Restless legs syndrome: Impact on sleep-related breathing disorders

Authors

  • FRANCOISE J. ROUX

    Corresponding author
    1. Section of Pulmonary and Critical Care Medicine and Yale Center of Sleep Medicine, Yale University School of Medicine, New Haven, Connecticut, USA
    Search for more papers by this author

  • The Author: Francoise Roux, MD, PhD is an Associate Professor of Medicine at Yale University School of Medicine in the Section of Pulmonary, Critical Care and Sleep Medicine. She is also the Medical Director of the Yale Center for Sleep Medicine and Program Director of the Sleep Fellowship at Yale University School of Medicine. Her main interests include the link between sleep-disordered breathing and adverse cardiovascular outcomes.

  • SERIES EDITORS: JOHN E HEFFNER AND DAVID CL LAM

Francoise J. Roux, Section of Pulmonary and Critical Care Medicine, Yale University School of Medicine, 333 Cedar Street, Post Office Box 208057, New Haven, CT 06520-8057, USA. Email: francoise.roux@yale.edu

ABSTRACT

Restless legs syndrome (RLS) is a common chronic sensory-motor neurological disorder that remains a clinical diagnosis. Most RLS patients present with sleep complaints in the form of initiation and/or maintenance insomnia as RLS has a circadian rhythmicity. An increased number of periodic leg movements during sleep (PLMS) is a supportive criterion in the diagnosis of RLS. Abnormalities in the central dopaminergic and iron systems are involved in the physiopathology of RLS. There is a higher prevalence of RLS and PLMS in sleep-disordered breathing patients, particularly those with obstructive sleep apnoea (OSA), the most common sleep disorder in western societies. The complex mechanisms underlying the association between OSA, RLS and PLMS remain unclear. Untreated OSA can lead to adverse cardiovascular consequences due to cardio-metabolic dysfunction. It remains controversial whether RLS could further adversely impact the cardiovascular consequences of OSA. The PLMS do not have an additive effect on the hypersomnia experienced by some sleep-disordered breathing patients. Continuous positive airway pressure (CPAP) therapy is the most effective therapy for OSA. The presence of PLMS during CPAP treatment could be a marker of an incomplete resolution of sleep-disordered breathing in the form of increased upper airway resistance syndrome, despite treatment. Dopaminergic agonists are the preferred agent for the treatment of RLS, and are indicated when RLS symptoms are frequent and affect quality of life. PLMS and RLS do not seem to contribute to the residual hypersomnia that can be observed in some sleep-disordered breathing patients despite adequate compliance and effective CPAP therapy.

Abbreviations:
CPAP

continuous positive airway pressure

OSA

obstructive sleep apnoea

PLMD

periodic limb movement disorder

PLMS

periodic leg movements during sleep

RLS

restless legs syndrome

INTRODUCTION

Restless legs syndrome (RLS) is a common neurological movement disorder, affecting up to 10% of the population but still remains an underdiagnosed condition.1 RLS patients present with sleep disturbance, namely initiation and/or maintenance of insomnia as RLS is more common during evening or night. RLS is very often associated with an increased number of periodic leg movements during sleep (PLMS). The physiopathology of RLS still remains elusive, although iron and dopamine have been implicated. Clinicians have observed for many years that RLS can be associated with sleep-disordered breathing, mostly obstructive sleep apnoea (OSA), but it is unclear whether there is a causal relationship between those two disorders. Nonetheless, the recognition of RLS among OSA patients has some important clinical implications as RLS can disrupt sleep despite adequate treatment of sleep-disordered breathing and should be treated accordingly. In this article, the current knowledge on RLS is reviewed, and the relationship between RLS/PLMS and sleep-disordered breathing is discussed. The evidence on the effect of treatment of RLS/PLMS on OSA and vice versa is also reviewed.

DEFINITIONS OF RLS, PLMS/PERIODIC LIMB MOVEMENT DISORDER AND SLEEP-DISORDERED BREATHING

RLS is a clinical diagnosis made by patient history and physical examination. The International Restless Legs Syndrome Study Group2 defined four criteria to allow a better recognition and standardization of this diagnosis: (i) RLS consists of an urge to move the legs in association with some paraesthesias and disagreeable sensations, such as crawling, aching or burning in the legs; (ii) the RLS sensations are relieved by movement; (iii) the symptoms have circadian rhythmicity and are worse at night; and (iv) the symptoms are also worse during periods of rest, especially long periods of inactivity.

The International Restless Legs Syndrome Study Group added supportive criteria, such as the presence of PLMS. The majority of RLS patients will also have PLMS, although PLMS are neither necessary nor sufficient on their own to make the diagnosis of RLS. PLMS are recorded at night during the polysomnography. PLMS consist of slow stereotyped, repetitive involuntary periodic movements of the leg(s) with ankle dorsiflexion and flexion of the great toe lasting about 0.5–5 s, and happening at a regular interval of 20–40 s; a rate of more than 15 episodes per hour of total sleep time is considered clinically significant.3 These PLMS can cause arousals, can occur in one or both legs, and might also less commonly involve the arms in severe cases. Periodic limb movement disorder (PLMD) is a sleep disorder characterized by an increased number of PLMS associated with insomnia, unrefreshing sleep or daytime hypersomnia after exclusion of other sleep disorders as a cause of these symptoms.

An apnoea is defined as cessation of airflow for at least 10 s in the presence of thoracoabdominal ventilatory efforts. A hypopnoea is a reduction in airflow of at least 30%, with a decrease in oxygen saturation of 2% or more for at least 10 s in the presence of thoracoabdominal ventilatory efforts. The apnoea–hypopnoea index is the sum of apnoeas and hypopnoeas per hour of sleep. An apnoea–hypopnoea index of ≥5/h is required to establish the diagnosis of OSA, according to the criteria of the American Academy of Sleep Medicine.4 However, there are various definitions of hypopnoea. A survey of American Academy of Sleep Medicine-accredited sleep centres found that no two laboratories used the same definition of hypopnoea,5 including various degrees of reduction in airflow, thoracoabdominal movement, associated oxygen desaturation and arousal. Such differences have important implications for both the diagnosis of OSA and the standardization of research results. A position paper was published in 2001 and defined hypopnoea as a 30% reduction in airflow or chest wall movement from baseline movement for at least 10 s and accompanied by oxygen desaturation of 4% or greater.6 Apnoeas can also be central in nature and present with a crescendo–decrescendo pattern indicative of Cheyne-Stokes breathing, which is most often associated with cardiac failure.

PREVALENCE OF RLS, PLMS/PLMD AND SLEEP-DISORDERED BREATHING

Epidemiological studies have been hampered by a lack of standardization in the diagnosis of RLS. The recent use of more stringent criteria to diagnose RLS by the International Restless Legs Syndrome Study Group has allowed a more accurate recognition of the syndrome. The RLS Epidemiology, Symptoms, and Treatment program was conducted in the United States and in five other European countries to better define the prevalence of RLS in the general population. This large multinational study involved 15 391 participants who answered either telephone or face-to-face interviews.7 A subpopulation of RLS sufferers was defined as having moderate or severe RLS symptoms that had occurred at least twice weekly in the last 12 months. Overall, the prevalence was 7.2% for RLS symptoms of any frequency and 5% for weekly symptoms; 2.7% of this population met the criteria for RLS sufferers. The prevalence of RLS was doubled among women and also increased with age up to 79 years old. A large cohort of adult primary care patients in Europe confirmed that RLS is a common disorder, with a prevalence at around 2–3%.8 Various studies have also shown that the prevalence of RLS is at least double in women compared with men. RLS symptoms are more frequent in the third trimester of pregnancy. There seems also to be a possible association between parity and the development of RLS.9

PLMS occurs more frequently in patients with RLS. Specifically, about 80–90% of RLS patients will have PMLS, although most studies to date have used older criteria defining clinically significant PLMS with a PLMS index of >5/h, in contrast to the more recent criteria of 15/h.10 The number of PLMS exhibits a significant night to night variability,11 especially among RLS patients, which is a complicating factor to assess its prevalence. A large epidemiological study was performed in Europe, involving about 18 980 subjects in the general population. These subjects were interviewed over the phone using diagnostic criteria proposed by the International Classification of Sleep Disorders to evaluate the prevalence of PLMD. This study reported the prevalence of PLMD to be 3.9%, higher in women than in men, and increasing with age.12 Another study found that the number of PLMS increased significantly after age 65 and tended to be equally distributed throughout the night.13

In Western countries, the prevalence of OSA is reported to be 24% in men and 15% in women; 4% of men and 2% of women have OSA with symptoms of daytime sleepiness.14

CLINICAL FEATURES AND CLASSIFICATION OF RLS

RLS is a common disorder and sleep disturbance is the most common reason for patient presentation, although it is not required for the diagnosis of RLS. The clinical criteria for RLS as defined by the International Restless Legs Syndrome Study Group has enabled more uniform diagnosis and improved recognition of the syndrome by general practitioners. The four diagnostic criteria described earlier are essential to make the diagnosis. Patients have a compelling need to move the legs, accompanied by disagreeable sensations such as crawling, aching, pulling, burning or creeping. These leg sensations are usually bilateral and can, in severe cases, also involve the upper limbs. These symptoms worsen with inactivity and are partially or totally relieved by walking and stretching. The symptoms are usually worse in the evening and at night than at daytime, but this circadian aspect can disappear in severe cases, affecting the patient 24 h a day. Some supportive criteria have been developed to help make the diagnosis, including a family history of RLS, the presence of associated PLMS, insomnia and an improvement in symptoms with dopaminergic therapy.2

RLS can be classified into primary RLS, which accounts for most cases, and secondary RLS. Primary RLS has an early onset, affecting patients less than 30 years old,15 tends to be familial,16 without any other associated medical conditions and is slowly progressive but less responsive to iron therapy. In contrast, secondary RLS has a late onset, affecting patients more than 40 years old, with a more rapid progression of symptoms. Secondary RLS is associated with various other medical conditions which, when treated, improves or resolves RLS symptoms. RLS is most commonly found in conjunction with iron deficiency states, such as iron deficiency anaemia, chronic renal failure17 and pregnancy. Other conditions, such as neuropathy, radiculopathy, spinal cord diseases, Parkinson's disease, narcolepsy, rheumatoid arthritis, fibromyalgia and Sjögren's syndrome, have all been linked to RLS.18,19 Some medications can worsen RLS. Selective serotonin re-uptake inhibitors increase serotonin transmission, which leads to inhibition of the function of the dopaminergic receptor. Neuroleptics are dopamine-blocker agents. Lithium can also decrease dopamine release and worsen RLS,20 as can antipsychotics and antihistamines.

Patients most often present with complaints of initiation or maintenance insomnia or sleep disruption. PLMS occur in the majority of RLS patients, especially in the elderly, and contribute to sleep disruption.21 However, PLMS are non-specific and can also occur in association with narcolepsy, neurodegenerative diseases, spinal cord lesions, neuroleptics and antidepressant medications.

Various conditions can share some features of RLS and need to be excluded.22 Differential diagnoses of RLS include hypnic myoclonus or sleep starts, fragmentary myoclonus, paraesthesia due to neuropathy, nocturnal leg cramps and neuroleptic-induced akathisia. Akathisia gives an inner sense of generalized restlessness but is not worse at rest and has no circadian pattern. Quiescegenic nocturnal dyskinesia is characterized by abnormal involuntary periodic movements with a circadian pattern but without any urge to move the legs—it might be a variant of RLS.23

PHYSIOPATHOLOGY OF RLS

Clinical experience has contributed to the understanding of the physiopathology of RLS. Clinicians have observed that in primary RLS, most patients develop the syndrome at a younger age and report a positive family history, suggesting a genetic predisposition. The initial approach in looking for a gene variant was to perform linkage studies in RLS patients. Linkage studies in Canadian and European RLS families have identified eight genome-wide significant loci, but no causally related sequence variant has been identified. Stefansson et al.24 analysed Icelandic and North American RLS patients and their first-degree relatives who were screened from the community by a validated RLS questionnaire and assessed for PLMS. The sequence variant (BTBD9) found in this study might be implicated in the motor manifestations of RLS, such as PLMS, and therefore appeared to be a gene for PLMS rather than for RLS. The same year, Winkelmann and colleagues25 performed a large genome-wide association study in German and French-Canadian RLS patients with a strong family history, and found the same sequence variant as Stefansson and colleagues, as well as two additional sequence variants. It is unclear whether these sequence variants are specific for RLS or PLMS as the authors did not measure the presence of PLMS in their study. However, it was notable that each genetic variant was associated with more than 50% increase in risk for RLS. These results explain the familial clustering found in RLS and PLMD. These genes appear to play a role in the development period; however, the precise function of these genes in the pathogenesis of RLS still needs to be identified.

In contrast, secondary RLS is characterized by more rapid progression of symptoms and later age of onset. Secondary RLS can be associated with iron deficiency anaemia, renal failure, pregnancy, peripheral neuropathy or chronic myelopathy, and some medications such as tricyclic antidepressants, selective serotonin re-uptake inhibitors, lithium, antipsychotics and dopamine antagonists. RLS can improve if these conditions resolve or are treated. In light of these observations, an abnormality in iron and dopamine metabolism has been suspected in the pathogenesis of RLS, especially as iron is a cofactor for tyrosine hydroxylase, a rate-limiting enzyme in dopamine metabolism. A small retrospective clinical study in RLS patients found a strong correlation between the severity of the RLS symptoms and the decrease in serum ferritin, the latter being a good reflection of iron stores.26 Interestingly, patients can have RLS despite a normal ferritin level. A later study was undertaken in RLS patients in whom iron deficiency had been excluded.27 These RLS patients had evidence for brain iron deficiency, with a low ferritin and a high transferrin level in the cerebrospinal fluid despite a normal serum ferritin level indicative of a normal systemic iron status.27 There is a decrease in brain iron concentration in the substantia nigra and the putamen among RLS patients which correlates with RLS severity.28 The low brain iron concentration in RLS patients may arise from dysfunction of iron transport and acquisition across the blood–brain barrier.29 Recently, Connor and colleagues found a decrease in myelin and in white matter in the brain of RLS patients based on autopsy findings.30

Abnormalities associated with dopamine are thought to be due to the involvement of the dopaminergic system in movement control. Dopamine levels are lower at night when RLS symptoms are worse. Dopamine agonists relieve RLS symptoms, whereas dopamine antagonists exacerbate. A lack of iron would affect dopamine synthesis. Autopsies from the brain of RLS sufferers have shown that a decrease of dopamine receptor density in the putamen of RLS patients is correlated with RLS severity.31 These findings support the notion that alteration of the dopaminergic system in the brain can contribute to the symptoms of RLS, and that iron and dopamine have a strong connection.

RLS/PLMS AND SLEEP-RELATED BREATHING DISORDERS

Prevalence of RLS/PLMS in sleep-disordered breathing patients

RLS and/or PLMS have been found in association with various sleep disorders, such as narcolepsy,32 rapid eye movement behaviour disorder33 and unexplained insomnia.34 An increase in PLMS has also been described in patients with mild sleep-disordered breathing, such as increased upper airway resistance syndrome.35 OSA is commonly encountered in clinical practice, and 4% of men and 2% of women have OSA, with symptoms of sleepiness.14

RLS/PLMS is also commonly seen in patients with sleep-disordered breathing, although very few studies have examined the prevalence of RLS/PLMS in this patient population. PLMS must be differentiated from non-periodic leg movements occurring during arousals at the end of each apnoeic episode. A small prospective study36 examined the prevalence of RLS in referred patients with confirmed OSA on diagnostic polysomnography. The OSA patients were screened for RLS by a questionnaire and by an interviewer using the criteria defined by the International Restless Legs Syndrome Study Group, requiring the incidence of RLS symptoms at least two to three times per week. This study found that clinically significant RLS occurred in 8.3% of OSA patients compared with 2.5% in the control group. There was no direct relationship between the severity of the RLS or the PLMS and the severity of the OSA. A large cross-sectional study was performed in Europe, interviewing 18 980 persons from the general population about RLS symptoms based on the criteria provided by the International Classification of Sleep Disorders.12 The study reported a prevalence of 3.9% and 5.5% for PLMS and RLS, respectively, and that having OSA was predictive of PLMS and RLS.12 However, this epidemiological study did not perform polysomnography, and the degree of severity of the RLS was not assessed.

PLMS are commonly seen during polysomnography in OSA patients. The current scoring criteria specify that PLMS should only be scored if the PLMS are spontaneous and not induced by abnormal respiratory events. A retrospective study examining the prevalence of PLMS in patients diagnosed with OSA found that 92% of patients referred for a polysomnography had OSA, and that 48% of newly diagnosed OSA patients had PLMS.37 OSA was a significant predictive factor of PLMS in this study. A cross-sectional study reported that OSA is common after cardiac transplantation due to weight gain.38 Among these OSA patients, 33% and 45% were also found to have PLMS and RLS, respectively,38 demonstrating a high prevalence of RLS and PLMS among sleep-disordered breathing patients. Nearly half of the patients with PLMS had RLS in this study.

PLMS can also be found in other types of sleep-disordered breathing. The prevalence of PLMS has been shown to be higher in cardiac failure patients who suffer from Cheyne-Stokes breathing compared with controls.39 In this study, the PLMS appeared to partially correlate with Cheyne-Stokes breathing.39 A recent study reported that PLMS were associated with Cheyne-Stokes breathing and disappeared with the resolution of Cheyne-Stokes breathing by continuous positive airway pressure (CPAP) treatment, whereas dopamine agonists had no effect on resolution of PLMS.40 The authors of this study suggested that Cheyne-Stokes breathing seems to play an active role in generating PLMS.

Physiopathology of RLS/PLMD among sleep-disordered patients

The precise mechanisms involved in the pathogenesis of RLS and PLMS have yet to be elucidated. It is not clear why RLS and/or PLMS are more frequent in patients with OSA. Studies have examined whether OSA could promote iron deficiency leading further to dopaminergic dysfunction that could trigger RLS/PLMS.41 Specifically, a prospective study measuring various indices of iron status, such as serum iron, iron-binding capacity, transferring saturation and serum ferritin, in patients who were referred for a diagnostic polysomnography found no correlation between ferritin level and the degree of OSA.41 Indeed, worse OSA with more severe oxygen desaturations seemed to predict a higher ferritin level. While some patients exhibited a low iron status, no association was found between low iron level and the degree of OSA, hypersomnia or the periodic leg movement index, despite the study having adequate power to detect such associations. A case report of patients with OSA and RLS has suggested that the brainstem may be involved in the pathology of RLS and OSA.42 Intermittent hypoxia associated with apnoeic events has also been linked with dysfunction of the dopaminergic pathway,43 suggestive of a possible relationship between OSA and RLS.

The prevalence of obesity in the United States has been increasing for decades. About 60 million adults (30% of the adult population) are obese, twice the percentage in 1980. Increasing weight has been associated with an increasing prevalence of OSA and can also worsen OSA.44 A large epidemiological study,45 using two cohorts, the Nurses' Health Study II and the Health Professional Follow-up Study, showed that overall obesity as well as abdominal obesity increased the likelihood of developing RLS in women and men. The odds ratio for RLS was 1.42 for a body mass index >30 and 1.60 for the presence of abdominal obesity, even after adjusting for age, ethnicity, physical activity, use of antidepressant and presence of a number of chronic diseases. Several studies have found that a dysfunctional dopaminergic pathway is implicated in the genesis of RLS, but it also seems to be affected in obese people. There is a lower dopamine D2 receptor availability in the striatum of obese people compared with normal-weight individuals.46 This deficit in striatal dopamine receptor in obese people could contribute to the observed link between RLS and OSA as most OSA patients are obese. However, other studies40 have reported that dopamine agonists do not impact the number of PLMS associated with Cheyne-Stokes breathing, suggesting that primary dopaminergic dysfunction may not play a major role in their physiopathology. In summary, the physiopathology of RLS/PLMS in sleep-disordered breathing appears to be heterogeneous and complex.

Clinical impact of RLS/PLMS on sleep-disordered breathing patients before CPAP treatment

Sleep-disordered breathing patients can present with varying degree of hypersomnia which can even be absent in some individuals. Patients with idiopathic RLS have been found to have fragmented sleep with prolonged sleep latencies, shorter total sleep time and a higher arousal index compared with age- and gender-matched controls using polysomnography.47 PLMS are traditionally found in about 80% of RLS patients. The PLMS could disrupt sleep and contribute to the symptoms of hypersomnia among OSA patients. A retrospective study found that the combination of PLMS and OSA did not result in worse hypersomnia compared with OSA patients without PLMS.37 A larger study confirmed those findings, namely that the rate of PLMS did not predict increased hypersomnia in suspected or diagnosed OSA patients and was even associated with less objective sleepiness.48 Haba-Rubio and colleagues49 evaluated in an objective and subjective fashion, using multiple sleep latency test and Epworth sleepiness scale, respectively, the degree of sleepiness among OSA patients having PLMS compared with OSA patients without PLMS. The PLMS did not contribute to any additive sleepiness among OSA patients compared with OSA patients without PLMS before CPAP treatment, as shown by multiple sleep latency tests and the Epworth sleepiness scale. Another large study, using multiple sleep latency tests to detect objective hypersomnia, did not show any association between the rate of periodic leg movements and subjective or objective sleepiness.48 Various studies have found that quality of life is altered50 in RLS patients, in part due to sleep disturbances, and can even be worse than in chronic medical conditions such as diabetes. The sleep loss induced by RLS might be responsible for the anatomical abnormalities noted in the prefrontal lobes51 in these patients. Frontal changes can be linked to the objective cognitive deficits found in RLS patients and have also been described in OSA patients52 before treatment with CPAP.

There are many cardiovascular consequences of OSA. OSA has been proposed as an independent risk factor for the development of hypertension because it can precede and predict the onset of hypertension.53 Most studies have found an independent association between OSA and adverse cardiovascular events, such as coronary artery disease54 and stroke.55 Although controversial, some reports have suggested that RLS might be a risk factor for developing cardiovascular diseases. Winkelman and colleagues,56 using the Sleep Heart Health study, examined whether those with RLS had a higher risk of cardiovascular diseases in about 3000 patients. They found that the RLS group, especially the moderate to severe RLS patients, had a much higher risk of coronary artery disease and cerebrovascular disease, compared with controls, even after controlling for multiple confounding factors, such as age, gender, body mass index, diabetes, systolic blood pressure, antihypertensive medication use, cholesterol and smoking history. RLS patients have more sleep disturbances and a higher arousal index at night compared with the general population, even in RLS patients with only intermittent symptoms and when sleep studies were performed in the home setting. Increased arousals at night or insufficient sleep can heighten the peripheral sympathetic tone and increase blood pressure, even among normal subjects. It has been suggested that the sleep disruption from RLS or from the increased number of nocturnal periodic leg movements associated with RLS could significantly contribute to the development of diurnal hypertension among RLS patients. A recent study found that women with RLS had a greater prevalence of hypertension, which increased with the severity of the RLS.57 However, other studies have shown no association between RLS and hypertension, which might be related to the absence of PLMS among these RLS patients. The association of RLS and OSA could be additive and worsen even further the predisposition of each disease to promote adverse cardiovascular events. However, a long-term prospective study58 on the natural course of RLS in untreated patients has shown that 40% of the RLS patients had persistent symptoms over a 2-year period, whereas RLS disappeared in the remaining 60% of patients. A higher frequency of symptoms was considered a risk factor for the perpetuation of RLS. Depression and subjective sleep disturbances were higher among patients with persistent RLS.

Impact of treatment of sleep-disordered breathing on RLS/PLMS

CPAP is considered the mainstay therapy for OSA. Treatment of OSA with CPAP plays an important role in improving cardiovascular outcomes, such as hypertension59 and left ventricular function, features that could be beneficial in these patients as RLS/PLMS might promote adverse cardiovascular effects. Periodic leg movements are commonly found with OSA or upper airway resistance syndrome but have also been associated with CPAP therapy despite correction of sleep-disordered breathing. Some investigators have suggested that CPAP therapy can even induce or worsen PLMS. Fry and colleagues examined the number of PLMS in OSA patients during the diagnostic polysomnography, the CPAP titration and after 7 months of CPAP therapy.60 They found that 27% of OSA patients had an increased number of PLMS during the diagnostic sleep study, with a PLMS index of 16.9/h which increased significantly to 39.3/h during the CPAP titration and even further to 42.9/h after prolonged CPAP use.60 Another study found that the effect of CPAP on the number of PLMS may be related to the severity of the OSA before CPAP treatment. Severe OSA patients experienced an increase in the number of PLMS during CPAP treatment, whereas a decrease in PLMS was observed among mild OSA patients during CPAP therapy. The investigators suggested that PLMS may have various aetiologies during CPAP therapy; some could be spontaneous in patients with underlying periodic leg movement disorder, and thus could be unmasked by CPAP therapy, whereas others could be induced by sleep-disordered breathing, and as such improve with CPAP treatment.61

In contrast, other studies have reported that the severity of OSA might not influence the number of PLMS as PLMS had decreased with CPAP treatment whether the patients had mild or severe OSA.62 Among patients with OSA and PLMS, another study found a decrease in the PLMS index after CPAP therapy.63 A recent study has demonstrated, using recordings of expiratory abdominal muscles to detect residual sleep-disordered breathing, that the presence of PLMS during CPAP titration represents persistence of sleep-disordered breathing.64 The severity of RLS symptoms has also been found to improve after 3 months of CPAP therapy in patients affected by the combination of RLS and OSA.65 Conversely, CPAP has been described to induce PLMS in patients who had neither OSA nor PLMS prior to CPAP therapy. The physiopathology of PLMS in that setting is incompletely understood and is most likely complex.66 The treatment of OSA not by CPAP but by mandibular advancement device can also induce PLMS in OSA patients who did not have a diagnosis of PLM at baseline.67 Clearly, the physiopathology of RLS/PLMS in relation to the treatment of sleep-disordered breathing is multifactorial and still poorly understood.

Impact of RLS/PLMS on the treatment of sleep-disordered breathing

CPAP therapy is very effective in treating sleep-disordered breathing. CPAP for OSA has also been shown to improve sleep architecture,68 decrease nocturnal desaturation events, decrease daytime somnolence69 and improve neurocognitive function.70 However, some sleep-disordered breathing patients suffer from residual sleepiness despite adequate compliance and therapeutic efficacy of CPAP therapy. There is not a good correlation between the degree of sleep-disordered breathing and the sleepiness, suggesting that other factors might be at play. Other associated sleep disorders could also cause hypersomnia despite effective CPAP therapy. Drigo and colleagues showed that the degree of PLMS did not affect the outcome of the CPAP titration. In their study, the OSA patients with PLMS were as effectively treated with CPAP therapy, with decreased sleep apnoea indices and arousals on CPAP, as the OSA patients without PLMS.71 However, it is still a common belief that RLS and PLMS could be a cause of residual hypersomnia among sleep-disordered breathing patients due to increased arousals. When the degree of sleepiness was assessed by objective means such as multiple sleep latency test, the PLMS were not associated with excessive daytime sleepiness.48 Other authors have also found that OSA patients with PLMS experienced disappearance of hypersomnia on CPAP treatment.72 Another study,49 also using multiple sleep latency test, showed that PLMS had no impact on residual objective sleepiness in adequately treated OSA patients on CPAP therapy for 1 year. In OSA patients with RLS, CPAP therapy led a normalization of the Epworth sleepiness score on CPAP therapy but to a lesser degree than in the OSA patients without RLS; however, the sleepiness was not assessed in an objective fashion.73 In contrast, the presence of RLS in OSA patients predicted worse residual fatigue compared with OSA patients without RLS after CPAP therapy.73

In summary, the available data indicate that RLS and PLMS do not contribute to the residual hypersomnia observed in some OSA patients despite effective CPAP therapy, and other causes of hypersomnia should be sought in these patients. RLS could be implicated in the residual fatigue experienced by some OSA patients despite effective CPAP therapy.

Treatment of RLS/PLMS in sleep-disordered breathing patients

The need for treatment of RLS is dictated by the degree of impairment and the frequency of symptoms. In patients with intermittent symptoms, lifestyle modifications and simple measures such as mild exercise, elimination of caffeine intake, leg massages, or cold or hot baths might be sufficient. Long-term pharmacotherapy is indicated in patients with daily or frequent symptoms that impair their quality of life and sleep efficiency. The dopaminergic agonists, targeting the dopamine-2 and dopamine-3 receptor-subtypes, are the first-line treatment for RLS and are Food and Drug Administration-approved in the United States for this indication. Dopaminergic agents relieve RLS symptoms in 90% of patients. Benzodiazepines, such as clonazepam, have been used to treat RLS but are not Food and Drug Administration-approved for this indication; they should be used with caution in OSA patients as benzodiazepines have the potential to worsen sleep-disordered breathing. Opiates have also been used to treat RLS but can promote onset of central apnoeas. In some reports, the severity of RLS as well as the number of periodic leg movements62 have been shown to improve after CPAP treatment in OSA patients.65 The prevalence of PLMS is very high in RLS patients but does not appear to play an additive role in the residual hypersomnia that can be encountered in sleep-disordered breathing patients, despite adequate CPAP therapy. As a consequence, treatment of PLMS in routine sleep-disordered breathing patients is probably not warranted. The presence of PLMS in sleep-disordered breathing patients with an obstructive component might be an indication that some residual mild sleep-disordered breathing remains present. In patients with Cheyne-Stokes breathing and PLMS, positive airway pressure should be attempted first as dopaminergic agents do not appear helpful in that setting.

CONCLUSIONS

RLS is a very common sensory-motor neurological disorder with lasting symptoms that can significantly impair quality of life and sleep quality. RLS is often associated with PLMS. The prevalence of RLS/PLMS is higher among sleep-disordered breathing patients than in the general population, but the physiopathology of RLS and PLMS remains poorly understood in these patients. The presence of PLMS in sleep-disordered breathing patients might be an indication of incomplete resolution of breathing abnormalities with CPAP therapy. RLS should be treated with dopaminergic agents when the symptoms are frequent. CPAP therapy remains the mainstay of therapy for sleep-disordered breathing and can occasionally improve the RLS symptomatology. RLS and PLMS do not appear to contribute to the residual hypersomnia observed in some OSA patients despite effective CPAP therapy, but RLS could explain, in part, the persistent fatigue experienced by some of these patients.

A better understanding of the RLS physiopathology in sleep-disordered breathing patients will allow tailored therapy and improved quality of life of these patients.

Ancillary