Abstract
Obstructive sleep apnoea (OSA) and restless legs syndrome (RLS) are two of the most prevalent sleep disorders and can coexist within the same patient. Nonetheless, the recognition of RLS among OSA patients has important clinical implications, since RLS can disrupt sleep despite adequate treatment of sleep disordered breathing and should be treated accordingly. Furthermore, the presence of OSA can also increase the severity of RLS. Therefore, it is important to be able to correctly identify both disorders and treat them effectively. The present article reviews our current knowledge on this comorbidity and discusses potential treatment options for RLS in the context of OSA.
Abstract
Treating restless legs syndrome in the context of sleep disordered breathing comorbidity http://bit.ly/2lUgFcT
Introduction
Both obstructive sleep apnoea (OSA) and restless legs syndrome (RLS) are considered to be among the most common sleep disorders [1–5]. However, it is unclear whether there is a causal association between both disorders. Identifying RLS in OSA patients is relevant, particularly because RLS can contribute to sleep fragmentation despite adequate treatment of sleep disordered breathing. In addition, the presence of RLS can make adjustment to continuous positive airway pressure (CPAP) treatment more difficult, and should be treated accordingly [6].
In this article, we will review current knowledge of RLS its relationship with sleep disordered breathing. The evidence on the effects of treatment of RLS on OSA, and treatment of OSA on RLS, are also discussed.
RLS in sleep disordered breathing
RLS is not only a common neurological movement disorder, but also one of the most common causes of sleep disturbance. It affects up to 10% of the general adult population and can be severe in up to 2–3%. Nevertheless, it still remains an underdiagnosed condition [7, 8].
RLS patients frequently present with difficulties initiating and/or maintaining sleep, as RLS is more common during the evening or night. It is mainly characterised by neurosensory symptoms in the form of strong urge to move the limbs frequently associated to distressing paraesthesia-like sensations [9, 10], and is typically associated to periodic leg movements of sleep (PLMS) when assessed by means of polysomnography [11].
RLS typically manifests when the patient is at rest; immobility or relaxation are associated with a greater likelihood of symptom occurrence. Conversely, symptoms usually improve or resolve when the patient initiates physical activity; when symptoms arise patients will experience an intense urge to move in order to relieve the discomfort felt.
Genetic, environmental and medical factors provide a variable expression of RLS symptoms. These symptoms vary considerably in frequency from being occasional to daily, while severity ranges from mildly annoying to fully disabling. Furthermore, symptoms fluctuate widely in its intensity over time and may even remit for differing periods of time.
The diagnostic definition of RLS is based on expert consensus and was first developed in 2003 following a workshop at the National Institutes of Health. However, a more recent update improves specificity by excluding mimics. Thus, RLS is currently diagnosed by ascertaining the presence of five essential clinical criteria (table 1) [9].
Diagnostic criteria for restless legs syndrome (RLS)
Surprisingly, not many studies have examined the prevalence of RLS patients diagnosed with OSA to date [12]. However, one small prospective study [13] examined the prevalence of RLS in patients who had already been diagnosed with OSA by means of polysomnography. Screening for RLS was performed by means of a questionnaire and by a face-to-face interview which used established diagnostic criteria. Clinically significant RLS was found in 8.3% of OSA patients in contrast to 2.5% in the control group. However, no direct relationship was found between the severities of both disorders. In addition, a large European cross-sectional study interviewed nearly 20 000 subjects among the general population for RLS symptoms and found prevalences of 3.9% for PLMS and 5.5% for RLS, respectively [14]. An OSA diagnosis appeared to be predictive of PLMS and RLS.
Excessive daytime sleepiness in patients with OSA and RLS
Polysomnographic studies have shown that RLS patients with idiopathic RLS have fragmented sleep with long sleep latencies, shorter total sleep time and a higher arousal index compared with controls [15].
Moreover, an increased PLMS index is present in ∼88% of RLS patients [11]. These PLMS can disrupt sleep and contribute to hypersomnia among OSA patients with comorbid RLS. It is conceivable that a combination of PLMS and OSA might lead to increased excessive daytime sleepiness. However, a retrospective study found that the combination of PLMS and OSA did not result in worsening hypersomnia compared with OSA patients without PLMS [16, 17].
One study, using both the multiple sleep latency test (MSLT) and the Epworth sleepiness scale, compared the severity of excessive daytime sleepiness among OSA patients with PLMS and without PLMS [18]. Before CPAP treatment, OSA with PLMS did not show greater sleepiness compared to OSA patients without PLMS. In another large study, which measured daytime sleepiness by means of MSLT, no association between the rate of periodic leg movements and subjective or objective sleepiness was found [17].
Periodic leg movements during sleep should be differentiated from movements caused at the end of respiratory events. Current international guidelines, based on preliminary empirical evidence, recommend to remove from scoring any candidate leg movement on which any part of the leg movement occurs between 2.0 s before and 10.25 s after the end of SDB events [19, 20].
Frontal changes and cardiovascular risk in RLS with OSA
Quality of life is reduced in patients with RLS, even when compared to other chronic medical conditions, and this is mainly due to the presence of sleep disturbance [21]. Indeed, sleep loss might be responsible for cognitive deficits caused by frontal lobe dysfunction [22]. Similar changes have been described in untreated OSA patients [23].
The risk of coronary artery disease and cerebrovascular disease is higher in patients with moderate-to-severe RLS compared with controls, even after controlling for age, sex, body mass index, diabetes, systolic blood pressure, antihypertensive medication use, cholesterol and smoking history [24].
Furthermore, RLS might be an independent risk factor for the development of cardiovascular diseases. It has been suggested that either the sleep disruption resulting from RLS or the increased number of periodic leg movements leads to arterial hypertension in RLS patients, thus facilitating an increase in the cardiovascular and/or cerebrovascular risk [25, 26]. The association of RLS and OSA could be additive and worsen even further the predisposition of each disease to promote adverse cardiovascular events.
Impact of treatment of sleep disordered breathing on RLS/PLMS
Periodic leg movements are commonly found both with OSA or upper airway resistance syndrome, and also following successful CPAP therapy. Indeed, it has been suggested that CPAP therapy can even induce or worsen PLMS [27], and the effect of CPAP on PLMS may be related to the severity of the OSA: thus, patients with severe OSA experience an increase in PLMS during CPAP treatment, while patients with mild OSA show a decrease in PLMS.
PLMS may have various aetiologies during CPAP therapy; some could be spontaneous in patients with underlying periodic limb movement disorder, and be unmasked by CPAP therapy, whereas others could be induced by sleep disordered breathing, and as such improve with CPAP treatment [28–31].
Taken together, although the physiopathology of RLS/PLMS in relation to the treatment of sleep disordered breathing is multifactorial and still poorly understood, treatment of OSA with CPAP improves cardiovascular outcomes, and this could be beneficial for RLS/PLMS, given its cardiovascular effects [32].
Impact of RLS/PLMS on the treatment of sleep disordered breathing
Persistence of residual sleepiness despite adequate treatment and CPAP compliance is not uncommon in OSA patients [33]. In these patients, other factors than OSA might be at play.
Available data indicate that neither RLS nor PLMS contribute to the daytime sleepiness, and other causes of hypersomnia should be sought in these patients. However, RLS could be implicated in the residual fatigue experienced by some OSA patients despite effective CPAP therapy [17, 18, 34, 35].
Treatment of RLS in the context of OSA
In general, treatment for RLS should be initiated when the symptoms impair the patient's quality of life, daytime functioning, social functioning or sleep. In the context of comorbidity with OSA, any RLS symptoms that interfere with the treatment of OSA, i.e. making the use of CPAP more difficult should induce the initiation of a specific treatment [12, 36, 37].
Before any RLS treatment is initiated, it is important to first normalise iron stores. Thus, any measures (including oral or intravenous iron administration and prevention of blood losses) should be taken to ensure that ferritin levels are raised to >75 ng·mL−1. Recent guidelines on iron treatment for RLS recommend oral iron treatment for ferritin levels ≤75 μg·L−1 [38]. Oral iron is non-efficacious in iron-sufficient subjects, but its benefit for patients with low peripheral iron status has not been adequately evaluated.
For patients with ferritin ≤100 μg·L−1, or whenever oral iron is not appropriate, intravenous supplementation should be considered. No iron treatment is recommended if transferrin saturation is >45%.
Additional conservative measures include elimination of any medications that exacerbate RLS symptoms, such as antihistamines, dopamine antagonists, antinausea medications, antidepressants and lithium [36].
Two classes of drugs are frequently used for the treatment of RLS: dopamine agonists and α2δ ligands [36]. The first group includes pramipexole, ropinirole and rotigotine, all of which are US Food and Drug Administration (FDA)- and European Medicines Agency (EMEA)-approved (table 2) [36]. Dopamine agonists are effective in the treatment of subjective sensitive symptoms and PLMS. Their main side-effects are transient nausea, headache and somnolence, and in the case of the rotigotine patch, skin reactions. Nevertheless, although their short-term efficacy is very clear in most cases, the main long-term complications of dopamine agonists are augmentation of RLS symptoms, tolerance and impulse control disorders (ICDs) [39].
Recommended doses for commonly used medications in restless legs syndrome
Augmentation is a worsening of RLS symptoms manifested by an earlier onset of symptoms in the afternoon or evening compared to before treatment initiation [40]. It consists of symptoms starting earlier in the afternoon, a spread of symptoms to previously unaffected body parts (i.e. the arms) or symptoms arising more quickly following immobility. It can occur at any time during treatment, but is usually more likely with higher doses and drugs with shorter-acting half-lives. One study found that up almost 75% of treated patients suffered from some degree of symptoms of augmentation [41].
ICDs have been reported to develop in up to 17% of patients with RLS undergoing treatment with dopamine agonists [42]. ICDs are likely to increase with higher doses of dopaminergic agents and in females.
Surprisingly, the effect of dopaminergic drugs on sleep disordered breathing has rarely been investigated. One study found a marginal improvement of the apnoea–hypopnoea index in patients undergoing a 3-week treatment with 0.09–0.54 mg pramipexole for sleep bruxism [43].
Given the potential for a worsening of RLS symptoms following long-term treatment with dopamine agonists, most current expert guidelines recommend keeping their dose at a minimum, and if possible, initiating treatment with a non-dopaminergic agent [36].
The main alternatives to dopamine agonists are α2δ ligands of the calcium channels, such as gabapentin or pregabalin. None of these drugs are FDA- or EMEA-approved, but gabapentin enacarbil, a prodrug of the former, is approved in the USA and in Japan. Their mechanism of action consists in an inhibition of central excitatory neurotransmitters, particularly glutamate. Both substances cause an improvement of subjective symptoms, sleep and, although to a lesser degree than dopamine agonists, PLMS. In contrast, over the long-term, α2δ ligands do not cause significant augmentation [44]. Side-effects include dizziness, somnolence, fatigue, headache and weight gain [44]. However, α2δ ligands are less effective in patients previously treated with dopamine agonists [45].
A small crossover study on eight elderly patients found that 300 mg gabapentin caused an increase in the apnoea–hypopnoea index [46]. In addition, pregabalin can cause significant respiratory depression in postoperative opium-induced analgesia [47].
Opioids such as oxycodone or methadone are widely used for the treatment of RLS resistant to dopamine agonists [36]. Although this class of drugs has been investigated in long-term studies [48, 49], few studies have investigated their effects on sleep disordered breathing. However, Walters et al. [50] performed polysomnography on seven out of 493 patients with RLS treated for an average of 7 years (range 1–15 years) with opioids as a monotherapy, and found a worsening of the respiratory disturbance index and minimum arterial oxygen saturation (SaO2) in two of them. Furthermore, a meta-analysis comprising seven studies with a total 803 long-term opioid users diagnosed with OSA found a small increase in the apnoea–hypopnoea index, but a moderate increase in the central sleep apnoea index [51].
Finally, benzodiazepines are frequently used for sleep stabilisation in RLS, although no convincing evidence exists of their therapeutic efficacy for the treatment of dysaesthesias or PLMS [36]. A meta-analysis involving five studies on a total of 233 patients with insomnia and chronic obstructive pulmonary disease (COPD) using benzodiazepines showed no significant increase (sd 0.15, 95% CI −0.25–0.56; p=0.46) in the percentage of time of SaO2 <90% and mean SaO2 in night sleep. However, in one study a mild increase in transcutaneous carbon dioxide tension (PCO2) during sleep was shown. Taken together, due to the possibility of increased maximum PCO2 during sleep, caution is recommended in COPD patients with hypercapnia [52].
Benzodiazepines have been used for RLS; the most studied is clonazepam. They have been shown to decrease the number of arousals due to periodic limb movements [53]. A recent review by the Cochrane group showed that there is insufficient evidence to support or refute the use of benzodiazepines to treat symptoms of RLS [54].
Conclusions
OSA and RLS are two common sleep disorders that frequently coexist within the same patient. Although these disorders do not share a common pathophysiological mechanism, the presence of comorbidity with the other increases the severity of both disorders. Furthermore, comorbidity makes treatment of either of the two more difficult.
Treatment of RLS in the context of sleep disordered breathing should be started, whenever possible, with conservative measures. If not sufficient, a pharmacological approach should be started.
Dopamine agonists are a safe and effective treatment for RLS in OSA. However, their main complication is long-term augmentation of RLS symptoms. In contrast, α2δ ligands are effective and generally a safe treatment for RLS in sleep disordered breathing. However, these treatments are less effective in patients previously treated with dopamine agonists. Opioids are effective in patients refractory to dopaminergic treatment, but, similarly to benzodiazepines, require caution regarding respiratory depression and central apnoeas. In addition, oral and intravenous iron treatments are generally safe and effective in the mid- to long-term.
Footnotes
Provenance: Commissioned article, peer reviewed
Conflict of interest: S. Romero-Peralta has nothing to disclose.
Conflict of interest: I. Cano-Pumarega has nothing to disclose.
Conflict of interest: C. Garcia-Malo has nothing to disclose.
Conflict of interest: L. Agudelo Ramos has nothing to disclose.
Conflict of interest: D. Garcia-Borreguero reports personal fees from American Regent and grants from Merck, during the conduct of the study.
- Received May 30, 2019.
- Accepted August 31, 2019.
- Copyright ©ERS 2019.
This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0.
References
