Sleep, 19(1):52-58

© 1996 American Sleep Disorders Association and Sleep Research Society

Movements and Sleep

Randomized, Double-Blind, Placebo-Controlled

Study of Clonidine in Restless Legs Syndrome

*M. L. Wagner, ttA. S. Walters, *R. G. Coleman, ttW. A. Hening,

§K. Grasing and ttiIS. Chokroverty

Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024

*Rutgers-The State University of New Jersey, College of Pharmacy,
Department of Pharmacy Practice and Administration, Piscataway, New Jersey, U.S.A.
tUMDNJ-Robert Wood Johnson Medical School, Department of Neurology,
New Brunswick, New Jersey, U.S.A.
:j:Veterans Administration Medical Center, Neurology Service, Lyons, New Jersey, U.S.A.
§UMDNJ-Robert Wood Johnson Medical School, Program in Clinical Pharmacology,
Clinical Research Center, New Brunswick, New Jersey, U.S.A.
liSt. Vincent's Hospital and Medical Center of New York, Department of Neurology, New York, U.S.A.

Summary: Ten patients with idiopathic restless leg syndrome (RLS) were asked to rate their symptoms at baseline
d~ring 2 weeks of placebo and 2 weeks of clonidine treatment by using a four-point scale. On two consecutive
nights of each treatment period, polysomnography (PSG) and actigraphic studies were performed. Patients subjec-
tively reported improvement in leg sensations (p = 0.02) and motor restlessness (p = 0.001) while receiving
clonidine (mean = 0.5 mg/day). On PSG testing, sleep onset occurred faster with clonidine (12 minutes) compared
with placebo (30 minutes) and baseline (47 minutes) (p = 0.006). Adverse findings associated with clonidine
treatment included decreased percent REM sleep in the clonidine group (4%) compared with placebo (16%) and
baseline (16%) (p = 0.001) and increased REM latency in the clonidine group (195 minutes) compared to the
placebo (70 minutes) and baseline groups (89 minutes) (p = 0.028). There were no significant changes in total
sleep time, stage 1 and 2 sleep, sleep efficiency, awakenings, arousals or periodic limb movements in sleep. There
was a nonstatistical trend toward and increase in stage 3 and 4 sleep and a decrease in motor activity as measured
by actigraphic recordings. Globally, seven out of 10 patients felt clonidine was more effective than placebo. Four
patients chose to continue clonidine after the study. Clonidine may be an effective treatment for RLS patients who
don't have large numbers of sleep-disrupting periodic limb movements but have delayed sleep onset due to leg
sensations and subsequent motor restlessness. Key Words: Restless legs syndrome-Periodic limb movements in
sleep-Clonidine-Sleep disorders.

Although restless legs syndrome (RLS) is a com-
mon condition, occurring in two to five percent of the
population, it is rarely studied and often fails to be
diagnosed (1-4). The diagnosis of RLS is based large-
lyon a detailed patient history, and some features of
the condition can be demonstrated by polysomno-
graphic studies (1,4). Clinical features present in RLS
have been previously well described and include par-
esthesias (creeping, crawling, tingling, burning or oth-
er sensations primarily in the legs), motor restlessness
(floor pacing, tossing and turning in bed), periodic
Accepted for publication September 5, 1995.
Address correspondence and reprint requests to M. L. Wagner,
M.S., Pharrn.D., Assistant Professor, Rutgers-The State University
of New Jersey, College of Pharmacy, Department of Pharmacy Prac-
tice and Administration, P.O. Box 789, Piscataway, NJ 08855-0789,
U.S.A.
52
limb movements of sleep (PLMS), involuntary move-
ments while awake that are similar in appearance to
PLMS and sleep disturbances (1,4-7). The symptoms
of RLS usually develop after prolonged sitting or lying
and usually occur mostly at night. This syndrome is
usually idiopathic and often familial (8), but secondary
forms of RLS are often associated with conditions
such as gastrectomy (2), diabetes mellitus (2), uremia
(9), vascular insufficiency (10), anemia (II) and caf-
feine abuse (12). Some secondary forms are associated
with peripheral neuropathy or radiculopathy. Patients
with secondary forms of RLS may respond differently
to treatment than do patients with idiopathic RLS;
thus, it is important to distinguish the type of RLS a
patient has when selecting patients for studies.
Current treatment of RLS includes monotherapy or
 CLONIDINE IN RESTLESS LEGS SYNDROME
poly therapy with carbidopa/levodopa (Sinemet®), do-
pamine agonists, opioids or benzodiazepines. Double-
blind, placebo-controlled studies have shown that
levodopa (13,14), dopamine agonists (15), carbama-
zepine (16), benzodiazepines (17), baclofen (18) and
opioids (19) are clinically effective in treating RLS
and, in some instances, PLMS. However, these medi-
cations have shown varying effectiveness on the dif-
ferent clinical symptoms of RLS and may cause sig-
nificant adverse effects. These five classes of medica-
tions (13-15,17-20) have been studied polysomno-
graphically. Clonidine has also been extensively
studied in RLS using subjective data but not polysom-
nography. Levodopa and opioids have a short duration
of action and often require repeat dosing during the
night. Another drawback with Levodopa therapy is re-
bound of RLS symptoms in the day time (21). Phy-
sicians or patients are often reluctant to use benzodi-
azepines and opioids on a chronic basis because they
are controlled substances with a risk for abuse or ad-
dition. Other risks with benzodiazepines include tol-
erance and daytime somnolence. Carbamazepine does
not improve PLMS (20) and baclofen (18) decreases
the amplitude but not the number of PLMS. The effect
of baclofen on paresthesias and motor restlessness has
not been studied. Thus, to date, no known optimal
pharmacological treatment is available for this disor-
der.
Several reports indicate that clonidine may effec-
tively treat RLS symptoms with minimal side effects
(5,22-25); however, one study indicated that clonidine
resulted in little improvement in RLS symptoms with
many adverse effects (26). None of these previous
studies used objective polysomnographic data to eval-
uate clonidine's effect on sleep architecture, actigraphy
to measure movements, detailed patient diaries to de-
scribe effects on individual RLS symptoms, or cloni-
dine blood concentrations to measure pharmacody-
namic effects. Only one of these studies used a double-
blind, placebo-controlled design, but it did not employ
a randomized design and did not test the effect of clo-
nidine in idiopathic RLS (25). Four studies used open-
labeled, uncontrolled methods and either tested sec-
ondary forms of RLS or did not state the type of pa-
tients studied (22-24,26). Bamford and Sandyk re-
ported that two out of seven patients had a worsening
of RLS symptoms and all patients had worsened in-
somnia; however, it is unclear which type of RLS pa-
tient was enrolled or which of the subjective RLS
symptoms worsened (26). Combining all the studies,
30 of the 41 patients (73%) treated with clonidine
(0.1-0.3 mg/day) had alleviation of RLS symptoms.
Low and possibly inadequate doses may have been
used in some of these studies. We found that higher
doses of clonidine (0.9 mg/day) may be needed to treat
53
RLS in certain patients, and spontaneous remission of
symptoms needs to be considered when evaluating the
effectiveness of treatment (5). The goal of this study
is to define the effectiveness of clonidine in idiopathic
RLS using polysomnography (PSG), actigraphy, de-
tailed patient diaries and clonidine blood concentra-
tions.
METHODS
Patient selection
Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024
Patients signed the Institutional Review Board ap-
proved informed consent form and were selected from
patients evaluated at the Department of Neurology
clinic at Robert Wood Johnson University Hospital,
New Brunswick, NJ, and the Veterans Administration
Medical Center, Lyons, NJ. All patients exhibited the
following four symptoms of RLS: (a) paresthesias, (b)
motor restlessness, both of which worsen at (c) night
and (d) rest (lying or sitting). All patients experienced
subjective complaints of sleep disturbances such as
difficulty with sleep onset and maintenance or exces-
sive daytime somnolence. Patients were not required
to have periodic limb movements of sleep (PLMS).
Patients with uncontrolled hypertension [systolic blood
pressure (SBP) > 160 mm Hg, diastolic blood pressure
(DBP) > 100 mm Hg], hypotension (SBP < 90 mm
Hg, DBP < 60 mm Hg), bradycardia «60 bpm), sec-
ondary forms of RLS due to peripheral neuropathy or
radiculopathy, history of drug abuse, sleep apnea or
allergy to clonidine were excluded from the study. Pa-
tients unable to discontinue RLS medications because
of intolerable worsening of RLS symptoms were also
excluded from the study.
Design
The study was a randomized, double-blind, placebo-
controlled, 3-period, crossover trial. For a period of at
least 3 days before the study, all patients were grad-
ually tapered off of all medications that could affect
RLS symptoms (for example Levodopa, dopamine ag-
onists, benzodiazepines, opioids, baclofen, carbama-
zepine or tricyclic antidepressants). Patients did not
receive any RLS medications for at least 2 days before
the baseline PSG studies. We did not adjust any of the
medications prescribed for other purposes, and at-
tempts were made to keep them constant throughout
the study. The first period consisted of a 4-day baseline
period, during which no RLS medications were taken.
The next two periods were 2-3 weeks in length de-
pending on scheduling. Before each of these periods,
patients were randomized to receive clonidine or pla-
cebo.
Sleep, Vol. 19, No.1, 1996
54 M. L. WAGNER ET AL.
At baseline and at the end of each of the subsequent
two periods, all patients were required to stay over-
night in the sleep laboratory for two nights at the Clin-
ical Research Center (CRC), UMDNJ-Robert Wood
Johnson Medical School. At this time, polysomnog-
raphy (PSG) and actigraphic studies were performed,
RLS symptoms were assessed and vital signs were
taken. Blood was also collected during the polysom-
nographic portion of periods 2 and 3 at bedtime and
upon awakening to measure clonidine concentrations.
Blood pressure was measured each sleep study night
in the sitting position. If patients reported symptoms
of dizziness, orthostatic hypotension was assessed by
measuring blood pressures in both the standing and
sitting position.
Dosing
Clonidine and matching placebo tablets were pro-
vided by Boehringer Ingelheim (PD 1172, PD09l5, re-
spectively). After COmpletion of two baseline sleep
studies without medication in period I, patients began
with an initial dose of one tablet (0.1 mg of clonidine
or lactose) and the dosage was gradually increased by
one tablet per day every 1-2 days to the maximum
therapeutic benefit or a maximum of 10 tablets. The
tablets were administered in divided dosages about 2
hours before the onset of symptoms based on the pa-
tient diary and at bedtime. After completion of the two
nights of sleep study in period 2, the period 2 tablets
were tapered to zero over a 3-day period, and simul-
taneously, the period 3 medications were titrated up-
ward. The average dose for each period for purposes
of data computation was the mean of the medication
received during the two nights of sleep study. The in-
vestigators telephoned patients daily to assess signs
and symptoms and to adjust dosages. If patients ex-
perienced an exacerbation of RLS symptoms or ad-
verse medication effects during the dose titration
phases, their dose was adjusted up or down or their
dosing interval was adjusted accordingly. Compliance
was assessed by counting the returned number of tab-
lets at the end of each treatment period.
Subjective data
Patients subjectively rated their symptoms daily on
a diary form. The following symptoms were rated: leg
sensations while awake, restless movements while
awake and daytime fatigue. The scale ranged from
zero to four (zero = none, one = mild, two = mod-
erate, three = severe, four = very severe). Patients
reported these ratings to one of the investigators daily
to determine if dosage adjustments were necessary. At
Sleep, Vol. 19, No.1, 1996
the end of the study, these data were averaged for the
two nights of sleep study for each period, and patients,
unaware of the treatment order, were asked which
treatment phase they preferred and why.
Polysomnographic data
Patients were monitored for approximately 8 hours
of the night. In the laboratory, sleep architecture was
recorded with a polysomnogram which included an
electroencephalogram (EEG) (C 3-A2 , CCA\), elec-
Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024
trooculogram (EOG) from the left and right outer can-
thi of each eye, chin electromyogram (EM G) of both
the mentalis and submentalis muscles to monitor REM
sleep and EMG of left and right tibialis anterior mus-
cles to measure PLMS (27). Respiration was moni-
tored by standard measures of air flow (oral/nasal ther-
mocouple; Rochester Electro-Medical, Tampa, FL), ef-
fort (RESP-EZ chest and abdominal belts containing
piezo-electric crystals; EPM Systems, Midlothian, VA)
and oxygen saturation (finger pulse oximetry; Ohmeda
Biox 3700, Boulder, CO). The following PSG param-
eters were monitored: sleep efficiency (minutes of
sleep time divided by the time in bed X 100), sleep
latency (the number of minutes for sleep onset to oc-
cur) and REM latency (the amount of time in minutes
measured from sleep onset to REM onset), PLMS in-
dex, awakening index, arousal index and sleep stages.
PLMS, awakening and arousal indices were calculated
by dividing the number of events by the total hours of
sleep. Awakenings and arousals were scored by pre-
viously published methods (28-30). Periodic and ape-
riodic limb movements while the patients were awake
and lying down were not quantified for this study. The
percent time spent in stage 1, 2, 3 and 4 sleep, as well
as REM sleep, was calculated as a percentage of sleep
period time (the total awake and sleep time from sleep
onset to final awakening). Comparisons were made by
combining stage 1 and 2 sleep and by combining stage
3 and 4 sleep (27). All PSGs were scored by a blinded
certified sleep technician in the CRe.
Actigraphy
Actigraphic monitors, which measure PLMS and
motor restlessness, were furnished by Individual Mon-
itoring Systems Inc. and applied to both ankles during
the PSG studies. Movements were counted during a
patient's total time in bed (awake and asleep), and an
activity index (counts per hour) was calculated for
each leg. The monitor was set to record the amount of
movement in 2-minute samples around the clock. The
actigraphic data was then correlated with the recorded
PSG data.
 CLONIDINE IN RESTLESS LEGS SYNDROME
Clonidine blood concentrations
Blood (4 ml) was obtained prior to sleep and again
upon awakening at each sleep study visit (periods 2
and 3) to measure clonidine concentrations. Blood
samples were promptly processed and the plasma frac-
tion stored at -70°C until the study code was broken.
Only samples from the clonidine phase were sent for
analysis by gas chromatography with mass spectros-
copy at National Medical Services, Inc., Willow
Grove, PA. The blood concentration range from sleep
onset to sleep awakening was correlated with the PSG
studies. Therefore, this pharmacodynamic analysis was
not affected by variations in dosage administration
times or sample collection times.
Statistical analysis
The subjective, actigraphic and PSG data were an-
alyzed by the Friedman nonparametric rank sums test.
This test was used to compare the differences between
the three treatment periods. The Wilcoxon signed rank
post hoc test was used to describe which treatment
periods were statistically different from others. Anal-
ysis of variance (ANOVA) was used to analyze
changes in blood pressure and heart rate. The Scheffe
post hoc test was used to determine which treatment
periods were statistically different from each other.
Spearman's rank correlation was used to test the re-
lationship between clonidine blood concentrations and
RLS symptoms (31,32).
RESULTS
From January to November 1993, 11 patients (three
women, eight men, ages 29-61 years) were enrolled
in the study. One patient dropped out for personal rea-
sons. The average age of the subjects was 44.5 years
of age (29-6l years old) with and average age of dis-
ease onset at 28.1 years of age (10-48 years old). Sev-
en out of 10 of the patients received concomitant med-
ications, although medications known to interfere with
RLS symptoms were discontinued prior to the sleep
study. Six patients had a first or second degree relative
affected with RLS.
Five patients received clonidine during the first
treatment period and the other five received it during
the second treatment period. Seven out of 10 patients
were considered responders to clonidine. A responder
was defined as a person who preferred clonidine to the
placebo or had more symptom-free days during the
clonidine treatment period. Patients were considered
compliant to medication since according to the drug
accountability record, tablet counts were at least 95%
accurate.
55
TABLE 1. Patient mean subjective reportsa
Parameter Baseline Placebo Clonidine p-value
Paresthesias 2.1 :!:: 0.9 1.9 :!:: 1.2 0.5 :!:: 0.5 0.021
Motor restlessness 2.2 :!:: 0.8 1.8 :±: 1.3 0.5 :!:: 0.6 0.001
Daytime fatigue 2.7 :!:: 0.7 2.1 :±: 1.2 1.0 :!:: 1.0 0.014
a Patients rated paresthesias, motor restlessness and daytime fa-
tigue on a daily basis under double-blind conditions on a 1--4 scale
(0 = no symptoms, I = mild, 2 = moderate, 3 = severe, and 4 =
very severe). Symptoms were rated during the baseline, placebo and
drug periods. Ratings are an average of those obtained from 2 sleep
study nights and were compared using a Friedman nonparametric
rank sums test with a post hoc Wilcoxon signed rank test.
Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024
Patients subjectively reported a statistically signifi-
cant decrease in leg sensations (p = 0.021), motor rest-
lessness (p = 0.001) and daytime fatigue (p = 0.014)
when receiving clonidine (Table 1). Patients generally
rated their RLS symptoms as moderate in severity pri-
or to clonidine and mild to none with clonidine. Over-
all, seven patients preferred clonidine over placebo;
however, three patients thought that the side effects
outweighed any benefit. Four of the seven patients
chose to continue treatment with clonidine after the
study was completed. Three of the 10 patients pre-
ferred to return to their previous RLS medications.
The severity of the RLS symptoms can change over
time, however, the severity of symptoms patients re-
ported in this study remained constant during the treat-
ment periods. Therefore, no period effect was ob-
served. The subjective symptoms of two patients im-
proved significantly from baseline to placebo; there-
fore, they may have experienced a placebo effect.
These patients, however, had a much greater decrease
in symptoms during the clonidine period. The order of
treatment was evenly distributed between patients and
an order effect was not observed.
PSG studies indicated a clinically and statistically
significant decrease in sleep latency to 12 minutes with
clonidine compared to placebo (30 minutes) and base-
line (47 minutes) (p = 0.006) (Table 2). Sleep latency
time normalized in six patients. Percent REM sleep
decreased to four percent with clonidine compared to
placebo (16%) and baseline (16%) (p = 0.001). REM
latency significantly increased to 195 minutes with
clonidine compared to placebo (70 minutes) and base-
line (89 minutes) (p = 0.028). No significant changes
in total sleep time, sleep efficiency, amount of time
spent in stage 1 and 2 sleep, amount of time spent in
stage 3 and 4 of sleep, awakening index, arousal index
or PLMS index was observed, but patients tended to
spend more time in stage 3 and 4 sleep (p = 0.067).
No change was observed in PLMS frequency per total
non-REM sleep time.
Because of technical difficulties, actigraphic record-
ings were collected for only seven of the 10 patients.
One patient had an increase in average motor activity
Sleep, Vol. 19. No.1. 1996
56 M. L. WAGNER ET AL.

TABLE 2. Polysomnographic recordings a

Parameter Baseline Placebo Clonidine p-value

Sleep latency (minutes) 47.4 ::t:
55.0 29.4 ::t: 23.8 11.9 ::t: 10.7 0.006
REM latency (minutes) 88.9 ::t:
53.9 70.1 ::t: 47.5 195.1 ::t: 79.1 0.028
Sleep efficiency (%) 72.0 ::t:
21.9 77.9 ::t: 11.3 76.9 ::t: 13.1 0.905
Awakening index (awakeningslhour) 3.7 ::t:
1.9 4.1 ::t: 1.8 4.2 ::t: 1.7 0.150
Arousal index (arousalslhour) 21.4 ::t:
12.9 21.9 ::t: 11.3 22.3 ::t: 9.0 0.497
Stage 1 & 2 sleep (%) 65.4 ::t:
15.3 69.2::t: 5.8 74.9 ± 12.9 0.407
Stage 3 & 4 sleep (%) 4.3 ::t:
3.4 2.2 ::t: 1.7 6.6 ::t: 4.7 0.067
Total sleep time (minutes) 272.2 ::t:
88.2 283.7::t: 56.9 283.0 ::t: 77.0 1.000
REM stage (%) 15.9 ::t:
9.0 15.8 ::t: 6.5 3.8 ::t: 4.3 0.001
PLMS index (movementslhour) 36.8 ± 40.1 31.0 ::t: 26.3 43.2 ± 36.4 0.741
a Patients were studied polysomnographically for two nights under double-blind conditions at the end of the baseline, placebo and c10nidine
treatment periods. The averages for the two nights of polysomnographic study for each period were compared using a Friedman nonpara-

Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024

metric rank sums test with a post hoc Wilcoxon signed rank test.

(motor restlessness and PLMS) but five patients had a Adverse effects included dry mouth, decrea.-;ed cog-
decrease in average motor activity. Mean activity nition, lightheadedness, sleepiness post dose, consti-
scores decreased by nearly 1h after clonidine (0.6 pation, decreased libido and headache (Table 3). In-
countslhour) when compared to baseline (1.1 counts/ creased sleepiness may be explained by four patients
hour) and placebo (0.9 countslhour); however, there who started with normal sleep latency, but had sleep
was no statistical difference between the three study latencies of 2.5-6 minutes after clonidine. The dose at
periods. which many of the adverse effects were initially de-
Mean plasma clonidine concentrations were 1.72 tected was variable among patients but the mean dose
ng/ml at bedtime and 1.54 ng/ml upon awakening. ranged from 0.3 to 0.4 mg/day. Although, a large num-
There was no correlation between clonidine concen- ber of patients experienced adverse reactions, these ad-
trations and control of RLS symptoms or sleep archi- verse effects were generally considered mild and de-
tecture except for REM sleep. Higher clonidine con- creased or subsided with dosage reduction. However,
centrations at either bedtime (p = 0.01) or upon awak- adverse drug reactions prevented seven patients from
ening (p = 0.01) were correlated to a greater reduction increasing their dose enough to completely relieve
in REM sleep. Concentrations increased linearly with their symptoms.
increasing doses. Doses ranged from 0.1 to 1.0 mg/ The average sitting SBP significantly decreased to
day (mean 0.5 mg/day), whereas corresponding morn- 119 mm Hg with clonidine compared to placebo (132
ing trough clonidine concentrations ranged from 0.34 mm Hg) and baseline (135 mm Hg; p = 0.0001). The
to 3.6 ng/ml. Variability in dose response to clonidine average DBP decreased to 74 mm Hg with clonidine
was significant. For example, one patient had adequate compared to placebo (84 mm Hg) and baseline (83.0
relief of RLS symptoms when taking only 0.1 mg/day mm Hg) (p = 0.058). The average heart rate was 76
and had adverse effects at 0.4 mg/day, whereas another beats/minute (bpm) with clonidine compared to pla-
patient had only minimal relief at 0.4 mg/day and ad- cebo (79 bpm) and baseline (77 bpm) (p = 0.543).
verse effects at 0.9 mg/day. Although six patients experienced lightheadedness
during the clonidine dosage titration, these symptoms
TABLE 3. Adverse effectsa subsided with dosage reduction, and no symptomatic
changes were noted at the time of the blood pressure
Average daily measurements.
c10nidine
dose
Number of when AE Number of
Adverse effects patients with occurred patients with DISCUSSION
(AE) c10nidine AE (mg) placebo AE
Dry mouth 8 0.3 0 This double-blind study suggests that clonidine is
Decreased cognition 6 0.4 0 effective in relieving the leg sensations, restless move-
Lightheadedness 6 0.4 2
Sleepy after dose 5 0.36
ments and morning drowsiness attributed to RLS. The
0
Constipation 4 0.4 I PSG data showed that 2-3 weeks of clonidine (mean
Decreased libido 2 0.35 0 = 0.5 mg/day) decreased sleep latency, increased
Headache 2 0.35 1
REM latency and decreased the amount of time spent
a n = 10. Although a large number of patients experienced AE,
in REM sleep. The first of these PSG results can be
the majority of these AE were considered mild and decreased with
dosage reduction. Rapid dose titration may have increased the in- considered a beneficial effect and the latter two dele-
cidence of AE. terious effects. Single-dose (33,34) and multiple-dose
Sleep, Vol. 19, No.1, 1996
 CLONIDINE IN RESTLESS LEGS SYNDROME
(35,36) clonidine studies in non-RLS patients also in-
dicate a decrease in percent REM sleep with a com-
pensatory increase in the percent of time spent in
stages 1 and 2 sleep (33,34). Although we did not
detect a dose-response relationship, Nicholson and
Pascoe found no effect on sleep parameters at 0.05 mg/
day but a decrease in REM sleep duration, prolonged
REM latency, and increased stage 2 sleep duration at
0.1 mg/day (33). Clonidine has sedative effects (34,35)
possibly mediated through activation of a 2-adrenore-
ceptors in the locus ceruleus (34) that results in a de-
crease in sleep latency. Kostis et al. also found an in-
crease in REM latency and decrease in sleep latency
after 3 months of treatment with clonidine (0.1-0.3 mg
twice a day) in 24 hypertensive males. The investi-
gators also reported a reduction in the total sleep time
and sleep maintenance (36). The long-term effects of
chronic clonidine administration on REM sleep have
not been elucidated and further studies, using PSG
data and drug concentrations, need to be conducted to
determine if the REM suppressant effect on sleep is
transient and related to drug pharmacokinetics.
Patients in our study had no improvement in total
sleep time, sleep efficiency, number of PLMS, number
of arousals or number of awakenings. The trend to-
ward an increased amount of time spent in stage 3 and
4 sleep, may possibly explain why our patients felt
more rested the next day without an objective im-
provement in sleep parameters.
Previous studies have found an improvement in
global subjective symptoms, but our study evaluated
each of the RLS symptoms individually and found a
statistically significant decrease in paresthesias, motor
restlessness and daytime fatigue. Each of the subjec-
tive variables in our study was measured while the
patients were awake prior to sleep onset. In contrast,
the PSG and actigraphic data were collected while the
patients were in bed both awake and asleep. Since ac-
tigraphy measures both motor restlessness and PLMS,
the nonsignificant decrease in motor activity on acti-
graphic recordings may be due to decreased motor
restlessness.
On the other hand, because many of our patients
experienced adverse effects with clonidine, they may
have guessed which treatment they were receiving,
which would compromise the subjective symptom re-
ports from the patient diaries. The failure to find any
dose relationship to symptom relief is suggestive of a
placebo effect. In addition, we cannot exclude the pos-
sibility that a nonspecific sedative effect may have
caused the decrease in sleep latency seen in our pa-
tients. Thus, it is important to combine the results from
the patient diaries, actigraphic recordings and the PSG
data when judging the effectiveness of clonidine in
RLS.
57
Because there is no animal model that evaluates the
etiology of RLS, hypothesized etiologies are based on
clinical theories and the mechanism of drugs that im-
prove RLS symptoms in humans. Clinical evidence
suggests that RLS symptoms may be mediated by in-
creased vascular tone and decreased blood flow to the
extremities. Patients with PLMS often have cold feet
and decreased peripheral pulses (37,38). Thermal bio-
feedback has decreased PLMS symptoms, presumably
by increasing blood flow to the legs (37) and Watanabe
et al. observed that two patients developed PLMS in
Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024
sleep during epidural or spinal anesthesia probably be-
cause of decreased blood flow to the legs (39). Phen-
oxybenzamine, a postsynaptic ai-adrenergic blocker
and vasodilator, restored peripheral pulses, decreased
PLMS and improved insomnia in two RLS patients
(38). From these results, the investigators hypothesized
that the sympathetic nervous system may mediate the
periodicity of sleep-related periodic leg movements.
Sympathetic nerve blockade with bupivacaine (0.25-
0.5%) also reversed PLMS-like movements during
wakefulness in two patients who probably had RLS
(40). Clonidine may improve peripheral vascular cir-
culation and RLS symptoms by activating inhibitory
presynaptic aT adrenergic receptors in the brainstem
(possibly at the level of the locus ceruleus) leading to
peripheral vasodilation (41). The beneficial effect of
phenyoxybenzamine on PLMS versus the lack of ef-
fect of clonidine on PLMS could be explained by the
differential effect of these two drugs on ai-postsyn-
aptic versus a 2-presynaptic adrenergic receptors. This
effect would suggest that the ai-adrenergic system is
more intimately involved than the aradrenergic sys-
tem in the pathogenisis of PLMS.
Based on cautious interpretation of the subjective
results, combined with the objective measurements
from the PSG and actigraphic recordings, clonidine
may be effective for RLS patients with delayed sleep
onset caused primarily by paresthesias and motor rest-
lessness. Clonidine may be a good alternative for RLS
patients that have hypertension, do not have a large
number of sleep-disrupting PLMS or who have failed
standard therapy (levodopa/dopamine agonists, ben-
zodiazepines or opioids). Subjectively, c10nidine did
not cause daytime somnolence or rebound daytime
symptoms, which is a disadvantage of some current
therapies. Clonidine should be reserved for second-line
therapy, initiated at a low dose and titrated up slowly
to effect. We would suggest starting at 0.1 mg several
hours prior to the onset of symptoms with an addi-
tional dose at bedtime if needed. A response is usually
observed within several days; however, case reports
indicate that patients may require up to a month to
show a response (24). Therefore, an adequate trial may
require several weeks of administration.
Sleep, Vol. 19, No.1, 1996
58 M. L. WAGNER ET AL.
Acknowledgements: We acknowledge Christine Dunne,
Karen De Russo, R. Psg T, Annette Zoe and Mitchell Ru-
binstein, R Psg T., for their technical assistance. The research
was supported in part by a grant from the American Asso-
ciation of Colleges of Pharmacy (M.L.W.) and the Veterans
Administration Merit Review Program (A.S.W.). W.A.H. is
supported by a grant from the Veterans Administration Ca-
reer Development Program. The results from this study have
been previously presented at the 8th Annual Northeastern
Sleep Society Meeting in March 1994, the American Acad-
emy of Neurology in May 1994 and the American College
of Clinical Pharmacy in August 1994. We would like to
thank Boehringer Ingelheim, Inc. for providing clonidine and
matching placebo tablets for the study.
REFERENCES
I. Walters AS, Hening W. Clinical presentation and neurophar-
macology of restless legs syndrome. Clin Neuropharmacol
1987;10:225-37.
2. Banerji NK, Hurwitz LJ. Restless legs syndrome, with particular
reference to its occurrence after gastric surgery. Br Med J
1970;4:774-5.
3. Ekbom KA. Restless legs. Acta Med Scand 1945;Suppl 158:1-
123.
4. Walters AS, Hening WA, Chokroverty S. Review and videotape
recognition of idiopathic restless legs syndrome. Movement Di-
sord 1991;6:105-10.
5. Zoe A, Wagner ML, Walters AS. High-dose clonidine in a case
of restless legs syndrome. Ann Pharmacother 1994;28:878-81.
6. Ekbom KA. Restless legs syndrome. Neurology 1960;10:868-
73.
7. Krueger BR. Restless legs syndrome and periodic movements
of sleep. Mayo Clin Proc 1990;65:999-1006.
8. Montplaisir J, Godbout R, Boghen D, DeChamplain J, Young
SN, Lapierre G, et al. Familial restless legs with periodic move-
ments in sleep: electrophysiologic biochemical, and pharmaco-
logic study. Neurology 1985;35:130-4.
9. Callaghan N. Restless legs syndrome in uremic neuropathy.
Neurology 1966; 16:359-61.
10. Harvey JC. Cholesterol crystal microembolization: a cause of
the restless leg syndrome. South Med J 1976;69:269-72.
II. Nordlander NB. Therapy in restless legs. Acta Med Scand
1953;145:453-7.
12. Lutz EG. Restless legs, anxiety and caffeinism. J CUn Psychi-
atry 1978;39:693-8.
13. Brodeur C, Montplaisir J, Godbout R, Marinier R. Treatment of
restless legs syndrome and periodic movements during sleep
with L-dopa: a double-blind, controlled study. Neurology
1988;38:1845-8.
14. Kaplan PW, Allen RP, Buchholz DW, Walters JK. A double-
blind, placebo-controlled study of the treatment of periodic limb
movements in sleep using carbidopa/levodopa and propoxyphe-
ne. Sleep 1993;16:717-23.
15. Walters AS, Hening WA, Kavey N, Chokroverty S, Gidro-Frank
S. A double-blind randomized crossover trial of bromocriptine
and placebo in the restless legs syndrome. Ann NeuroI1988;24:
455-8.
16. Telstad W, Sprensen 0, Larsen S, Lillevold PE, Stensrud P, Ny-
berg-Hansen R. Treatment of the restless legs syndrome with
carbamazepine: a double blind study. Br Med J 1984;288:
444-6.
17. Peled R, Lavie P. Double-blind evaluation of clonazepam on
periodic leg movements in sleep. J Neurol Neurosurg Psychi-
atry 1987;50: 1679-81.
Sleep, Vol. 19, No.1, 1996
18. Guilleminault C, Flagg W. Effect of baclofen on sleep-related
periodic leg movements. Ann Neurol 1984;15:234-9.
19. Walters AS, Wagner ML, Hening WA, Grasing K, Mills R,
Chokroverty S, et al. Successful treatment of idiopathic restless
legs syndrome in a randomized, double-blind trial of oxycodone
versus placebo. Sleep 1993;16:327-32.
20. Zucconi M, Coccagna G, Petronelli R, Gerardi R, Mondini S,
Cirignotta F. Nocturnal myoclonus in restless legs syndrome ef-
fect of carbamazepine treatment. Functional Neurology 1989;4:
263-71.
21. Guilleminault C, Cetel M, Philip P. Dopaminergic treatment of
restless legs and rebound phenomenon. Neurology 1993;43:445.
22. Handwerker JV, Palmer RF. Clonidine in the treatment of "rest-
less leg syndrome". N Engl J Med 1985 ;313: 1228-9.
23. Steiner JC. Clonidine helps in restless legs syndrome. Neurology
Downloaded from https://academic.oup.com/sleep/article/19/1/52/2749779 by guest on 03 January 2024
1987;37(Suppl 1):278. [Abstract.)
24. Bastani B, Westervelt FB. Am J Kidney Dis 1987;10:326-7.
[Letter.)
25. Ausserwinkler M, Schmidt P. Erfolgreiche Behandlung des rest-
less legs Syndroms bei chronischer Niereninsuffizienz mit Clon-
idin. Schweiz Med Wschr 1989;119:184-6.
26. Bamford CR, Sandyk R. Failure of clonidine to ameliorate the
symptoms of restless legs syndrome. Sleep 1987; 10:398-9. [Let-
ter.)
27. Carskadon MA, Rechtschaffer A. Monitoring and staging hu-
man sleep. In: Kryger MH, Roth T, Dement WC, eds. Principles
and practices of sleep medicine. Philadelphia: WB Saunders,
1994:943-60.
28. Downey R, Bonner M, Von Pohle W, Wilms DJ. Comparison
of sleep stage and arousal indices in normal young adult sleep-
ers, normal elderly adult sleepers and elderly adult insomniacs.
Sleep Res 1989;18:137.
29. Hoelscher TJ, Ware JC, McBrayer RH. Insomnia with periodic
EEG arousals occurring in the absence of apnea and myoclonus.
Sleep Res 1989;18:245.
30. Atlas Task ·Force. EEG arousals: scoring rules and examples.
ASDA report. Sleep 1993;15:173-84.
31. Rosner B. Fundamentals of biostatistics, 2nd edition. Boston:
Duxbury Press, 1986:278-93.
32. Keppel G. Design and analysis: a researchers handbook, 2nd
edition. Englewood Cliffs NJ: Prentice Hall Inc., 1982:382-406.
33. Nicholson AN, Pascoe PA. Presynaptic Ci 2-adrenoreceptor func-
tion and sleep in man: studies with c10nidine and idazoxan.
Neuropharmacol 1991 ;30:367-72.
34. Carskadon MA, Cavallo A, Rosekind MR. Sleepiness and nap
sleep following a morning dose of clonidine. Sleep 1989;12:
338-44.
35. Issa BG. Effect of clonidine in obstructive sleep apnea. Am Rev
Respir Dis 1992;145:435-9.
36. Kostis JB, Rosen RC, Holzer BC, Randolph C, Taska LS, Miller
MH. CNS side effects of centrally-active antihypertensive
agents: a prospective placebo-controlled study of sleep, mood
state, and cognitive and sexual function in hypertensive males.
Psychopharm 1990;102:163-70.
37. Ancoli-Israel S, Seifert AR, Lemon M. Thermal biofeedback
and periodic movements in sleep: patients' SUbjective reports
and a case study. Biofeedback and Self Reg 1986; 11: 177-88.
38. Ware JC, Blumoff R, Pittard JT. Peripheral vasoconstriction in
patients with sleep related periodic leg movements. Sleep
1988; 11: 182-7.
39. Watanabe S, Ono A, Naito H. Periodic leg movements during
either epidural or spinal anesthesia in an elderly man without
sleep-related (nocturnal) myoclonus. Sleep 1990;13:262-6.
40. Uchihara T, Ichikawa T, Furukawa T, Tsukagoshi H. Myoclonus
with burning sensation in legs that remits with sympathetic
blockade. J Neural Sci 1990;100:161-4.
41. Gerber JG, Nies AS. Antihypertensive agents and the drug ther-
apy of hypertension. In: Gilman AG, Ral TW, Nies AS, Taylor.
P, eds. The pharmacological basis of therapeutics, 8th edition.
New York: Pergamon Press, 1990:784-813.