INTRODUCTION — Restless legs syndrome (RLS) refers to symptoms of spontaneous, continuous leg movements associated with unpleasant paresthesias. These sensations occur only at rest and are relieved by movement. Sleep disturbance and a frequent association with involuntary, jerking movements of the legs during sleep, known as periodic leg movements of sleep (PLMS), are common. Nocturnal leg cramps are a separate disorder.

EPIDEMIOLOGY — Mild symptoms of RLS occur in 5 to 15 percent of the population; prevalence figures vary widely depending upon the population surveyed and severity of symptoms required for inclusion.

In a population study of RLS that used standard diagnostic criteria, 15,391 subjects (18-years-old) from the United States and five European countries (France, Germany, Italy, Spain, and the United Kingdom) completed questionnaires and the following observations were made: RLS symptoms of any frequency were reported by 7.2 percent. RLS symptoms occurred at least twice a week and were reported as moderately or severely distressing by 2.7 percent. The prevalence of RLS was approximately twice as high for women compared with men. The prevalence of RLS increased with age up to age 79, and then declined.

A previous survey also found that the prevalence of restless legs (with symptoms experienced on five or more nights per month) increased with age, from 3 percent of participants ages 18 to 29, compared with 10 percent of those ages 30 to 79 and 19 percent of those ages 80 or older. Increasing parity may be associated with an increased risk of RLS.

RLS also occurs in children. In a questionnaire study of patients with RLS, 25 percent experienced their first symptoms between ages 11 and 20. Misdiagnoses such as "growing pains" and attention deficit hyperactivity disorder were common, and medical attention was often not sought until after age 40 when symptoms frequently begin to progress. In a retrospective study, 32 of 538 children and adolescent subjects (5.9 percent) presenting to a sleep disorders clinic had restless legs syndrome. Low serum ferritin (<50 g/L) was found in 20 of 24 subjects (83 percent) who had ferritin levels. A family history of RLS was present in 23 of 32 (72 percent) subjects, and mothers were nearly threefold more likely to be affected than fathers.

PATHOGENESIS — In most cases RLS is a primary idiopathic disorder, but it also can be associated with a variety of underlying medical disorders.

Primary RLS — The cause of primary, idiopathic RLS is unknown. A family history consistent with dominant inheritance is present in more than 40 percent of patients with idiopathic RLS. This observation in combination with several large kindreds with RLS that have been reported suggest a genetic basis for the disorder. A genetic component is further supported by a twin study in which 10 of 12 monozygotic twin pairs were concordant for RLS, although there was considerable variance in age of onset and symptom severity.

Genetic linkage studies have mapped at least four susceptibility loci for RLS: chromosome 12q13-23, chromosome 14q13-21, chromosome 9p24-22, and chromosome 20p13.

There is a high incidence of RLS in spinocerebellar ataxia type 3 (Machado-Joseph disease), which is unrelated to the incidence of peripheral neuropathy in that disorder.

Several neurophysiologic changes have been identified in patients with RLS, including reduced motor cortex inhibition, spinal flexor reflex hyperactivity, and brainstem reflex abnormalities, but these are nonspecific findings, the significance of which is currently undetermined.

It has been assumed that RLS is a peripheral disorder, but studies of brain dopamine metabolism raise the possibility of a central nervous system cause. One hypothesis holds that RLS arises from dysfunction of hypothalamic dopaminergic cells that are the source of spinal cord dopamine. Data from functional imaging with single photon emission computed tomography (SPECT) and positron emission tomography (PET) are conflicting, having demonstrated reduced, increased, or unchanged basal ganglia dopamine receptor binding and 18F-Dopa uptake in patients with RLS compared with control subjects. This is of particular interest since there is probably an increased incidence of RLS in Parkinson's disease. Brain iron metabolism may also play a role.

Hypocretins are hypothalamic neuropeptide transmitters that participate in the normal control of the sleep wake cycle and are depleted in the pathophysiology of narcolepsy. Hypocretins increase arousal and interact with the dopamine system. A small clinical study also found that increases in hypocretin levels in the cerebrospinal fluid were associated with an increased incidence of restless legs syndrome, particularly early onset disease.

Secondary RLS — RLS can occur secondary to a number of disorders including iron deficiency, uremia, diabetes mellitus, rheumatic disease, and venous insufficiency, among others.

Iron deficiency — Iron deficiency has been considered a possible cause of RLS since the original publications on this disorder. Subsequently, the relationship between iron status and RLS has been examined in several studies: In a report of 18 elderly patients with RLS and 18 matched controls, the serum ferritin concentration was significantly lower in patients with RLS (33 versus 59 mcg/L) and was inversely correlated with the severity of symptoms. Although the hemoglobin concentration in patients and controls was similar, treatment with ferrous sulfate for two months reduced RLS symptoms in 14 of 15 patients. In a second study, patients with RLS had a lower spinal fluid ferritin concentration than age-matched controls (1.11 versus 3.50 ng/mL) and a higher spinal fluid transferrin concentration (26.4 versus 6.71 ng/mL). However, there was no difference in serum ferritin and transferrin between the two groups. A study that performed MRI estimates of brain iron concentration in a small number of patients with RLS found significantly reduced iron in the substantia nigra. Neuropathological studies found altered iron management protein levels that were consistent with iron insufficiency in substantia nigra neuromelanin cells from patients with RLS. Quantitative analyses showed decreased levels of ferritin, divalent metal transporter 1 (DMT1), ferroportin, transferrin receptor, and iron regulatory protein 1 (IRP1) compared with controls.

While these findings are not conclusive, they warrant the measurement of serum ferritin levels in patients with RLS and a trial of oral iron therapy when ferritin levels are low. Oral iron therapy may also be considered when ferritin levels are within normal range, as some patients without iron deficiency may still respond.

End-stage renal disease — RLS is common among dialysis patients, with a reported incidence of 6 to 60 percent. Anemia may play a role in these circumstances, since low-dose erythropoietin therapy significantly reduced RLS in one report. In addition, iron therapy among dialysis patients with functional, but not absolute, iron deficiency was beneficial in another study.  A third study, however, found that a low serum parathyroid hormone concentration, but not serum hemoglobin, was associated with the presence of RLS in 136 dialysis patients.

Diabetes mellitus — RLS can be a prominent feature of diabetic neuropathy. RLS and other sensory symptoms of neuropathy often improve following successful pancreatic-kidney transplantation.

Parkinson's Disease (PD) — Both RLS and PD appear to involve disturbances in the dopaminergic neurotransmitter system, suggesting a common mechanism. However, few studies have examined the relationship between these two disorders, and progress is hampered by a lack of understanding regarding the pathophysiology of RLS. In contrast to PD, preliminary neuropathologic evidence suggests that the pathogenesis of RLS does not involve neuronal degeneration of nigrostriatal pathways.

Estimates of the prevalence of RLS among patients with Parkinson's disease vary widely from 0 to 20.8 percent; some but not all studies have found that the prevalence of RLS is higher in patients with PD than in the general population. Akathisia is also common in PD, with prevalence rates ranging from 26 to 45 percent and the potential for overlap with RLS. One series found that patients with both PD and RLS were older at RLS onset, were less likely to have a family history of RLS, and had lower serum ferritin levels compared with patients with idiopathic RLS. When both disorders were present, PD preceded RLS in 68 percent of patients.

Pregnancy — Pregnancy appears to be a risk factor for the occurrence or worsening of RLS. In a study of 626 pregnant women admitted to a single center, the diagnosis of RLS was determined by the four International RLS Study group criteria; the prevalence of RLS in this group of women was 10 percent before pregnancy and increased to 27 percent during pregnancy. The highest rates were seen in the third trimester and dropped quickly after delivery. Older studies have reported lower rates of RLS during pregnancy, but none of them had used or strictly applied the four standard diagnostic criteria. The cause of the increased frequency of RLS during pregnancy is unclear, but possible causes include iron deficiency, folate deficiency, and hormonal changes.

Rheumatic disease — The relationship between RLS and rheumatic disease is unclear. In one study, RLS was present in 25 percent of patients with rheumatoid arthritis compared with 4 percent of controls with osteoarthritis or seronegative arthropathy and was associated with greater disease activity and severity. A higher than expected prevalence of RLS has also been reported in patients with Sjögren's syndrome, and in a study of 135 patients with fibromyalgia, 42 had RLS. In contrast, another study carried out in a neurology specialty clinic found no evidence of clinical rheumatologic disease among 68 patients with RLS, and only four patients had positive serologies for rheumatic diseases.

Venous insufficiency — Varicose veins have been associated with RLS and treatment of varicose veins and chronic venous insufficiency may be helpful in some patients. As an example, one study found that 312 of 1397 patients seeking treatment for varicose veins had symptoms of RLS on a screening questionnaire and interview. Sclerotherapy was performed in 113 of the patients with RLS; 98 percent reported initial relief of RLS. Symptom relief was maintained in 72 percent at two years of follow-up.

Treatment of chronic venous insufficiency with hydroxyethylrutoside also appears to be beneficial in patients with RLS but not as striking or dramatic as sclerotherapy. A meta-analysis of 15 trials that included a total of 1973 patients found that RLS improved in 36 percent of those treated with an average dose of 1000 mg/day for a minimum of four weeks compared with 26 percent of placebo controls.

Other miscellaneous conditions — Other disorders that are reported to be associated with RLS include other peripheral neuropathies, vitamin deficiencies, lumbosacral radiculopathy, spinal stenosis, excess caffeine intake, administration of mianserin, hypoglycemia, and hypothyroidism.

CLINICAL MANIFESTATIONS — Although the subjective symptoms of RLS are often difficult to describe, the clinical features are highly stereotyped. The hallmark of RLS is a marked discomfort in the legs that occurs only at rest and is immediately relieved by movement. The abnormal feelings are typically deep seated and localized below the knees. Distribution is usually bilateral, but some asymmetry may occur and the arms can be affected in more severe cases.

Terms that patients use to describe the symptoms include crawling, creeping, pulling, itching, drawing, or stretching, all localized to deep structures rather than the skin. Pain and tingling paresthesia of the type that occurs in painful peripheral neuropathy is usually absent, and there is no sensitivity to touching of the skin.

Symptoms typically worsen towards the end of the day and are maximal at night, when they appear within 15 to 30 minutes of reclining in bed. In severe cases symptoms may occur earlier in the day while the patient is seated, thereby interfering with attending meetings, sitting in a movie theater, and similar activities. In milder cases patients will fidget, move in bed, and kick or massage their legs for relief. Patients with more severe symptoms feel forced to get out of bed and pace the floor to relieve symptoms.

Periodic leg movements of sleep — PLMS are sudden jerking leg movements that commonly accompany RLS. The patient is usually unaware of these movements. The prevalence of PLMS increases with age, and PLMS are identified in the vast majority of patients with RLS during sleep laboratory evaluations.

Treatment is not necessary if PLMS occur without sleep complaints. However, in some patients, the periodic limb movements may cause partial or total arousal from sleep and provide an additional cause of insomnia and excessive daytime drowsiness. Periodic limb movement disorder (PLMD) of sleep is the term used to characterize the association of PLMS and hypersomnolence.

Although data are limited, treatment of symptomatic PLMD is approached using the same drugs and regimens as used to treat RLS, particularly when PLMD is accompanied by RLS. Benzodiazepines and pramipexole are often used to treat PLMD in the absence of RLS.

DIAGNOSIS — The diagnosis of RLS is often delayed or missed, especially when the symptoms are relatively mild or non-specific. The diagnosis of primary RLS rests on typical symptoms in the presence of a normal neurologic examination. Patients with secondary forms of RLS due to peripheral neuropathy may have related sensory and reflex abnormalities. The International Restless Legs Study Group proposed the following four features as minimal criteria for the diagnosis of RLS: Desire to move the extremities, often associated with paresthesias or dysesthesias. Motor restlessness. Worsening of symptoms at rest with at least partial and temporary relief during activity. Worsening of symptoms in the evening or at night

Polysomnography in a sleep laboratory is not necessary for the diagnosis, but it may be helpful, especially when RLS is resistant to treatment. In such cases coexisting PLMS may be identified in the majority.

The differential diagnosis begins with separating primary from secondary RLS. Iron deficiency and renal failure are particularly important to exclude; a normal hemoglobin does not rule out iron deficiency if the serum ferritin is low. Peripheral neuropathy, lumbosacral radiculopathy, and ordinary leg cramps, all of which are typically more painful conditions, should also be considered.

RLS should be differentiated from akathisia, a common side effect of the phenothiazine antipsychotic drugs and selective serotonin reuptake inhibitor (SSRI) antidepressants. Akathisia (from the Greek "not to sit") is a more constant and generalized feeling of motor restlessness unassociated with subjective discomfort localizing to the legs. In contrast to akathisia, RLS has a circadian rhythm (worse at night when the patient is sitting or in bed) and is commonly accompanied by paresthesias and myoclonic jerks during the waking state. The syndrome of "painful legs and moving toes" is a rare disorder of unknown cause characterized by more prominent involuntary toe movements and leg pain.

TREATMENT — A number of treatments for RLS have been studied in primarily small, randomized, controlled trials. Placebo controlled studies are particularly important since RLS is normally characterized by fluctuations and remissions.

Early reports described the benefit of vasodilators, iron, and vitamins, but currently these agents are rarely used except for iron in demonstrated iron deficiency anemia. Stretching exercises for the posterior leg muscles may be helpful before retiring for patients with persistent symptoms.

Pharmacotherapy for idiopathic RLS with benzodiazepines, dopaminergic drugs, or, in resistant cases, opioids has been successful in many patients. Dopaminergic drugs such as levodopa (L-dopa) and dopamine agonists appear to be more effective than benzodiazepines in patients with RLS, although there are no direct comparison studies.

Expert panel recommendations — An algorithm for the management of RLS devised by an expert panel (the Medical Advisory Board of the Restless Legs Syndrome Foundation) was published in July 2004 and highlights three particular types of RLS (intermittent, daily, and refractory). Recommendations for these three types are excerpted here.

Intermittent RLS — Intermittent RLS is defined as RLS that is troublesome enough when present to require treatment but that is not sufficiently frequent to require regular daily medication use. Treatment options include: Nonpharmacologic therapy. Levodopa. Dopamine agonists. Low potency opioids or opioid agonists. Benzodiazepines or benzodiazepine agonists.

Daily RLS — Daily RLS is defined as RLS that is frequent and troublesome enough to require daily treatment. Treatment options include: Nonpharmacologic therapy. Dopamine agonists Gabapentin. Low potency opioids or opioid agonists

Refractory RLS — Refractory RLS is defined as daily RLS treated with a dopamine agonist but with a poor response. The response is considered poor if there is an inadequate initial response despite adequate doses of medication or if the response becomes inadequate with time despite increasing doses. Augmentation (the onset of symptoms earlier in the day or extension of symptoms to arms or trunk) that is not controllable with additional earlier doses of the drug also qualifies RLS as refractory, as does intolerable adverse effects.

Referral to a specialist for RLS management should be considered for these patients. Four different pharmacologic treatment approaches are recommended: Change to gabapentin. Change to a different dopamine agonist. Add a second agent such as gabapentin, a benzodiazepine, or an opioid. Change to a high potency opioid or tramadol

Levodopa — The efficacy of L-dopa has been demonstrated in small randomized trials: In one double-blind, crossover study lasting four weeks, an evening dose of L-dopa 200 mg with benserazide (a peripheral dopa decarboxylase inhibitor used in Europe and equivalent to carbidopa used in the United States) significantly reduced the frequency of awakenings and duration of waking periods in 13 patients with RLS compared with placebo. In a second double-blind trial in which 20 patients with RLS were given L-dopa, 50 to 200 mg daily with benserazide or lactose on alternate days for an unspecified period of time, 17 patients reported complete relief with L-dopa. L-dopa was also effective in a double-blind study of uremic RLS. Tachyphylaxis (a decreasing response with continual treatment) to L-dopa has been reported, but in one long-term study lasting two years, 26 of 30 patients initially responding to L-dopa maintained a good response.

Three different problems may occur with patients on L-dopa therapy: Augmentation is a worsening of RLS symptoms earlier in the day after an evening dose of medication, including earlier onset of symptoms, increased intensity of symptoms, or spread of symptoms to the arms. It occurs in up to 70 percent of patients taking daily L-dopa. Rebound is the recurrence of RLS early in the morning. It occurs in 20 to 35 percent of patients taking L-dopa. Recurrence of symptoms in the second half of the night may occur.

An expert advisory panel on RLS recommends L-dopa only for intermittent RLS. The risk of augmentation may be decreased with intermittent use of dopaminergic drugs, although this has not been well studied. The expert panel recommends discontinuation of the drug if augmentation occurs. Controlled release L-dopa (CR) combined with standard L-dopa may help sleep quality during the second half of the night for patients who experience recurrence.

For best absorption, L-dopa should not be taken with high protein foods. Suggested agents include: Carbidopa/levodopa 25 mg/100 mg, one-half or one tablet, can be used for intermittent RLS that occurs during the evening, at bedtime, or on waking during the night. Additionally, this medication may be helpful for RLS associated with specific triggers including lengthy travel by auto or airline, or spectator events with prolonged sitting. Controlled release (CR) carbidopa/levodopa, 25 mg/100 mg before bed may be helpful for RLS that awakens the patient during the night.

Dopamine agonists — Dopamine agonists belong to a class of drugs that directly stimulate dopamine receptors and have a longer half-life (four to six hours) than levodopa (90 minutes). They are generally superior to levodopa for the treatment of daily RLS.

Pramipexole and ropinirole — The non-ergot dopamine agonists, pramipexole and ropinirole, are less likely to cause side effects compared with other dopamine agonists. These agents are considered to be the drugs of choice in most patients with daily RLS. They may also be helpful in patients with intermittent RLS. Both ropinirole and pramipexole are approved by the the United States Food and Drug Administration (FDA) for the treatment of RLS. The benefit of ropinirole for RLS was demonstrated in three similar randomized controlled trials. As an example, one of these trials enrolled 284 patients with moderate to severe RLS; improvement on the primary outcome measure of symptoms as assessed by the International Restless Legs Scale (IRLS) was significantly greater for ropinirole treatment (average dose 1.13 mg/day) than for placebo. The benefit of pramipexole was demonstrated in a 12-week double-blind trial that assigned 344 patients with moderate or severe RLS to one of three pramipexole regimens (0.25, 0.50, and 0.75 mg/daily) or to placebo. The initial pramipexole dose was 0.125 mg/daily for the first week, and the dose was titrated up to the assigned dose over the first three weeks of the trial. All three doses of pramipexole were superior to placebo on the two primary outcome measures (patient-rated symptom severity improvement measured by the IRLS, and clinician-rated improvement). Pramipexole was well tolerated; nausea and somnolence were the most frequent side effects associated with its use.

An earlier, small (n = 10) double-blind, crossover study lasting two months found that pramipexole markedly reduced RLS symptoms for 24 hours following a single evening dose and also reduced the frequency of periodic leg movements of sleep (PLMS) to normal levels. The initial pramipexole dose was 0.375 mg/day, and the dose was increased to 0.75 mg after one week and 1.5 mg/day after the second week. Seven of the 10 patients who participated in this study were followed for an additional mean period of 7.8 months with no decrease in benefit, increase in dose requirement, or shift of symptoms to daytime. The optimal pramipexole dosage was 0.25 mg for one patient, 0.5 mg for five patients and 0.75 mg for one patient.

Adverse effects with pramipexole and ropinirole have been mild and transient and limited to nausea, lightheadedness, and fatigue; these usually resolve within 10 to 14 days. Less frequent side effects include nasal stuffiness, constipation, insomnia, and leg edema; these are reversible if the medication is stopped. Sudden, unexpected sleep attacks, as described in patients with Parkinson's disease (PD) on higher doses of pramipexole, do not appear to occur in patients on low-dose pramipexole (mean of 0.37 mg/day) for RLS.

The onset of action for dopamine agonists is typically 90 to 120 minutes after intake. Therefore, these medications should be started two hours before RLS symptoms start. The recommended starting doses are: Pramipexole 0.125 mg once daily. The dose may be increased by 0.125 mg every two to three days until relief is obtained. In a clinical trial, all three doses of pramipexole (0.25, 0.50, and 0.75 mg/daily) were equally effective, and some patients responded to the initial dose of 0.125 mg daily. However, side effects were more common with the 0.50 mg and 0.75 mg daily. Therefore, it is expected that 0.25 mg daily has the best therapeutic margin. Most patients require 0.5 mg or less, but doses up to 2 mg may be needed. Ropinirole 0.25 mg once daily. The dose may be increased by 0.25 mg every two to three days until relief is obtained. Most patients require 2 mg or less, but doses up to 4 mg or higher may be needed.

Augmentation is less common with these drugs than with levodopa, but it has been reported in up to one-third of patients taking pramipexole for two years. The risk of augmentation with ropinirole is unknown. In one study, augmentation was significantly more common in patients with a family history of restless legs syndrome and in those who had no evidence of neuropathy on electromyography or nerve conduction studies.

Unlike levodopa, additional doses of a dopamine agonist earlier in the day can often reduce the risk and intensity of augmentation. The development of augmentation with one dopamine agonist does not necessarily predict augmentation with a different agonist. In addition, ropinirole and pramipexole can be substituted for one another. However, therapy should be changed to a different class of medications if augmentation develops with a second dopamine agonist.

Other dopamine agonists — Other dopamine agonists may also be effective in RLS but are used infrequently. Cabergoline (0.5 to 4 mg as an evening dose) is a dopamine agonist with a long half-life (65 hours) that may benefit patients who experience rebound symptoms with shorter half-life medications.

In a five-week randomized clinical trial of 85 patients with RLS, cabergoline at doses of 0.5, 1.0, and 2.0 mg/day (taken at least three hours before bedtime) was effective compared with placebo in treating RLS symptoms and sleep disturbance. In an open-label, one-year extension, a mean daily dose of cabergoline 2.2 mg/day was required to maintain improvement of RLS symptoms. A later five-week randomized controlled trial of 40 patients with moderate to severe RLS found that cabergoline 2 mg once daily in the evening was significantly more effective than placebo by polysomnography for reducing PLMS-related arousal.

Side effects in the first trial including nausea, constipation, headache, dizziness, fatigue, and drowsiness caused treatment discontinuation in 11 patients (13 percent). Augmentation (the onset of symptoms earlier in the day or extension of symptoms to arms or trunk) was observed in six treated patients (9 percent) in the long-term phase. In a study of patients with Parkinson's disease, high cumulative dose and long-term treatment with cabergoline was associated with an increased risk of cardiac valvulopathy.

Cabergoline is approved by the FDA only for the treatment of hyperprolactinemic disorders, and its use for RLS is off-label. Several controlled clinical trials have demonstrated that pergolide is effective for relieving symptoms of RLS. However, pergolide was associated with restrictive cardiac valve disease in 33 percent of patients when used in high doses to treat PD, and this risk would likely preclude its use in RLS. In addition, pergolide is associated with an increased risk of pleuropulmonary fibrosis.

Benzodiazepines — Benzodiazepines are useful in mild cases of RLS, particularly in younger patients. Diazepam has been used in RLS for many years, although there have been no controlled trials. In a small randomized, double-blind, crossover trial, treatment with clonazepam, 1.0 mg daily, was superior to placebo in six patients with RLS. In an open trial, 14 of 15 patients with RLS due to uremia responded to clonazepam, 1 to 2 mg daily.

An RLS expert panel algorithm recommends benzodiazepines or benzodiazepine agonists for intermittent RLS, especially if the patient has another cause of poor sleep in addition to RLS. Short-acting agents can be helpful for sleep onset insomnia caused by RLS; these agents include: Triazolam 0.125 to 0.5 mg Zolpidem 5 to 10 mg Zaleplon 5 to 10 mg. For RLS that awakens the patient later in the night, the expert panel recommends intermediate acting agents such as temazepam 15 to 30 mg.

Although most trials have been performed with clonazepam, its long duration of action may result in more adverse effects, such as nocturnal unsteadiness and drowsiness or cognitive impairment in the morning. However, at least one study has shown a low rate of adverse effects from clonazepam in elderly patients. Long-term maintenance treatment with benzodiazepines is limited by tolerance in many patients, but abuse appears to be low in this disorder.

Opioids — A variety of opioids, including codeine, methadone, and propoxyphene, have been reported to be helpful for RLS in uncontrolled trials. In a double-blind, four week crossover study of 11 patients, oxycodone at a mean dose 15.9 mg/day, was superior to placebo with regard to number of sleep arousals, PLMS frequency, and sleep efficiency. One long-term retrospective study of 113 patients treated with opioids has shown persistent benefit, but a small number of patients developed sleep apnea. Similar to benzodiazepines, opioid abuse potential is low in patients with RLS. Nevertheless, we typically restrict use to patients with more severe symptoms who fail to respond to benzodiazepines or dopaminergic drugs.

The expert RLS panel recommends more liberal use of low potency opioids or opioid agonists for intermittent RLS and as an alternative for daily RLS. These drugs are usually taken before bed and include: Propoxyphene napsylate 100 to 200 mg. Propoxyphene hydrochloride 65 to 130 mg. Codeine 30 to 60 mg, usually in combination preparations with acetaminophen. Tramadol 50 to 100 mg before bed or during the night.

For patients with refractory RLS, the same panel recommends change to a high potency opioid or tramadol as one of four different management approaches. High potency opioids may be effective in the management of RLS, and escalation of doses and dependence are uncommon in the absence of a history of substance abuse. High potency opioids may be used one to three times a day depending on timing of symptoms; agents include: Oxycodone 5 to 15 mg. Hydrocodone 5 to 15 mg. Methadone 5 to 10 mg. Tramadol 50 to 100 mg

Gabapentin — Gabapentin is an alternative choice for patients with daily RLS. Gabapentin, mean dose 733 mg/day, was effective in a four-week open-label study of nine patients with idiopathic RLS, and at a much higher mean dose (1,855 mg) in a randomized, placebo-controlled trial of 22 patients. Among dialysis patients, gabapentin (200 to 300 mg three times weekly after hemodialysis) was effective in a 12-week, double-blind crossover study.

Although the trial data above suggest that mean doses of 1800 mg a day are needed, many patients appear to benefit from a lower dose. The expert panel suggests beginning treatment with 100 to 300 mg per dose because of the tendency of the drug to cause somnolence and gait unsteadiness.

Gabapentin may be particularly useful in patients whose symptoms are less intense and in patients in whom RLS is perceived as painful. It may also be useful when RLS occurs in the setting of a painful peripheral neuropathy or an unrelated chronic pain syndrome. Gabapentin may also be considered in the treatment of RLS in association with neurodegenerative disorders, such as PD or dementia.

Other drugs — Other drugs that may be useful in RLS but have been reported in only a small number of mainly open studies include carbamazepine (mean dose 236 mg/day), clonidine (0.05 mg/day), propranolol (40 to 120 mg/day), and amantadine (up to 300 mg/day).

Nonpharmacologic therapy — Nonpharmacologic therapy is recommended by the expert panel for patients with intermittent or daily RLS. The therapy involves the following components: Iron replacement therapy. Mental alerting activities. Avoidance of aggravating factors.

Iron replacement — Administer iron replacement if the serum ferritin level is lower than 45 to 50 mcg/mL. The suggested regimen is ferrous sulfate (325 mg three times a day) in combination with vitamin C (100 to 200 mg) with each dose of ferrous sulfate to enhance absorption. Ferritin levels should be checked after three to four months of therapy and then every three to six months until the serum ferritin level is greater than 50 mcg/mL and iron saturation is greater than 20 percent.

Mental alerting activities — Mental alerting activities, such as video games or crossword puzzles, may reduce symptoms at times of boredom.

Aggravating drugs — Caffeine, nicotine, and alcohol may aggravate RLS symptoms. A trial of abstinence is reasonable in many patients.

Antidepressants, neuroleptic agents, dopamine-blocking antiemetics such as metoclopramide, or sedating antihistamines (including those found in nonprescription medications) may contribute to RLS symptoms. However, discontinuation may not be possible without causing patient harm. If antidepressants are necessary, the symptoms of secondary RLS can usually be treated in the same way as primary RLS. Bupropion is an alternative antidepressant that may be less likely to induce or worsen RLS.

RECOMMENDATIONS — The following sequence of pharmacologic therapy in patients with RLS is recommended: Begin with pramipexole (0.125 mg) or ropinirole (1.0 mg) approximately one hour before the usual time of symptom onset; the dose is titrated upward according to response. These drugs are effective and well tolerated by most patients. Pramipexole doses above 0.75 mg/day and ropinirole doses above 4 mg/day are of unproven benefit. An alternative is levodopa/carbidopa, 25/100 mg given before symptom onset in patients unable to tolerate a dopamine agonist. In some cases a controlled release formulation of levodopa may be necessary to get the patient through the night or to avoid rebound daytime symptoms. Clonazepam, 0.5 mg before sleep can be used alone or as adjunctive treatment. Gabapentin may be a good additional option because of its relative lack of adverse and sedative effects. Opioids should be reserved as a final option, either alone or in conjunction with other medications. A trial of oral iron therapy is indicated for all patients with RLS, particularly premenopausal women, as some patients without evidence of iron deficiency may still respond favorably.

Combination therapy is often useful, and periodic changes in choice of medication to deal with tolerance are frequently necessary. In some patients, additional doses of medication may be necessary during the day to manage diurnal symptoms.

Cost of medications — The treatment plan should include consideration of prescription costs. Typical monthly costs will be highest for cabergoline; intermediate for pramipexole, ropinirole, and gabapentin; lower for levodopa/carbidopa; and lowest for carbamazepine.

Treatment in pregnancy — The treatment of RLS in pregnant women is complicated by the fact that nearly all of the drugs used for RLS are considered pregnancy risk factor C or D. The one major exception is pergolide, which has other serious side effects. The following options are therefore suggested: Opioids can be used in the second or third trimester if treatment of RLS symptoms is necessary. The lowest effective dose of opioids should be used for the shortest period of time to reduce the risk of neonatal withdrawal. In most patients, RLS begins in the third trimester of pregnancy and disappears toward the end of the pregnancy. Oral iron supplements should be maintained since it appears that women with RLS in pregnancy have lower hemoglobin and mean corpuscular volume than healthy subjects. In exceptional cases, parenteral iron therapy is a consideration. Nonpharmacologic therapy should be tried if RLS symptoms occur earlier than the third trimester. Leg stretching before sleep and use of elastic stockings may also be helpful. Elastic stockings or sclerotherapy should be considered if varicose veins are prominent. The patient should be counseled regarding the benign nature of the problem.

Most drugs for RLS are not recommended or are not rated by the American Academy of Pediatrics (AAP) for use during breast feeding. Codeine and propoxyphene are considered compatible with breast feeding but should be used with caution.