DR WEEKS’ COMMENTS: WARNING: SLEEP MEDICATIONS AND SSRI ANTIDEPRESSANTS BOTH DISRUPT STAGE 3,4 DEEP SLEEP TERMED REM SLEEP DISORDER (RSD) WHICH POTENTIALLY CAUSE SERIOUR OR FATAL BEHAVIOR TERMED REM SLEEP BEHAVIOR (RSB)
Conclusions: Subjects taking serotonergic antidepressants had more
EMG activity in the submental lead during REM sleep than did controls.
This correlated with measures of REM suppression and age. Individuals
taking such medications may be at increased risk of developing REM
sleep behavior disorder, particularly with increasing age.
. . . there are substantial potential public health implications
of REM sleep abnormalities in individuals taking serotonergic
antidepressants. Nearly 10 million people in the United States are taking
these medications on a routine basis. Increased awareness of RBD
among physicians who see individuals with sleep disorders, and among
those who prescribe serotonergic antidepressants, will allow for an accurate
estimate of sleep-related behavioral abnormalities observed as a
result of serotonergic antidepressants.
http://www.journalsleep.org/Articles/270219.pdf
SLEEP, Vol. 27, No. 2, 2004 317 Serotonergic Antidepressants and REM Sleep””Winkelman and James
Serotonergic Antidepressants are Associated with REM Sleep Without Atonia
PARASOMNIAS
John W. Winkelman, MD, PhD1; Lynette James2
1Divisions of Psychiatry and Sleep Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass 02459, USA; 2School of
Biomedical and Molecular Sciences, University of Surrey, Guildford, Surrey, GU2 7XH, UK
Study Objectives: Rapid eye movement (REM) sleep behavior disorder
(RBD) is generally observed in older men and in individuals with specific
neurologic diseases. There are case reports of RBD in individuals taking
serotonergic antidepressants. Our objective was to assess electromyogram
(EMG) activity during REM sleep in individuals taking serotonergic
antidepressants and in a matched control group not on such medication.
Design: Chart review of clinical and polysomnographic data.
Setting: Sleep laboratory affiliated with a general hospital.
Participants: 15 subjects taking a serotonergic antidepressant and 15
age-matched individuals not on such medication.
Measurements: Submental and anterior tibialis tonic and phasic EMG
activity during REM sleep, REM latency, time in REM, apnea-hypopnea
index, periodic leg movements of sleep index, and sleep-architecture
measures.
Results: Tonic, but not phasic, submental EMG activity during REM sleep
was significantly more common in the antidepressant-treated group than
in the control group (P < .02). Tonic REM submental EMG activity correlated
with REM latency (r =.42, P = .02) and inversely with REM time (r =
-.36, P = .05). Subject age correlated with tonic REM submental EMG
activity (r = .58, P = .02) in the antidepressant group There were also
trends for more phasic activity in the anterior tibialis (P = .09) and submental
(P = .07) EMG in REM sleep in the antidepressant group than in
the control group.
Conclusions: Subjects taking serotonergic antidepressants had more
EMG activity in the submental lead during REM sleep than did controls.
This correlated with measures of REM suppression and age. Individuals
taking such medications may be at increased risk of developing REM
sleep behavior disorder, particularly with increasing age.
Key Words: REM sleep, antidepressants, serotonergic, REM sleep
behavior disorder, EMG activity
Citation: Winkelman JW; James L. Serotonergic antidepressants are
associated with REM sleep without atonia. SLEEP 2004;27(2):317-21.
Disclosure Statement
No significant financial interest/other relationship to disclose.
Submitted for publication October 2003
Accepted for publication December 2003
Address correspondence to: John W. Winkelman, MD, PhD, Brigham and
Women’s Hospital, Sleep Health Center, 1400 Centre Street, Suite 109,
Newton Center, MA 02459; Tel: 617 527 2227; Fax: 617 527 2098;
E-mail: jwinkelman@sleephealth.com
INTRODUCTION
ATONIA OF SKELETAL MUSCLES IS ONE OF THE CARDINAL
FEATURES OF RAPID EYE MOVEMENT (REM) SLEEP.
Superimposed on this atonia is intermittent activity in both axial and
limb muscles. REM sleep behavior disorder (RBD) is characterized by
excessive motor activity during REM sleep with acting out of dreams.1
The diagnosis of RBD is made by the appearance of elevated submental
electromyogram (EMG) tone during REM and/or excessive phasic submental
or anterior tibialis EMG activity, combined with polysomnographic
documentation or a history of frank movements during REM
sleep.2 RBD is more common in elderly men, and at least half of those
followed for 10 years develop Parkinson disease.3
Muscle-tone abnormalities in REM sleep may consist along a spectrum,
with maintenance of full atonia at one end and full RBD at the
other end. REM sleep without atonia has been described as an intermediate
condition, in which REM sleep atonia is reduced on polysomnography,
in the absence of reports of abnormal behaviors by the patient or
bed partner. This polysomnographic finding has also been called “subclinical”
RBD. Eisensehr’s recent report4 demonstrating that those
patients with subclinical RBD have an intermediate reduction of striatal
dopamine transporters, roughly halfway between normal individuals and
those with RBD, establishes the potential importance of this disorder.
Antidepressants have substantial effects on REM sleep. Many studies
show that they prolong REM sleep latency and suppress REM sleep
time.5 They are also associated with reports of “vivid” dreams.6 In addition,
case reports dating back 30 years show that antidepressants can
induce RBD7 or reduce REM sleep atonia.8 In fact, medications with a
wide variety of mechanisms of action have been implicated in producing
loss of REM sleep atonia, including serotonergic reuptake blockers
such as fluoxetine,9 monoamine oxidase inhibitors,10 β-adrenergic
receptor blockers,11 the noradrenergic and 5-HT1A-mediated serotonergic
enhancer mirtazapine,12 and the tricyclic antidepressants.13 However,
no study has systematically assessed EMG tone during REM sleep in
individuals chronically taking antidepressants. Given the number of
individuals taking these medications, this issue is potentially of substantial
public health importance.
The objective of this study was to compare tonic and phasic EMG
during REM sleep in individuals without a complaint of abnormal
behavior during sleep who were taking serotonergic antidepressants with
the REM characteristics of matched controls not taking such medications.
We hypothesize that serotonergic antidepressants will increase
tonic and phasic submentalis and anterior tibialis EMG activity during
REM sleep compared to the control population not taking such medications.
METHODS
Subjects were recruited from the polysomnography database of Sleep
Health Centers, Newton Center, Mass. All sleep studies between June
2001 and August 2003 were reviewed and excluded if any of the following
features were present: apnea-hypopnea index > 15 per hour;
REM-related apnea-hypopnea index > 10; continuous positive airway
pressure use during the sleep study; complaint of abnormal behavior
during sleep or abnormal behavior on polysomnogram; duration of REM
sleep < 20 minutes; active neurologic disease (other than migraine); or
benzodiazepine, antipsychotic, or anticonvulsant use.
All subjects who met these criteria and were taking a serotonergic
antidepressant were included as the antidepressant group (n = 15). Five
subjects were taking fluoxetine (20-50 mg per day), 3 were taking
paroxetine (15-40 mg per day), 3 were taking citalopram (20-40 mg per
day), 3 were taking sertraline (100-225 mg per day), and 1 was taking
venlafaxine (400 mg per day). Two subjects in the antidepressant group
were taking bupropion (200 mg) in the morning in addition to their sero-
tonergic antidepressant. Duration of antidepressant treatment was
unknown, though subjects had been taking such medications for at least
2 weeks (based upon questionnaire data). Four of the 15 subjects in the
antidepressant group reported a history of depression only, and 4
described a history of an anxiety disorder only; 7 described a history of
both an anxiety and a depressive disorder. Fluoxetine equivalents were
calculated for antidepressant doses of all subjects by the following equation14:
fluoxetine = 5; sertraline = 1.2; paroxetine = 5; citalopram = 3.33;
venlafaxine = 1.
An age- and sex-matched sample fulfilling the inclusion and exclusion
criteria and not taking an antidepressant or any other centrally acting
agent was identified as the control group. No subjects in the control
group reported a history of either depressive or anxiety disorders. Fiftythree
percent (8/15) of subjects in the control group and 40% (6/15) in
the serotonergic antidepressant group were women. All subjects were
referred to rule-out obstructive sleep apnea. Data from an extensive
sleep, psychiatric, and medical history questionnaire were entered into a
database for all subjects.
All polysomnograms were performed in the same laboratory using
Alice 3 and 4 digitizing software (Respironics, Murrysville, Penn)
according to the following standard methods: left and right central and
occipital electroencephalogram (EEG) leads referenced to the opposite
ear; bilateral electrooculogram, submental EMG, bilateral anterior tibialis
EMG, and cardiorespiratory recordings consisting of nasal pressure
monitoring, nasal-oral thermistors, abdominal and chest effort, pulse
oximetry from the digit, and electrocardiogram.
Sleep staging was performed according to standard criteria,15 though
scoring of REM sleep was modified according to the method of Lapierre
and Montplaisir.16 In this modification, a REM epoch is terminated for
an EEG arousal but not as a result of increased EMG submental tone.
Each REM period for each subject was assessed for both tonic and phasic
EMG activity. REM epochs in which an EEG arousal (scored according
to standard guidelines), snore artifact in the submental EMG, periodic
leg movement (in a group of 4, with a stable intermovement interval),
or a hypopnea was present were eliminated from all further analyses.
Tonic EMG activity for each 30-second REM epoch was scored as
present (or put another way, was scored as absence of atonia) if greater
than 50% of the epoch had submental EMG activity greater than 4 times
the lowest level in that REM period. The percentage of epochs without
atonia was computed for each REM period and averaged for each subject.
Phasic EMG was scored in 2-second bins separately for the submental
and bilateral anterior tibialis leads according to the method of
Lapierre and Montplaisir.16 Each 2-second bin containing EMG activity
lasting 0.1 to 5.0 seconds, which exceed 4 times the lowest EMG activity
in that epoch, was counted as a bin with phasic activity. The percentage
of bins with phasic activity in the anterior tibialis and submental
leads was computed for each REM period and then averaged for each
subject. Phasic activity was also scored by the method of Eisensehr,4 in
which “long” EMG phasic activity was quantified. EMG bursts were
defined as “long” when they exceeded 0.5 seconds. A 10-second epoch
of REM was considered to have “long” EMG activity when the total of
such long bursts exceeded 1.0 seconds (eg, either at least 2 bursts lasting
0.5 seconds or 1 burst exceeding 1.0 seconds). The percentage of such
10-second epochs was determined for each subject for each REM period
and then averaged for each subject.
Statistical analyses were performed with the Student t test in normally
distributed data. The rank-sum test was used for variables that were
not normally distributed.
RESULTS
The 2 study groups did not differ in age, sex, body mass index, or
complaint that initiated the sleep study (see Table 1). On polysomnography,
subjects taking antidepressants had less REM time, longer REM
latency, greater sleep latency, a higher percentage of stage 2 sleep, and a
higher periodic limb movements of sleep index (see Table 1). No statistically
significant differences in apnea-hypopnea index (total or REMrelated)
or arousal index were noted between groups.
Subjects taking antidepressants had significantly more 30-second
REM epochs without submental atonia (with submental tone) than control
subjects (P = 0.02) (Table 2). There were significant correlations
between the submental EMG tone during REM and the degree of REM
suppression in the total sample, such that REM latency was positively
correlated with submental EMG tone (r = .42, P = .02) (see Figure 1),
and REM time was negatively correlated with submental EMG tone (r =
-.36, P = .05). There was a significant correlation between age and submental
EMG tone during REM in the antidepressant group (r = .58, P =
.02) (see Figure 2). This association was not significant in the control
group. There was no correlation between submental EMG tone during
REM and antidepressant dose (in fluoxetine equivalents).
There were trends for the subjects taking antidepressants to have more
2-second epochs in REM with phasic EMG activity in both the submental
(P = .07) and anterior tibialis (P = .09) leads than the control group
(see Table 2). There was a negative correlation between such phasic
activity in the anterior tibialis and REM time (r = -0.42, P = .02). There
was no correlation between either phasic submental or anterior tibialis
EMG activity in REM and medication dose (in fluoxetine equivalents).
SLEEP, Vol. 27, No. 2, 2004 318 Serotonergic Antidepressants and REM Sleep””Winkelman and James
Table 1””Demographic and Polysomnographic Features of
Antidepressant and Control Groups
Demographic or Control Serotonergic P value
Polysomnographic Feature Antidepressant
No. 15 15
Age (range), y 42.0 ± 14.1 (18-63) 45.5 ± 10.8 (26-60)
Men, no. (%) 8 (53) 6 (40)
BMI, kg/m2 25.0 ± 3.4 27.1 ± 5.5
Arousal index, arousals/h 15.3 ± 4.8 18.9 ± 9.7
Sleep efficiency, % 84.9 ± 11.9 81.7 ± 9.3
Sleep latency, min 13.0 ± 12.7 24.7 ± 14.4 .03
REM latency, min 68.8 ± 20.1 185.7 ± 73.7 < .001
PLM index, PLM/h 3.6 ± 6.3 18.8 ± 19.8 .08
Sleep stage, %
1 8.3 ± 5.9 9.05 ± 5.4
2 62.6 ± 6.8 69.6 ± 9.5 .03
3 6.1 ± 4.0 5.3 ± 3.7
4 7.2 ± 7.8 4.9 ± 7.4
REM 21.0 ± 4.8 14.4 ± 5.3 .001
REM time, min 79.1 ± 26.5 49.4 ± 21.3 .002
AHI, events/h 4.0 ± 2.5 4.7 ± 2.7
AHI during REM, events/h 5.6 ± 4.4 7.1 ± 5.4
Data are presented as mean ± SD, unless otherwise noted. All P values are not significant
unless otherwise noted.
BMI refers to body mass index; REM, rapid eye movement; PLM, periodic leg movement,
AHI, apnea-hypopnea index.
Table 2””Submental and Anterior-Tibialis Characteristics in
Antidepressant and Control Groups
Epochs, % Control Serotonergic P value
(n = 15) Antidepressant
(n = 15)
30-second with
submental EMG tone* 2.36 ± 3.88 9.54 ± 9.06 .02
2-second with phasic EMG”
Submental 5.63 ± 5.31 10.74 ± 9.16 .07
Anterior tibialis 9.72 ± 8.64 16.82 ± 14.69 .09
10-second with long EMG”¡
Submental 6.71 ± 6.06 13.39 ± 11.62 .03
Anterior tibialis 2.98 ± 2.63 8.94 ± 12.59 .06
Data are presented as mean ± SD, unless otherwise noted.
*Electromyogram (EMG) tone considered present if more than 50% of the epoch had submental
EMG activity greater than 4 times the lowest level in that rapid eye movement
(REM) period.
” Phasic EMG considered present if EMG activity lasted 0.1 to 5.0 seconds and exceeded 4
times the lowest EMG activity in that epoch.
”¡EMG considered present if the total of “long” bursts ( > 0.5 seconds) exceeded 1.0
seconds.
The antidepressant group had significantly more 10-second REM
epochs with “long” phasic activity than the control group in both the
submental (P = .03) and anterior tibialis (P = .06) leads. REM latency
correlated with submental “long” EMG activity for the entire sample (r
= .52, P = .003).
\The REM-period number (ie, 1 vs 2 vs 3) did not influence the degree
of EMG tone during REM in the submental lead or the extent of phasic
activity in the anterior tibialis or submental recordings.
DISCUSSION
Our results demonstrate that serotonergic antidepressants are associated
with a statistically significant and persistent reduction in REMsleep
atonia, even in individuals without overt clinical features of RBD.
We have also demonstrated that the degree of REM sleep without atonia
is correlated with other evidence of antidepressant effects on REM sleep
(suppression of REM time and prolongation of REM latency). Previous
case reports have described RBD in individuals taking antidepressants
for depression,17-18 narcolepsy,19 or Parkinson disease.12 Two previous
reports describe absence of atonia in REM sleep with the use of the tricyclic
antidepressant clomipramine.20-21 Guilleminault20 reported that
EMG atonia was “generally absent” in his narcoleptic subjects taking
clomipramine. Niyama21 identified this sleep stage as 1-REM in his normal
control subjects given single doses of 25 to 50 mg of clomipramine.
This is a retrospective study, and future studies of EMG tone after
medication treatments should address issues that we were unable to,
given this design. For instance, data on length of antidepressant treatment
and details regarding dream emotional quality and motor activity
would be of great interest. Further, increased numbers of subjects,
preferably in an age range that might be more vulnerable to REM sleep
without atonia (over 60 years), would also increase the power of such
studies. In addition, prospective studies of EMG tone before and after
chronic administration of a single serotonergic antidepressant are recommended
to confirm our findings and to better establish the precise
nature of this relationship.
A number of limitations of our data exist, which should be considered.
We did not evaluate the sleep of individuals prior to medication administration
and, thus, cannot definitely conclude that the serotonergic
antidepressants were responsible for the elevation in EMG activity during
REM sleep. Three of the subjects in the antidepressant group were
taking medication with effects beyond the serotonergic system: 2 were
taking bupropion, which enhances dopaminergic neurotransmission, and
1 was taking venlafaxine, which, in addition to its serotonergic properties,
produces noradrenergic reuptake blockade. It is possible that some
of our results may be a consequence of these other biologic effects. It is
also possible that depression or anxiety disorders themselves produced
these findings. It should be noted, however, that these findings have been
demonstrated acutely in normal volunteers.21 Similarly, these findings
were observed in our subjects treated for both depression and anxiety
disorders. Our subjects were not a random sample of individuals taking
serotonergic antidepressants but were recruited from individuals referred
for sleep study. To minimize this referral bias, we excluded individuals
with a description of behavioral abnormalities during sleep. All of our
subjects were referred to rule-out sleep apnea. Finally, we excluded subjects
taking medications such as benzodiazepines and anticonvulsants to
eliminate the potential effects of these medications on the polysomnogram
and to avoid a potential referral bias, as these medications may
have been used to treat sleep disruption resulting from the use of antidepressants.
This restriction may thus in fact have reduced the observed
prevalence of REM sleep abnormalities.
For a diagnosis of RBD, the International Classification of Sleep
Disorders2 requires both (1) abnormal behavior and (2) “excessive” submental
EMG tone or “excessive” phasic submental or limb twitching
during polysomnography. Although the behavioral markers for RBD
may be relatively clear,22 the polysomnographic criteria for what constitutes
“excessive” submental or anterior tibialis EMG tone during REM
sleep have not been established. Gagnon et al23 suggested that absence
of atonia (requiring 50% of the epoch with elevated tone) in greater than
20% of REM epochs is abnormal. In their study, 19 of 33 (57%) subjects
with Parkinson disease exceeded this degree of REM sleep without atonia,
whereas only 1 of 16 (6%) normal subjects exceeded this threshold.
By comparison, 2 of our 15 (13.3%) subjects taking antidepressants
exceed this criterion, whereas none of our control subjects did.
Eisensehr4 defined the upper limit of normal motor activity during
REM sleep as 15% of 10-second REM epochs containing at least 1 second
of elevated submental EMG activity (counting only “long” EMG
bursts, as described above). No unselected normative data were cited to
support the validity of this figure. Nevertheless, 8 of our 15 subjects taking
antidepressants (53%) exceeded this threshold in either the anterior
tibialis or submental lead, compared to only 1 of our 15 controls (7%).
Gagnon et al23 recently demonstrated the increased sensitivity of submental
EMG tone compared to anterior tibialis EMG tone in distinguishing
patients with Parkinson disease with RBD from both patients
with Parkinson disease without RBD and controls. In our data as well,
submental EMG tone over 30-second REM epochs was more sensitive
than either submental or anterior tibialis leads over shorter REM epoch
durations in distinguishing antidepressant from control groups. When 2-
second REM epochs were used, submental and anterior tibialis phasic
EMG were roughly equivalent in distinguishing subjects taking antidepressants
from the control subjects.
Figure 2””Correlation between submental electroencephalogram (EMG) tone and age in
the group taking antidepressants (r = 0.58; P = .02).
Figure 1””Correlation between submental electroencephalogram (EMG) tone and rapid eye
movement (REM) sleep latency (r = 0.42; P = .02).
Integrity of motor atonia during REM sleep is maintained by a number
of neuronal systems and, thus, may be disrupted by lesions or biochemical
interventions at a variety of sites.24 In fact, based on animal
experiments, separate systems, potentially colocalized at some points,
have been postulated to control the atonia and phasic locomotor aspects
of REM.25 Gilman et al’s26 recent demonstration of anatomic distinctions
between areas subserving atonia and those underlying phasic motor activation
in REM in subjects with RBD associated with multiple system
atrophy is further evidence of this. Our data demonstrating an effect of
serotonergic antidepressants on submental motor tone, in the absence of
robust effects on phasic activity, are consistent with other clinical reports
indicating a similar dissociation.7 The absence of reported abnormal
nocturnal behaviors in the majority of individuals taking serotonergic
antidepressants (including our subjects) may thus be due to the fact that
serotonergic antidepressants primarily disrupt tonic rather than phasic
components of motor activity during REM sleep.
The pathophysiology of RBD and REM sleep without atonia, as suggested
above, are likely complex. Dopaminergic mechanisms have
recently been suggested by imaging studies in patients with RBD and
Parkinson disease or multiple system atrophy.4,26-27 On the other hand,
basic research on motor control during REM sleep implicates glycinergic,
GABAergic, noradrenergic, and serotonergic transmitter systems.28-
30 In an animal model of RBD, Trulson et al28 found that raphe neurons,
which are usually quiet in REM, became tonically active. Similarly,
Lai’s30 recent finding that electrical or acetylcholine stimulation of the
pontine inhibitory area produces both motor-tone suppression and reductions
in serotonergic (and noradrenergic) activity further emphasizes the
importance of serotonergic inputs on spinal motor units in REM sleep.
Serotonergic antidepressants could thus influence motor tone during
REM sleep indirectly at brainstem levels (pedunculopontine nucleus or
pontine inhibitory area), or directly at spinal levels, producing REM
sleep without atonia.
The clinical status of REM sleep without atonia is ambiguous.
Although it is not listed in the International Classification of Sleep
Disorders nosology, it appears to be common in populations vulnerable
to RBD. In a recent study, 58% of patients with Parkinson disease
demonstrated atonia in REM sleep on polysomnography, 42% of whom
had no history of behavioral manifestations.31 In our series of consecutive
subjects without RBD taking antidepressants, 15% to 53% had evidence
of REM sleep without atonia, depending upon the definition.
REM sleep without atonia may be a “sentinel” finding on polysomnography,
expressing a vulnerability to overt RBD.31 From this perspective,
it may be a form fruste of early or evolving RBD. The evolution of RBD
into Parkinson disease in a high percentage of patients suggests that
EMG activity during REM sleep may be a sensitive indicator of early
central nervous system dysfunction. Finally, the distinction between
REM sleep without atonia and RBD may be blurred, as some individuals
with the former may in fact have behavioral manifestations of RBD
that are missed or ignored by patients and their bed partners and/or are
not present on a single night of polysomnography. In summary, it is
unclear whether elevated REM tone is just a polysomnographic finding
or whether it represents an important clinical prognostic finding.
Longitudinal studies of patients with Parkinson disease probably represent
the best opportunity to address this question scientifically.
If REM sleep without atonia is an early stage of RBD, it will be
important to understand the mediators of this response to antidepressants.
Our data suggests that age, in agreement with the increase in idiopathic
RBD in the elderly,1 is one such potential mediator. Older subjects
taking serotonergic antidepressants were more vulnerable to antidepressant-
related disinhibition of submental EMG tone in REM sleep. It is
unclear whether age is a surrogate for other factors that mediate this relationship
(central nervous system damage, antidepressant-receptor binding
or metabolism, etc.). However, we are aware of no other data that
demonstrate an influence of age on antidepressant effects on sleep.
Serotonergic antidepressant suppression of REM sleep (increased
REM latency and decreased REM time) was also a marker of the degree
of REM sleep without atonia in our subjects. Although no such correlations
have been demonstrated for patients with idiopathic (or Parkinsonrelated)
RBD, percentage of REM time is no different between those
with idiopathic RBD and normal controls.16 This may suggest that the
mechanisms producing abnormalities in EMG tone in REM sleep are
different in patients with idiopathic RBD and those given serotonergic
antidepressants. In both RBD and “idiopathic” REM sleep without atonia
(subclinical RBD), there are striatal presynaptic dopamine-transporter
deficits.4 On the other hand, serotonergic agonism may be more
relevant to REM suppression and increased EMG tone in our antidepressant
group.32-34\
Other potential vulnerability markers were not of value in predicting
REM sleep without atonia. For instance, male sex is an important risk
factor for idiopathic RBD.1 We did not find an increased vulnerability to
REM sleep atonia with male sex in our antidepressant group. Similarly,
we did not find a relationship between antidepressant dose (in fluoxetine
equivalents) and inhibition of REM sleep atonia. The relationship
between REM latency and antidepressant serum level has only been documented
for discontinuation of fluoxetine after subchronic use.35
Whether this is true at steady state after chronic dosing is unclear. One
important mediator on which we did not have data was length of treatment.
It is not clear whether length of time on an antidepressant may predispose
the individual to developing REM sleep without atonia. Future
studies of antidepressant effects on sleep should address this issue.
Although the clinical significance of REM sleep without atonia has
not been established, there are substantial potential public health implications
of REM sleep abnormalities in individuals taking serotonergic
antidepressants. Nearly 10 million people in the United States are taking
these medications on a routine basis. Increased awareness of RBD
among physicians who see individuals with sleep disorders, and among
those who prescribe serotonergic antidepressants, will allow for an accurate
estimate of sleep-related behavioral abnormalities observed as a
result of serotonergic antidepressants.
REFERENCES
1. Schenck CH, Mahowald MW. REM sleep behavior disorder: clinical, developmental,
and neuroscience perspectives 16 years after its formal identification in SLEEP. Sleep
2002;25:120-38.
2. International Classification of Sleep Disorders, Revised: Diagnostic and Coding
Manual. Rochester: American Sleep Disorders Association; 1997.
3. Schenck CH, Bundlie SR, Mahowald MW. REM Behavior Disorder (RBD): delayed
emergence of parkinsonism and/or dementia in 65% of older men initially diagnosed
with idiopathic RBD, and an analysis of the minimum and maximum tonic and/or phasic
electromyographic abnormalities found during REM sleep. Sleep 2003;26:A316.
4. Eisensehr I, Linke R, Tatsch K, et al. Increased muscle activity during rapid eye movement
sleep correlates with decrease of striatal presynaptic dopamine transporters. IPT
and IBZM SPECT imaging in subclinical and clinically manifest idiopathic REM sleep
behavior disorder, Parkinson’s disease, and controls. Sleep. 2003;26:507-12.
5. Sharpley AL, Cowen PJ. Effect of pharmacologic treatments on the sleep of depressed
patients. Biol Psychiatry 1995;37:85-98.
6. Pace-Schott EF, Gersh T, Silvestri R, et al. SSRI treatment suppresses dream recall frequency
but increases subjective dream intensity in normal subjects. J Sleep Res
2001;10:29-42.
7. Mahowald MW, Schenck CH. REM sleep parasomnias. In: Kryger M, Roth T, Dement
W, eds. Principles and Practice of Sleep Medicine. 3rd ed. Philadelphia: WB Saunders;
2000:724-41.
8. Guilleminault C, Raynal D, Takahashi S, et al. Evaluation of short-term and long-term
treatment of the narcolepsy syndrome with clomipramine hydrochloride. Acta Neurol
Scand 1976;54:71-87.
9. Schenck CH, Mahowald MW, Kim SW, et al. Prominent eye movements during NREM
sleep and REM sleep behavior disorder associated with fluoxetine treatment of depression
and obsessive-compulsive disorder. Sleep 1992;15:226-35.
10. Louden MB, Morehead MA, Schmidt HS. Activation by selegiline (Eldepryle) of REM
sleep behavior disorder in parkinsonism. W V Med J 1995;91:101.
11. Iranzo A, Santamaria J. Bisoprolol-induced rapid eye movement sleep behavior disorder.
Am J Med 1999;107:390-2.
12. Onofrj M, Luciano AL, Thomas A, et al. Mirtazapine induces REM sleep behavior disorder
(RBD) in parkinsonism. Neurology 2003;60:113-5.
13. Passouant P, Cadilhac J, Ribstein M. Sleep privation with eye movements using antidepressive
agents. Rev Neurol 1972;127:173-92.
14. Bollini P, Pampalllona S, Tibaldi G, Kupelnick B, Munizza C. Effectiveness of antidepressants:
meta-analysis of dose-effect relationships in randomized clinical trials. Br J
Psychiatry 1999;174:297-303.
SLEEP, Vol. 27, No. 2, 2004 320 Serotonergic Antidepressants and REM Sleep””Winkelman and James
15. Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques, and
Scoring System for Sleep Stages of Human Subjects. Washington, DC: Public Health
Service, US Government Printing Office; 1968.
16. Lapierre O, Montplaisir J. Polysomnographic features of REM sleep behavior disorder:
development of a scoring method. Neurology 1992;42:1371-4.
17. Schutte S, Doghramji K. REM behavior disorder seen with venlafaxine (Effexor). Sleep
Res 1996;25:364.
18. Nofzinger EA, Reynolds CF 3rd. REM sleep behavior disorder. JAMA 1994;27:820.
19. Bental E, Lavie P, Sharf B. Severe hypermotility during sleep in treatment of cataplexy
with clomipramine. Israel J Med Sci 1979;15:607-9.
20. Guilleminault C, Raynal D, Takahashi S, Carskadon M, Dement W. Evaluation of shortterm
and long-term treatment of the narcolepsy syndrome with clomipramine hydrochloride.
Acta Neurol Scand 1976;54:71-87.
21. Niiyama Y, Shimizu T, Abe M, Hishikawa Y. Cortical reactivity in REM sleep with tonic
mentalis EMG activity induced by clomipramine: an evaluation by slow vertex response.
Electroencephalogr Clin Neurophysiol 1993;86:247-51.
22. Bologna, Genova, Parma and Pisa Universities group for the study of REM Sleep
Behaviour Disorder (RBD) in Parkinson’s disease. Interobserver reliability of ICSD-R
criteria for REM sleep behaviour disorder. J Sleep Res 2003;12:255-7.
23. Gagnon J, Bedard MA, Fantini ML, et al. Comparison between submental and anterior
tibialis phasic EMG activity during REM sleep in Parkinson’s disease with and without
REM sleep behavior disorder. Sleep 2003;26:A337.
24. Lai YY, Siegel JM. Medullary regions mediating atonia. J Neurosci 1988;8:4790-6.
25. Lai YY, Siegel JM. Muscle tone suppression and stepping produced by stimulation of
midbrain and rostral pontine reticular formation. J Neurosci 1990;10:2727-34.
26. Gilman S, Koeppe RA, Chervin RD, et al. REM sleep behavior disorder is related to striatal
monoaminergic deficit in MSA. Neurology 2003;61:29-34.
27. Eisensehr I, Linke R, Noachtar S, Schwarz J, Gildehaus Fj, Tatsch K. Reduced striatal
dopamine transporters in idiopathic rapid eye movement sleep behavior disorder.
Comparison with Parkinson’s disease and controls. Brain 2000:123:1155-60.
28. Trulson ME, Jacobs BL, Morrison AR. Raphe unit activity during REM sleep in normal
cats and in pontine lesioned cats displaying REM sleep without atonia. Brain Res
1981;226:75-91.
29. Kubin L, Kimura H, Tojima H, Davies RO, Pack AI. Suppression of hypoglossal
motoneurons during the carbachol-induced atonia of REM sleep in not caused by fast
synaptic inhibition. Brain Res 1993;611:300-12.
30. Lai YY, Kodama T, Siegel JM. Changes in monoamine release in the ventral horn and
hypoglossal nucleus linked to pontine inhibition of muscle tone: an in vivo microdialysis
study. J Neurosci 2001;21:7384-91.
31. Gagnon JF, Bedard MA, Fantini ML, et al. REM sleep behavior disorder and REM sleep
without atonia in Parkinson’s disease. Neurology 2002;59:585-9.
32. Rush AJ, Armitage R, Gillin JC, et al. Comparative effects of nefazodone and fluoxetine
on sleep in outpatients with major depressive disorder. Biol Psychiatry 1998;44:3-14.
33. Seifritz E, Stahl SM, Gillin JC. Human sleep EEG following the 5-HT1A antagonist pindolol:
possible disinhibition of raphe neuron activity. Brain Res 1997;759:84-91.
34. Landolt HP, Kelsoe JR, Rapaport MH, Gillin JC. Rapid tryptophan depletion reverses
phenelzine-induced suppression of REM sleep. J Sleep Res 2003;12:13-8.
35. Feige B, Voderholzer U, Riemann D, Dittmann R, Hohagen F, Berger M. Fluoxetine and
sleep EEG: effects of a single dose, subchronic treatment, and discontinuation in healthy
subjects. Neuropsychopharmacology 2002;26:246-58.
SLEEP, Vol. 27, No. 2, 2004 321 Serotonergic Antidepressants and REM Sleep””Winkelman and James