ADVERTISEMENT

Issue Navigator

Volume 12 No. 12
Earn CME
Accepted Papers
Classifieds





Scientific Investigations

Depth and Distribution of Symptoms in Restless Legs Syndrome/ Willis-Ekbom Disease

Paul Yeh, MD, MPH1; William G. Ondo, MD2; Daniel L. Picchietti, MD3; J. Steven Poceta, MD4; Richard P. Allen, PhD5; Charles R. Davies, MD, PhD3; Lily Wang, PhD6; Yaping Shi, PhD6; Kanika Bagai, MD7; Arthur S. Walters, MD7
1Department of Psychiatry, University of Texas Health Science Center at San Antonio, TX; 2Department of Neurology, Methodist Neurological Institute, Houston, TX; 3Carle Neuroscience Institute, University of Illinois College of Medicine at Urbana-Champaign, Urbana, IL; 4Viterbi Family Sleep Center, Scripps Clinic, La Jolla, CA; 5RLS Center, Johns Hopkins University School of Medicine, Baltimore, MD; 6Department of Biostatistics, Vanderbilt School of Medicine, Nashville, TN; 7Department of Neurology, Vanderbilt University Medical Center, Nashville, TN

ABSTRACT

Study Objectives:

To determine the depth and distribution of sensory discomfort in idiopathic restless legs syndrome/Willis-Ekbom disease (RLS) and RLS concurrent with other leg conditions, specifically peripheral neuropathy, sciatica, leg cramps, and arthritis.

Methods:

RLS subjects (n = 122) were divided into 71 idiopathic RLS and 51 RLS-C, or Comorbid, groups. All subjects were examined by an RLS expert, answered standardized RLS questionnaires, and received a body diagram to draw the location and depth of their symptoms.

Results:

Age was 63.04 ± 12.84 years, with 77 females and 45 males. All patients had lower limb involvement and 43/122 (35.25%) also had upper limb involvement. Of the 122 subjects, 42.62% felt that the RLS discomfort was only deep, 9.84% felt that the discomfort was only superficial, and 47.54% felt both superficial and deep discomfort. There were no defining characteristics in depth or distribution of RLS sensations that differentiated those patients with idiopathic RLS from those patients with RLS associated with other comorbid leg conditions. The sensation of arthritis was felt almost exclusively in the joints and not in the four quadrants of the leg, whereas the exact opposite was true of RLS sensations.

Conclusions:

Depth and distribution cannot be used as a discriminative mechanism to separate out idiopathic RLS from RLS comorbid with other leg conditions. Although seen in clinical practice, the total absence of patients with non-painful RLS only in the joints in the current study attests to the rarity of this presentation and raises the possibility of misdiagnosis under these circumstances. We recommend that such patients not be admitted to genetic or epidemiological studies.

Citation:

Yeh P, Ondo WG, Picchietti DL, Poceta JS, Allen RP, Davies CR, Wang L, Shi Y, Bagai K, Walters AS. Depth and distribution of symptoms in restless legs syndrome/Willis-Ekbom disease. J Clin Sleep Med 2016;12(12):1669–1680.


INTRODUCTION

Restless legs syndrome, also known as Willis-Ekbom disease (RLS) is a sensorimotor disorder that affects an estimated 5% to 15% of the adult population with 2% to 3% of adults having symptoms severe enough to impact quality of life and require medical intervention.1 RLS is diagnosed by the presence of an urge to move the limbs (typically the legs), which is usually but not necessarily accompanied by uncomfortable and unpleasant sensations. This urge to move will develop and intensify in periods of inactivity, worsen in the evening or night, is alleviated by some form of movement (such as stretching one's legs or walking), and is not solely accounted for by another medical or behavioral diagnosis (such as arthritis or leg cramps).2,3 Typically, patients with RLS report abnormal and unpleasant sensations that are deep and bilateral.46 However, some studies have found that RLS patients can diverge from the depth and distribution of the classical presentation. Clinical progression of RLS typically begins in the calves or thighs and may spread to other parts of the lower limbs, such as the foot.79 Sensations have also been noted in the upper limbs in rare cases as the initial manifestation of RLS,10,11 but usually RLS in the upper extremities appears later in the course of typical patients who had RLS only in the legs.1214 Other studies report RLS sensations in the abdominal region,15 groin and genital areas,1622 buttocks,23,24 back,25 chest,25 neck,26 and face.26,27

BRIEF SUMMARY

Current Knowledge/Study Rationale: Although previous studies have looked at the depth and distribution of idiopathic RLS, there have been no prior studies that have compared idiopathic RLS to patients with RLS and other conditions that cause lower extremity sensations. Therefore, we compared the depth and distribution of RLS sensations in patients with idiopathic RLS versus RLS patients with an accompanying comorbid leg condition; additionally, we compared the depth and distribution of the sensations of the comorbid leg conditions themselves against idiopathic RLS sensations.

Study Impact: Neither the depth nor the distribution of RLS sensory discomfort can be used to discriminate between idiopathic RLS and RLS associated with other causes of leg discomfort. However, arthritis sensations were almost exclusively found in isolated joints, which was never true of RLS sensations in general; thus, patients who meet all 4 criteria for RLS but have only isolated joint involvement should be suspected of not having true RLS and therefore should not be enrolled in RLS genetic and epidemiological studies where only idiopathic RLS cases are desired.

A major question in RLS research is whether the depth and the topographic distribution in which RLS sensory discomfort is felt can discriminate between idiopathic RLS and RLS associated with other leg conditions such as peripheral neuropathy, sciatica, leg cramps and arthritis. Only two studies have specifically evaluated the topography of RLS patients' sensory symptoms.4,28 The study by Karroum et al. analyzed the topography of RLS symptoms in idiopathic RLS patients being treated with medications,28 and the study by Koo et al. analyzed drug-naÏve RLS patients.4 However, these studies analyzed the distribution of RLS symptoms solely in patients with idiopathic RLS. Neither the Karroum nor the Koo papers analyzed RLS patients with other conditions that affect the lower extremities, such as neuropathy, sciatica, leg cramps and arthritis. Thus, this study sought to determine if RLS patients with a comorbid condition of their lower extremities have a different topography and depth of RLS sensations than patients with idiopathic RLS. We also sought to examine the depth and distribution of any accompanying neuropathy, sciatica, leg cramps and arthritis.

The issues raised in this study would be particularly important in epidemiology studies where non-clinicians interview the general public in telephone surveys where only patients with primary RLS are desired. If the study shows that the depth and distribution of idiopathic RLS symptoms is different than the depth and distribution of RLS symptoms of patients with RLS with another leg disorder or from the depth and distribution of comorbid leg conditions (neuropathy, radiculopathy, leg cramp, and arthritis sensations) then telephone interviewers in epidemiology studies could employ questions about depth and distribution in interviews of patients as a way of excluding patients who do not have RLS.

METHODS

Inclusion/Exclusion Criteria

Patients were evaluated by RLS experts to verify the diagnosis with face-to-face assessment and neurological examination. All patients met the 2003 International Restless Legs Syndrome Study Group (IRLSSG) criteria for RLS: An urge to move the limbs (typically the legs) which is usually accompanied by uncomfortable and unpleasant sensations; this urge to move will develop and intensify in periods of inactivity; worsens at night; and is alleviated by some form of movement (such as stretching one's legs or walking).2Peripheral neuropathy was diagnosed by a stocking distribution of pinprick and/or light-touch decrement and electromyogram (EMG) and nerve conduction studies (NCVs). Nocturnal leg cramps were diagnosed by a current history of intermittent leg spasms in which the muscle tensed (Charlie horse) that was relieved by placing brief pressure on the limb. Sciatica was diagnosed by a current history of back pain with radiation down the lower extremity and consistent EMG/NCV. Arthritis was determined by a previous history of arthritis from another medical professional and by discomfort in the joints that was painful and when the patient could distinguish between the symptoms of arthritis and RLS.

Patients were included if they had leg sensations at least once in the week prior to entry to the study. Additionally, the survey of each patient's sensations assessed only those experienced in the past week prior to the date of the telephone or face-to-face assessment. This enhances the ability for patients to accurately and reliably report their sensation's depth and distribution and minimizes the possible confounding effect of recall bias. All patients included said that they could readily distinguish any neuropathy, radiculopathy, leg cramp, or arthritis symptoms from their RLS symptoms.

Interview

We partnered with colleagues and RLS researchers from six different American institutions: Vanderbilt University Medical Center (VUMC), Scripps Clinic (SC), Carle Foundation Hospital (CFH), Baylor College of Medicine (BCM), the University of Texas at Houston (UTH), and the University of Texas at Dallas (UTD). Together we conducted interviews with 122 patients. In total, we had 40 patients from VUMC, 25 from SC, 19 from CFH, 14 from BCM, 20 from UTH, and 4 from UTD. Seventy-one subjects had idiopathic RLS and 51 had RLS in conjunction with other comorbid conditions that affect the lower extremities, including neuropathy, osteoarthritis, leg cramps, or sciatica.

All patients were then given an adapted Lund-Browder body burn diagram29 (see Figures 13). The patients were asked to circle the areas affected by their RLS symptoms to determine their distribution. Moreover, the patients also indicated on these diagrams the depth of the RLS symptoms by marking any purely superficial sensations with horizontal lines, any purely deep sensations with vertical lines, and deep plus superficial sensations with a combination vertical and horizontal line (see Figures 1Figure 23). All body diagrams were then scanned into the computer. The percent body surface area calculated for each patient was estimated based upon the Lund-Browder % BSA assigned to each body part. Using the BSA dictated by the Lund-Browder diagram, we tabulated the % BSA affected by RLS symptoms in the “lower extremity” (defined as everything below the waist, excluding the buttocks) and the % BSA affected by RLS symptoms in the “upper extremity” (defined as solely the upper extremities below the shoulders including hands and fingers). For RLS patients with a comorbid condition that affects the legs in addition to RLS, separate Lund-Browder body diagrams were filled out so that a separate analysis of the comorbid condition's depth and distribution of sensory discomfort could be made and compared to the depth and distribution of RLS sensory discomfort. At the time of administration of the Lund-Browder body diagram, we also performed the International RLS Rating Scale (IRLS)30 to classify the severity of each patient's RLS symptoms over the past week.

Distribution and depth diagram for RLS and any associated comorbid leg conditions.

Adapted from the Lund-Browder body burn diagram. Participants were asked to circle and then shade in the areas that were affected by RLS sensation within the past week. The numbers on the diagram corresponds to the percent of the total body surface area (%BSA) of that corresponding body part. Area “A,” designated as the front or the back of the head is defined as 3.5% of BSA. Area “B,” designated as the front or rear (thigh) of one upper leg is defined as 4.75% of BSA. Area “C,” designated as the shin or calf of one lower leg is defined as 3.5% of BSA. For RLS patients with a comorbid leg condition, a separate Lund-Browder diagram, identical to the one presented in Figure 1 was filled out for the comorbid condition.

jcsm.12.12.1669a.jpg

jcsm.12.12.1669a.jpg
Figure 1

Distribution and depth diagram for RLS and any associated comorbid leg conditions.

(more ...)

Example of filled RLS depth and distribution diagram.

Adapted from the Lund-Browder body burn diagram. Participants were asked to circle and then shade in the areas that were affected by RLS sensation within the past week. The numbers on the diagram corresponds to the percent of the total body surface area (%BSA) of that corresponding body part. Area “A,” designated as the front or the back of the head is defined as 3.5% of BSA. Area “B,” designated as the front or rear (thigh) of one upper leg is defined as 4.75% of BSA. Area “C,” designated as the shin or calf of one lower leg is defined as 3.5% of BSA. For RLS patients with a comorbid leg condition, a separate Lund-Browder diagram, identical to the one presented in Figure 1 was filled out for the comorbid condition.

jcsm.12.12.1669b.jpg

jcsm.12.12.1669b.jpg
Figure 2

Example of filled RLS depth and distribution diagram.

(more ...)

Example of filled non-RLS depth and distribution diagram.

Adapted from the Lund-Browder body burn diagram. Participants were asked to circle and then shade in the areas that were affected by RLS sensation within the past week. The numbers on the diagram corresponds to the percent of the total body surface area (%BSA) of that corresponding body part. Area “A,” designated as the front or the back of the head is defined as 3.5% of BSA. Area “B,” designated as the front or rear (thigh) of one upper leg is defined as 4.75% of BSA. Area “C,” designated as the shin or calf of one lower leg is defined as 3.5% of BSA. For RLS patients with a comorbid leg condition, a separate Lund-Browder diagram, identical to the one presented in Figure 1 was filled out for the comorbid condition.

jcsm.12.12.1669c.jpg

jcsm.12.12.1669c.jpg
Figure 3

Example of filled non-RLS depth and distribution diagram.

(more ...)

We also interviewed each patient, either by telephone or in person and administered other surveys including an RLS Health Information and Demographic Questionnaire, the Johns Hopkins telephone diagnostic interview31 and the International RLS Severity Scale (IRLS) To assist in the data-collection process, all of the interview forms, traditionally done in paper format, were programmed into an electronic survey form via Vanderbilt University's Research Electronic Data Capture System (REDCap).32

Data Analysis

The data were split into demographic and clinical subgroups (Table 1). We dichotomized patients based on demographic variables such as medication status, family history of RLS, severity of RLS, age of onset, and gender. For medication status, we compared RLS patients who were medication-naÏve (de novo patients) with those currently on medication for RLS such as dopamine drugs. Family history was contingent upon knowledge of a first-degree relative with RLS. For severity, we segregated the patients based upon the four IRLS severity categories into a combined mild and moderate group (IRLS scores ≤ 20) versus a combined severe and very severe group (scores ≥ 21).30 Age of onset was considered early if less than 45 and late-onset if over 45.33 We also asked patients to report their estimated time of onset of their RLS symptoms as well as the estimated proportion of the time in a typical day their symptoms are felt lying down versus sitting.

Demographics and clinical characteristics of the 122 patients with restless legs syndrome.

jcsm.12.12.1669.t01.jpg

table icon
Table 1

Demographics and clinical characteristics of the 122 patients with restless legs syndrome.

(more ...)

Patients were split into subgroups according to their diagnosis. Patients with only RLS formed the idiopathic RLS group. Patients who had RLS and a comorbid health condition that affected the lower extremities were assigned to the RLSC (Comorbid group). In RLS-comorbid patients, the depth and distribution of their RLS symptoms and their comorbid lower extremity conditions were separately documented and analyzed.

In addition, we had an additional category of patients, “RLS plus other non-diagnosed sensory symptoms” which was defined as those subjects who reported an undiagnosed sensory symptom such as numbness or vague leg pain during ambulation which did not meet specific criteria for RLS, leg cramps, or arthritis, and whose neurological examination and testing did not provide evidence of a peripheral neuropathy or radiculopathy.

We analyzed both the distribution and the depth of the RLS sensations within the demographic and clinically based subgroups as described. To assess the depth of RLS, we tabulated the number of patients within each subgroup who reported either superficial sensations only (category 1), deep sensations only (category 2), or both superficial and deep sensations (category 3). We further split this latter category by tabulating the patients who reported having both superficial and deep sensation in the same part of the body (category 3a) and those having a deep sensation in one body part and a superficial in another part of the body (category 3b). Lastly, we calculated within each subgroup how many individuals had any form of superficial sensations (thus categories 1 + 3) and how many individuals had any form of deep sensations in their body (categories 2 + 3).

To measure the distribution of symptoms within our subgroups, the body diagrams were used to determine how many patients in each subgroup felt sensations in different parts of the body. The lower extremity was split into four generic “quadrants” so as to enhance any potential distribution differences seen between various subgroups: The anterior thigh region, posterior thigh region, anterior lower leg region, and the calves. Isolated hip, knee, ankle, and foot joints were tabulated separately and not included in the four leg quadrants. Isolated shoulder, elbow, and wrist joint sensations were documented as well. Lastly, any sensations felt in the trunk or in other non-extremity body parts (back, shoulder, neck, face, and buttocks) were tabulated.

Statistical Analysis

We utilized the SAS analysis software program to calculate statistics for the patient data. Due to the multiple comparisons we used across our various demographic and clinical subgroups, we used the “Q-value” false discovery rate (FDR) as the metric to analyze for statistical differences between the depth and distribution of the various demographic and clinical-based subgroups. The FDR statistical test produces a q-value in which we established a FDR q-value of < 0.05 as the threshold for significance. For our statistical comparison, we considered the reference group to be idiopathic RLS patients against which the other patient subgroups were compared. We did not power the study based on any assumptions.

RESULTS

Sample Characteristics

Table 1 highlights the demographic characteristics of the 122 patients with RLS. Mean age was 63.04 years ± 12.84 with 77 females. Ethnicity, age of onset, duration of RLS symptoms, percentage of patients with early-onset RLS, percent of patients with positive RLS family history, RLS severity, time of usual daily onset of RLS symptoms, distribution of RLS symptoms, and effect of lying vs. sitting are also listed. Of note, patients experienced exacerbation of RLS symptoms on a typical day for a similar amount of time sitting (45.11% of time) as lying down (54.89% of time), with a tendency toward lying down perhaps due to the later time of onset of RLS symptoms (mean 4:38 pm). Of particular interest, patients experienced RLS sensations over 18.59% ± 13.71% of the total body surface area (BSA). All patients experienced discomfort in the lower limbs with 39.46% ± 26.16% of the surface area of the lower limbs affected. In the 43 patients who reported additional upper limb involvement, 27.63% ± 21.52% of the surface area of the upper limbs was affected.

Table 2 illustrates the depth of sensations, and Table 3 illustrates the distribution of sensation felt by RLS patients as distributed across the demographic subgroups. There were no striking differences in the depth or distribution of RLS sensations between the demographic subgroups that would be clinically useful.

Depth of RLS patients' sensations by demographics.

jcsm.12.12.1669.t02.jpg

table icon
Table 2

Depth of RLS patients' sensations by demographics.

(more ...)

Distribution of RLS patients' sensations by demographics.

jcsm.12.12.1669.t03.jpg

table icon
Table 3

Distribution of RLS patients' sensations by demographics.

(more ...)

Table 4 shows the depth of RLS sensations and the depth of sensations of other associated conditions (neuropathy, arthritis, leg cramps, sciatica, and non-diagnosed sensory symptoms). There were 71 patients with idiopathic RLS and 51 patients with RLS that was associated with one or more comorbid disorders. There were no significant differences in the depth of RLS leg discomfort when accompanied by any of the comorbid conditions as analyzed separately compared to the depth of the leg discomfort in idiopathic RLS. With the exception of leg cramps, there were also no significant differences between the depth of the discomfort from any of the comorbid condition themselves as analyzed separately versus the depth of the RLS leg discomfort in idiopathic RLS. When leg cramps accompanied RLS, the leg cramps sensations were more commonly felt as uniquely deep (84.62% of RLS patients with leg cramps) than the sensations of idiopathic RLS (35.21%). However, when idiopathic RLS patients who have both deep and superficial RLS sensations in the same patient are combined with those who have only deep sensations, the percentage of idiopathic RLS patients who have some deep component to their RLS discomfort rises to 90.14%. A substantial proportion of all RLS patient subgroups also felt their RLS sensations as being deep and therefore the depth of leg cramp sensations is probably not a clinically useful way to distinguish leg cramps from RLS.

Depth of RLS patients' sensations with associated conditions.

jcsm.12.12.1669.t04.jpg

table icon
Table 4

Depth of RLS patients' sensations with associated conditions.

(more ...)

Table 5 demonstrates the distribution of RLS sensations as well as the distribution of the other associated conditions. There were no significant differences in the distribution of RLS leg discomfort when accompanied by any of the comorbid conditions as analyzed separately compared to the distribution of the leg discomfort in idiopathic RLS. With the exception of arthritis, there were also no significant differences between the distribution of the discomfort from any of the comorbid conditions themselves as analyzed separately versus the distribution of the RLS leg discomfort in idiopathic RLS. Of note, the sensations of arthritis were felt almost exclusively in the joints and not in the anterior thigh, posterior thigh, anterior lower leg, or calves whereas the exact opposite was true of RLS sensations. However, RLS sensations could commonly cross joint spaces. In addition, patients who had RLS and accompanying other non-diagnosed sensory symptoms felt their other non-diagnosed sensory symptoms more commonly in the isolated hips, knees, and the foot, similar to the accompanying arthritis sensations in RLS patients. These data suggest that these subjects may have undiagnosed arthritis.

Distribution of RLS patients' body part sensations with associated conditions.

jcsm.12.12.1669.t05.jpg

table icon
Table 5

Distribution of RLS patients' body part sensations with associated conditions.

(more ...)

Combined Analysis

In the RLS-Comorbid group, the depth and distribution of the RLS sensations are fairly similar for RLS patients with peripheral neuropathy (n = 11), leg cramps (n = 13), and sciatica (n = 5). For RLS-C patients the depth and distribution of the sensations of the accompanying peripheral neuropathy, leg cramps, and sciatica are also fairly similar (Table 4 and Table 5), Therefore, we performed an analysis of this combined comorbid group versus our 71 idiopathic RLS patients. We did four separate analyses. We first compared the depth of the RLS sensations in our combined comorbid group to the depth of the RLS sensations in the idiopathic group. We next compared the distribution of the RLS sensations in our combined comorbid group to the distribution of RLS sensations in the idiopathic group. We then compared the depth of the discomfort felt from any accompanying peripheral neuropathy, leg cramps, and sciatica in the combined comorbid group to the depth of the RLS sensations in the idiopathic group. Lastly we compared the distribution of the discomfort felt from any accompanying peripheral neuropathy, leg cramps, and sciatica to the distribution of the RLS sensations in the idiopathic group. There were no significant differences in any of these comparisons (data not shown). Thus depth and distribution do not allow us to distinguish RLS discomfort in idiopathic RLS from the RLS discomfort in peripheral neuropathy, leg cramps, and sciatica. In addition, depth and distribution do not allow us to distinguish RLS discomfort from the discomfort experienced from any accompanying peripheral neuropathy, leg cramps, and sciatica.

DISCUSSION

We studied the depth and distribution of sensory symptoms in 122 patients with RLS of which 71 had idiopathic RLS and 51 had RLS-associated neuropathy, arthritis, leg cramps, and lumbosacral radiculopathy (sciatica). Our study is unique in this regard and thus is different from two previous studies that examined these parameters only in RLS subjects without other comorbid leg conditions.4,28 All patients included said that they could readily distinguish any neuropathy, arthritis, leg cramp, or sciatic symptoms from their RLS symptoms. The most important finding is that the depth and distribution of RLS symptoms cannot be used as a way of discriminating between the sensations of idiopathic RLS and the sensations of RLS concurrent with other causes of leg discomfort. However, patients with arthritis and accompanying RLS experienced their arthritis almost exclusively in isolated joints whereas this was never true of the RLS sensations experienced in idiopathic RLS, nor the RLS sensations experienced in RLS associated with other causes of leg discomfort including arthritis. Although we have seen occasional patients in clinical practice with isolated non-painful joint involvement that we suspect had true RLS, the total absence of such patients in the current study raises the possibility of mis-diagnosis. Thus we would recommend that such patients not be enrolled in RLS genetic and epidemiological studies.

The inability of the current questionnaire and its accompanying diagrams to distinguish the symptoms of RLS from those of peripheral neuropathy, radiculopathy or leg cramps by means of depth and distribution alone, however, does not mean that these disorders cannot be distinguished by means of a classic history and neurological examination. It is well known that patients with a peripheral neuropathy usually experience constant numbness in the legs distally and a neurological examination will reveal a distal stocking distribution to light touch, pin prick and often proprioception with decreased ankle reflexes. Patients with radiculopathy will often have back pain with pain shooting down the legs and when the examiner raises the leg, additional pain will often be elicited. Patients with leg cramps will describe a brief spasm of the legs, which will be relieved by putting pressure on the leg.

Comparison with Previous Studies

In terms of body surface area coverage of RLS sensations, Koo et al.4 and Karroum et al.28 found that RLS sensations were felt, on average, over 20.8% (SD 21.0) and 12% (SD 13.2) of the entire body's total surface area (BSA), respectively, which is similar to our finding of 18.59% (SD 13.71, median 14.58) of the entire body. This suggests that, on average, about 15% to 20% of the entire body is affected by RLS sensations. All of our patients (n = 122) had RLS discomfort in the legs. Out of our 122 patients, 35.25% had upper extremity involvement, in between the 57% found in Karroum's study group28 and the 20% in Koo's group.4 All of these figures indicate that although leg involvement is virtually universal, arm discomfort is not rare. Previous studies suggest that higher severity of RLS has been associated with the spreading of sensations to the upper extremity.12,13,34 However, despite having a wide variance in upper extremity involvement between the three studies, the RLS severity of the populations in either this study, Koo, or Karroum did not significantly differ from each other, with a mean of 19.69 (SD 7.72), 25.9 (SD 7.6), and 21.3 (SD 9.8) on the standardized IRLS scale,31 respectively. Although not specifically analyzed in our study, previous studies have noted that arm involvement is more common in longer duration RLS.12,35

The RLS patients in our study have a deep component to their RLS sensations in 90.16% of all patients if one combines patients with purely deep sensations and those that have both deep and superficial sensations. This is in line with previous studies4,28 that also show that RLS manifests more commonly as a deeper sensation rather than a superficial one. Unlike earlier documentation that characterized RLS only as deep sensations,8 our study population indicates the true clinical presentation of RLS is broader. Pure superficial sensations were not rare among the RLS population, as 9.84% of the total patient population (Table 2) and 9.86% of the idiopathic RLS patients (Table 4) reported only superficial sensations. Patients with deep and superficial sensations account for the majority 57.38% of our total study population (Table 2) and 64.79% of our idiopathic RLS patients (Table 4). Therefore, superficial sensations, though not part of the classical description of RLS symptoms, should not exclude diagnostic consideration of RLS.

Possible Mechanisms for Depth and Distribution of RLS Symptoms

A hypothesis to explain the propensity of RLS to affect the lower limbs over other parts of the body (upper extremities and very rarely, neck/trunk areas) may be due to a recent discovery of susceptibility genes for RLS which are coincidentally involved with limb development.28,36,37 In addition, the dopaminergic diencephalospinal tract, which has been suspected as part of the pathological process for RLS, has its longest axons projecting to the dorsal horn of the spinal cord.38 These longer axons are more susceptible to stress, thus suggesting any damage to the tracts would occur in a caudal to rostral direction, as degeneration would first affect the longest axons and then move upward toward shorter axons.38 RLS patients also have abnormal activation of the gastrocnemius muscles, found in the upper calf, during walking when compared to non-RLS patients thus suggesting that RLS patients may have impaired supraspinal dopaminergic control over the muscle.39 This may be related to the observation that RLS patients have some impairment of the inhibitory descending dopaminergic pathway stemming from the A11 cell group in the periventricular posterior hypothalamus.40 Without the proper inhibitory actions of the dopaminergic pathway, consistent activation of the leg muscles may lead to high threshold muscle afferents in the legs, causing uncomfortable sensations and the need to move to relieve those sensations.28,38

Limitations of the Study

This study is a cross-sectional rather than a longitudinal study. As such, this study only takes a snapshot of our patient's RLS sensations at the exact time of the patient interview. Thus, the progression of symptoms either for idiopathic RLS or RLS-C patients was not tracked in this study. Future studies should take a longitudinal approach to RLS sensations to better examine the sequential progression of RLS sensations and determine if sensations in certain areas are precursors or predictors for spread to other areas of the body. Moreover, a potentially valuable metric that may help discriminate between idiopathic RLS from RLS-C patients may be to determine if there is a significant difference between the two groups in the interval between time of sleep onset and time of RLS symptom onset. Though our study asked participants to estimate their time of RLS symptom onset, we did not ask for the estimated interval between sleep and symptom onset.

Additionally, there was a small sample size for some of our subgroups with accompanying comorbid leg conditions other than RLS. Correcting for multiple comparisons minimized power to detect significant differences in the relatively smaller sizes of some of the subgroups. A larger sample size is thus necessary to elucidate clearer differences in the depth and/or distribution of RLS and RLS-related conditions in RLS-C patients from that of idiopathic RLS patients. On the other hand, when we did a combined analysis of all cases of neuropathy, sciatica, and leg cramps (a large sample size of 29 cases) versus the 71 idiopathic RLS patients, results still did not show any clinically important distinguishing features.

Lastly, this study did not use a computerized program to measure and analyze the specific topography of RLS sensations in each patient. Therefore, the percent body surface area was a calculation based upon visual inspection obtained from what the patient circled on the body diagram, rather than an exact percentage of body surface area coverage, such as those calculated in Karroum's study.28 On the other hand, our estimates closely resemble those of Karroum et al.28

Summary

Neither the depth nor the distribution of RLS sensory discomfort can be used to discriminate between idiopathic RLS and RLS associated with other causes of leg discomfort. Patients who meet all 4 criteria for RLS but have only isolated joint involvement should be suspected of not having true RLS.

DISCLOSURE STATEMENT

This was not an industry-supported study. This study was financed by internal research funding at Vanderbilt University Medical Center specific to Dr. Walters. Dr. Ondo has grant/research support from Cynapsus and Lundbeck as well as being on the Speakers' Bureau of Xenoport, IMPAX, Avanir, UCB Pharma, and Lundbeck. Dr. Picchietti is an unpaid consultant to UCB Pharma and has royalties related to 2 Uptodate chapters unrelated to the content in this manuscript. Dr. Allen has grant/research support from AMAG Pharmaceuticals, is a consultant to Luitpold, and is a stockholder of an investigational device/drug from Orablas, LLC unrelated to this article. Dr. Wang has grant/research support from UCB Pharma. Dr. Bagai has grant/research support from Xenoport, unrelated to this article or its study. Dr. Walters has grant/research support from and is a consultant to UCB Pharma, Mundi Pharma, and Xenoport. The other authors have indicated no financial or commercial relationship conflicts of interest.

ABBREVIATIONS

BCM

Baylor College of Medicine

BSA

body surface area

CFH

Carle Foundation Hospital

EMG

electromyogram

FDR

false discovery rate

IRLS

International RLS Severity Scale

IRLSSG

International Restless Legs Syndrome Study Group

n

sample size

NCV

nerve conduction study

RedCAP

Research Electronic Data Capture System

REF

reference group

RLS

restless legs syndrome/Willis-Ekbom disease

RLS-C

restless legs syndrome patient(s) with comorbid lower extremity conditions

SC

Scripps Clinic

SD

standard deviation

UTD

University of Texas at Dallas

UTH

University of Texas at Houston

VUMC

Vanderbilt University

ACKNOWLEDGMENTS

Dr Poceta wishes to recognize Scripps Clinical Research Services and Ms. Crystal Sanchez.

REFERENCES

1 

Yeh P, Walters AS, Tsuang JW. Restless legs syndrome: a comprehensive overview on its epidemiology, risk factors, and treatment. Sleep Breath. 2012;16:987–1007. [PubMed]

2 

Allen RP, Picchietti DL, Hening WA. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology. A report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med. 2003;4:101–19. [PubMed]

3 

Allen RP, Picchietti DL, Garcia-Borreguero D, et al. Restless legs syndrome/Willis-Ekbom disease diagnostic criteria: updated International Restless Legs Syndrome Study Group (IRLSSG) consensus criteria- history, rationale, description, and significance. Sleep Med. 2014;15:860–73. [PubMed]

4 

Koo YS, Lee GT, Lee SY, Cho YW, Jung KY. Topography of sensory symptoms in patients with drug-naïve restless legs syndrome. Sleep Med. 2013;14:1369–74. [PubMed]

5 

Ekbom KA. Restless legs syndrome. Neurology. 1960;10:868–73. [PubMed]

6 

Walters AS. Toward a better definition of the restless legs syndrome. The International Restless Legs Syndrome Study Group. Mov Disord. 1995;10:634–42. [PubMed]

7 

Ondo W, Jankovic J. Restless legs syndrome: clinicoetiologic correlates. Neurology. 1996;47:1435–41. [PubMed]

8 

Ekbom KA. Restless legs: clinical study of hitherto overlooked disease in legs characterized by peculiar paresthesia (“Anxietas tibiarum”), pain and weakness and occurring in two main forms, asthenia crurum paresthetica and asthenia crurum dolorosa. Acta Med Scand. 1945;158 Suppl.:1–123.

9 

Winkelmann J, Wetter TC, Collado-Seidel V, et al. Clinical characteristics and frequency of the hereditary restless legs syndrome in a population of 300 patients. Sleep. 2000;23:597–602. [PubMed]

10 

Horvath J, Landis T, Burkhard PR. Restless arms. Lancet. 2008;371:530. [PubMed]

11 

Freedom T, Merchut MP. Arm restlessness as the initial symptom in restless legs syndrome. Arch Neurol. 2003;60:1013–5. [PubMed]

12 

Michaud M, Chabli A, Lavigne G, Montplaisir J. Arm restlessness in patients with restless legs syndrome. Mov Disord. 2000;15:289–93. [PubMed]

13 

Sevim S, Dogu O, Camdeviren H, et al. Unexpectedly low prevalence and unusual characteristics of RLS in Mersin, Turkey. Neurology. 2003;61:1562–9. [PubMed]

14 

Xiong L, Montplaisir J, Desautels A, et al. Family study of restless legs syndrome in Quebec, Canada: clinical characterization of 671 familial cases. Arch Neurol. 2010;67:617–22. [PubMed]

15 

Perez-Diaz H, Iranzo A, Rye DB, et al. Restless abdomen: a phenotypic variant of restless legs syndrome. Neurology. 2011;77:1283–6. [PubMed]

16 

Waldinger MD, Venema PL, van Gils AP, de Lint GJ, Schweitzer DH. Stronger evidence for small fiber sensory neuropathy in restless genital syndrome: two case reports in males. J Sex Med. 2011;8:325–30. [PubMed]

17 

Waldinger MD, de Lint GJ, Venema PL, van Gils AP, Schweitzer DH. Successful transcutaneous electrical nerve stimulation in two women with restless genital syndrome: the role of adelta- and C-nerve fibers. J Sex Med. 2010;7:1190–9. [PubMed]

18 

Waldinger MD, Venema PL, van Gils AP, Schutter EM, Schweitzer DH. Restless genital syndrome before and after clitoridectomy for spontaneous orgasms: a case report. J Sex Med. 2010;7:1029–34. [PubMed]

19 

Waldinger MD, Schweitzer DH. Persistent genital arousal disorder in 18 Dutch women: part II, A syndrome clustered with restless legs and overactive bladder. J Sex Med. 2009;6:482–97. [PubMed]

20 

Waldinger MD, Venema PL, van Gils AP, Schweitzer DH. New insights into restless genital syndrome: static mechanical hyperesthesia and neuropathy of the nervus dorsalis clitoridis. J Sex Med. 2009;6:2778–87. [PubMed]

21 

Akcali A, Ferini-Strambi L, Kaynak H, Karadeniz D, Akcali C. Genital restlessness (vulvodynia) events accompanying restless legs syndrome. Sleep Med. 2009;10:395–6. [PubMed]

22 

Lombardi C, Provini F, Vertrugno R, Plazzi G, Lugaresi E, Montagna P. Pelvic movements as rhythmic motor manifestation associated with restless legs syndrome. Mov Disord. 2003;18:110–3. [PubMed]

23 

Valko PO, Siccoli MM, Bassetti CL. Unilateral RLS with predominately ipsilateral PLMS and variable response to dopaminergic drugs: a variant of idiopathic RLS? Eur J Neurol. 2009;16:430–2. [PubMed]

24 

Bornstein B. Restless legs. Psychiatr Neurol. 1961;141:165–201.

25 

Umehara H, Sumitani S, Ohmori T. Restless legs syndrome with chest and back restlessness as the initial symptom. Psychiatry Clin Neurosci. 2010;64:211.

26 

Buchfuhrer MJ. Restless legs syndrome (RLS) with expansion of symptoms to the face. Sleep Med. 2008;9:188–90. [PubMed]

27 

Fukunishi I, Kitaoka T, Shirai T, Kino K. Facial paresthesias resembling restless legs syndrome. Sleep Med Rev. 2010;14:339–46. [PubMed]

28 

Karroum EG, Leu-Semenescu S, Arnulf I. Topography of the sensations in primary restless legs syndrome. J Neurol Sci. 2012;320:26–31. [PubMed]

29 

Artz CP, Moncrief JA. The treatment of burns. edition 2. Philadelphia: WB Saunders; 1969.

30 

Walters AS, LeBrocq C, Dhar A, et al. Validation of the International Restless Legs Syndrome Study Group rating scale for restless legs syndrome. Sleep Med. 2003;4:121–32. [PubMed]

31 

Hening WA, Allen RP, Washburn M, Lesage S, Earley CJ. Validation of the Hopkins telephone diagnostic interview for restless legs syndrome. Sleep Med. 2008;9:283–9. [PubMed]

32 

Harris PA, Taylor R, Conde JG, et al. Research Electronic Data Capture (REDCap)- A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Informatics. 2009;42:377–81.

33 

Hornyak M, Scholz H, Kiemen A, Kassubek J. Investigating the response to intravenous iron in restless legs syndrome: an observational study. Sleep Med. 2012;13:732–5. [PubMed]

34 

Tison F, Crochard A, Leger D, Bouee S, Lainey E, El Hasnaoui A. Epidemiology of restless legs syndrome in French adults: a nationwide survey: the INSTANT Study. Neurology. 2005;65:239–46. [PubMed]

35 

Whittom S, Dauvilliers Y, Pennestri MH, et al. Age-at-onset in restless legs syndrome: a clinical and polysomnographic study. Sleep Med. 2007;9:54–59. [PubMed]

36 

Mignot E. A step forward for restless legs syndrome. Nat Genet. 2007;39:938–9. [PubMed]

37 

Winkelmann J, Schormair B, Lichtner P, et al. Genome-wide association study for restless legs syndrome identifies common variants in three genomic regions. Nat Genet. 2007;39:1000–6. [PubMed]

38 

Clemens S, Rye D, Hochman S. Restless legs syndrome: revisiting the dopamine hypothesis from the spinal cord perspective. Neurology. 2006;67:125–30. [PubMed]

39 

Paci D, Lanuzza B, Cosentino Fl, et al. Subclinical abnormal EMG activation of the gastrocnemii during gait analysis in the restless legs syndrome: a preliminary report in 13 patients. Sleep Med. 2009;10:312–6. [PubMed]

40 

Benarroch EE. Descending monoaminergic pain modulation: bidirectional control and clinical relevance. Neurology. 2008;71:217–21. [PubMed]