EEG sleep in primary depression

Journal of Affective Disorders, 1 (1979) 131-138 @ ElsevierlNorth-Holland Biomedical Press

EEG SLEEP IN PRIMARY

131

DEPRESSIQN

A Longitudinal Placebo Study

PATRICIA RICHARD Department Pittsburgh

A. COBLE, DAVID J. McPARTLAND

J. KUPFER,

DUANE

G. SPIKER,

of Psychiatry, Western Psychiatric Institute and Clinic, School of Medicine, Pittsburgh, PA 15261 (U.S.A.)

JOHN

F. NEIL and

University

of

SUMMARY Characteristic EEG sleep changes in depression are highlighted by a sleep continuity disturbance, delta sleep reduction, and a shortened REM latency. Since these findings have been derived primarily from only a few baseline recordings, questions regarding their persistence and/or variability have not been previously addressed. As part of an extensive set of investigations of EEG sleep in depression, we examined nightly the sleep of 12 hospitalized, non-delusional, primary depressives who were involved in a program of active psychosocial treatment intervention and received only placebo during a 5-week study period. EEG sleep findings revealed a relative lack of change across time, particularly in those parameters reported to be associated with a primary or ‘biologic’ depressive episode. While some degree of clinical improvement was noted, the group failed to achieve a state of remission or even partial remission as determined by the Hamilton Rating Scale. It appears that the major sleep alterations associated with such disorders persist for up to at least 5 weeks in the absence of pharmacologic or other somatic intervention.

INTRODUCTION

While clinicians have long noted the prevalence of sleep disturbances in their depressed patients, recent technological advances in the measurement of this physiologic state have now made it possible to describe these disturbances objectively. EEG sleep studies have demonstrated a sleep profile in depression which is distinguishable from the sleep of normal and nondepressed populations and is characteristically highlighted by sleep discontinuity, delta sleep reduction, and shortened REM latency (Diaz-Guerrero et al. 1946; Hawkins and Mendels 1966; Mendels and Hawkins 1967a,b; Hartmann 1968; Snyder 1968, 1972a,b; Kupfer and Foster 1972; Hauri and Supported in part by National Institute of Mental Health Grants MH-25642 and MH30915. Correspondence to: David J. Kupfer, M.D., Western Psychiatric Institute and Clinic, 3811 O’Hara Street, Pittsburgh, PA 15261, U.S.A.

132

Hawkins 1973; Kupfer et al. 1973). These REM sleep alterations have been strongly associated with the primary affective disorders, more specifically the endogenous depressions, and have even been proposed as representing a psychobiological marker for such disorders (Kupfer 1976). Such studies have also confirmed a relationship between the degree of sleep disturbance and the severity of the depression (Snyder 1972b; Spiker et al. 1978). The question of whether a sleep disorder exists in depression is no longer disputed although work in this area continues with studies to ascertain the more discrete features of those disturbances associated with the various subtypes of depression (Giannelli et al. 1968; Kupfer et al. 1972, 1978a; Kupfer and Foster 1975; Coble et al. 1976; Foster et al. 1976). Of more current interest are questions regarding the persistence and/or variability of these sleep features over time. We know, for instance, in assessing the relationship of sleep, drug, and clinical course, that major and immediate changes occur in these sleep measures when an antidepressant medication is administered to a depressed patient (Zung 1969; Wyatt et al. 1971) and that these changes occur differentially in drug responders and non-responders (Kupfer et al. 1976); but what happens to these sleep features if the treatment intervention involves non-pharmacologic modalities? How variable are they in the absence of clinical change? It was questions such as these, which have not been previously well explored, that we addressed when we designed our investigations of EEG sleep in depression. METHOD

All of the subjects studied were inpatients on the Clinical Research Unit (CRU) of Western Psychiatric Institute and Clinic (WPIC). Informed consent was obtained from them and their families. This facility has been specifically designed for the comprehensive clinical assessment and treatment of adults with depressive syndromes as well as for the pursuit of a variety of research and educational endeavors and operates as a therapeutic community in which all patients participate in a wide range of consistent psychosocial treatment modalities. At the time of admission, in addition to a standard psychiatric interview and physical examination, collateral information was obtained from each patient’s family and from case records of previous hospitalizations. During a two-week drug-free period, patients underwent a series of routine laboratory tests including thyroid function tests, an electroencephalograph, and other tests as indicated based on their history or physical examination. After the drug-free period, the Schedule for Affective Disorders and Schizophrenia (SADS) was filled out by-their ward psychiatrist. The SADS, a structured research interview which collects data necessary to make diagnoses using the Research Diagnostic Criteria (RDC) (Spitzer et al. 1978), was completed using information obtained from the initial interview, the case record, collateral information from relatives, observation on the CRU, and when necessary, a second interview with the patient. If the level of

133

severity remained sufficiently high [a minimum score of 30 on the 17-item Hamilton Rating Scale (HRS) using the sum of the two raters] patients entered the actual protocol and were subsequently rated twice weekly for severity of depression using the HRS throughout the investigation. All patients received 4 idential “study capsules” daily during the 35-day protocol, so that both the patient and the staff remained “blind” to the medication. The first 7 days represented a placebo period for all subjects. During the succeeding 4 weeks, patients were randomly assigned to receive either placebo or amitriptyline. EEG sleep was recorded nightly and as in our previous investigations, all sleep records were scored blindly. Further, in addition to the standard classification of sleep by stages, all records were scored for REM activity on a scale of O-8 for each minute of REM sleep. This report will deal only with those 12 non-delusional depressed patients receiving placebo throughout the entire 35-day protocol. As shown in Table 1, the group included 8 females and 4 males with a mean age of 34.5 years. Their mean Hamilton score was 40.2 + 3.6 at the beginning of the protocol and 26.8 + 3.2 at the end of the study. Other ratings including the Brief Psychiatric Rating Scale (BPRS) and the Global Raskin Scale confirmed the initial degree of severity in this sample. RDC diagnostic characteristics are also shown in Table 1.

TABLE

1

SAMPLE CHARACTERISTICS ON PLACEBO FOR 5 WEEKS Clinical

OF 12 NON-DELUSIONAL

PRIMARY

DEPRESSIVES

characteristics

Age Sex Initial Hamilton Rating Final Hamilton Rating Initial Self-Rated Depression (KDS 1 + 2) Final Self-Rated Depression (KDS 1 + 2) Global Raskin Score (initial) Brief Psychiatric Rating Scale (initial) Diagnostic characteristics Primary Unipolar Bipolar Endogenous Recurrent Incapacitating Agitated Retarded

34.5 * 3.2 years 8 females, 4 males 40.2 * 3.6 26.3 * 3.2 33.0 * 1.9 26.1 * 2.6 10.1 + 0.9 12.3 * 1.9

a 12 11 1 11 5 7 3 8

a Diagnostic Characteristics as defined by the Research Diagnostic Criteria (Spitzer, R.C., Endicott, J. and Robins, E., Research Diagnostic Criteria, Arch. Gen. Psychiat., 35 (1978) 773-782).

134 DATA

ANALYSIS

For purposes of this report, in which all subjects received placebo throughout the protocol period, data were examined on the basis of weekly means. Each data point represented the average of one subject for one period. Major testing was performed by repeated measures ANOVA on the 12 patients, followed by the appropriate post-hoc tests in the presence of significant F-ratios as well as specified contrasts to determine the strength of the linear components. In this manner, within group differences during baseline versus subsequent weeks were examined. In addition, the application of repeated ANOVA measures allowed examination of any underlying dynamics of change over time. RESULTS

As can be seen in Table 2, measures of sleep continuity and non-REM sleep architecture showed very little change in this group of patients over the course of the 5-week protocol. The only exceptions were a significant increase in Stage 2 percent (P < 0.01) and a significant decrease in the percent of Stage 2-REM (P < 0.05). REM sleep measures, however, did show some alterations over time (see Table 3). Stage l-REM percent along with REM activity, average REM activity, and REM density all showed significant decreases from Week 1 to the end of the protocol (P < 0.001, < 0.01, < 0.001 and < 0.05, respectively). In contrast, REM latency and the number of REM periods remained relatively constant over time. Clinical data pointed to a significant decrease in the HRS from Week 1 to Week 2 (40.3 + 3.6 to 31.3 + 2.9, P < 0.01) with further improvement by Week 4 (26.3 + 3.8, P < -0.05). Since the most pronounced clinical changes appeared to have occurred between Weeks 1 and 2, the protocol was divided into two periods: the first 14 days and the subsequent 21 days. When comparing these two time periods, the HRS improved from 35.8 f 2.5 to 27.2 + 1.9 (P < 0.01) and selfrated depression (KDS) showed similar improvement (31.0 + 1.8 to 26.4 + 1.5, P < 0.05), however, there were no significant differences in any of the sleep measurements when comparing Weeks 1 and 2 to Weeks 3-5. As described above, 4 of the REM and two of the non-REM sleep measurements had tested as significant on repeated measures. To follow-up the ANOVA design, the linear trend was abstracted for each variable which accounted for most of the between period variance. Since each test for linear trend was significant, leaving a non-significant residual, the strength of the linear trends over the 5-week comparison could be interpreted in a straightforward way (see Table 4). DISCUSSION

While EEG sleep studies have confirmed the existence of a sleep disorder in depression and continue to contribute valuable information in furthering

2

f + + ?

1.2 2.2 0.5 1.0

8.8 64.5 1.5 1.8

+ + c t

1.4 1.4 0.5 0.7

86.6 ? 1.6

0.2 0.2

48.3 f 6.8 24.9 ?- 2.0 136.2 f 22.7 0.41 f 0.6

REM latency (min) Stage l-REM % REM activity Units Average REM activity (RA/TSA) REM density (RA/RT) Number of REM periods

1.55 f 3.5 *

Week 1

0.2 0.2

+ 8.3 f 1.5 + 18.9 * 0.6

1.62 ? 3.4 +

50.2 23.4 132.7 0.40

Week 2

IN 12 NON-DELUSIONAL

8.9 62.5 1.6 2.1

2 2.7

Sleep parameter

REM SLEEP ARCHITECTURE

3

1 % 2 % 2-REM % (Stage 3 + 4) %

TABLE

Stage Stage Stage Delta

84.3

5.0 + 1.0

6.0 f 1.4

%)

32.2 f 5.8 3.2 + 1.3

35.2 + 7.6 7.7 + 3.6

Sleep latency (min) Early morning wakefulness (min) Intermittent wakefulness (A/TSA %) Sleep efficiency (TSA/TRP + + * *

2.1 1.4 0.4 0.9

0.2 0.2

+ 5.6 f 1.6 + 16.5 + 0.5 1.52 + 3.3 +

47.6 22.7 120.7 0.36

Week 3

PRIMARY

9.4 64.9 1.2 1.8

85.9 + 1.9

4.9 i 1.2

32.1 f 8.0 7.2 + 2.3

Week 3

ARCHITECTURE

Week 2

SLEEP

Week 1

NON-REM

Sleep parameter

SLEEP CONTINUITY AND PLACEBO FOR 5 WEEKS

TABLE 12

* + t ?

1.7 1.3 0.3 0.6

1.45 * 3.3 f

0.2 0.2

+ + * *

1.6 0.7 0.3 0.6

* 1.9 0.19 4.30 3.25 1.07

0.46

0.85

0.31 1.50

0.1 0.2

r 5.1 i 1.6 r 14.4 f 0.4 1.41 + 3.4 r

54.6 21.8 107.9 0.32

Week 5

FOR

2.69 1.16

1.53 6.65 3.58 5.84

Repeated measures F-ratio

5 WEEKS

ON

0.05 NS

NS 0.001 0.01 0.001

Probability

0.01 0.05 NS

NS

NS

NS

NS

NS

Probability

DEPRESSIVES

measures

PRIMARY

Repeated F-ratio

ON PLACEBO

9.4 66.5 1.0 1.3

85.6

5.2 ? 1.0

31.8 + 7.2 6.6 * 3.0

56.6 i 6.9 21.1 5 1.7 108.8 + 16.4 0.33 f 0.5

Week 4

DEPRESSIVES

9.0 67.8 1.0 1.2

85.9 * 1.7

3.8 * 0.7

31.1 i 6.1 10.9 * 4.7

Week 5

NON-DELUSIONAL

Week 4

IN

z ul

136 TABLE

4

SPECIAL CONTRASTS OVER PRESSIVES ON PLACEBO

5 WEEKS

IN 12 NON-DELUSIONAL

Linear component F-ratio (degrees of freedom

1,44)

PRIMARY

DE-

Non-linear component F-ratio (degrees of freedom 3,44)

REM sleep parameters REM activity Average REM (RA/TSA) REM density Stage l-REM

units activity

11.22 13.59

c d

1.03 a 1.59 a

(RA/RT) percent (RT/TSA)

6.07 b 18.60 d

1.56 a 2.67 a

11.33 c 9.47 c

1.95 a 1.18 a

Non-REM sleep parameters Stage 2 percent Stage 2-REM percent aPNS. b P < 0.05. c P < 0.01. d P < 0.001.

our current understanding of the affective illnesses, most such findings have been derived from only a few baseline recordings. Well controlled longitudinal studies have been sparse, and of these, most have involved single case studies of psychotically depressed patients. Within the context of an intensive, double-blind, drug-placebo investigation of BEG sleep in depression, we had the unique opportunity of examining the sleep of a group of carefully diagnosed hospitalized primary depressives who received no active drug treatment over the course of a 5week study period. This present report has dealt only with those patients who were non-delusional and who completed the entire 35 day study. Our findings demonstrate a relative lack of major change in the sleep parameters across time, particularly in those sleep measures which have been reported to be characteristic of a ‘biological’ depression. The group, as a whole, did not exhibit a very marked sleep continuity disturbance which may in part have been due to the sample characteristics; that is, this was a relatively young group of non-delusional depressives with a preponderance of retarded versus agitated features. Nevertheless, although not markedly impaired at the onset of the study, there were no significant changes in measures of sleep continuity across time. The group did, however, exhibit the characteristic delta sleep reduction and shortened REM latency from the beginning of the protocol period and these findings persisted throughout the study despite vigorous psychosocial intervention and placebo. In fact, all the patients in this study ultimately required pharmacologic intervention in order to be discharged from the hospital.

137

Some changes in sleep, specifically in the REM sleep measures, were noted. Gradual but significant decreases in REM sleep percent and in various measures of REM activity (the ‘phasic’ component of REM sleep) were particularly interesting in view of the clinical improvement evidenced during the protocol period. Although these clinical changes did not reach a state of clinical remission or for that matter even a partial remission which we had defined as requiring an HRS score of <20 (sum of two raters), there was nevertheless, a significant decrease in both HRS and self-rated depression from the beginning to the end of the protocol. EEG sleep studies have reported changes in REM sleep, particularly in the phasic aspects of REM sleep, to be associated with more discrete changes in mood or clinical state (Ham-i and Hawkins 1971; Kupfer and Heninger 1972; Kupfer and Foster 1973). Studies of the major EEG sleep changes associated with complete clinical remission have almost exclusively involved patients who were being treated with a tricyclic antidepressant. In these studies, patients who were considered drug responders showed decreased sleep discontinuity as well as a sustained prolongation of REM latency and marked decreases in REM sleep percent (Kupfer et al. 1976). Recent work by our group also suggests that while REM latency and the number of REM periods appear to remain suppressed by a tricyclic antidepressant in drug responders, REM sleep percent and measures of REM activity show a partial, and in the case of REM activity, a near complete tolerance to this suppressant effect over time (Kupfer et al. 1978b). The findings of this study would, therefore, tend to support the hypothesis that depressed patients with sleep changes characteristic of a ‘biologic’ depressive disorder do not respond in a dramatic manner to non-drug treatment interventions. While there may be some clinical improvement and some changes in sleep, particularly in the phasic aspects of REM sleep, the major sleep variables associated with these disorders, namely a shortened REM latency and delta sleep reduction, show essentially no change over a 5-week period of intensive non-pharmacologic intervention. ACKNOWLEDGEMENTS The authors gratefully acknowledge the assistance of Richard the nursing staff of the Clinical Research Unit at WPIC.

Ulrich,

David Shaw, and

REFERENCES Coble, P.A., Foster, F.G. and Kupfer, D.J., EEG sleep diagnosis of primary depression, Arch. Gen. Psychiat., 33 (1976) 1124-1127. Diaz-Guerrero, R., Gottlieb, J.S. and Knott, J.R., The sleep of patients with manic depression, Brit. J. Psychiat., 112 (1946) 1263-1267. Foster, F.G., Kupfer, D.J., Coble, P.A. and McPartland, R.J., Rapid eye movement sleep density - An objective indicator in severe medical-depressive syndromes, Arch. Gen. Psychiat., 33 (1976) 1119-1123.

138 Giannelli, A., Penati, G. and Pietropolli-Charmet, G., Some considerations on the psychiatric aspects of insomnia. In: H. Gastaut. E. Lugaresi, G.B. Ceroni and G. Coccogna (Eds.), The Abnormalities of Sleep in Man, Au10 Gaggi Editore, Bologna, 1968, pp. 127-134. Hartmann, E., Longitudinal studies of sleep and dream patterns in manic-depressive patients, Arch. Gen. Psychiat., 19 (1968) 312-329. Hauri, P. and Hawkins, D.R., Phasic REM, depression, and the relationship between sleeping and waking, Arch. Gen. Psychiat., 25 (1971) 56-63. Hauri, P. and Hawkins, D.R., Individual differences in the sleep of depression. In: U.J. Jovanovic (Ed.), The Nature of Sleep, Gustav Fischer Verlag, Stuttgart, 1973, pp. 193-197. Hawkins, D.R. and Mendels, J., Sleep disturbances in depressive syndromes, Amer. J. Psychiat., 123 (1966) 682-690. Kupfer, D.J. and Heninger, G.R.,REM activity as a correlate to mood changes throughout the night (EEG sleep patterns in a patient with a 48 hour cyclic mood disturbance), Arch. Gen. Psychiat., 27 (1972) 368-373. Kupfer, D.J. and Foster, F.G., Interval between onset of sleep and rapid eye movement sleep as an indicator of depression, Lancet, 2 (1972) 684-686. Kupfer, D.J., Himmelhoch, J., Swartzburg, M., Anderson, C., Byck, R. and Detre, T.P., Hypersomnia in manic depressive disease, Dis. Nerv. Syst., 33 (1972) 720-724. Kupfer, D.J., Foster, F.G. and Detre, T.P., Sleep continuity changes in depression, Dis. Nerv. Syst., 34 (1973) 192-195. Kupfer, D.J. and Foster, F.G., Sleep and activity in a psychotic depression, J. Nerv. Ment. Dis., 156 (1973) 341-348. Kupfer, D.J. and Foster, F.G., The sleep of psychotic patients: Does it all look alike? In: D.X. Freedman (Ed.), The Biology of the Major Psychosis - A Comparative Analysis, Raven Press, New York, 1975, pp. 143-164. Kupfer, D.J., REM latency - A psychobiological marker for primary depressive disease, Biol. Psychiat., 2 (1976) 159-174. Kupfer, D.J., Foster, F.G., Reich, L., Thompson, K. and Weiss, B.L., EEG sleep changes as predictors in depression, Amer. J. Psychiat., 133 (1976) 622-626. Kupfer, D.J., Spiker, D.G., Coble, P.A. and Shaw, D.H., EEG sleep and depression in the elderly, J. Amer. Geriat. Sot., 26 (1978a) 53-57. Kupfer, D.J., Spiker, D.G., Coble, P.A. and McPartland, R.J., Amitriptyline and EEG sleep in depressed patients, Part 1 (Drug effect), Sleep, l/2 (197813) 149-159. Mendels, J. and Hawkins, D.R., Sleep and depression - A controlled EEG study, Arch. Gen. Psychiat., 15 (1967a) 744-754. Mendels, J. and Hawkins, D.R., Sleep and depression - A follow-up study, Arch. Gen. Psychiat., 16 (196713) 536-542. Snyder, F., Electroencephalographic studies of sleep in depression. In: N.S. Kline and E. Lasha (Eds.), Computers and Electronic Devices in Psychiatry, Grune and Stratton, Inc., New York, 1968, pp. 272-301. Snyder, F., NIH studies of EEG sleep in affective illness. In: T. Williams, M. Katz and J. Shields (Eds.), DHEW, Washington, 1972a. Snyder, F., Electroencephalographic studies of sleep in psychiatric disorders. In: M.H. Chase (Ed.), The Sleeping Brain, Brain Information Service/Brain Research Institute, Los Angeles, 1972b, pp. 376-393. Spiker, D.G., Coble, P.A., Cofsky, J., Foster, F.G. and Kupfer, D.J., EEG sleep and severity of depression, Biol. Psychiat., 13 (1978) 485-488. Spitzer, R.C., Endicott, J. and Robins, E., Research Diagnostic Criteria, Arch. Gen. Psychiat., 35 (1978) 773-782. Wyatt, R., Fram, D., Kupfer, D.J. and Snyder, F., Total prolonged drug-induced rapid eye movement sleep suppression in anxious-depressed patients, Arch. Gen. Psychiat., 24 (1971) 145-155. Zung, W.W.K., Effect of antidepressant drugs on sleeping and dreaming, Part 3 (On the depressed patient), Biol. Psychiat., 1 (1969) 283-287.