www.elsevier.com/locate/sleep

Sleep Medicine 8 (2007) 491–497

Original article

First night efficacy of pramipexole in restless legs syndromeand periodic leg movements

Mauro Manconi a,*, Raffaele Ferri b, Marco Zucconi a, Alessandro Oldani a,Maria Livia Fantini a, Vincenza Castronovo a, Luigi Ferini-Strambi a

a Sleep Disorders Center, Department of Neurology, Scientific Institute and University Ospedale San Raffaele, Vita-Salute University, Milan, Italyb Sleep Research Centre, Department of Neurology I.C. Oasi Institute (IRCCS), Troina, Italy

Received 14 July 2006; received in revised form 19 September 2006; accepted 14 October 2006Available online 18 May 2007

Abstract

Objective: Restless legs syndrome (RLS) seems to improve immediately after a single dose of dopamine-agonists (DA). The aim ofthe present study was to investigate the acute effects of a low standard dose of pramipexole in RLS drug-naı̈ve patients.Methods: A single-blind placebo-controlled study in 32 consecutive idiopathic RLS de-novo patients was carried out. Patients whomet the standard criteria for RLS, with a PLMS index greater than 10 as well as an RLS rating scale score greater than 20 underwentclinical and neurophysiological evaluation, hematological screening and two consecutive full-night polysomnographies. On the sec-ond night, all patients received 0.25 mg of pramipexole or placebo at 9:00 p.m. Acute symptom response was assessed by a visualanalogical scale (VAS).Results: Eighteen patients received pramipexole and 14 patients received placebo. Compared to placebo, the single low dose(0.25 mg) of pramipexole significantly improved RLS symptoms (VAS: from 7.4 ± 1.68 to 1.3 ± 1.62, p < 0.00001) and stronglyreduced PLMS index (from 45.8 ± 33.56 to 9.4 ± 11.40, p < 0.0002). A significant increase in the percentage of stage 2 non-rapideye movement (NREM) sleep was also observed in the pramipexole group (from 38.7 ± 10.50 to 50.6 ± 12.13, p < 0.02).Conclusions: A low dose of pramipexole was effective in treatment-naı̈ve patients with RLS from the first night of administration.These results support a direct involvement of the dopaminergic system in RLS pathogenesis and might have important implicationsfor a possible future pramipexole administration on-demand, as well as for a pharmacological test to confirm diagnosis in clinicallycomplex cases.� 2006 Elsevier B.V. All rights reserved.

Keywords: Restless legs syndrome; Periodic leg movements; Pramipexole; Sleep; Dopamine; On-demand Therapy

1. Introduction

Restless legs syndrome (RLS) is a common sleep-re-lated movement disorder characterized by uncomfort-able sensations in the limbs appearing or becomingworse at evening/night rest and alleviated by motoractivity [2]. Most RLS patients present periodic limbmovements during sleep (PLMS) [3]. The diagnosis of

1389-9457/$ - see front matter � 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.sleep.2006.10.008

* Corresponding author. Tel.: +39 2 2643 3358; fax: +39 2 26433394.

E-mail address: manconi.mauro@hsr.it (M. Manconi).

RLS is currently based on the ability of the patient todescribe his or her symptoms and of the physician tomatch these symptoms to the clinical diagnostic criteria[2,3]. As in other neurological diseases not associatedwith objective reliable markers, the response to standardtreatments often helps in confirming the diagnosticassessment.

Because of their remarkable efficacy and tolerability,dopamine-agonists are nowadays considered the firstchoice for treatment of RLS [4]. At the present time, car-bidopa/levodopa is not frequently used because of itsshort half-life and the consequent high incidence of

492 M. Manconi et al. / Sleep Medicine 8 (2007) 491–497

symptom rebound and/or augmentation [5]. Further-more, typical side effects associated with ergot-derivatemolecules partially limit the use of bromocriptine, per-golide and cabergoline, and often require a concomitantmedication with the peripheral dopamine-antagonistdomperidone [6]. Due to their tolerability and half-life,low evening doses of the D3-agonists pramipexole andropinirole have become the first of line treatment inRLS [4]. However, follow-up studies demonstrated anaugmentation effect also following the use of D3non-ergoline agonists in about one-third of patients,appearing after at least six months of therapy [7]. Somecontrolled studies, supported by polysomnographic(PSG) recordings, proved that ropinirole and pramipex-ole are noticeably effective in reducing both symptomsand PLMS [8–10,1,8,11]. In one of these studies [10], theefficacy of ropinirole has been demonstrated after a singledrug administration in RLS patients not naı̈ve totreatment.

Response to dopaminergic medications, together withpositive family history and presence of PLMS, are con-sidered to be supportive criteria for the diagnosis ofRLS [4]. In clinical practice, it is common that pramip-exole is effective for RLS from the first days of treat-ment. These empirical data have never beendemonstrated by an experimental procedure.

The aim of the present investigation was to evaluate theinitial response to a standard low dose of pramipexole, ina cohort of drug-naı̈ve patients affected by idiopathicRLS, by means of a controlled clinical PSG study.

2. Subjects and methods

2.1. Subjects

A prospective single-blind controlled study was car-ried out in consecutive subjects affected by idiopathicRLS. According to the International RLS Study Group,the minimal criteria for the diagnosis of RLS were legrestlessness, usually accompanied or caused by uncom-fortable and unpleasant sensations in the legs; beginningor worsening of this unpleasant sensation during rest orinactivity such as lying or sitting; partial or total relief ofthe unpleasant sensations by movement; and worseningor occurrence of the unpleasant sensations in the even-ing or night, compared to daytime [11]. To be includedin the study, the mean frequency of symptoms duringthe last six months had to be greater than two timesper week, with a score of at least 20 on the InternationalRLS Study Group rating scale (corresponding to severeRLS) at the time of enrollment [11]. Only patients free ofmedication at the time of the evaluation, and never pre-viously treated for RLS (dopaminergic agents, benzodi-azepines, opioids and anticonvulsants), were included.Additional inclusion criteria were being between 18and 70 years old and having a PLMS index greater than

10 during the baseline PSG (see below). Patients with anapnea/hypopnea index >5 were excluded.

Results of neurological examination in all patientswere unremarkable. Routine blood tests (includingserum iron and ferritin, B12 vitamin, and folate), as wellas electromyography (EMG) and electroneurography ofthe lower limbs, were also normal. Patients sufferingfrom known causes of secondary RLS (e.g., renal fail-ure, anemia with iron-deficiency, pregnancy, rheuma-toid arthritis, or clinical peripheral neuropathy), othersleep disorders (e.g., narcolepsy, parasomnia and sleepbreathing disorder), other movement disorders, or anymedical conditions that would affect the assessment ofRLS (e.g., fibromyalgia syndrome) were also excluded.

Subjects were randomly assigned to two groups(treated and placebo) and underwent an adaptationnight in the lab, followed by two nocturnal PSG record-ings. No medication was administered before the firstnight recording (baseline). Before the second night ofrecording, subjects included in the treatment groupreceived a single oral dose of 0.25 mg pramipexole at9:00 p.m., while the remaining patients received placebo.All patients gave their written consent for these proce-dures and were unaware of the content (drug or placebo)of the medication. In the morning, after each PSGrecording, all patients evaluated the severity of their pre-vious night symptoms by means of the Visual Analogi-cal Scale (VAS) [12]. The local ethical committeeapproved the study.

2.2. Nocturnal polysomnographic (PSG) studies

Nocturnal PSGs were carried out after a night ofadaptation in a standard sound-attenuated (noise levelto a maximum of 30 dB nHL) sleep laboratory room.Subjects were not allowed to drink caffeinated beveragessix hours before the beginning of each PSG and wereallowed to sleep until their spontaneous awakening inthe morning. Lights-out time was based on the individ-ual’s usual bedtime and ranged between 10:30 and 11:30p.m. The following signals were recorded: electroen-cephalogram (EEG) (six channels, including C3 or C4and O1 or O2, referred to the controlateral mastoid);electrooculogram (EOG; electrodes placed 1 cm abovethe right outer cantus and 1 cm below the left outer can-tus and referred to the mastoid); EMG of the submen-talis muscle; EMG of the right and left tibialis anteriormuscles (bipolar derivations with two electrodes placed2 cm apart on the anterior tibialis muscle of each leg,impedance was kept less than 10 KX, according to theAmerican Sleep Disorders Association (ASDA) scoringcriteria) [13]; and electrocardiogram (ECG; one deriva-tion). Sleep signals were sampled at 200 Hz and storedon hard disk in European data format (EDF) [14] forfurther analysis. EMG signals, in particular, were digi-tally band-pass filtered at 10–100 Hz, with a notch filter

M. Manconi et al. / Sleep Medicine 8 (2007) 491–497 493

at 50 Hz. The sleep respiratory pattern of each patientwas monitored using oral and nasal airflow thermistorsand/or nasal pressure cannula, thoracic and abdominalrespiratory effort strain gauge and by monitoring oxy-gen saturation (pulse-oximetry). This was performed inall subjects in a previous recording (within 1 week) orduring the study recording.

2.3. Sleep scoring and detection of leg movements

Prior to any recording, we verified that the EMGamplitude recorded from the two tibialis anterior mus-cles was below 2 lV at rest and exceeded 7–10 lV forsmall voluntary dorsiflexions of the foot. EMG ampli-tude at maximal deflection was also measured for theapplication of the ASDA scoring criteria [13]. Sleepstages were scored following standard criteria [15] on30-s epochs using the sleep analysis software Hypnolab1.2 (SWS Soft, Italy). Leg movements (LM) during sleepwere first detected by the same software, which allowsfor computer-assisted detection. With this software,detection of LMs is performed using a human-super-vised automatic approach controlled by the scorer fol-lowing ASDA criteria [13]. The performance of thissystem has been recently evaluated and validated [16],but in this study one scorer visually edited, epoch byepoch, the detections proposed by the automatic analy-sis, before computing a final result. In particular, thetotal LM index was calculated to represent the totalnumber of LM per hour of sleep. The PLMS indexwas calculated as the number of LM included in a seriesof four or more, separated by more than five and lessthan 90 s, per hour of sleep. Analysis included thePLMS index of the entire night, and separately ofREM and NREM sleep, as well as the total number ofLMs and number of PLMS sequences. Finally, thechange in PLMS index observed between the baselineand the treatment nights was normalized for its baselinevalue by means of the calculation of the following ratio:

ðbaseline PLMS index

� treatment PLMS indexÞ=baseline PLMS index

that will be indicated here as ‘‘PLMS change index’’.Mathematically, this index can vary from +1 (completedisappearance of PLMS in the treatment night) to �1(appearance of PLMS in the treatment night).

Table 1Demographic data of the patients and results of the subjective evaluation of RRLS Study Group rating scale

Males (n) Females (n)

Total 9 23Pramipexole group 4 14Placebo group 5 9

2.4. Statistical analysis

The comparison between the different sleep and LMparameters obtained before and after the administrationof pramipexole or placebo was carried out by means ofthe factorial analysis of variance (ANOVA), with ‘‘group’’(pramipexole or placebo) and ‘‘night’’ (baseline or treat-ment) as factors, which was followed by post hoc analysisfor the individual group or night comparisons. For thisanalysis, the significance level was set at p < 0.05 andthe commercially available Statistica software package(StatSoft, Inc., 2001. STATISTICA data analysissoftware system, version 6. www.statsoft.com) was used.

3. Results

Thirty-two consecutive untreated patients were includ-ed in this study (mean age 58.4 ± 11.6, 9 males and 23females); 18 patients received pramipexole and 14 weregiven placebo. Table 1 reports the demographic data ofthe patients and the results of the subjective evaluation ofthe severity of the RLS symptoms measured before treat-ment by the International RLS Study Group rating scale.

One patient treated with pramipexole and anotherwho was administered placebo reported mild morningnausea; no other significant side effects were reportedby any of the rest of the patients.

The mean VAS score before and after treatmentchanged from 7.4 ± 1.68 standard deviation (SD) to1.3 ± 1.62 SD (p < 0.00001) in the pramipexole group,and from 6.8 ± 1.72 SD to 5.4 ± 2.33 SD (not signifi-cant) in the placebo group. The VAS score improvedin all subjects and reached the value of 0 in 10 out of18 patients treated with pramipexole, while it did notchange in two patients and became worse in anotherthree who took the placebo.

Table 2 shows the comparison between the differentsleep scoring parameters obtained before and after theadministration of pramipexole or placebo. In this table,the ANOVA was significant only for the factor ‘‘night’’and for the percentage of sleep stage 2 which wasincreased in the pramipexole treatment night. Otherparameters, in addition, such as time in bed and sleepefficiency, tended to show evident modifications in thepramipexole treatment night but these differences didnot reach statistical significance.

LS symptom severity, measured before treatment by the International

Age, mean (SD) IRLSSG score, mean (SD)

58.4 (11.6) 27.0 (4.85)59.8 (11.2) 27.3 (5.19)56.6 (12.3) 26.6 (4.55)

Table 2Comparison between the different sleep scoring parameters obtained before and after the administration of pramipexole or placebo

1st Night 2nd Night ANOVA Post hoc

Pramipexole Placebo Pramipexole Placebo Baseline vs. Treatment Pramipexole vs.Placebo

Mean SD Mean SD Mean SD Mean SD Night Group Pramipexole Placebo Baseline Treatment

TIB (min) 512.9 100.02 509.3 89.97 525.6 44.87 481.6 97.79 NS NSSPT (min) 473.4 102.64 473.3 87.87 499.9 45.76 453.3 93.72 NS NSTST (min) 343.5 131.20 378.4 87.60 414.3 57.86 354.1 76.44 NS NSSOL (min) 27.1 34.98 22.7 24.06 12.4 8.14 17.3 21.93 NS NSFRL (min) 134.4 79.45 100.5 40.93 140.8 99.15 122.2 67.81 NS NSSS/hour 11.0 4.23 12.3 3.35 14.2 4.48 12.4 4.16 NS NSAWN/hour 4.8 2.95 4.6 2.44 5.7 3.71 4.6 2.33 NS NSSleep efficiency, % 65.6 21.00 74.3 10.02 78.9 8.36 74.3 11.07 NS NSWASO (min) 129.9 78.27 94.9 61.31 85.6 40.16 99.2 71.27 NS NSS1 (min) 26.1 24.01 30.5 19.71 33.3 26.47 24.8 22.74 NS NSS2 (min) 184.4 66.79 199.8 66.71 253.6 66.01 191.9 46.61 NS NSSWS (min) 66.2 41.04 69.7 37.66 66.2 35.99 74.0 31.60 NS NSREM (min) 66.8 41.38 78.5 29.39 61.3 28.48 63.4 21.46 NS NSWASO, % 29.7 19.73 19.8 11.99 17.2 7.99 21.0 12.05 NS NSS1(%) 5.5 5.02 6.3 3.72 6.6 4.93 5.5 4.64 NS NSS2 (%) 38.7 10.50 41.9 10.70 50.6 12.13 43.4 10.85 0.02 NS 0.0005 NSSWS (%) 13.1 7.64 15.4 9.70 13.5 7.84 16.3 6.46 NS NSREM (%) 13.1 7.69 16.5 5.23 12.2 5.18 13.8 3.98 NS NS

TIB, time in bed; SPT, sleep period time; TST, total sleep time; SOL, sleep onset latency; FRL, first REM latency; SS, stage shifts; AWN, awakeningsnumber; WASO, wakefulness after sleep onset;S1, stage 1; S2, stage 2; SWS, slow-wave sleep; REM, REM sleep.

494 M. Manconi et al. / Sleep Medicine 8 (2007) 491–497

Table 3 reports the comparison between the differ-ent LM scoring parameters obtained before and afterthe administration of pramipexole or placebo. In thiscomparison, whole-night PLMS index and NREMsleep total LM and PLMS index were significantlylower during the pramipexole treatment night thanduring the baseline recording. The same parametersshowed a statistically significant difference betweenthe two patient groups, after treatment but not atbaseline. No detectable changes followed the placebotreatment.

Fig. 1 shows, for example, the hypnogram of thebaseline and treatment nights in a patient who wasadministered pramipexole; in this figure it is possibleto see the important changes in LM activity (a detail isshown in the upper panel, in baseline conditions) andof sleep structure occurring after treatment.

Fig. 2 reports the individual and average (with SD)values of the PLMS change index in the two groups ofpatients; it is possible to notice that the pramipexolegroup shows significantly higher values of this indexthan the placebo group (p < 0.00001). Moreover, 13out of 18 subjects receiving pramipexole show a PLMSchange index higher than the highest observed in theplacebo group (0.726); the same figure also shows theindividual values of the PLMS index in both groups ofpatients, at baseline and after treatment. Finally, no sig-nificant correlation was found between the PLMSchange index and the magnitude of VAS change for sub-jects receiving pramipexole (Pearson r = 0.140) or place-bo (Pearson r = 0.106).

4. Discussion

The acute administration of a low dose (0.25 mg) ofpramipexole in treatment-naı̈ve, idiopathic, severe RLSpatients markedly improved the symptoms from the firstnight of therapy in all treated subjects, leading to a com-plete disappearance of symptoms in more than half ofthe cases.

The most remarkable objective effect of pramipexoleis represented by the immediate drop of the PLMSindex, clearly evident in all subjects who took the drug.Despite the inclusion of severe RLS patients withPLMS, our unambiguous results may be also an effectof the fact that only patients who had never been treatedfor RLS previously were included. At the same time it isimportant to note that, as expected because of theiroccurrence during sleep, PLMS do not seem to be influ-enced by the placebo effect [11]. The abrupt response ofsymptoms and PLMS to pramipexole stands for a pos-sible common dopaminergic origin of both of thesepathological phenomena [17]. Our results support thehypothesis of a direct involvement of D3 subtype recep-tors in the pathogenesis of PLMS and RLS symptoms[18–20]. Pramipexole is a potent dopamine auto-recep-tor agonist which interacts with dopamine Gprotein-coupled D2 subfamily receptors (D2, D3, andD4 subtypes) showing a weaker bind to D1 subfamilyreceptors (D1 and D5 subtypes) and almost absent toadrenergic or serotonergic receptors [21]. Furthermore,pramipexole has a preferential affinity to D3 receptorswith a selectivity 15 times more potent at D3 than D2

Tab

le3

Co

mp

aris

on

bet

wee

nth

ed

iffer

ent

sco

rin

gp

aram

eter

sfo

rle

gm

ove

men

ts,

ob

tain

edb

efo

rean

daf

ter

the

adm

inis

trat

ion

of

pra

mip

exo

leo

rp

lace

bo

1st

Nig

ht

2nd

Nig

ht

AN

OV

AP

ost

ho

c

Pra

mip

exo

leP

lace

bo

Pra

mip

exo

leP

lace

bo

Bas

elin

evs

.T

reat

men

tP

ram

ipex

ole

vs.

Pla

ceb

o

Mea

nS

DM

ean

SD

Mea

nS

DM

ean

SD

Nig

ht

Gro

up

Pra

mip

exo

leP

lace

bo

Bas

elin

eT

reta

men

t

NR

EM

slee

p

To

tal

LM

ind

ex58

.134

.47

57.4

27.9

419

.614

.15

64.2

39.2

10

.03

70

.00

40

.00

03

NS

NS

0.0

000

7

PL

MS

ind

ex50

.432

.84

49.2

28.6

310

.413

.12

55.0

37.9

40

.02

10

.00

40

.00

00

7N

SN

S0

.00

00

5

RE

Msl

eep

To

tal

LM

ind

ex35

.640

.55

27.1

17.6

017

.48.

8932

.027

.61

NS

NS

PL

MS

ind

ex24

.343

.72

15.9

17.1

13.

84.

6621

.431

.62

NS

NS

Tota

l

To

tal

LM

ind

ex54

.334

.61

51.1

25.0

319

.312

.25

58.4

37.6

90

.01

4N

SN

S0

.00

03

PL

MS

ind

ex45

.833

.56

42.3

25.7

39.

411

.40

48.8

36.7

90

.03

60

.01

30

.00

02

NS

NS

0.0

002

PL

MS

Seq

uen

cen

um

ber

11.4

4.89

11.9

5.86

7.1

7.79

10.5

4.44

NS

NS

M. Manconi et al. / Sleep Medicine 8 (2007) 491–497 495

receptors [22]. D3 receptors appear to be concentrated inthe limbic system, which is more related to cognitiveaspects, and in the dorsal horns and intermediolateral-gray matter of the spinal cord, likely to be involved innociceptive modulation [23–26]. Spinal D3 receptorsreceive dopaminergic projections from the A11 nuclei,located within the lateral hypothalamus, which are sus-pected to play a role in the pathogenesis of RLS, and areclosely related to the central circadian pacemaker of thesuprachiasmatic nucleus [27].

In comparison to the effect on the subjective RLScomponents and on PLMS, the pharmacological effecton sleep macrostructure is less clear-cut but leanstoward being a general improvement of hypnic parame-ters. From a panoramic vision of the PSG results, a ten-dency towards a quantitative sleep improvement (sleepeffiency, total sleep time), especially for the light NREMstages, after acute pramipexole treatment, becomes evi-dent. Part of these results did not reach statistical signif-icance probably because of the high variance of the data.Interestingly, this rise in sleep quantity does not seem tobe accompanied by a similar improvement in sleep con-tinuity, with a propensity toward sleep fragmentation(increased number of awakenings/hour and stageshifts/hour) and an inefficacy in consolidating slow wavesleep. Probably, a longer period of treatment would benecessary for a more effective stabilization of sleep con-tinuity. Moreover, PLMS might not be directly respon-sible for the instability in EEG microstructure and thedisappearance of LMs is not automatically followedby a similar reduction in sleep fragmentation indexes.Saletu et al. [10] performed an acute (night-to-night)treatment evaluation in RLS patients by using 0.5 mgof ropinirole and found a significant ‘‘decrease in arous-als due to PLMS, but an increase in spontaneous arous-als after medication’’. This effect may also be interpretedas a selective action of ropinirole on PLMS, without adirect influence on the electro-cortical arousal activitythat remains unchanged after treatment.

The demonstration of an immediate efficacy of ropin-irole and pramipexole is extremely important, not onlybecause it is an obvious benefit to patients able to obtainrapid therapeutic results, but also because it opens thedoors to two other possibly crucial drug applications,already discussed, in the near future for idiopathicRLS: on-demand therapy and the use of a standard sin-gle dopaminergic agent as a diagnostic ex-juvantibus test[28,29]. On-demand therapy may have significant resultsfor patients with moderate or intermittent symptomsand for patients (treated and untreated) who are goingto face forced immobility for a long time, such as inair, train or car travel, social events during the eveningor sedentary working postures. In these cases, it isrecommended that patients take the drug about 40–60 min before the onset of the desired therapeutic windowand choose middle/short half-life dopamine-agonists

Fig. 1. Examples of the hypnogram at baseline and treatment nights (bottom panels) in a patient who was administered pramipexole; a short PSGsegment of the baseline recording is shown in the top panel, for example. Abbreviations: LOC, ROC, left and right electro-oculogram; A1, A2, leftand right earlobes; ECG, electrocardiogram; W, wakefulness; R, REM sleep; S1, S2, S3, S4, sleep stages 1, 2, 3, and 4.

Fig. 2. Individual and average (with S.D., whiskers) values of the PLMS change index in the two groups of patients (middle panel); individual valuesof the PLMS index in patients treated with pramipexole (left panel) and in those treated with placebo (right panel), at baseline and after treatment.

496 M. Manconi et al. / Sleep Medicine 8 (2007) 491–497

that do not need a pre-medication [28]. Regarding thesecond possible application, notwithstanding the factthat all of our patients showed an improvement inPLMS and symptoms (high sensitivity), the diagnosticvalue of a single-shot drug test might be decreased bythe existence of false responders (low specificity), simi-larly to patients who show improved PLMS index orsymptoms as a result of physiological night-to-night var-iability. We must also note that a possible limit of our

study was the absence of a longer period of baselineevaluation. On the other hand, a diagnostic test mightbe useful even if it can only exclude the diagnosis ofRLS in the case of worsening of PLMS and/or symp-toms; in fact, we never observed in our study an increasein PLMS index in subjects treated with pramipexole.The real feasibility of this hypothetical test should bealso evaluated on the basis of cost, because the patientsneed to undergo two consecutive full-night PSG studies;

M. Manconi et al. / Sleep Medicine 8 (2007) 491–497 497

however, this might be acceptable in consideration oflong-term therapy needed in RLS. Moreover, the repeat-ed application of the Suggested Immobilization Testinstead of the full-night PSG could reduce diagnosticcosts and time of execution [30].

In conclusion, our investigation demonstrates theimmediate effectiveness and the excellent tolerability ofa low standard dose of pramipexole in RLS patients,suggesting interesting pathogenetic speculations andimportant consequences on clinical practice for apossible on-demand treatment. The use of pramipexolein a diagnostic test needs further evidence in terms ofreliability and suitability.

References

[1] Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS,Montplaisi J. Restless legs syndrome: diagnostic criteria, specialconsiderations, and epidemiology. A report from the restless legssyndrome diagnosis and epidemiology workshop at the NationalInstitutes of Health. Sleep Med 2003;4:101–19.

[2] Walters AS. Toward a better definition of the restless legssyndrome. The International Restless Legs Syndrome StudyGroup. Mov Disord 1995;10:634–42.

[3] Thorpy MJ. New paradigms in the treatment of restless legssyndrome. Neurology 2005;64:S28–33.

[4] Allen RP, Earley CJ. Augmentation of the restless legs syndromewith carbidopa/levodopa. Sleep 1996;19:205–13.

[5] Walters AS, Hening WA, Kavey N, Chokroverty S, Gidro-FrankS. A double-blind randomized crossover trial of bromocriptineand placebo in restless legs syndrome. Ann Neurol 1988;24:455–8.

[6] Ondo W, Romanyshyn J, Vuong KD, Lai D. Long-termtreatment of restless legs syndrome with dopamine agonists. ArchNeurol 2004;61:1393–7.

[7] Allen R, Becker PM, Bogan R, Schmidt M, Kushida CA, Fry JM,et al. Ropinirole decreases periodic leg movements and improvessleep parameters in patients with restless legs syndrome. Sleep2004;27:907–14.

[8] Manconi M, Casetta I, Govoni V, Cesnik E, Ferini-StrambiL, Granieri E. Pramipexole in Restless Legs syndrome.Evaluation by suggested immobilization test. J Neurol2003;250:1494–5.

[9] Saletu M, Anderer P, Saletu B, Hauer C, Mandl M, OberndorferS, et al. Sleep laboratory studies in restless legs syndrome patientsas compared with normals and acute effects of ropinirole. 2.Findings on periodic leg movements, arousals and respiratoryvariables. Neuropsychobiology 2000;41:190–9.

[10] Saletu M, Anderer P, Saletu-Zyhlarz G, Hauer C, Saletu B. Acuteplacebo-controlled sleep laboratory studies and clinical follow-upwith pramipexole in restless legs syndrome. Eur Arch PsychiatryClin Neurosci 2002;252:185–94.

[11] Walters AS, LeBrocq C, Dhar A, Hening W, Rosen R, Allen RP,et al. Validation of the International Restless Legs SyndromeStudy Group rating scale for restless legs syndrome. Sleep Med2003;4:121–32.

[12] Wewers ME, Lowe NK. A critical review of visual analogue scalesin the measurement of clinical phenomena. Res Nurs Health1990;13:227–36.

[13] American Sleep Disorders Association. Recording and scoring legmovements. The atlas task force. Sleep 1993;16:748–59.

[14] Kemp B, Varri A, Rosa AC, Nielsen KD, Gade J. A simpleformat for exchange of digitized polygraphic recordings. Electro-encephalogr Clin Neurophysiol 1992;82:391–3.

[15] Rechtschaffen A, Kales A. A manual of standardized terminology,techniques, and scoring system for sleep stages of human subjects.Washington Public Health service. Washington: US GovernmentPrinting Office; 1968.

[16] Ferri R, Zucconi M, Manconi M, Bruni O, Miano S, Plazzi G,et al. Computer-assisted detection of nocturnal leg motor activityin patients with restless legs syndrome and periodic leg move-ments during sleep. Sleep 2005;28:998–1004.

[17] Trenkwalder C, Paulus W. Why do restless legs occur atrest? – Pathophysiology of neuronal structures in RLSNeurophysiology of RLS (part 2). Clin Neurophysiol2004;115:1975–88.

[18] Garcia-Borreguero D, Larrosa O, de la LY, Granizo JJ, Allen R.Correlation between rating scales and sleep laboratory measure-ments in restless legs syndrome. Sleep Med 2004;5:561–5.

[19] Allen RP, Earley CJ. Validation of the Johns Hopkins restless legsseverity scale. Sleep Med 2001;2:239–42.

[20] Montplaisir J, Lorrain D, Godbout R. Restless legs syndrome andperiodic leg movements in sleep: the primary role of dopaminergicmechanism. Eur Neurol 1991;31:41–3.

[21] Piercey MF. Pharmacology of pramipexole, a dopamine D3-preferring agonist useful in treating Parkinson’s disease. ClinNeuropharmacol 1998;21:141–51.

[22] Sautel F, Griffon N, Levesque D, Pilon C, Schwartz JC,Sokoloff P. A functional test identifies dopamine agonistsselective for D3 versus D2 receptors. Neuroreport1995;6:329–32.

[23] Austin J. Schizophrenia: an update and review. J Genet Couns2005;14:329–40.

[24] Bordet R. Central dopamine receptors: general considerations(Part 1). Rev Neurol (Paris) 2004;160:862–70.

[25] Nieoullon A, Amalric M. Dopaminergic receptors: structuralfeatures and functional implications. Rev Neurol (Paris) 2002;158.Spec no 1.

[26] Levant B, McCarson KE. D(3) dopamine receptors in rat spinalcord: implications for sensory and motor function. Neurosci Lett2001;303:9–12.

[27] Ondo WG, He Y, Rajasekaran S, Le WD. Clinical correlates of 6-hydroxydopamine injections into A11 dopaminergic neurons inrats: a possible model for restless legs syndrome. Mov Disord2000;15:154–8.

[28] Silber MH, Ehrenberg BL, Allen RP, Buchfuhrer MJ, Earley CJ,Hening WA, et al. An algorithm for the management of restlesslegs syndrome. Mayo Clin Proc 2004;79:916–22.

[29] Tribl GG, Sycha T, Kotzailias N, Zeitlhofer J, Auff E. Apomor-phine in idiopathic restless legs syndrome: an exploratory study. JNeurol Neurosurg Psychiatry 2005;76:181–5.

[30] Montplaisir J, Boucher S, Nicolas A, Lesperance P, Gosselin A,Rompre P, et al. Immobilization tests and periodic leg move-ments in sleep for the diagnosis of restless leg syndrome. MovDisord 1998;13:324–9.