Suppl. 1), S125-S139. Thase, M.E., Reynolds III, C.F., Frank, E., Jennings, J.R., Nofziger, E., Fasiczka, B.A., Garamoni, G., Kupfer, D. 1994. Polysomnographic studies of unmedicated depressed men before and after cognitive behavioral therapy. Am. J. Psychiatry 151, 1615-1622. Thase, M.E., Buysse, D.J., Frank, E., Cherry, C.R., Cornes, C.L., Mallinger, A.G., Kupfer, D.J., 1997. Which depressed patients will respond to interpersonal psychotherapy? The role of abnormal EEG sleep profiles. Am. J. Psychiatry 154, 502-509. Zarcone, V.P., Benson, K.L., Berger, P.A., 1987. Abnormal rapid eye movement latencies in schizophrenia. Arch. Gen. Psychiatry 44, 45-48. |
