Parkinsonism and Related Disorders 20S1 (2014) S5–S9
Contents lists available at SciVerse ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Restless legs syndrome in Parkinson’s disease Roselyne M. Rijsman *, Louise F. Schoolderman, Rob S. Rundervoort, Maartje Louter Center of Sleep and Wake Disorders and Department of Neurology, Medical Center Haaglanden, The Hague, The Netherlands
article info
summary
Keywords: Restless legs syndrome RLS Parkinson’s disease PD prevalence mimics sleep profile impact quality of life
The Restless legs syndrome (RLS) and Parkinson’s disease (PD) are two disorders that can co-exist, whether or not they share a common pathophysiology. If, and to what extent RLS and PD share the same pathophysiology, is still under debate. Sleep disturbances are prevalent in PD, and as PD progresses, nocturnal disturbances become even more evident, in association not only with motor symptoms but also with non-motor symptoms. Alertness to, and recognition of, RLS in PD patients with sleep disorders could improve customized treatment and quality of life of these patients. In this article the prevalence of RLS in PD, the clinical profile of RLS in PD, the PD profile of patients with RLS, RLS mimics specifically related to PD and impact of RLS in PD will be reviewed.
1. Introduction Restless legs syndrome (RLS) is a common neurological disorder, with an estimated prevalence of 4–10% in the general population. Although the precise pathophysiology remains unknown, it is assumed that dopaminergic mechanisms play a central role. In the past few years, several studies have been done which suggest an association between RLS and Parkinson’s disease (PD). Both RLS and PD respond to dopaminergic treatment, making the assumption of a common pathophysiology, and therefore an association between the two disorders, fairly plausible [1,2]. The RLS symptoms seem to result from an abnormal sensorimotor integration, due to dysinhibition at the spinal level which, in turn, results from reduced inhibitory mechanisms at the supraspinal level. The descending supraspinal diencephalospinal dopaminergic neurons, originating in the hypothalamus (A11), are proposed to play an important role in the pathophysiology of RLS. It is postulated that, in the course of the PD disease, these neurons degenerate, along with nigrostriatal neurons, leading to the development of RLS, and, therefore, to a higher prevalence of RLS among PD patients. But shared pathophysiology is still under debate [1,2]. Unlike in the case of PD, no neuronal degeneration or Lewy body deposition has ever been recognized in RLS post-mortem studies. Furthermore, until now, no shared disease loci have been identified. Except for the provocation of both RLS and PD by neuroleptics, and possibly lower ferritin levels in PD patients with RLS, the secondary forms of PD and RLS do not form an argument for a common pathophysiology [2]. * Corresponding author. Medical Center Haaglanden, Center for Sleep and Wake Disorders, Lijnbaan 32, 2501 CK The Hague, The Netherlands 1353-8020/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
© 2013 Elsevier Ltd. All rights reserved.
Even though the relationship between RLS and PD is still controversial, RLS has revealed itself to be co-existing with PD. Sleep disturbances are prevalent in PD, and as PD progresses, nocturnal disturbances become even more evident, in association not only with motor symptoms but also with non-motor symptoms. Thus, alertness to, and recognition of, RLS in PD patients with sleep disorders could improve customized treatment and quality of life of these patients. That is why, in this article, we will specifically look at studies on the prevalence and on the clinical profile of RLS in PD, as well as on the impact that RLS has on PD patients. 2. Prevalence of RLS in PD Good comparisons of the different estimates of RLS prevalence in PD studies are rather difficult due to variations in methodology, and so, definite generalisations are hard to make. In total, 18 cross-sectional studies on the prevalence of RLS in PD, published in English, were found [3–20]. In all these studies, the international IRLSSG essential criteria for the diagnosis of RLS were used, but some studies used the criteria of 1995, whereas other (more recent) studies used the revised IRLSG diagnostic criteria of 2003. The methodology, however, for the assessment of these criteria in the subjects was not identical in all studies. In addition, the background of the interviewers varied from study to study, which might influence, of course, the focus, the wording and the interpretation of the RLS histories that were collected. In most studies, the definite diagnosis of RLS was based only on the constellation of the four minimal IRLSSG criteria, but in some other studies, thresholds on severity, frequency, and on the impact on quality of life were included. From a purely epidemiological and pathophysiological point of view, the former procedure may be
S6
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
regarded defensible and thoughtful, but from a clinical point of view, inclusion of the threshold criteria could obviously help to better define PD patients who actually need treatment for RLS. The appearance of secondary RLS is usually related with a late onset of RLS (above 45 years of age). As a result, there may be more secondary forms of RLS in PD cohorts, besides purely PDrelated RLS, simply because PD is also a disease that becomes more frequent with age. Some studies, therefore, deliberately excluded these secondary forms of RLS from their data [3,16], to avoid confusion with this age-related variable in their assessment of the true relation between RLS and PD. From a clinical point of view, one may question whether such an exclusion of these ‘secondary’ forms of RLS from the investigation of the co-morbidity of any sort of RLS in PD is appropriate, because, in doing this, we can obviously not assess their impact and, by implication, also not their possible implications for therapeutic intervention. There was also variation in the size and composition of the PD cohorts, and in the way they were treated. And last but not least, a few studies excluded PD-specific RLS mimics, such as motor symptoms, akathisia, non-motor symptoms, as pain, and other sensory symptoms, motor fluctuation, such as dystonia, and nonmotor fluctuation symptoms, even though the possible overlap of these symptoms with ‘true’ RLS has been discussed explicitly by some authors. The prevalence of RLS in the 18 PD cohorts varied from 0% to 50%. Two of the studies were performed in Latin America; the one from Brazil showed a very high PD-related RLS prevalence of 50%, whereas the other one found 18.8% [5,10]. Eight of the 18 studies were performed in Asian countries, with PD-related RLS prevalence from 0.98% to 16% [4,11–14,17,19,21]. The PD-related RLS prevalence in Western countries was generally higher than those in the Asian countries, namely from 5.5% to 27% in Europe [3,6,8,9,16,18,20], and 20.8% in the USA [15]. Also, the prevalence of RLS in the ‘general population’ is typically lower in Asian countries than in European and North American ones, namely from 0.9% to 12.1% in Asia, against from 3.9% to 18.8% in Europe and the USA [22]. In addition, the correlation with age is clearly different in these two continents: in Asia it does not seem to go up with age, whereas in Europe and the USA, the RLS prevalence in the general population is found to double every 20 years, with a peak at the age of 65 [22]. Since the mean age of all the PD cohorts in which RLS prevalence was examined was at least 59, we may say that the factor age was probably a less confounding factor for the assessment of RLS prevalence in the Asian PD cohorts than in the non-Asian ones. To show a real association between RLS and PD, one should find a significant increase of RLS prevalence in the PD cohorts relative to that in the general population. Seven of the 18 studies [3,6,8,13,17,19,20] did not find such a significant increase and did not support the hypothesis of a similar pathophysiology between RLS and PD. The actual percentages of RLS in these negative studies were between 0% and 5.5% in the Asian countries, and between 5.5% and 12.7% in the Western countries. The majority of all the RLS-prevalence studies in PD were done on patient groups with relatively advanced forms of disease, treated which dopaminergic medication. So there was a serious risk of including RLS confounders and secondary forms of RLS, stemming from other causes, in the assessment. Lee and colleagues mention a prevalence of RLS, not different from that in the general population, in only one subgroup of patients who were not given dopamine medication, and who were still in an early stage of PD [12]. After multivariate analysis of the total PD cohort, only duration of dopaminergic treatment appeared to be a variable that correlated significantly with the presence of RLS. Their conclusion was that RLS in PD is associated with long-term dopaminergic treatment, rather than with PD itself. In 2011, two RLS-prevalence studies were
performed in a cohort with only de novo and dopaminergic naive PD patients. Angelini and colleagues [3] found that RLS prevalence in their dopamine-free PD cohort (5.5%) was not significantly different from that in their control group (2.3%), and concluded that RLS is not really related to PD pathophysiology. The question whether an increase in risk of co-morbidity of RLS with the progression of PD is a result of dopaminergic neuron degeneration, or of dopaminergic treatment, is not answered by this study. Gjerstadt et al. [8] performed a case–control study in an early-phase and dopamine un-medicated PD cohort of 200 patients. Patients who had an urge to move their legs according to the Johns Hopkins diagnostic interview for RLS were categorized as ‘true’ RLS if they fulfilled the four essential IRLSSG criteria, and as ‘Leg Movement Restlessness’ (LMR) if they did not fulfil all four IRLSSG criteria. They found similar RLS prevalence in PD patients and controls. But they demonstrated that LMR grows to almost to a 3-fold higher risk in early PD, as weighted against controls. The authors speculated if RLS and akathisia (as one of the possible interpretations of LMR) may represent overlapping features within the same spectrum of motor restlessness in PD. Suzuki et al. [17] found significantly higher nocturnal restlessness (and not RLS), as measured by the Parkinson disease sleepiness scale (PDSS) sub-items 4 and 5, in PD compared to controls. This nocturnal restlessness was associated with the PDSS total score, but not with disease severity and treatment duration, or total levo-dopa equivalent (LDE) dose. The nocturnal restlessness was therefore interpreted as being the result of an endogenous dopamine deficit during night-time, rather than as a motor complication (i.e., wearing off phenomenon) due to dopaminergic treatment which can mimic RLS. By contrast, however, ten studies [4,5,9–12,14–17], did find higher RLS frequencies in PD, and, thus, support the hypothesis that RLS is part of the symptomatic profile of PD, and not just an incidental cooccurrence of RLS in PD. RLS prevalence in these ten studies varied from 7.9% to 16% in the Asian countries, against from 20.8% to 27% in the European countries, and from 19.9% to 50% in the SouthAmerican countries. 3. Clinical profile of RLS in PD 3.1. Comparison of PD with and without RLS: risk factor in PD? All but one [5] of the prevalence studies considered in this article have to some extent compared the PD profile of patients with RLS with that of patients without RLS. Several, although not necessarily consistent, significant differences between the two groups were found. Again, since the studies differ in methodology, as well as in composition of PD cohorts, it is not justified to draw definite conclusions. The following is an overview of the major results. Onset of RLS in relation to onset of PD was evaluated in several studies [11,13–16,20]. In all of them, a clear majority of patients (76–100%) showed an onset of RLS only after, or together with the onset of PD. Ondo et al. [15] found that, for PD patients with a positive RLS family history, the RLS onset preceded the PD onset more often (52%) than for PD patients without a positive RLS family history (29%). Given such a result, one might be inclined to say that RLS without a positive family history is more consistent with the ‘secondary’ forms of RLS in PD, but these results could not be replicated in later studies [20]. Though lateralization of RLS is found to be rather high in the PD population (range 35–70%), no correlations were found between side of PD onset and dominant side of the RLS symptoms [4,11,14] which immediately counters the suggestion, of course, that RLS in PD is related to the degeneration of the dopaminergic neurons in the substantia nigra. The mean severity of RLS, expressed in terms of the international RLS severity scale (IRLSS), varied from 11.9±6.3 to 21.3±6.3,
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
or from mild to moderately severe RLS [9,10,13,14,20]. Little or no information was provided about the frequency of RLS in PD patients. Only one study mentioned that most of the patients (24%) experienced RLS 2–4 times a month [20]. In some studies, RLS was associated with younger age at time of investigation of the PD patients [14,16], whereas in other studies it was associated with older age, as compared with PD patients without RLS [11]. The possibility of an association between RLS and severity of PD was addressed in many of the studies with the Hoehn and Yahr scale [3,10–12,15,18,20,23] and the UPDRS [4,9–11,18]. Only in two of them, however, a significant, but opposite association was found [12,18]. The possible association between PD duration and RLS was specifically addressed in only one study [12], which reported a positive association: the longer the PD, the more RLS. After multivariate analysis, however, in which several variables were included in the analysis, only duration of treatment was found to be a significant predictor of the presence or absence of RLS (OR = 1.199, 95% CL 1.014–1.419, P = 0.0334). The authors stated that the more frequent RLS in PD patients with longer dopaminergic treatment could be the result of the same mechanisms as those which induce an augmentation of RLS in prolonged levodopa exposure. Although these augmentation mechanisms are not fully clarified yet, Paulus and Trenkwalder have proposed that the augmentation could be a result of overstimulation by the dopaminergic medication of the D 1 dopamine receptors, in comparison with the D 2 receptors, in the spinal cord. The subsequent finding, in another study, that the RLS severity was decreased after subthalamic stimulation concomitant with dopaminergic medication dose reduction, endorses the hypothesis that RLS in PD can be a result of dopamine receptor over-stimulation by dopaminergic medication [24]. The idea that RLS symptoms are correlated with dopamine treatment, rather than with PD pathology itself, is also supported by some other observations, such as: (1) more PD patients without RLS, in contrast to PD patients with RLS, used no levodopa at all (resp. 6.8% and 0%) [9], (2) the same prevalence of RLS as in the general population, in a small subgroup of dopamine-medication free PD patients, compared to an increased prevalence of RLS in the levodopa-treated PD patients [12], (3) equal levels of RLS prevalence in various dopamine-free PD cohorts [3,8]. Besides dopamine treatment duration, also level of dopamine dose might be a factor that leads to overstimulation of the dopaminergic neurons. Although the dopamine therapy dose of PD patients with RLS was compared with that of PD patients without RLS in nine studies [9,11–17,20], only one showed a significant, though not positive but negative, correlation between dopamine dose and RLS. These authors suggest that higher dopaminergic medication could suppress the RLS symptoms. Also Verbaan et al. [20] discuss the possibility that the relatively high mean dopaminergic dose in their cohort could have resulted in an underestimation of RLS patients. This hypothesis was further supported by the observation of emerging RLS in 11 of 195 PD patients after subthalamic stimulation with a concurrent 75% reduction of the mean dopaminergic medication. This can be explained by an unmasking effect of RLS symptoms, which otherwise would be suppressed by a higher dose of dopaminergic medication [25]. Also iron deficiency has been associated with RLS. In two studies [11,15], lower levels of ferritin were found in PD RLSpositive patients than in PD RLS-negative patients. Ondo and colleagues, therefore, speculate that low ferritin levels could, besides reducing the dopaminergic system function, also be a risk factor for RLS symptoms in PD. In two other studies [10,14],
S7
no such statistically significant differences in level of ferritin were found, suggesting that not iron deficiency but dysfunction of the central dopaminergic system itself, due to PD, might be the underlying pathophysiology of RLS in PD. 3.2. Comparison of the RLS profile in PD with that of idiopathic RLS It is well known that the prevalence of RLS in the general population is higher among women than among men. Consistent with this pattern, also several studies on PD patients found a higher RLS prevalence among women than among men (range 60–75%) [6,9,12–14,20]. In only two of these studies, however, the proportion of women was significantly higher in the PD RLS-positive group than in the PD RLS-negative group [9,20]. In other studies on PD patients, however, this typical pattern of gender distribution was not found [3,4,8,10,11,16]. All together, we cannot say that there is a really consistent pattern of gender distribution among PD patients with RLS. Parkinson-related RLS has a remarkably lower rate of RLS family history (range 0–33%) than is usually seen in idiopathic RLS (>50%) [3,9,11,13–15,20]. The age of RLS onset in PD tends to be relatively high, and goes from 41.7±3.3 years to 62.4±7.3 years [3,9,11,13,14,16,20]. This is consistent with the onset of other ‘secondary’ RLS symptoms (typically after the age of 45). There has been no case–control study in which the severity of RLS in PD has been compared with that in controls by means of the IRLSSG severity scale. Only one study mentioned the complete distribution, showing that 59% of the patients reported no or mild RLS in the week before the severity scale was filled out, 38% reported moderate to very severe RLS, and 4% very severe RLS [20]. On average, the severity of RLS in PD seems to be milder than in the general population. This finding can be biased by the fact that, in most studies, the PD patients used dopaminergic medication, and therefore RLS severity could be suppressed. Frequency of pain sensation, but not pain intensity, was significantly higher for PD patients with RLS than for PD patients without RLS, just like in the case of idiopathic RLS versus healthy controls [26]. On the basis of this result, the hypothesis was coined that the presence of pain in PD patients can be exacerbated by RLS or, stated otherwise, that the presence of RLS may exacerbate pain in patients who have PD. 4. RLS mimics in PD A variety of conditions can ‘mimic’ RLS by satisfying the four diagnostic criteria. Conditions which can do this, and are not PD specific, include cramps, (poly)neuropathy, positional discomfort, and local leg pathology. Besides PD-non-specific, also PD-specific RLS mimics should be differentiated from ‘real’ RLS in PD patients. This is important, not only to avoid overestimation of RLS frequency in prevalence studies, but also and most importantly, from a clinical viewpoint, to avoid misdiagnosis, and to optimize therapeutic discussions and interventions. For example, some nonmotor sensory systems, such as motor- and non-motor sensory fluctuations in PD, could mimic RLS. Fluctuating state sensory symptoms, like numbness or paraesthesias without objective sensory loss, are described to be experienced more often in the legs and arms, and rarely in the face or neck [26]. This is a distribution which follows the distribution of RLS symptoms, and, thus, can mimic RLS. Akathisia in PD is an important mimic of idiopathic RLS, but even more so for PD-related RLS. The term akathisia is used to describe a condition in which the patient has an inner restlessness with an urge to move. Although exposure to neuroleptics is the
S8
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
most common underlying cause for akathisia, this condition is also often reported by PD patients (with and without dopaminergic treatment), with a prevalence from 25% to 45% [27–29]. The motor restlessness in akathisia differs from RLS, by the fact that it is usually a generalized whole-body sensation, without sensory discomfort, without a strong circadian rhythm, and without relief by movement. Although this is certainly helpful for a proper differentiation, it is not a hundred percent guarantee. In the case of severe RLS the circadian rhythmicity, aggravation by rest and suppression by movement fades away. Moreover, also some overlap between akathisia symptoms and RLS symptoms in PD is mentioned in the literature. For example, Comella [30] mentions that 40% of their patients had akathisia only at certain times of the day, suggesting an overlap with RLS. Or, as mentioned already earlier in this text, Gjerstadt and colleagues [8] wonder if, and to what extent, RLS and akathisia are really two different and separated conditions, or represent overlapping features within the same spectrum of motor restlessness in PD. The finding of Peralta et al. [16] that 61% of all RLS positives mentioned that their RLS symptoms were associated with a wearing off, also strengthens the possibility of RLS mimics in fluctuating patients with PD. Therefore true RLS symptoms, and RLS-like symptoms resulting from levodopa, deserve further investigation, as stated earlier by several authors. Whether or not certain leg motor restlessness symptoms that do not fulfill all RLS criteria can later develop into full-blown RLS in PD patients is unclear, and is also a subject matter for further investigation [7]. 5. Impact of RLS in PD patients Sleep and wake problems are a very common among PD patients, as high as 70%. Many factors are thought to play a qualifying role in the sleep and wake complaints of these patients. Several studies have focussed on the attributive role of RLS in the sleep and wake complaints of PD patients [8,9,11,13,15,18,20,23]. Nomura et al. [23] found a significantly higher (meaning lower sleep quality) Pittsburgh Sleep Questionnaire Index (PSQI) in PD patients with RLS than in PD patients without RLS. In another study on PD patients, RLS had no influence on the PSQI [13]. Excessive daytime sleepiness, as measured with the Epworth Sleepiness Scale (ESS), was also not found to be higher (meaning more sleepiness) among PD patients with RLS than among PD patients without RLS [9,11,15]. If sleep scales are used that are specially designed and validated for PD populations, then some studies find significantly more sleep or wake complaints in PD patients with RLS than in PD patients without RLS [9,20]. In the cohort of Gomez-Esteban et al. [9], the Parkinson’s Disease Sleepiness Scale (PDSS) score was significantly lower (= worse) for PD patients with RLS than for PD patients without RLS, suggesting more sleep disturbances in the former group, as indicated by the responses on the following sub-items: ‘nocturnal restlessness of legs or arms at night or in the evening, causing sleep disruption’, ‘numbness or tingling of arms or legs with an awaken effect’, ‘fidgeting in bed’, and ‘painful muscle cramps in arms or legs whilst sleeping at night’. However, because these four sub-items describe elements that can be part, not only of ‘true’ RLS, but also of RLS mimics, the authors suggest that it is thinkable that some of their patients with akathisia or with nocturnal motor problems other than RLS, may have been wrongly diagnosed as RLS. In the cohorts of Gjerstad et al. [8] and of Bhalsing et al. [4] similar results were found. In the Norwegian PD population study [18], it was found that three variables were significantly associated with PDSS score, namely poor mental health, fatigue and RLS. The significant relationship between PDSS and RLS persisted, even if the scores on the 4th item (restlessness of legs or arms at night or in the evening causing sleep
disruptions) was removed from the analysis. In this study, ‘true’ RLS seemed to have an negative effect on the sleep quality. The impact on quality of life (QoL) of PD is tremendous. The question of the attributive role of RLS in this impact, however, has only been addressed explicitly in one study. Gomez-Estaban and colleagues [9] compared the scores on the Parkinson’s Disease Questionnaire (PDQ-39) of PD patients with RLS with those of PD patients without RLS, and found no difference. Some other studies have looked at the effect of RLS on mental health in PD patients, as an isolated parameter [8,11,16,20]. In one study it was found that not the presence of RLS as such, but rather the severity of RLS was correlated with depressive symptoms [20]. In another study, depression was not correlated with RLS, but was significantly more prevalent in PD patients with RLS (40%) than in PD patients without RLS (10.3%) [16]. In the study of Krishnan et al. [11], pain, as a possible factor for QoL, was found to be significantly higher in PD patients with RLS than in PD patients without RLS. The same was found in a study of Rana et al. [26], but not in another one by Peralta et al. [16]. 6. Summary and conclusion RLS and PD can co-exist, whether or not they share a common pathophysiology. If, and to what extent, RLS and PD share the same pathophysiology, is still under debate. There are arguments for and arguments against. As the observed disease profiles of PD patients with RLS were not really consistent over the various studies, no clear risk factors for RLS in PD can be pointed out unequivocally. One might be inclined to argue that dopaminergic medication is a risk factor, but more research is needed before one can give a clear answer to this hypothesis. The RLS profile in PD, however, seems to be different from that in the general population. On average, RLS in PD seems to be less severe than in the general population. One study suggests that the sensory entity ‘pain’ is more prevalent in PD-associated RLS than in non-PD-associated RLS. We also see that RLS in PD patients starts relatively late, the majority after PD onset, and that most of these patients have no positive family history. The gender distribution of RLS in most PD cohorts does not follow the classical gender distribution of RLS in the general population. Impact on sleep, and psychological aspects of RLS in PD patients, are found in some studies, but seem to be more correlated with the severity of RLS than with RLS as such. Although sensory symptoms and fluctuating symptoms in relation to dopaminergic medication may mimic RLS, overlap with real RLS seems to go with these symptoms as well. Awareness of the possible impact of RLS, whether or not as a co-morbid disorder, or as part of the profile of PD, could optimize the therapy strategy in PD. Until now, no studies about what is the best strategy to treat RLS in PD have been done. The current choice of individualized treatment strategies, therefore, can only be based on our knowledge from RLS studies in the idiopathic domain. A point of particular interest in the future, therefore, should be the construction and validation of RLS criteria which are specifically designed for the PD population, so as to exclude as much as possible the confusion with RLS mimics which are typically seen in PD. With this in hand, better PD profiling, RLS profiling, neuropathological studies, quality of life studies in the full range of quality, and treatment studies will become possible in the future. Conflict of interests The authors have no conflicts of interest to declare. References [1] Garcia-Borreguero D, Odin P, Serrano C. Restless legs syndrome and PD: a review of the evidence for a possible association. Neurology 2003 Sep 23;61:S49–55.
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
[2] Moller JC, Unger M, Stiasny-Kolster K, Oertel WH. Restless legs syndrome (RLS) and Parkinson’s disease (PD) – related disorders or different entities? J Neurol Sci 2010 Feb 15;289:135–7. [3] Angelini M, Negrotti A, Marchesi E, Bonavina G, Calzetti S. A study of the prevalence of restless legs syndrome in previously untreated Parkinson’s disease patients: absence of co-morbid association. J Neurol Sci 2011 Nov 15;310:286–8. [4] Bhalsing K, Suresh K, Muthane UB, Pal PK. Prevalence and profile of restless legs syndrome in Parkinson’s disease and other neurodegenerative disorders: a case–control study. Parkinsonism Relat Disord 2013 Apr;19:426–30. [5] Braga-Neto P, da Silva-Junior FP, Sueli MF, de Bruin PF, de Bruin VM. Snoring and excessive daytime sleepiness in Parkinson’s disease. J Neurol Sci 2004 Jan 15;217:41–5. [6] Calzetti S, Negrotti A, Bonavina G, Angelini M, Marchesi E. Restless legs syndrome and Parkinson’s disease. Mov Disord 2010 Jul 15;25:1314–5. [7] Gama RL, Tavora DG, Bomfim RC, Silva CE, de Bruin VM, de Bruin PF. Sleep disturbances and brain MRI morphometry in Parkinson’s disease, multiple system atrophy and progressive supranuclear palsy – a comparative study. Parkinsonism Relat Disord 2010 May;16:275–9. [8] Gjerstad MD, Tysnes OB, Larsen JP. Increased risk of leg motor restlessness but not RLS in early Parkinson disease. Neurology 2011 Nov 29;77:1941–6. [9] Gomez-Esteban JC, Zarranz JJ, Tijero B, Velasco F, Barcena J, Rouco I, et al. Restless legs syndrome in Parkinson’s disease. Mov Disord 2007 Oct 15;22:1912–6. [10] Guerreiro TM, Nishikawa DR, Ferreira LC, Melo HA, Prado RC. Restless legs syndrome in Parkinson’s disease: clinical characteristics and biochemical correlations. Arq Neuropsiquiatr 2010 Dec;68:869–72. [11] Krishnan PR, Bhatia M, Behari M. Restless legs syndrome in Parkinson’s disease: a case-controlled study. Mov Disord 2003 Feb;18:181–5. [12] Lee JE, Shin HW, Kim KS, Sohn YH. Factors contributing to the development of restless legs syndrome in patients with Parkinson disease. Mov Disord 2009 Mar 15;24:579–82. [13] Loo HV, Tan EK. Case–control study of restless legs syndrome and quality of sleep in Parkinson’s disease. J Neurol Sci 2008 Mar 15;266:145–9. [14] Nomura T, Inoue Y, Miyake M, Yasui K, Nakashima K. Prevalence and clinical characteristics of restless legs syndrome in Japanese patients with Parkinson’s disease. Mov Disord 2006 Mar;21:380–4. [15] Ondo WG, Vuong KD, Jankovic J. Exploring the relationship between Parkinson disease and restless legs syndrome. Arch Neurol 2002 Mar;59:421–4.
S9
[16] Peralta CM, Frauscher B, Seppi K, Wolf E, Wenning GK, Hogl B, et al. Restless legs syndrome in Parkinson’s disease. Mov Disord 2009 Oct 30;24:2076–80. [17] Suzuki K, Miyamoto M, Miyamoto T, Tatsumoto M, Watanabe Y, Suzuki S, et al. Nocturnal disturbances and restlessness in Parkinson’s disease: using the Japanese version of the Parkinson’s disease sleep scale-2. J Neurol Sci 2012 Jul 15;318:76–81. [18] Svensson E, Beiske AG, Loge JH, Beiske KK, Sivertsen B. Sleep problems in Parkinson’s disease: a community-based study in Norway. BMC Neurol 2012;12:71. [19] Tan EK, Lum SY, Wong MC. Restless legs syndrome in Parkinson’s disease. J Neurol Sci 2002 Apr 15;196:33–6. [20] Verbaan D, van Rooden SM, van Hilten JJ, Rijsman RM. Prevalence and clinical profile of restless legs syndrome in Parkinson’s disease. Mov Disord 2010 Oct 15;25:2142–7. [21] Jagota P, Asawavichienjinda T, Bhidayasiri R. The low prevalence of primary restless legs syndrome in Thai Parkinson’s disease patients at Chulalongkorn University Hospital. J Med Assoc Thai 2012 Feb;95:175–80. [22] Ohayon MM, O’Hara R, Vitiello MV. Epidemiology of restless legs syndrome: a synthesis of the literature. Sleep Med Rev 2012 Aug;16:283–95. [23] Nomura T, Inoue Y, Nakashima K. Clinical characteristics of restless legs syndrome in patients with Parkinson’s disease. J Neurol Sci 2006 Dec 1;250:39–44. [24] Chahine LM, Ahmed A, Sun Z. Effects of STN DBS for Parkinson’s disease on restless legs syndrome and other sleep-related measures. Parkinsonism Relat Disord 2011 Mar;17:208–11. [25] Kedia S, Moro E, Tagliati M, Lang AE, Kumar R. Emergence of restless legs syndrome during subthalamic stimulation for Parkinson disease. Neurology 2004 Dec 28;63:2410–2. [26] Rana AQ, Siddiqui I, Mosabbir A, Athar A, Syed O, Jesudasan M, et al. Association of pain, Parkinson’s disease, and restless legs syndrome. J Neurol Sci 2013 Apr 15;327:32–4. [27] Comella CL, Goetz CG. Akathisia in Parkinson’s disease. Mov Disord 1994 Sep;9:545–9. [28] Lang AE, Johnson K. Akathisia in idiopathic Parkinson’s disease. Neurology 1987 Mar;37:477–81. [29] Poewe W, Hogl B. Akathisia, restless legs and periodic limb movements in sleep in Parkinson’s disease. Neurology 2004 Oct 26;63:S12–6. [30] Comella CL. Sleep disturbances in Parkinson’s disease. Curr Neurol Neurosci Rep 2003 Mar;3:173–80.
Contents lists available at SciVerse ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Restless legs syndrome in Parkinson’s disease Roselyne M. Rijsman *, Louise F. Schoolderman, Rob S. Rundervoort, Maartje Louter Center of Sleep and Wake Disorders and Department of Neurology, Medical Center Haaglanden, The Hague, The Netherlands
article info
summary
Keywords: Restless legs syndrome RLS Parkinson’s disease PD prevalence mimics sleep profile impact quality of life
The Restless legs syndrome (RLS) and Parkinson’s disease (PD) are two disorders that can co-exist, whether or not they share a common pathophysiology. If, and to what extent RLS and PD share the same pathophysiology, is still under debate. Sleep disturbances are prevalent in PD, and as PD progresses, nocturnal disturbances become even more evident, in association not only with motor symptoms but also with non-motor symptoms. Alertness to, and recognition of, RLS in PD patients with sleep disorders could improve customized treatment and quality of life of these patients. In this article the prevalence of RLS in PD, the clinical profile of RLS in PD, the PD profile of patients with RLS, RLS mimics specifically related to PD and impact of RLS in PD will be reviewed.
1. Introduction Restless legs syndrome (RLS) is a common neurological disorder, with an estimated prevalence of 4–10% in the general population. Although the precise pathophysiology remains unknown, it is assumed that dopaminergic mechanisms play a central role. In the past few years, several studies have been done which suggest an association between RLS and Parkinson’s disease (PD). Both RLS and PD respond to dopaminergic treatment, making the assumption of a common pathophysiology, and therefore an association between the two disorders, fairly plausible [1,2]. The RLS symptoms seem to result from an abnormal sensorimotor integration, due to dysinhibition at the spinal level which, in turn, results from reduced inhibitory mechanisms at the supraspinal level. The descending supraspinal diencephalospinal dopaminergic neurons, originating in the hypothalamus (A11), are proposed to play an important role in the pathophysiology of RLS. It is postulated that, in the course of the PD disease, these neurons degenerate, along with nigrostriatal neurons, leading to the development of RLS, and, therefore, to a higher prevalence of RLS among PD patients. But shared pathophysiology is still under debate [1,2]. Unlike in the case of PD, no neuronal degeneration or Lewy body deposition has ever been recognized in RLS post-mortem studies. Furthermore, until now, no shared disease loci have been identified. Except for the provocation of both RLS and PD by neuroleptics, and possibly lower ferritin levels in PD patients with RLS, the secondary forms of PD and RLS do not form an argument for a common pathophysiology [2]. * Corresponding author. Medical Center Haaglanden, Center for Sleep and Wake Disorders, Lijnbaan 32, 2501 CK The Hague, The Netherlands 1353-8020/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
© 2013 Elsevier Ltd. All rights reserved.
Even though the relationship between RLS and PD is still controversial, RLS has revealed itself to be co-existing with PD. Sleep disturbances are prevalent in PD, and as PD progresses, nocturnal disturbances become even more evident, in association not only with motor symptoms but also with non-motor symptoms. Thus, alertness to, and recognition of, RLS in PD patients with sleep disorders could improve customized treatment and quality of life of these patients. That is why, in this article, we will specifically look at studies on the prevalence and on the clinical profile of RLS in PD, as well as on the impact that RLS has on PD patients. 2. Prevalence of RLS in PD Good comparisons of the different estimates of RLS prevalence in PD studies are rather difficult due to variations in methodology, and so, definite generalisations are hard to make. In total, 18 cross-sectional studies on the prevalence of RLS in PD, published in English, were found [3–20]. In all these studies, the international IRLSSG essential criteria for the diagnosis of RLS were used, but some studies used the criteria of 1995, whereas other (more recent) studies used the revised IRLSG diagnostic criteria of 2003. The methodology, however, for the assessment of these criteria in the subjects was not identical in all studies. In addition, the background of the interviewers varied from study to study, which might influence, of course, the focus, the wording and the interpretation of the RLS histories that were collected. In most studies, the definite diagnosis of RLS was based only on the constellation of the four minimal IRLSSG criteria, but in some other studies, thresholds on severity, frequency, and on the impact on quality of life were included. From a purely epidemiological and pathophysiological point of view, the former procedure may be
S6
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
regarded defensible and thoughtful, but from a clinical point of view, inclusion of the threshold criteria could obviously help to better define PD patients who actually need treatment for RLS. The appearance of secondary RLS is usually related with a late onset of RLS (above 45 years of age). As a result, there may be more secondary forms of RLS in PD cohorts, besides purely PDrelated RLS, simply because PD is also a disease that becomes more frequent with age. Some studies, therefore, deliberately excluded these secondary forms of RLS from their data [3,16], to avoid confusion with this age-related variable in their assessment of the true relation between RLS and PD. From a clinical point of view, one may question whether such an exclusion of these ‘secondary’ forms of RLS from the investigation of the co-morbidity of any sort of RLS in PD is appropriate, because, in doing this, we can obviously not assess their impact and, by implication, also not their possible implications for therapeutic intervention. There was also variation in the size and composition of the PD cohorts, and in the way they were treated. And last but not least, a few studies excluded PD-specific RLS mimics, such as motor symptoms, akathisia, non-motor symptoms, as pain, and other sensory symptoms, motor fluctuation, such as dystonia, and nonmotor fluctuation symptoms, even though the possible overlap of these symptoms with ‘true’ RLS has been discussed explicitly by some authors. The prevalence of RLS in the 18 PD cohorts varied from 0% to 50%. Two of the studies were performed in Latin America; the one from Brazil showed a very high PD-related RLS prevalence of 50%, whereas the other one found 18.8% [5,10]. Eight of the 18 studies were performed in Asian countries, with PD-related RLS prevalence from 0.98% to 16% [4,11–14,17,19,21]. The PD-related RLS prevalence in Western countries was generally higher than those in the Asian countries, namely from 5.5% to 27% in Europe [3,6,8,9,16,18,20], and 20.8% in the USA [15]. Also, the prevalence of RLS in the ‘general population’ is typically lower in Asian countries than in European and North American ones, namely from 0.9% to 12.1% in Asia, against from 3.9% to 18.8% in Europe and the USA [22]. In addition, the correlation with age is clearly different in these two continents: in Asia it does not seem to go up with age, whereas in Europe and the USA, the RLS prevalence in the general population is found to double every 20 years, with a peak at the age of 65 [22]. Since the mean age of all the PD cohorts in which RLS prevalence was examined was at least 59, we may say that the factor age was probably a less confounding factor for the assessment of RLS prevalence in the Asian PD cohorts than in the non-Asian ones. To show a real association between RLS and PD, one should find a significant increase of RLS prevalence in the PD cohorts relative to that in the general population. Seven of the 18 studies [3,6,8,13,17,19,20] did not find such a significant increase and did not support the hypothesis of a similar pathophysiology between RLS and PD. The actual percentages of RLS in these negative studies were between 0% and 5.5% in the Asian countries, and between 5.5% and 12.7% in the Western countries. The majority of all the RLS-prevalence studies in PD were done on patient groups with relatively advanced forms of disease, treated which dopaminergic medication. So there was a serious risk of including RLS confounders and secondary forms of RLS, stemming from other causes, in the assessment. Lee and colleagues mention a prevalence of RLS, not different from that in the general population, in only one subgroup of patients who were not given dopamine medication, and who were still in an early stage of PD [12]. After multivariate analysis of the total PD cohort, only duration of dopaminergic treatment appeared to be a variable that correlated significantly with the presence of RLS. Their conclusion was that RLS in PD is associated with long-term dopaminergic treatment, rather than with PD itself. In 2011, two RLS-prevalence studies were
performed in a cohort with only de novo and dopaminergic naive PD patients. Angelini and colleagues [3] found that RLS prevalence in their dopamine-free PD cohort (5.5%) was not significantly different from that in their control group (2.3%), and concluded that RLS is not really related to PD pathophysiology. The question whether an increase in risk of co-morbidity of RLS with the progression of PD is a result of dopaminergic neuron degeneration, or of dopaminergic treatment, is not answered by this study. Gjerstadt et al. [8] performed a case–control study in an early-phase and dopamine un-medicated PD cohort of 200 patients. Patients who had an urge to move their legs according to the Johns Hopkins diagnostic interview for RLS were categorized as ‘true’ RLS if they fulfilled the four essential IRLSSG criteria, and as ‘Leg Movement Restlessness’ (LMR) if they did not fulfil all four IRLSSG criteria. They found similar RLS prevalence in PD patients and controls. But they demonstrated that LMR grows to almost to a 3-fold higher risk in early PD, as weighted against controls. The authors speculated if RLS and akathisia (as one of the possible interpretations of LMR) may represent overlapping features within the same spectrum of motor restlessness in PD. Suzuki et al. [17] found significantly higher nocturnal restlessness (and not RLS), as measured by the Parkinson disease sleepiness scale (PDSS) sub-items 4 and 5, in PD compared to controls. This nocturnal restlessness was associated with the PDSS total score, but not with disease severity and treatment duration, or total levo-dopa equivalent (LDE) dose. The nocturnal restlessness was therefore interpreted as being the result of an endogenous dopamine deficit during night-time, rather than as a motor complication (i.e., wearing off phenomenon) due to dopaminergic treatment which can mimic RLS. By contrast, however, ten studies [4,5,9–12,14–17], did find higher RLS frequencies in PD, and, thus, support the hypothesis that RLS is part of the symptomatic profile of PD, and not just an incidental cooccurrence of RLS in PD. RLS prevalence in these ten studies varied from 7.9% to 16% in the Asian countries, against from 20.8% to 27% in the European countries, and from 19.9% to 50% in the SouthAmerican countries. 3. Clinical profile of RLS in PD 3.1. Comparison of PD with and without RLS: risk factor in PD? All but one [5] of the prevalence studies considered in this article have to some extent compared the PD profile of patients with RLS with that of patients without RLS. Several, although not necessarily consistent, significant differences between the two groups were found. Again, since the studies differ in methodology, as well as in composition of PD cohorts, it is not justified to draw definite conclusions. The following is an overview of the major results. Onset of RLS in relation to onset of PD was evaluated in several studies [11,13–16,20]. In all of them, a clear majority of patients (76–100%) showed an onset of RLS only after, or together with the onset of PD. Ondo et al. [15] found that, for PD patients with a positive RLS family history, the RLS onset preceded the PD onset more often (52%) than for PD patients without a positive RLS family history (29%). Given such a result, one might be inclined to say that RLS without a positive family history is more consistent with the ‘secondary’ forms of RLS in PD, but these results could not be replicated in later studies [20]. Though lateralization of RLS is found to be rather high in the PD population (range 35–70%), no correlations were found between side of PD onset and dominant side of the RLS symptoms [4,11,14] which immediately counters the suggestion, of course, that RLS in PD is related to the degeneration of the dopaminergic neurons in the substantia nigra. The mean severity of RLS, expressed in terms of the international RLS severity scale (IRLSS), varied from 11.9±6.3 to 21.3±6.3,
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
or from mild to moderately severe RLS [9,10,13,14,20]. Little or no information was provided about the frequency of RLS in PD patients. Only one study mentioned that most of the patients (24%) experienced RLS 2–4 times a month [20]. In some studies, RLS was associated with younger age at time of investigation of the PD patients [14,16], whereas in other studies it was associated with older age, as compared with PD patients without RLS [11]. The possibility of an association between RLS and severity of PD was addressed in many of the studies with the Hoehn and Yahr scale [3,10–12,15,18,20,23] and the UPDRS [4,9–11,18]. Only in two of them, however, a significant, but opposite association was found [12,18]. The possible association between PD duration and RLS was specifically addressed in only one study [12], which reported a positive association: the longer the PD, the more RLS. After multivariate analysis, however, in which several variables were included in the analysis, only duration of treatment was found to be a significant predictor of the presence or absence of RLS (OR = 1.199, 95% CL 1.014–1.419, P = 0.0334). The authors stated that the more frequent RLS in PD patients with longer dopaminergic treatment could be the result of the same mechanisms as those which induce an augmentation of RLS in prolonged levodopa exposure. Although these augmentation mechanisms are not fully clarified yet, Paulus and Trenkwalder have proposed that the augmentation could be a result of overstimulation by the dopaminergic medication of the D 1 dopamine receptors, in comparison with the D 2 receptors, in the spinal cord. The subsequent finding, in another study, that the RLS severity was decreased after subthalamic stimulation concomitant with dopaminergic medication dose reduction, endorses the hypothesis that RLS in PD can be a result of dopamine receptor over-stimulation by dopaminergic medication [24]. The idea that RLS symptoms are correlated with dopamine treatment, rather than with PD pathology itself, is also supported by some other observations, such as: (1) more PD patients without RLS, in contrast to PD patients with RLS, used no levodopa at all (resp. 6.8% and 0%) [9], (2) the same prevalence of RLS as in the general population, in a small subgroup of dopamine-medication free PD patients, compared to an increased prevalence of RLS in the levodopa-treated PD patients [12], (3) equal levels of RLS prevalence in various dopamine-free PD cohorts [3,8]. Besides dopamine treatment duration, also level of dopamine dose might be a factor that leads to overstimulation of the dopaminergic neurons. Although the dopamine therapy dose of PD patients with RLS was compared with that of PD patients without RLS in nine studies [9,11–17,20], only one showed a significant, though not positive but negative, correlation between dopamine dose and RLS. These authors suggest that higher dopaminergic medication could suppress the RLS symptoms. Also Verbaan et al. [20] discuss the possibility that the relatively high mean dopaminergic dose in their cohort could have resulted in an underestimation of RLS patients. This hypothesis was further supported by the observation of emerging RLS in 11 of 195 PD patients after subthalamic stimulation with a concurrent 75% reduction of the mean dopaminergic medication. This can be explained by an unmasking effect of RLS symptoms, which otherwise would be suppressed by a higher dose of dopaminergic medication [25]. Also iron deficiency has been associated with RLS. In two studies [11,15], lower levels of ferritin were found in PD RLSpositive patients than in PD RLS-negative patients. Ondo and colleagues, therefore, speculate that low ferritin levels could, besides reducing the dopaminergic system function, also be a risk factor for RLS symptoms in PD. In two other studies [10,14],
S7
no such statistically significant differences in level of ferritin were found, suggesting that not iron deficiency but dysfunction of the central dopaminergic system itself, due to PD, might be the underlying pathophysiology of RLS in PD. 3.2. Comparison of the RLS profile in PD with that of idiopathic RLS It is well known that the prevalence of RLS in the general population is higher among women than among men. Consistent with this pattern, also several studies on PD patients found a higher RLS prevalence among women than among men (range 60–75%) [6,9,12–14,20]. In only two of these studies, however, the proportion of women was significantly higher in the PD RLS-positive group than in the PD RLS-negative group [9,20]. In other studies on PD patients, however, this typical pattern of gender distribution was not found [3,4,8,10,11,16]. All together, we cannot say that there is a really consistent pattern of gender distribution among PD patients with RLS. Parkinson-related RLS has a remarkably lower rate of RLS family history (range 0–33%) than is usually seen in idiopathic RLS (>50%) [3,9,11,13–15,20]. The age of RLS onset in PD tends to be relatively high, and goes from 41.7±3.3 years to 62.4±7.3 years [3,9,11,13,14,16,20]. This is consistent with the onset of other ‘secondary’ RLS symptoms (typically after the age of 45). There has been no case–control study in which the severity of RLS in PD has been compared with that in controls by means of the IRLSSG severity scale. Only one study mentioned the complete distribution, showing that 59% of the patients reported no or mild RLS in the week before the severity scale was filled out, 38% reported moderate to very severe RLS, and 4% very severe RLS [20]. On average, the severity of RLS in PD seems to be milder than in the general population. This finding can be biased by the fact that, in most studies, the PD patients used dopaminergic medication, and therefore RLS severity could be suppressed. Frequency of pain sensation, but not pain intensity, was significantly higher for PD patients with RLS than for PD patients without RLS, just like in the case of idiopathic RLS versus healthy controls [26]. On the basis of this result, the hypothesis was coined that the presence of pain in PD patients can be exacerbated by RLS or, stated otherwise, that the presence of RLS may exacerbate pain in patients who have PD. 4. RLS mimics in PD A variety of conditions can ‘mimic’ RLS by satisfying the four diagnostic criteria. Conditions which can do this, and are not PD specific, include cramps, (poly)neuropathy, positional discomfort, and local leg pathology. Besides PD-non-specific, also PD-specific RLS mimics should be differentiated from ‘real’ RLS in PD patients. This is important, not only to avoid overestimation of RLS frequency in prevalence studies, but also and most importantly, from a clinical viewpoint, to avoid misdiagnosis, and to optimize therapeutic discussions and interventions. For example, some nonmotor sensory systems, such as motor- and non-motor sensory fluctuations in PD, could mimic RLS. Fluctuating state sensory symptoms, like numbness or paraesthesias without objective sensory loss, are described to be experienced more often in the legs and arms, and rarely in the face or neck [26]. This is a distribution which follows the distribution of RLS symptoms, and, thus, can mimic RLS. Akathisia in PD is an important mimic of idiopathic RLS, but even more so for PD-related RLS. The term akathisia is used to describe a condition in which the patient has an inner restlessness with an urge to move. Although exposure to neuroleptics is the
S8
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
most common underlying cause for akathisia, this condition is also often reported by PD patients (with and without dopaminergic treatment), with a prevalence from 25% to 45% [27–29]. The motor restlessness in akathisia differs from RLS, by the fact that it is usually a generalized whole-body sensation, without sensory discomfort, without a strong circadian rhythm, and without relief by movement. Although this is certainly helpful for a proper differentiation, it is not a hundred percent guarantee. In the case of severe RLS the circadian rhythmicity, aggravation by rest and suppression by movement fades away. Moreover, also some overlap between akathisia symptoms and RLS symptoms in PD is mentioned in the literature. For example, Comella [30] mentions that 40% of their patients had akathisia only at certain times of the day, suggesting an overlap with RLS. Or, as mentioned already earlier in this text, Gjerstadt and colleagues [8] wonder if, and to what extent, RLS and akathisia are really two different and separated conditions, or represent overlapping features within the same spectrum of motor restlessness in PD. The finding of Peralta et al. [16] that 61% of all RLS positives mentioned that their RLS symptoms were associated with a wearing off, also strengthens the possibility of RLS mimics in fluctuating patients with PD. Therefore true RLS symptoms, and RLS-like symptoms resulting from levodopa, deserve further investigation, as stated earlier by several authors. Whether or not certain leg motor restlessness symptoms that do not fulfill all RLS criteria can later develop into full-blown RLS in PD patients is unclear, and is also a subject matter for further investigation [7]. 5. Impact of RLS in PD patients Sleep and wake problems are a very common among PD patients, as high as 70%. Many factors are thought to play a qualifying role in the sleep and wake complaints of these patients. Several studies have focussed on the attributive role of RLS in the sleep and wake complaints of PD patients [8,9,11,13,15,18,20,23]. Nomura et al. [23] found a significantly higher (meaning lower sleep quality) Pittsburgh Sleep Questionnaire Index (PSQI) in PD patients with RLS than in PD patients without RLS. In another study on PD patients, RLS had no influence on the PSQI [13]. Excessive daytime sleepiness, as measured with the Epworth Sleepiness Scale (ESS), was also not found to be higher (meaning more sleepiness) among PD patients with RLS than among PD patients without RLS [9,11,15]. If sleep scales are used that are specially designed and validated for PD populations, then some studies find significantly more sleep or wake complaints in PD patients with RLS than in PD patients without RLS [9,20]. In the cohort of Gomez-Esteban et al. [9], the Parkinson’s Disease Sleepiness Scale (PDSS) score was significantly lower (= worse) for PD patients with RLS than for PD patients without RLS, suggesting more sleep disturbances in the former group, as indicated by the responses on the following sub-items: ‘nocturnal restlessness of legs or arms at night or in the evening, causing sleep disruption’, ‘numbness or tingling of arms or legs with an awaken effect’, ‘fidgeting in bed’, and ‘painful muscle cramps in arms or legs whilst sleeping at night’. However, because these four sub-items describe elements that can be part, not only of ‘true’ RLS, but also of RLS mimics, the authors suggest that it is thinkable that some of their patients with akathisia or with nocturnal motor problems other than RLS, may have been wrongly diagnosed as RLS. In the cohorts of Gjerstad et al. [8] and of Bhalsing et al. [4] similar results were found. In the Norwegian PD population study [18], it was found that three variables were significantly associated with PDSS score, namely poor mental health, fatigue and RLS. The significant relationship between PDSS and RLS persisted, even if the scores on the 4th item (restlessness of legs or arms at night or in the evening causing sleep
disruptions) was removed from the analysis. In this study, ‘true’ RLS seemed to have an negative effect on the sleep quality. The impact on quality of life (QoL) of PD is tremendous. The question of the attributive role of RLS in this impact, however, has only been addressed explicitly in one study. Gomez-Estaban and colleagues [9] compared the scores on the Parkinson’s Disease Questionnaire (PDQ-39) of PD patients with RLS with those of PD patients without RLS, and found no difference. Some other studies have looked at the effect of RLS on mental health in PD patients, as an isolated parameter [8,11,16,20]. In one study it was found that not the presence of RLS as such, but rather the severity of RLS was correlated with depressive symptoms [20]. In another study, depression was not correlated with RLS, but was significantly more prevalent in PD patients with RLS (40%) than in PD patients without RLS (10.3%) [16]. In the study of Krishnan et al. [11], pain, as a possible factor for QoL, was found to be significantly higher in PD patients with RLS than in PD patients without RLS. The same was found in a study of Rana et al. [26], but not in another one by Peralta et al. [16]. 6. Summary and conclusion RLS and PD can co-exist, whether or not they share a common pathophysiology. If, and to what extent, RLS and PD share the same pathophysiology, is still under debate. There are arguments for and arguments against. As the observed disease profiles of PD patients with RLS were not really consistent over the various studies, no clear risk factors for RLS in PD can be pointed out unequivocally. One might be inclined to argue that dopaminergic medication is a risk factor, but more research is needed before one can give a clear answer to this hypothesis. The RLS profile in PD, however, seems to be different from that in the general population. On average, RLS in PD seems to be less severe than in the general population. One study suggests that the sensory entity ‘pain’ is more prevalent in PD-associated RLS than in non-PD-associated RLS. We also see that RLS in PD patients starts relatively late, the majority after PD onset, and that most of these patients have no positive family history. The gender distribution of RLS in most PD cohorts does not follow the classical gender distribution of RLS in the general population. Impact on sleep, and psychological aspects of RLS in PD patients, are found in some studies, but seem to be more correlated with the severity of RLS than with RLS as such. Although sensory symptoms and fluctuating symptoms in relation to dopaminergic medication may mimic RLS, overlap with real RLS seems to go with these symptoms as well. Awareness of the possible impact of RLS, whether or not as a co-morbid disorder, or as part of the profile of PD, could optimize the therapy strategy in PD. Until now, no studies about what is the best strategy to treat RLS in PD have been done. The current choice of individualized treatment strategies, therefore, can only be based on our knowledge from RLS studies in the idiopathic domain. A point of particular interest in the future, therefore, should be the construction and validation of RLS criteria which are specifically designed for the PD population, so as to exclude as much as possible the confusion with RLS mimics which are typically seen in PD. With this in hand, better PD profiling, RLS profiling, neuropathological studies, quality of life studies in the full range of quality, and treatment studies will become possible in the future. Conflict of interests The authors have no conflicts of interest to declare. References [1] Garcia-Borreguero D, Odin P, Serrano C. Restless legs syndrome and PD: a review of the evidence for a possible association. Neurology 2003 Sep 23;61:S49–55.
R.M. Rijsman et al. / Parkinsonism and Related Disorders 20S1 (2014) S5–S9
[2] Moller JC, Unger M, Stiasny-Kolster K, Oertel WH. Restless legs syndrome (RLS) and Parkinson’s disease (PD) – related disorders or different entities? J Neurol Sci 2010 Feb 15;289:135–7. [3] Angelini M, Negrotti A, Marchesi E, Bonavina G, Calzetti S. A study of the prevalence of restless legs syndrome in previously untreated Parkinson’s disease patients: absence of co-morbid association. J Neurol Sci 2011 Nov 15;310:286–8. [4] Bhalsing K, Suresh K, Muthane UB, Pal PK. Prevalence and profile of restless legs syndrome in Parkinson’s disease and other neurodegenerative disorders: a case–control study. Parkinsonism Relat Disord 2013 Apr;19:426–30. [5] Braga-Neto P, da Silva-Junior FP, Sueli MF, de Bruin PF, de Bruin VM. Snoring and excessive daytime sleepiness in Parkinson’s disease. J Neurol Sci 2004 Jan 15;217:41–5. [6] Calzetti S, Negrotti A, Bonavina G, Angelini M, Marchesi E. Restless legs syndrome and Parkinson’s disease. Mov Disord 2010 Jul 15;25:1314–5. [7] Gama RL, Tavora DG, Bomfim RC, Silva CE, de Bruin VM, de Bruin PF. Sleep disturbances and brain MRI morphometry in Parkinson’s disease, multiple system atrophy and progressive supranuclear palsy – a comparative study. Parkinsonism Relat Disord 2010 May;16:275–9. [8] Gjerstad MD, Tysnes OB, Larsen JP. Increased risk of leg motor restlessness but not RLS in early Parkinson disease. Neurology 2011 Nov 29;77:1941–6. [9] Gomez-Esteban JC, Zarranz JJ, Tijero B, Velasco F, Barcena J, Rouco I, et al. Restless legs syndrome in Parkinson’s disease. Mov Disord 2007 Oct 15;22:1912–6. [10] Guerreiro TM, Nishikawa DR, Ferreira LC, Melo HA, Prado RC. Restless legs syndrome in Parkinson’s disease: clinical characteristics and biochemical correlations. Arq Neuropsiquiatr 2010 Dec;68:869–72. [11] Krishnan PR, Bhatia M, Behari M. Restless legs syndrome in Parkinson’s disease: a case-controlled study. Mov Disord 2003 Feb;18:181–5. [12] Lee JE, Shin HW, Kim KS, Sohn YH. Factors contributing to the development of restless legs syndrome in patients with Parkinson disease. Mov Disord 2009 Mar 15;24:579–82. [13] Loo HV, Tan EK. Case–control study of restless legs syndrome and quality of sleep in Parkinson’s disease. J Neurol Sci 2008 Mar 15;266:145–9. [14] Nomura T, Inoue Y, Miyake M, Yasui K, Nakashima K. Prevalence and clinical characteristics of restless legs syndrome in Japanese patients with Parkinson’s disease. Mov Disord 2006 Mar;21:380–4. [15] Ondo WG, Vuong KD, Jankovic J. Exploring the relationship between Parkinson disease and restless legs syndrome. Arch Neurol 2002 Mar;59:421–4.
S9
[16] Peralta CM, Frauscher B, Seppi K, Wolf E, Wenning GK, Hogl B, et al. Restless legs syndrome in Parkinson’s disease. Mov Disord 2009 Oct 30;24:2076–80. [17] Suzuki K, Miyamoto M, Miyamoto T, Tatsumoto M, Watanabe Y, Suzuki S, et al. Nocturnal disturbances and restlessness in Parkinson’s disease: using the Japanese version of the Parkinson’s disease sleep scale-2. J Neurol Sci 2012 Jul 15;318:76–81. [18] Svensson E, Beiske AG, Loge JH, Beiske KK, Sivertsen B. Sleep problems in Parkinson’s disease: a community-based study in Norway. BMC Neurol 2012;12:71. [19] Tan EK, Lum SY, Wong MC. Restless legs syndrome in Parkinson’s disease. J Neurol Sci 2002 Apr 15;196:33–6. [20] Verbaan D, van Rooden SM, van Hilten JJ, Rijsman RM. Prevalence and clinical profile of restless legs syndrome in Parkinson’s disease. Mov Disord 2010 Oct 15;25:2142–7. [21] Jagota P, Asawavichienjinda T, Bhidayasiri R. The low prevalence of primary restless legs syndrome in Thai Parkinson’s disease patients at Chulalongkorn University Hospital. J Med Assoc Thai 2012 Feb;95:175–80. [22] Ohayon MM, O’Hara R, Vitiello MV. Epidemiology of restless legs syndrome: a synthesis of the literature. Sleep Med Rev 2012 Aug;16:283–95. [23] Nomura T, Inoue Y, Nakashima K. Clinical characteristics of restless legs syndrome in patients with Parkinson’s disease. J Neurol Sci 2006 Dec 1;250:39–44. [24] Chahine LM, Ahmed A, Sun Z. Effects of STN DBS for Parkinson’s disease on restless legs syndrome and other sleep-related measures. Parkinsonism Relat Disord 2011 Mar;17:208–11. [25] Kedia S, Moro E, Tagliati M, Lang AE, Kumar R. Emergence of restless legs syndrome during subthalamic stimulation for Parkinson disease. Neurology 2004 Dec 28;63:2410–2. [26] Rana AQ, Siddiqui I, Mosabbir A, Athar A, Syed O, Jesudasan M, et al. Association of pain, Parkinson’s disease, and restless legs syndrome. J Neurol Sci 2013 Apr 15;327:32–4. [27] Comella CL, Goetz CG. Akathisia in Parkinson’s disease. Mov Disord 1994 Sep;9:545–9. [28] Lang AE, Johnson K. Akathisia in idiopathic Parkinson’s disease. Neurology 1987 Mar;37:477–81. [29] Poewe W, Hogl B. Akathisia, restless legs and periodic limb movements in sleep in Parkinson’s disease. Neurology 2004 Oct 26;63:S12–6. [30] Comella CL. Sleep disturbances in Parkinson’s disease. Curr Neurol Neurosci Rep 2003 Mar;3:173–80.
