Horst Gann, Bernd Feige, Fritz Hohagen, Dietrich van Calker, Dagmar Geiss, and. Riemann Dieter. Background: The present study investigated polysomno-.
Sleep and the Cholinergic Rapid Eye Movement Sleep Induction Test in Patients with Primary Alcohol Dependence Horst Gann, Bernd Feige, Fritz Hohagen, Dietrich van Calker, Dagmar Geiss, and Riemann Dieter Background: The present study investigated polysomnographically assessed sleep parameters in alcohol-dependent patients after withdrawal and in healthy control subjects during baseline and after a cholinergic stimulation paradigm. The aim of the study was to test whether sleep parameters, especially rapid eye movement (REM) sleep variables, may serve as predictors for relapse in alcohol-dependent patients. Methods: Forty patients diagnosed with alcohol dependence were admitted to a specialized ward for alcohol withdrawal and were investigated by polysomnography at three time points: 2–3 weeks after withdrawal (T0) and at follow-up investigations 6 (T1) and 12 (T2) months after discharge from the hospital. A subgroup of patients (n ⫽ 17) was studied at T0 after challenge with galanthamine, a reversible cholinesterase inhibitor (cholinergic REM induction test, CRIT). Patients were compared with two control groups: a) 30 healthy control subjects (matched for age- and gender-distribution) for comparison at baseline conditions; and b) 17 age- and gender-matched control subjects for comparison with the CRIT. Results: At baseline the patients showed significant disturbances of sleep continuity and sleep architecture (decreased slow-wave sleep, SWS) and exhibited an increase of “REM sleep pressure” (a combined index of REM latency, REM density, and REM sleep percent). Galanthamine provoked significant alterations of sleep continuity, sleep architecture (reduced SWS), and increased most of the components of REM pressure, taking patients and control subjects together. Apart from SWS %SPT (sleep period time) no significant drug– group interactions occurred. Patients who remained abstinent (n ⫽ 11) for at least 6 months at follow-up exhibited significantly less abnormalities of REM sleep at T0 compared to the group of patients that relapsed at 6 months follow-up.
From the Department of Psychiatry and Psychotherapy of the University of Freiburg (HG, BF, DvC, DG, RD), and the Department of Psychiatry and Psychotherapy, University of Lu¨beck (FH), Germany. Address reprint requests to Horst Gann, MD, Senior Psychiatrist, University of Freiburg, Psychiatric Department, Hauptstr. 5, Freiburg D-79104, Germany. Received November 1, 2000; revised April 4, 2001; accepted April 11, 2001.
© 2001 Society of Biological Psychiatry
Conclusions: It is concluded that increased REM sleep pressure after alcohol withdrawal is a robust predictor of vulnerability to relapse. Thus, a subgroup of alcoholic patients appears to exhibit distinct neurobiological abnormalities assessable by polysomnography that are related to an increased vulnerability for alcoholism and early relapse. Biol Psychiatry 2001;50:383–390 © 2001 Society of Biological Psychiatry Key Words: Alcoholism, relapse, REM sleep
Introduction
C
omplaints of sleep disturbances are common among alcohol-dependent patients during periods of drinking, as well as during acute withdrawal or subacute and chronic abstinence (Allen et al 1971; Drummond et al 1997; Gross and Hastey 1976; Williams and Rundell 1981). Recovered patients may show persistent sleep abnormalities for months or even years (Wagman and Allen 1975). Gillin et al (1990) reported that compared with normal control subjects, primary alcoholics after detoxification exhibited prolonged sleep onset latencies, reduced total sleep time, and impaired sleep efficiency. Alcoholic patients also showed reduced non–rapid eye movement (NREM) sleep. Particularly prominent were reductions in stage 2, stage 4 sleep, and total delta sleep. REM (rapid eye movement) sleep abnormalities, such as shortened REM latency and a higher REM density of the first REM period, were also important significant findings in the patient sample. In a longitudinal study of sleep in primary alcoholism Gillin et al (1986) provided evidence that profound disturbances of sleep, such as short REM latency, persisted in about half of the patients with primary alcoholism during a period of abstinence of 3 months. In nondepressed patients with primary alcoholism, REM sleep abnormalities like shortened REM latency and increased REM percentage (“REM pressure”) at the time of admission 0006-3223/01/$20.00 PII S0006-3223(01)01172-6
384
H. Gann et al
BIOL PSYCHIATRY 2001;50:383–390
were of predictive value for relapses at 3 months after discharge from the hospital (Gillin et al 1994). Additionally, disturbances in self-rated subjective sleep quality were found in some studies (Brower et al 1998; Foster and Peters 1999; Skoloda et al 1979) to predict later relapse. These findings suggest that sleep abnormalities may be a characteristic feature of patients suffering from alcoholism. The sleep abnormalities during abstinence indicate that this is not merely an epiphenomena of acute intoxication or withdrawal or associated depressive symptoms. Sleep abnormalities may be a relevant predictor for relapses in currently abstinent patients, which could be clinically relevant. In recent years our group has investigated the impact of several cholino-mimetics on the sleep of healthy volunteers (Berger et al 1983; Hohagen et al 1993; Riemann et al 1988, 1994a, 1996). The cholinergic REM induction test (CRIT) was also applied in patients with several psychiatric disorders (Berger et al 1989; Gann et al 1992; Lauer et al 1988; Riemann et al 1991a, 1991b, 1994b, 1994c), but up to now studies in alcohol-dependent patients have not been performed. The CRIT provokes changes in sleep continuity and REM sleep distribution in healthy subjects, which are even more pronounced in psychiatric patients, especially depressed samples, arguing for a cholinergic involvement in REM sleep regulation and some forms of psychopathology. The present study was undertaken to assess the predictive value of sleep abnormalities in alcoholics regarding relapses and to investigate potential neurochemical mechanisms underlying the increased “REM pressure.” For the latter purpose we administered the CRIT. To this end we examined 1) Clinical and sleep parameters in a longitudinal design at 2–3 weeks of abstinence as well as 6 and 12 months after discharge from the hospital; and 2) the influence of the cholinesterase inhibitor galanthamine hydrobromide on the sleep of patients with primary alcohol dependence at 2–3 weeks of abstinence.
Methods and Materials Subjects Forty patients (age 44 ⫾ 9 years; range, 24 – 62 years; 11 females: 44 ⫾ 9 years; 29 males: 43 ⫾ 9 years) with primary alcohol dependence (see Table 1) were studied. They were admitted for treatment of alcoholism to a specialized ward that offers a 21-day inpatient treatment program for alcohol withdrawal and motivational therapy. The patients were asked and consented to participate in a multidisciplinary longitudinal investigation. As part of this study, subjects were asked to spend 3 consecutive nights in the sleep laboratory while they were in the third week of hospitalization (T0), and at follow-up investigations 6 (T1) and 12 (T2) months after discharge. A subgroup of these 40 patients (n ⫽ 17, age 44 ⫾ 8 years; 5 females, 12 males)
Table 1. Clinical and Demographic Characteristics of Patients (n ⫽ 40) with Primary Alcoholism (Means ⫾ SD) Age Sex (males in %) 21-Hamilton Depression Rating Scale Duration of alcoholism (years) Free of medication before T0 (days) Secondary psychiatric diagnosis Positive family history for alcoholism Number of days since last drink before investigation Number of days with alcohol use in the 6 months (180 days) prior to admission Daily alcohol consumption per drinking day (grams) in the last 6 months GGT, IU (admission) GOT, IU (admission) GPT, IU (admission)
44 ⫾ 9 72.5 4⫾5 13 ⫾ 9 15 ⫾ 6 n ⫽ 1 (dysthymia) 53.3% 20 ⫾ 8 122 ⫾ 71 187 ⫾ 130 163 ⫾ 241 37 ⫾ 30 31 ⫾ 23
GGT,; GOT,; GPT,; IU,.
was additionally studied with galanthamine as cholinergic challenge (CRIT) at T0. Patients were compared with two healthy control groups: a) 30 age- and gender-matched subjects (age 43 ⫾ 9 years; range, 24 – 61 years; 10 females: 42 ⫾ 7 years, 20 males: 43 ⫾ 10 years) for comparison at baseline conditions; and b) 17 age- and gender-matched subjects (age 42 ⫾ 7 years; 5 females, 12 males) for comparison after the CRIT. The control group b) includes data of healthy subjects published previously (Riemann et al 1994a). Family and personal history of alcohol dependence or abuse was exclusionary criteria for control subjects. All patients and control subjects underwent a medical and psychiatric investigation and physical examination: routine clinical and hematologic laboratory examinations, urinanalysis, electrocardiogram, electroencephalogram, and magnetic resonance imaging (only patients). All patients met DSM-III-R criteria for alcohol dependence (diagnoses were made with a structured clinical interview, SCID, German version [Wittchen et al 1989]). Eligible patients were required to have no other psychiatric illness or significant medical problem before the onset of alcoholism. Thus, we excluded patients with psychotic features, clinically significant cognitive impairment, antisocial personality disorder, substance abuse other than alcohol, and major medical problems (i.e., cirrhosis of the liver). Patients with secondary major depression or dysthymia were not excluded. All patients were free of psychoactive medication for a minimum of 7 days before the investigation. The study was approved by the local ethics committee. Each patient and control subject signed written informed consent before inclusion.
Drug Galanthamine hydrobromide, an alkaloid isolated from the snowdrop galanthus nivalis, is a reversible and selective inhibitor of acetylcholinesterase. Maximal plasma concentrations of the drug are reached 2 hours after oral ingestion. The elimination half-life time is 5.3 (SD ⫽ 4.2) hours (for detailed information, see Riemann et al 1994a).
Sleep in Alcoholism
Design Patients and control subjects slept in the sleep laboratory for 3 nights. In the first (adaptation) night they were screened for sleep apnea and periodic leg movements during sleep (PLMS). Subjects with an apnea-hypopnea index ⱖ 10/hour or relevant PLMS were excluded from further study. Seventeen patients and 17 control subjects participated in the CRIT: At 9:00 PM on nights 2 and 3 a single oral dose of 10 mg galanthamine or placebo (baseline night) was administered in a double-blind randomized crossover design. Patients with contraindications against cholinergic drugs (n ⫽ 23) received placebo on both nights. The clinical outcome was determined at follow-up investigations 6 and 12 months after discharge from the hospital. Relapse was defined as any consumption of alcohol between hospital discharge and follow-up, evaluated on the basis of standardized interviews. Additionally, measurements of blood alcohol level and hepatic enzyme levels, including carbonhydrat-deficient transferrin (CDT), were performed. Any elevated hepatic enzyme level was defined as relapse. Patients abstinent at 6 (T1) and 12 (T2) months after hospital discharge were again assessed in the sleep laboratory (n ⫽ 11).
Statistics Arithmetic mean and SD were calculated within groups and the significance of group differences assessed by t tests (either for dependent or independent samples) or, for nongaussian variables, by Fisher’s Exact Test for independent samples. The time course of parameters across T0, T1, and T2 and a statistical analysis of the CRIT was assessed by repeated measurements analysis of variance (ANOVA) (with Greenhouse-Geisser correction). The variable “REM pressure” was calculated by means of a factorial analysis (principal component analysis, PCA) based on the parameters REM density in REM sleep period 1, REM percent sleep period time (SPT), and REM latency as suggested by Gillin et al (1994). One PCA was performed to derive the coefficients for REM pressure and this same formula was used for all subjects. The PCA was performed within the patient group only (40 patients), because the hypothesis was that REM parameters were altered in this group relative to the control group: PCA would probably describe as first factor the group difference between alcoholic patients and control subjects when using both groups, instead of just extracting the covariation in parameters for REM propensity. The formula was: REM pressure ⫽ 0.188 ⫻ ([first REM density ⫺ 29.97]/15.0975) ⫺ 0.587 ⫻ ([REM latency ⫺ 54.775]/ 27.5057) ⫹ 0.579 ⫻ ([REM %SPT ⫺ 23.395)/5.8828). Discriminant function analysis with cross-validation by a leave-one-out scheme was used to quantify the predictive power of REM pressure for later relapse.
Sleep Recording and Scoring For a detailed description of the algorithms used for recording of sleep parameters in our laboratory, see Riemann et al (1994a). Briefly, sleep was recorded between “lights out” (11:00 PM) and “lights on” (7:00 AM) using standard procedures (electroencephalogram: C 4-A1; C 3-A2; horizontal EOG, submental electro-
BIOL PSYCHIATRY 2001;50:383–390
385
myogram). Records were scored by raters blind to experimental conditions according to standardized criteria (Rechtschaffen and Kales 1968). Polysomnograms were then evaluated for parameters of sleep continuity, architecture, and REM sleep. REM latency was defined as time from sleep onset (10 min continuous stage 2, interrupted only for 1 min by either stage “wake” or stage 1), to the first epoch of 3 consecutive min of REM sleep minus intermittent wake time. REM %SPT was defined as REM sleep in percent of sleep period time (SPT). The “Pittsburgh Sleep Quality Index” (PSQI; Buysse et al 1989) was used to ascertain the subjective sleep quality at each assessment point (T0, T1, T2) and in healthy subjects.
Results Sleep Parameters at the Third Week of Hospitalization (T0) BASELINE CONDITIONS. Data from 40 patients were compared with 30 healthy control subjects matched for age- and gender-distribution (see Table 2). Patients showed significant disturbances in sleep continuity (decreased sleep efficiency, increase in number of awakenings, and total time awake) and in sleep architecture (decreased stage 2 percent and slow-wave sleep percent [SWS], increased stage 1 percent). Furthermore, patients exhibited an increased REM sleep pressure (Figure 1) with elevated total REM density, enhanced REM density of the first four REM periods, and decreased REM latency. The disturbances in the sleep of patients was also evident from the PSQI values (Table 2). CHOLINERGIC REM INDUCTION TEST (CRIT). The CRIT was administered to 17 patients and 17 age- and gender-matched control subjects. No side effects of galanthamine were observed. Table 3 summarizes the data for patients and control subjects and results of the ANOVA. Group differences were in the same direction as observed in the analysis of baseline data. Drug effects (galanthamine vs. placebo) included a significant prolongation of S2 latency, an increase of time awake (%SPT), a reduction of stages 2 and SWS (%SPT) and an increase of stage 1 (%SPT). For REM sleep parameters, galanthamine significantly prolonged the first REM period. It also enhanced the REM density of the first REM period and total REM density significantly. Only one significant treatment ⫻ group interaction emerged for SWS %SPT. The significant result for SWS (%SPT) is probably due to a “floor” effect in the patients.
Predictive Value of Sleep Variables under Baseline Conditions at T0 for Outcome at 6 Months At the 6 months follow-up (T1), 23 patients had relapsed, and 11 had remained abstinent. No information on clinical
386
H. Gann et al
BIOL PSYCHIATRY 2001;50:383–390
Table 2. Comparison (Two-Tailed t Test) of Sleep Variables under Baseline Conditions and Values of the Pittsburgh Sleep Quality Index (PSQI) during the Third Week of Hospitalization (T0) between 40 Patients with Primary Alcohol Dependence and 30 Healthy Control Subjects Mean ⫾ SD Patients (n ⫽ 40) Sleep variables Sleep continuity Sleep efficiency (%) S-2 latency (min) No. of awakenings Time awake (%SPT) Sleep architecture Stage 1 (%SPT) Stage 2 (%SPT) SWS (%SPT) REM Sleep REM latency (min) REM (%SPT) REM period 1 duration (min) 1. REM density (%) 2. REM density (%) 3. REM density (%) 4. REM density (%) Total REM density (%) REM pressure index PSQI score
Mean
SD
Control subjects (n ⫽ 30) Mean
SD
t test (two-tailed) t
p
84.20 18.20 25.23 9.98
8.21 12.77 8.86 5.95
89.31 12.92 15.10 6.00
6.74 10.88 10.17 5.67
2.775 1.865 4.440 2.823
.007 .067 ⬍.001 .006
11.16 54.11 1.06
4.31 5.85 3.31
7.37 58.69 4.89
4.27 6.92 5.51
3.658 2.994 3.378
⬍.001 .004 .002
53.86 23.28 18.21 30.01 39.70 39.88a 39.57b 38.03 0.000 6.8
27.93 5.85 14.10 15.04 14.57 14.23 14.14 11.57 1.000 3.3
77.08 22.40 19.30 19.53 23.12 26.05 27.36c 25.55 ⫺0.692 3.6
34.02 5.00 15.35 9.81 10.18 8.34 9.43 7.05 0.875 2.1
3.047 0.664 0.307 3.522 5.603 5.050 4.039 5.576 3.018 4.75
.004 .509 .759 .001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 .004 .001
n ⫽ 39 n ⫽ 35 c n ⫽ 26 SPT, sleep period time; REM, rapid eye movement sleep; PSQI, Pittsburgh Sleep Quality Index. a b
status was obtainable for the remaining 6 patients. The demographic and clinical variables of relapsed and abstinent patients did not differ significantly (data not shown); however, relapsed patients exhibited significantly (p ⫽ .03, two-tailed t test) higher REM pressure index at T0 than abstinent patients (Figure 2), due to higher REM %SPT (Relapsers: 24.57 ⫾ 4.96; Abstainers: 19.51 ⫾ 6.32; p ⫽ .016, two-tailed t test) and higher REM density in the first REM period (Relapsers: 35.47 ⫾ 14.14; Abstainers: 23.89 ⫾ 14.04; p ⫽ .032, two-tailed t test). A comparison of the REM sleep pressure index found in relapsed and abstinent patients revealed that the group of the relapsed but not of the abstinent patients exhibited significantly elevated values of REM sleep pressure at T0 (Figure 2). Discriminant function analysis with intrinsic cross-validation showed that within our sample, 17 of 23 relapsers and 7 of 11 abstainers could be correctly classified by a discriminant REM pressure index of ⫺0.03 (70% correct classification). The parametric tests and the discriminant analysis were recalculated without the patient with very low REM pressure (see Figure 2) in the abstinent group. In the parametric tests, this did not effect the direction of the results, because the outlier increased the variance in the
Figure 1. Comparison of rapid eye movement sleep pressure index under baseline conditions between 40 patients with primary alcohol dependence in the third week of hospitalization (T0) and 30 healthy control subjects (horizontal bars indicate mean values).
Sleep in Alcoholism
BIOL PSYCHIATRY 2001;50:383–390
387
Table 3. Statistical Analysis (ANOVA) of the Results at the Cholinergic REM Induction Test (CRIT) at the Third Week of Hospitalization (T0) of 17 Patients with Primary Alcohol Dependence Compared to Healthy Control Subjects ANOVA
Means ⫾ SD
Sleep continuity Sleep efficiency (%) Placebo Gala S-2 latency (min) Placebo Gala No. of awakenings Placebo Gala Time awake (%SPT) Placebo Gala Sleep architecture Stage 1 (%SPT) Placebo Gala Stage 2 (%SPT) Placebo Gala SWS (%SPT) Placebo Gala REM (%SPT) Placebo Gala REM latency (min) Placebo Gala REM density (%) Placebo Gala REM period 1 duration (min) Placebo Gala 1. REM density (%) Placebo Gala 2. REM density (%) Placebo Gala 3. REM density (%) Placebo Gala 4. REM density (%) Placebo Gala
Group ⫻ treatment (df ⫽ 1)
Treatment (df ⫽ 1)
Group (df ⫽ 1)
Patients
Controls
F
p
F
p
F
p
83.1 ⫾ 9.6 81.9 ⫾ 8.9
91.2 ⫾ 7.0 85.7 ⫾ 9.9
5.40
.027
3.86
.058
0.87
.357
19.1 ⫾ 15.6 30.2 ⫾ 27.7
13.2 ⫾ 10.2 24.6 ⫾ 17.6
1.44
.239
6.84
.013
0
.973
23.1 ⫾ 9.2 26.1 ⫾ 12.4
12.2 ⫾ 8.6 14.7 ⫾ 8.1
15.68
.000
2.44
.128
0.03
.869
10.3 ⫾ 5.1 11.9 ⫾ 6.3
4.0 ⫾ 4.6 7.6 ⫾ 6.6
10.87
.002
5.54
.025
0.86
.361
10.6 ⫾ 4.9 12.2 ⫾ 6.0
7.5 ⫾ 3.5 8.8 ⫾ 5.8
3.94
.056
4.76
.037
0.08
.780
55.6 ⫾ 6.7 51.8 ⫾ 7.7
58.9 ⫾ 4.6 55.3 ⫾ 6.6
3.18
.084
11.08
.002
0.01
.927
0.1 ⫾ 0.3 0.1 ⫾ 0.2
5.7 ⫾ 6.4 3.5 ⫾ 4.7
11.81
.02
7.4
.010
7.16
.012
22.8 ⫾ 4.3 23.5 ⫾ 5.5
23.1 ⫾ 4.3 24.5 ⫾ 6.7
0.18
.672
1.01
.323
0.09
.76
56.0 ⫾ 20.0 56.4 ⫾ 43.1
68.4 ⫾ 27.3 49.2 ⫾ 32.3
0.09
.768
1.95
.172
2.12
.155
36.9 ⫾ 11.6 40.0 ⫾ 9.7
20.7 ⫾ 6.9 24.0 ⫾ 8.1
32.36
.000
5.08
.031
0.01
.937
15.9 ⫾ 12.0 29.6 ⫾ 20.6
17.0 ⫾ 7.5 29.4 ⫾ 13.8
0.01
.908
21.94
.000
0.06
.809
26.8 ⫾ 16.1 41.0 ⫾ 14.1
14.5 ⫾ 6.5 24.0 ⫾ 11.6
18.02
.000
20.66
.000
0.82
.373
37.8 ⫾ 11.5 33.1 ⫾ 12.4
21.5 ⫾ 8.8 21.4 ⫾ 10.2
20.73
.000
1.37
.251
1.17
.287
40.4 ⫾ 16.7 35.8 ⫾ 16.7
18.0 ⫾ 8.6 21.7 ⫾ 14.2
23.84
.000
0.02
.883
1.64
.210
29.5 ⫾ 22.9 37.7 ⫾ 18.6
24.7 ⫾ 9.6 19.5 ⫾ 14.3
6.28
.018
0.17
.687
3.38
.075
ANOVA, analysis of variance; REM, rapid eye movement sleep; SPT, sleep period time; SWS, slow-wave sleep; Gala, galanthamine.
abstinent group, balancing the larger group difference: The significance of the difference, relapsers versus abstainers, changed from p ⫽ .03 to p ⫽ .007. The discriminant
function analysis, when repeated without the outlier, yielded a discriminant value for REM pressure of 0.1 instead of – 0.03 and scored 15 (instead of 17) of 23
388
H. Gann et al
BIOL PSYCHIATRY 2001;50:383–390
Sleep Variables under Baseline Conditions during the Course of Abstinence Eleven patients remained abstinent during the 1-year follow-up period and were thus available for study at the 6 and 12 months follow-up investigations. A repeated measures ANOVA revealed no significant alterations of the sleep parameters during the course of abstinence with the exception of a decrease in the REM density in the second REM period (Table 4, Figure 3). There is, however, a nonsignificant tendency toward normalization of variables of sleep continuity and REM latency. In contrast, subjective sleep quality as determined by the PSQI improved significantly with prolonged abstinence (Table 4).
Discussion
Figure 2. Comparison of rapid eye movement sleep pressure index under baseline conditions in the third week of hospitalization (T0) between 23 alcoholic relapsers and 11 abstainers at the 6 months follow-up (T1).
relapsers and 7 of 10 (instead of 7 of 11) abstainers correctly.
Pronounced disturbances of sleep continuity, sleep architecture, and REM sleep are well documented for alcoholic patients (Gillin et al 1986, 1990, 1994). The goal of the present study was twofold: To verify the potential predictive value of REM sleep abnormalities, especially REM pressure, for early relapse (Gillin et al 1994) in a large independent sample, and to assess potential neurobiological alterations associated with these disturbances. For the latter purpose, we administered the cholinergic REM induction test, a paradigm that has
Table 4. Statistical Analysis (ANOVA) of the Course of Sleep Variables under Baseline Conditions and Values of the Pittsburgh Sleep Quality Index (PSQI) during Abstinence (T0: at the Third Week of Hospitalization; T1: 6 Months after Discharge; T2: 12 Months after Discharge) of 11 Patients with Primary Alcohol Dependence Who Remained Abstinent during the 1-Year Follow-Up T0 Sleep Variables Sleep continuity Sleep efficiency (%) S-2 latency (min) No. of awakenings Time awake (%SPT) Sleep architecture Stage 1 (%SPT) Stage 2 (%SPT) SWS (%SPT) REM Sleep REM latency (min) REM (%SPT) REM period 1 duration (min) 1. REM density (%) 2. REM density (%) 3. REM density (%)a 4. REM density (%)b Total REM density (%) REM pressure index PSQI score
Mean
T1 SD
Mean
b
SD
Mean
SD
F
p
83.68 19.09 29.82 11.31
8.60 14.61 11.69 7.27
86.56 17.32 32.18 9.49
7.26 15.40 11.38 7.02
88.29 13.68 32.73 8.06
5.16 6.33 12.51 4.37
2.629 0.572 0.330 1.341
.110 .536 .682 .281
14.01 55.35 0.32
4.91 6.15 0.54
12.25 59.15 0.53
4.76 6.52 1.02
15.17 57.61 0.57
6.79 6.69 0.94
2.808 1.564 1.698
.091 .234 .218
61.05 18.82 12.50 29.06 34.39 39.46 33.93 34.90 ⫺0.613 7.08
35.63 5.85 6.38 16.20 15.78 17.49 20.42 14.72 1.337 3.42
65.32 18.45 22.45 25.40 24.51 30.12 33.87 29.29 ⫺0.779 4.83
28.19 4.10 13.35 15.80 10.07 16.00 8.11 9.93 0.691 2.41
71.77 18.42 16.64 29.16 31.09 33.35 30.54 32.45 ⫺0.876 4.00
25.64 6.68 8.35 16.81 9.19 18.92 12.49 10.20 1.130 1.28
0.370 0.030 3.585 0.383 4.367 1.011 0.370 1.648 0.202 6.09
.695 .936 .056 .673 .032 .372 .625 .225 .800 .011
n ⫽ 10 n⫽7 ANOVA, analysis of variance; SPT, sleep period time; REM, rapid eye movement sleep. a
ANOVA (df ⫽ 2)
T2
Sleep in Alcoholism
Figure 3. Rapid eye movement sleep pressure under baseline conditions during the course of abstinence (T0, at third week of hospitalization; T1, 6 months after discharge; T2, 12 months after discharge) of 11 patients who remained abstinent during the 1-year follow-up.
already revealed important insight into the neurobiology of REM sleep abnormalities in other psychiatric conditions, especially depression (see, for example, Berger et al 1989). Similar to previous studies (Gillin et al 1986, 1990, 1994), the alcoholic patients in our sample exhibited diminished sleep efficiency, an increase in the number of awakenings and in total time awake, as well as decreased SWS and increased REM sleep pressure in comparison to healthy subjects. In additon, we found that patients who remained abstinent for at least 6 months retrospectively exhibited significantly less abnormalities at T0 in the sub-components of REM pressure compared to the group of patients with relapse at 6 months follow-up. Therefore, an increased REM pressure 2–3 weeks after withdrawal is predictive of a higher risk for relapse within 6 months of abstinence with a specificity of 70%. Regulation of NREM/REM sleep cycles is postulated to occur via the reciprocal interaction of cholinergic and aminergic neurotransmission in neuronal pathways originating in the brain stem (Hobson et al 1986). Increased REM pressure is accordingly conceptualized to be caused by an imbalance of aminergic and cholinergic neurotransmission, caused by either cholinergic “overdrive” or insufficient aminergic activity. Depressive patients often exhibit signs of increased REM pressure, such as shortened REM latency and increased REM density, and are more sensitive than control subjects to the enhancement of
BIOL PSYCHIATRY 2001;50:383–390
389
REM pressure evoked by cholinergic stimulation (Berger et al 1989; Riemannn et al 1994b). The basal sleep profile of the alcoholic patients investigated in the present study resembled that of depressive patients in various aspects, including diminished SWS and increased REM pressure, although they were not depressed as evident from their normal mean HRSD score. Surprisingly, however, the reaction of these patients to stimulation with galanthamine differed from that of depressives. Whereas in our own studies depressives showed an exaggerated response to cholinergic stimulation (see Berger et al 1989; Riemann and Berger 1989, 1992; Riemann et al 1994b), the response of alcoholic patients to galanthamine was not significantly different from that of age- and gendermatched control subjects. These differential responses to cholinergic stimulation of depressive and alcoholic patients probably reflect distinct neurobiological alterations in these two groups. According to the reciprocal interaction model, the increased REM pressure in alcoholics might be explained by a severely compromised aminergic neurotransmission and not primarily by an increased cholinergic tone. Based on results obtained in both patients and animals, a compromised serotonergic neurotransmission has indeed been suggested as one potential biological abnormality in chronic alcoholic patients (for review see Roy et al 1987; Sellers et al 1992). Although both sleep continuity and REM latency in alcoholics appeared to normalize somewhat during abstinence and subjective sleep quality clearly improved, these disturbances nevertheless persisted for extended time periods. Thus, withdrawal might induce a protracted syndrome of serotonergic dysfunction in alcoholics, causing the persisting sleep abnormalities in these patients. Because only a subgroup of alcoholic patients exhibited pronounced abnormalities in REM pressure at T0, and because an increased REM pressure was found to be linked to the susceptibility for later relapse, it is tempting to speculate that a perhaps genetically determined vulnerability of the serotonergic system might pose an increased risk of relapse to a subgroup of alcoholic patients. In this case it might be possible to identify these relapse-prone patients by routine polysomnography early in the course of their illness, possibly by using ambulatory monitoring. This subgroup of patients might even benefit from a targeted treatment by REM-sleep suppressing drugs (e.g., selective serotonin reuptake inhibitors). In conclusion, REM pressure appears to be a robust predictor of relapse in a subgroup of alcoholic patients. The cholinergic REM induction test differentiates alcoholic subjects from depressive patients and thus points to distinct underlying neurobiological abnormalities in depressed and alcoholic patients, though their baseline REM sleep abnormalities appear to be very similar.
390
BIOL PSYCHIATRY 2001;50:383–390
This study was supported by the German ministry for Education and Research (BMBF; FKZ 01 EB 9413) A preliminary report from this study was published in Sleep Research Online, 1998, 1, 92-95
References Allen RP, Faillance LA, Wagman A (1971): Recovery time for alcoholics after prolonged alcohol intoxication. Johns Hopkins Med J 128:158 –164. Berger M, Lund R, Bronisch T, von Zerssen D (1983): REM latency in neurotic and endogenous depression and the cholinergic REM induction test. Psychiatry Res 10:113–123. Berger M, Riemann D, Hochli D, Spiegel R (1989): The cholinergic REM induction test with RS 86: State or trait marker of depression. Arch Gen Psychiatry 46:421– 428. Buysse DJ, Reynolds CF, Monte TH, Berman SR, Kupfer DJ (1989): The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Res 28:193–213. Brower KJ, Aldrich MS, Hall JM (1998): Polysomnographic and subjective sleep predictors of alcoholic relapse. Alcohol Clin Exp Res 22:1864 –1871. Drummond SPA, Gillin JC, Smith TL, Demodena A (1997): Sleep changes and recovery over 27 months in pure primary alcoholic patients. Sleep Res 26:289. Foster JH, Peters TJ (1999): Impaired sleep in alcohol misusers and dependent alcoholics and the impact upon outcome. Alcohol Clin Exp Res 23:1044 –1051. Gann H, Riemann D, Hohagen F, Dressing H, Mu¨ller WE, Berger M (1992): The sleep structure of patients with anxiety disorders in comparison to that of healthy control subjects and depressive patients under baseline conditions and after cholinergic stimulation. J Affect Disord 26:179 –190. Gillin JC, Kripke DF, Butters N, Grant I, Irwin M, Naimie MD, et al (1986): A longitudinal study of sleep in primary alcoholism. Sleep Res 15:92. Gillin JC, Smith T, Irwin M, Kripke DF, Brown S, Schuckit M (1990): Short REM latency in primary alcoholic patients with secondary depression. Am J Psychiatry 147:106 –109. Gillin JC, Smith TL, Irwin M, Buttos N, Demodena A, Schuckit M (1994): Increased pressure for rapid eye movement sleep at time of hospital admission predicts relapse in non-depressed patients with primary alcoholism at 3-month follow-up. Arch Gen Psychiatry 51:189 –197. Gross MM, Hastey JM (1976): Sleep disturbances in alcoholism. In: Sugarman A, editor. Alcoholism: Interdisciplinary Approaches to an Enduring Problem. Reading, MA: AddisonWesley, 257–309. Hobson AJ, Ludic R, Bagdhoyan HA (1986): Evolving concepts of sleep cycle generation from brain centers to neuronal populations. Behav Brain Sci 9:371– 448. Hohagen F, Spiegel R, Holzhauer M, Riemann D, Berger M (1993): The influence of the cholinergic agonist SDZ-210 – 086 on sleep in healthy subjects. Neuropsychopharmacology 9:225–232. Lauer CH, Zulley J, Krieg J-CH, Riemann D, Berger M (1988):
H. Gann et al
EEG sleep and the cholinergic REM-induction test in anorexic and bulimic patients. Psychiatry Res 26:171–181. Rechtschaffen A, Kales A (1968): A Manual of Standarized Terminology Techniques and Scoring System for Sleep Stages of Human Subjects. Washington, DC: Washington U.S. Government Printing Office (Public Health Service). Riemann D, Berger M (1989): EEG sleep in depression and remission and the cholinergic REM induction test with RS 86. Neuropsychopharmacology 2:145–152. Riemann D, Berger M (1992): REM sleep, age, depression and the cholinergic REM induction test with RS 86. Prog Neuropsychopharmacol Biol Psychiatry 16:311–316. Riemann D, Gann H, Dressing H, Mu¨ller WE, Aldenhoff JB (1994a): The influence of the cholinesterase inhibitor galanthamine hydrobromide on normal sleep. Psychiatry Res 51:253–267. Riemann D, Gann H, Fleckenstein PF, Olbrich R, Berger M (1991a): Effect of RS 86 on REM latency in schizophrenia Psychiatry Res 38:89 –92. Riemann D, Hohagen F, Bahro M, Berger M (1994b): Sleep in depression: The in- fluence of age, gender and diagnostic subtype on baseline sleep and the cholinergic REM induction test with RS 86. Eur Arch Psychiatry Clin Neurosci 243:279 – 290. Riemann, D, Hohagen F, Gann H, Olbrich R, Wark HJ, Bohus M, et al (1994c): The REM sleep response to cholinergic stimulation Indicator of muscarinic supersensitivity in schizophrenia? J Psychiatr Res 28:195–210. Riemann D, Joy D, Hochli D, Lauer CH, Zulley J, Berger M (1988): The influence of the cholinergic agonist RS 86 on normal sleep with regard to sex and age. Psychiatry Res 24:137–147. Riemann D, Lauer CJ, Hohagen F, Berger M (1991b): Longterm evolution of sleep in depression. In: Smirne S, editor. Sleep and Aging. Milan: Masson Press, 195–204. Riemann D, Lis S, Fritsch-Montero R, Krieger S, Meier S, Hohagen F, Berger M (1996): The effect of tetrahydroaminoacridine on sleep in healthy subjects. Biol Psychiatry 39:796 – 802. Roy A, Vikkunen M, Linnoila M (1987): Reduced central serotonin turnover in a subgroup of alcoholics? Neuropsychopharmacol Biol Psychiatry 11:173–177. Sellers EM, Higgins GA, Sobell MB (1992): 5-HT and alcohol abuse. Trends Pharmacol Sci 13:69 –75. Skoloda TE, Alterman AI, Gottheil EL (1979): Sleep quality reported by drinking and non-drinkung alcoholics. In: Gottheil EL, editor. Addiction Research and Treatment: Converging Trends. New York: Pergamon, 102–112. Wagmann AMI, Allen RP (1975): Effects of alcohol ingestion and abstinence on slow wave sleep on alcoholics. In: Gross MM, editor. Alcohol Intoxication and Withdrawal. Experimental Studies II. New York: Plenum Press. Williams HL, Rundell OH Jr (1981): Altered sleep physiology in chronic alcoholic reversal with abstinence. Alcoholism (NY) 5:318 –325. Wittchen HU, Sa H, Zaudig M, Koehler K (1989): Diagnostisches und statistisches Manual psychischer Sto¨rungen DSMIII-R. Weinheim-Basel: Beltz.