Abstract
Study Objectives
To investigate the prevalence of concurrent periodic limb movements during sleep (PLMS) and restless leg syndrome (RLS), as well as the prevalence of PLMS and RLS separately. Additionally, we document these prevalences by age, race/ethnicity, sex, and obesity status.
Methods
Cross-sectional data from 2041 Multi-Ethnic Study of Atherosclerosis (MESA) Sleep ancillary study participants were used. PLMS (>15 periodic limb movements per hour of sleep) was measured by polysomnography. RLS symptoms were assessed using the 2009 International Restless Legs Syndrome Study Group clinical criteria.
Results
The prevalence of RLS with PLMS was 6.7%, RLS alone 16.1%, and PLMS alone 21.2%. RLS with PLMS was prevalent in 7.0% of whites, 4.9% of blacks, 10.1% of Hispanics, and 3.3% of Chinese-Americans. In adjusted models, odds of RLS with PLMS was higher for those older than 67 years versus those younger (odds ratio [OR] [95% confidence interval [CI]] = 1.62 [1.09–2.40]). Relative to white participants, the prevalence of RLS with PLMS tended to be lower among blacks (0.56 [0.32–0.96]). The prevalence of concurrent RLS and PLMS did not statistically differ by sex or obesity status. RLS alone was more common in women
Conclusions
Approximately 7% of our sample had RLS with PLMS (“electro-clinical RLS”). This condition was more common among older individuals, did not vary by sex, and was less common among blacks. The findings provide some of the first information about the prevalence of concurrent RLS and PLMS in a community-based sample and show distinct sex and race associations for RLS versus electro-clinical RLS.
Keywords: restless leg syndrome, period leg movements, racial differences, Multi-Ethnic Study of Atherosclerosis
Statement of Significance
Periodic limb movements during sleep (PLMS) and restless leg syndrome (RLS) are highly correlated, but the prevalence of their co-occurrence is unclear. The consideration of both PLMS and RLS may yield insights in underlying phenotypic differences in movement disorders. In this racially/ethnically diverse community-based sample of older adults, approximately 7% of PLMS and RLS symptoms occurred concurrently; this co-occurrence was more common among those who were older, and less common among participants self-identifying as blacks. These data provide some of the first information about the prevalence of RLS with PLMS in a community-based setting, and how the prevalence of RLS with PLMS varies across population demographics.
Introduction
Restless legs syndrome (RLS), also known as Willis-Ekbom disease, is a neurological disease characterized by an unrelenting urge to move at night. RLS symptoms are often described as a “creeping” or “crawling” feeling located within the legs, which worsens during times of rest or inactivity and are relieved by movement [1]. In addition to causing disordered sleep, RLS has been associated with lower quality of life, hypertension and cardiovascular disease [2–4]. Diagnostic criteria for RLS have been developed by the International Restless Legs Syndrome Study Group (IRLSSG) and other groups in order to standardize the diagnosis of RLS [5, 6].
Periodic limb movements during sleep (PLMS) are characterized by “repetitive” and “forceful” contraction of muscles within the leg and foot during sleep and occur in up to 80%–90% of RLS cases [7, 8]. Risk factors for PLMS are similar to those of RLS [9]. PLMS has been recognized by the IRLSSG as a clinical feature supportive of the diagnosis of RLS and their frequency correlates to the severity of RLS [5, 10]. The clinical association of RLS and PLMS is further supported by genome-wide association studies which have identified associations between both phenotypes and variants in the genes, PTPRD, BTBD9, and MEIS1 in independent population samples [11–14], suggesting that PLMS may be an endophenotype for RLS. Due to these shared characteristics, the co-occurrence of RLS and PLMS, and concerns about the subjectivity of traditional RLS definitions, it has been proposed that a more accurate diagnosis of RLS can be achieved through the incorporation of objective measures of PLMS when defining RLS [9, 10]. This condition has been termed electro-clinical RLS [9, 10].
Prior epidemiological studies have yielded divergent estimates for the prevalence of RLS in the general population, with estimates ranging from 1% to 15% in North American and European populations [15, 16]. The variation in prevalence estimates may be due to differences in the methodology for identifying RLS (e.g. including the subjective nature of RLS and changes in diagnosis criteria), as well as differences in sample characteristics, study designs, and/or data collection methods [10]. In general, the prevalence of RLS increases with age and is higher among individuals who are Caucasian, female, and obese [3, 15–17]. To date few, if any, studies have utilized a definition of RLS, which incorporates the PLMS condition, which may identify a subset of individuals with unique features.
Using data from the community-based Multi-Ethnic Study of Atherosclerosis (MESA), we report the prevalence of electro-clinical RLS in a racially/ethnically diverse sample. We hypothesize that the prevalence of electro-clinical RLS will differ by race/ethnicity, and will be higher in individuals who are older, female, and/or obese.
Methods
Study population
The MESA is a prospective cohort study started in 2000 to evaluate cardiovascular disease risk factors in a racially/ethnically diverse study population [18]. This cohort is comprised of 6814 men and women aged 45–84 years who were free of cardiovascular disease at a baseline examination. Participant recruitment sites, and races/ethnicities recruited from each site, are as follows: Baltimore, Maryland (whites and blacks); Chicago, Illinois (whites, Chinese-Americans, and blacks); Forsyth County, North Carolina (whites, blacks, and Hispanics); Los Angeles, California (all races/ethnic groups); Manhattan, New York (whites, blacks, and Hispanics); and St. Paul, Minnesota (whites and Hispanics). A total of five clinical examinations have now been completed. The study protocols have been approved by all participating institutions, and participants provided written informed consent.
A total of 4655 participants took part in MESA Exam 5 (2010–2012) and were invited to participate in a sleep exam, consisting of an in-home overnight polysomnogram, 7 days wrist actigraphy, and sleep questionnaire administration. Of 4077 Exam 5 participants approached to take part in the sleep study, 147 (6.5%) were ineligible (95 due to a history of use of positive airway pressure or mandibular advancement devices to treat sleep apnea (2%); 4 due to use of an alternate oral appliance; and 4 due to oxygen use); and 141 participants lived too far away to participate. Of the remaining 3789 participants, 2261 participated in the MESA Sleep ancillary study (59.7%). Individuals with missing data for RLS and/or PLMS were excluded for the purposes of this analysis, leaving a final sample size of 2041.
Quantification of RLS and PLMS
RLS was assessed using a four-item questionnaire developed by the IRLSSG 2003 diagnostic criteria [19]. Individuals were considered to have RLS if all four questionnaire items were positively indicated (Table 1).
Table 1.
MESA definitions of RLS, PLMS, and electro-clinical RLS
| RLS | Designated as a case if “yes” indicated for each of the following questionnaire items: |
| 1. Do you ever experience a desire to move your legs because of discomfort or disagreeable sensations in your legs? | |
| 2. Do you sometimes feel the need to move to relieve the discomfort, for example by walking, or by rubbing your legs? | |
| 3. Are these symptoms worse when you are at rest, with at least temporary relief by activity? | |
| 4. Are these symptoms worse later in the day or at night? | |
| PLMS | >15 Periodic limb movements / hour of sleep. |
| Electro-clinical RLS | Having both RLS and PLMS conditions per MESA definitions. |
PLMS, as well as other physiologic sleep data, were measured using type II unattended in-home polysomnography (Compumedics Somte Systems; Compumedics Ltd., Abbostville, Australia). The recording montage included electroencephalography (central C4-M1, occipital Oz-Cz, and frontal Fz-Cz channels), chin electromyography, bilateral electro-oculography, airflow measured by thermistry and nasal pressure, thoracoabdominal respiratory inductance plethysmography, finger pulse oximetry, electrocardiography, and bilateral leg movements piezoelectric sensors. Sleep stages and arousals were scored using standard criteria [20]. Only leg movements that lasted between 0.5 to 5 seconds and had a clear difference from baseline amplitude were scored. This approach compares well to methods using leg EMG. A blinded in-laboratory validation study in 51 subjects to compare piezoelectric and electromyographic leg sensors in PLMS detection found high levels of agreement, r = 0.81.
Leg movements were considered periodic if a minimum of four movements occurred successively in intervals ranging from 5 to 90 seconds and were not exclusively associated with the termination or respiratory events.
Based on the polysomnography data, two indices were created. A periodic limb movement index (PLMI) was calculated by summing the total number of individual periodic limb movements and dividing by the total hours of sleep. The periodic limb movement arousal index (PLMAI) represents the total number of periodic limb movements per hour of sleep where arousal occurred within 3 seconds of the termination of the leg movement. A PLMI of >15 was used as the threshold for PLMS and a threshold of PLMAI >5 was used in a separate analysis [21]. Participants were considered to have electro-clinical RLS if requirements for both RLS and PLMS were met (Table 1).
Demographics, body mass index and covariates
Information about age, sex, and race/ethnicity were obtained from self-completed questionnaires. For analysis, age was treated as a dichotomous variable divided at the median value, 67 years. Race/ethnicity was categorized as white, black, Hispanic, or Chinese. At the Exam 5 clinic visit, height was measured to the nearest 0.1 cm and weight to the nearest 0.5 kg. Body mass index (BMI) was calculated from height and weight measurements as weight (kg) divided by height squared (m2). BMI was categorized as normal (18.5<25 kg/m2), overweight (25<30 kg/m2), and obese (≥30 kg/m2).
Other information collected at MESA Exam 5 included medical history for diabetes mellitus, physical activity level, depressive symptoms, medication use, education, income, smoking status, and alcohol consumption. Diabetes mellitus was categorized as yes if indicated as such during examination. The MESA Typical Week Physical Activity Survey was utilized to assess physical activity. A physical activity score was calculated based on the summation of reported time performing physical tasks for a typical week multiplied by the metabolic equivalent level for the tasks performed. Physical activity was log-transformed due to non-normal distribution of data. Depressive symptoms were assessed with the Center for Epidemiological Studies Depression Scale (CES-D) and were modeled dichotomously with a cutpoint at 16; scores at this point or higher are typically indicative of clinically significant symptomatology [22]. Antidepressant use was categorized as yes based on the use of tricyclic antidepressants, selective serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors and other antidepressants. Level of education was categorized as high school or less, some college, or bachelors/graduate degree. Annual income was grouped into four categories: <$25000, $25000–$39999, $40000–$74000, and >$74000. Smoking status was identified as never, former, or current. Alcohol consumption was dichotomized as current or not current. Information on dopaminergic medications used to treat RLS or other neurological conditions was also obtained (i.e. Mirapex [pramipexole], Neupro [rotigotine], Requip [ropinirole], Sinemet/Parcopa [carbidopa and levodopa], Levodopa) from participant interview.
Statistical methods
Characteristics of the cohort were compared across racial/ethnic groups. Prevalences and 95% confidence intervals (CIs) are also provided according to the primary outcome definition, which was a categorical definition using information on RLS and PLMS as indicated by a PLMI of >15: electro-clinical RLS (+RLS/+PLMS), RLS alone (+RLS/−PLMS), PLMS alone (−RLS/+PLMS), and neither RLS nor PLMS (−RLS/−PLMS). Secondary outcomes were RLS (yes/no) defined per questionnaire, PLMS, (>15 versus ≤15 movements per hour) and PLMAI (>5 versus ≤5 movements per hour of sleep with arousals).
The primary analyses employed multinomial regression because the outcome had more than two discrete categories. Results are presented as odds ratios with 95% confidence limits. We used a series of models to analyze the association between the demographics, BMI categories, and RLS/PLMS. The first model included the following characteristics: race/ethnicity, age, sex, BMI, and site. When not the exposure of interest, the model was adjusted for the remaining variables. Model 2 additionally adjusted for education, income, dichotomized CES-D depression score, smoking status, alcohol consumption, physical activity, and medication use. In the secondary analysis, we reported results for the RLS and PLMS (defined via both PLMI and PLMAI) conditions separately using prevalence ratios calculated via log-binomial regression. Additionally, in sensitivity analyses we explored a) counting individuals taking dopaminergic medications (n = 10) as having RLS and PLMS, and b) excluding these individuals. SAS version 9.3 was used to analyze the data (SAS Institute; Cary, North Carolina).
Results
The mean age in the analytic sample was 68.4 (SD = 9.1) years and 53.6% were female. Demographic, anthropometric, and other characteristics of the cohort stratified by race/ethnicity are provided in Table 2. Among the participants, 36.2% self-identified as white, 28.0% as black, 23.9% as Hispanic, and 11.9% as Chinese-American.
Table 2.
Distribution of participant characteristics by race/ethnicity*: The Multi-Ethnic Study of Atherosclerosis 2011–2013
| Participant characteristics | Overall | White | Black | Hispanic | Chinese-American |
|---|---|---|---|---|---|
| n = 2041 | 739 | 572 | 487 | 243 | |
| Demographics | |||||
| Age, years | 68.4 ± 9.1 | 68.4 ± 9.1 | 68.8 ± 9.0 | 68.4 ± 9.4 | 67.7 ± 9.0 |
| ≤67 | 1026 (50.3) | 378 (51.1) | 279 (48.8) | 245 (50.3) | 124 (51.0) |
| >67 | 1015 (49.7) | 361 (48.9) | 293 (51.2) | 242 (49.7) | 119 (49.0) |
| Sex, n (%) | |||||
| Males | 947 (46.4) | 343 (46.4) | 256 (44.8) | 229 (47.0) | 119 (49.0) |
| Female | 1094 (53.6) | 396 (53.6) | 316 (55.2) | 258 (53.0) | 124 (51.0) |
| Education, n (%) | |||||
| High school or less | 638 (31.3) | 129 (17.5) | 137 (24.0) | 296 (60.8) | 76 (31.3) |
| College/associates | 597 (29.3) | 179 (24.3) | 226 (39.7) | 135 (27.7) | 57 (23.5) |
| Bachelors/graduate | 802 (39.4) | 429 (58.2) | 207 (36.3) | 56 (11.5) | 110 (45.3) |
| Annual income, n (%) | |||||
| <$25000 | 538 (27.1) | 85 (11.8) | 132 (24.0) | 222 (46.9) | 99 (41.3) |
| $25000–$39999 | 361 (18.2) | 99 (13.8) | 131 (23.8) | 101 (21.4) | 30 (12.5) |
| $40000–$74000 | 522 (26.3) | 222 (30.8) | 148 (26.9) | 102 (21.6) | 50 (20.8) |
| >$74000 | 563 (28.4) | 314 (43.6) | 140 (25.4) | 48 (10.2) | 61 (25.4) |
| Behavioral and physiologic characteristics | |||||
| BMI, kg/m2 | |||||
| <25 | 555 (27.2) | 229 (31.1) | 95 (16.6) | 76 (15.6) | 155 (63.8) |
| 25–29.9 | 769 (37.7) | 291 (39.5) | 200 (35.0) | 201 (41.3) | 77 (31.7) |
| ≥30 | 714 (35.0) | 217 (29.4) | 276 (48.3) | 210 (43.1) | 11 (4.5) |
| Physical activity (MET-Min/Wk) | 2739 ± 5.5 | 3161 ± 4.1 | 2834 ± 6.2 | 2088 ± 8.1 | 2807 ± 3.4 |
| Smoking status, n (%) | |||||
| Never | 955 (47.1) | 306 (41.5) | 235 (41.6) | 230 (47.5) | 184 (76.0) |
| Former | 928 (45.8) | 386 (52.4) | 265 (46.9) | 226 (46.7) | 51 (21.0) |
| Current | 145 (7.2) | 45 (6.1) | 65 (11.5) | 28 (5.8) | 7 (2.9) |
| Use alcohol, n (%) | 878 (43.3) | 453 (61.5) | 237 (41.9) | 150 (31.0) | 38 (15.7) |
| CES-D score | 8.1 ± 7.6 | 7.8 ± 7.3 | 7.9 ± 7.5 | 9.5 ± 8.3 | 6.7 ± 6.5 |
| CES-D | |||||
| <16 | 1715 (85.5) | 633 (87.0) | 486 (86.8) | 381 (79.9) | 215 (89.6) |
| ≥16 | 290 (14.5) | 95 (13.0) | 74 (13.2) | 96 (20.1) | 25 (10.4) |
| Medication use, n (%)† | |||||
| No | 1752 (86.0) | 567 (76.8) | 521 (91.2) | 432 (88.9) | 232 (95.9) |
| Yes | 285 (14.0) | 171 (23.2) | 50 (8.8) | 54 (11.1) | 10 (4.1) |
Binary outcomes as number and frequency. Continuous variables are expressed as means ± SD.
*There were statistically significant differences by race at p < 0.05 for all participant characteristics except age (p = 0.59), gender (p = 0.90), and smoking status (p = 0.88).
†Use of benzodiazepines, tricyclic antidepressants, non-tricyclic antidepressants other than MAOI, or norepinephrine-dopamine reuptake inhibitors.
The prevalence of participant characteristics stratified by our four primary outcome categories (i.e. electro-clinical RLS [+RLS/+PLMS], RLS without PLMS [+RLS/−PLMS], PLMS without RLS [−RLS/+PLMS], and no PLMS or RLS [−RLS/−PLMS]) are provided in Table 3. Supplementary Table 1 provides the number of individuals in our sample by outcome category and participant characteristics. Overall, 465 participants (22.8%) had RLS and 570 participants (27.9%) had PLMS, while 6.7% met the definition for electro-clinical RLS. Among those with RLS (n = 465), 29.5% also had PLMS (n = 17). Electro-clinical RLS was identified in 7.0% of whites (n = 52), 4.9% of blacks (n = 28), 10.1% of Hispanics (n = 49), and 3.3% of Chinese-Americans (n = 8). Among those with electro-clinical RLS, 21.2% (29/137) were identified as using antidepressant medications (i.e. benzodiazepines, non-tricyclic antidepressants other than monoamine oxidase inhibitor [MAOI], tricyclic antidepressants other than MAOI, or norepinephrine-dopamine reuptake inhibitor). Of those with electro-clinical RLS, 3.0% (4/137) were using dopaminergic medications.
Table 3.
Prevalence ([%] and 95% confidence intervals) of participants by RLS and PLMS categories: The Multi-Ethnic Study of Atherosclerosis 2011–2013
| Participant characteristics | Electro-clinical RLS | RLS, no PLMS | PLMs, no RLS | No RLS, No PLMS |
|---|---|---|---|---|
| N* | 137 | 328 | 433 | 1143 |
| Overall prevalence | 6.7 (5.6–7.8) | 16.1 (14.5–17.7) | 21.2 (19.4–23.0) | 56.0 (53.8–58.2) |
| Demographics | ||||
| Race | ||||
| White | 7.0 (5.2–8.9) | 15.3 (12.7–17.9) | 26.4 (23.2–29.6) | 51.3 (47.7–54.9) |
| Black | 4.9 (3.1–6.7) | 20.3 (17.0–23.6) | 11.4 (8.8–14.0) | 63.5 (59.5–67.4) |
| Hispanic | 10.1 (7.4–12.7) | 17.5 (14.1–20.8) | 21.8 (18.1–25.4) | 50.7 (46.3–55.2) |
| Chinese-American | 3.3 (1.0–5.6) | 5.8 (2.8–8.7) | 27.6 (21.9–33.2) | 63.4 (57.3–69.5) |
| Age | ||||
| ≤67 years | 5.7 (4.2–7.1) | 16.9 (14.6–19.2) | 15.2 (13.0–17.4) | 62.3 (59.3–65.3) |
| >67 years | 7.8 (6.1–9.4) | 15.3 (13.1–17.5) | 27.3 (24.5–30.0) | 49.7 (46.6–52.7) |
| Sex | ||||
| Male | 6.2 (4.7–7.8) | 11.6 (9.6–13.7) | 25.2 (22.5–28.0) | 56.9 (53.8–60.1) |
| Female | 7.1 (5.6–8.7) | 19.9 (17.6–22.3) | 17.7 (15.5–20.0) | 55.2 (52.3–58.2) |
| Education | ||||
| High school or less | 8.2 (6.0–10.3) | 16.6 (13.7–19.5) | 23.4 (20.1–26.6) | 51.9 (48.0–55.8) |
| College/associates | 7.5 (5.4–9.7) | 19.4 (16.2–22.6) | 17.8 (14.7–20.8) | 55.3 (51.3–59.3) |
| Bachelors/graduate | 4.9 (3.4–6.4) | 13.0 (10.6–15.3) | 22.1 (19.2–24.9) | 60.1 (56.7–63.5) |
| Annual income | ||||
| <$25000 | 7.4 (5.2–9.7) | 16.2 (13.0–19.3) | 24.2 (20.5–27.8) | 52.2 (48.0–56.5) |
| $25000–$39999 | 8.6 (5.7–11.5) | 21.3 (17.1–25.6) | 19.4 (15.3–23.5) | 50.7 (45.5–55.9) |
| $40000–$74000 | 4.6 (2.8–6.4) | 13.4 (10.5–16.3) | 21.6 (18.1–25.2) | 60.3 (56.1–64.6) |
| >$74000 | 6.7 (4.7–8.8) | 14.0 (11.2–16.9) | 18.8 (15.6–22.1) | 60.4 (56.3–64.4) |
| Behavioral and physiologic characteristics | ||||
| BMI | ||||
| <25 kg/m2 | 5.2 (3.4–7.1) | 13.2 (10.3–16.0) | 23.4 (19.9–27.0) | 58.2 (54.1–62.3) |
| 25–29.9 kg/m2 | 7.2 (5.3–9.0) | 16.8 (14.1–19.4) | 21.8 (18.9–24.8) | 54.2 (50.7–57.8) |
| ≥30 kg/m2 | 7.4 (5.5–9.4) | 17.6 (14.8–20.5) | 18.6 (15.8–21.5) | 56.3 (52.7–59.9) |
| Smoking status | ||||
| Never | 5.9 (4.4–7.4) | 15.2 (12.9–17.5) | 19.8 (17.3–22.3) | 59.2 (56.0–62.3) |
| Former | 7.7 (5.9–9.4) | 16.2 (13.8–18.5) | 24.2 (21.5–27.0) | 51.9 (48.7–55.2) |
| Current | 6.2 (2.2–10.2) | 22.1 (15.2–28.9) | 12.4 (7.0–17.8) | 59.3 (51.2–67.4) |
| Use alcohol | 6.5 (4.9–8.1) | 15.6 (13.2–18.0) | 22.8 (20.0–25.6) | 55.1 (51.8–58.4) |
| CES-D | ||||
| <16 | 6.6 (5.5–7.8) | 14.8 (13.1–16.5) | 21.7 (19.7–23.6) | 56.9 (54.5–59.2) |
| ≥16 | 6.9 (4.0–9.8) | 23.8 (18.9–28.7) | 18.6 (14.1–23.1) | 50.7 (44.9–56.5) |
| Medication use† | ||||
| No | 6.1 (4.9–7.2) | 15.5 (13.8–17.2) | 21.0 (19.1–22.9) | 57.4 (55.1–59.7) |
| Yes | 10.2 (6.6–13.7) | 19.6 (15.0–24.3) | 22.8 (17.9–27.7) | 47.4 (41.5–53.2) |
*N total = 2041.
†Use of benzodiazepines, tricyclic antidepressants, non-tricyclic antidepressants other than MAOI, or norepinephrine-dopamine reuptake inhibitors.
Associations of demographics and BMI with electro-clinical RLS
The odds of the electro-clinical RLS, relative to no RLS/PLMS, can be found in Table 4. Blacks had lower odds of having electro-clinical RLS compared to whites in the minimally-adjusted model [OR (95% CI) = 0.50 (0.31–0.82)]; however, the association was attenuated slightly in the fully-adjusted [0.56 (0.32–0.96)] model. Similarly, relative to whites, participants of Chinese ancestry tended toward a lower odds of electro-clinical RLS in the minimally adjusted model [0.45 (0.20–1.03)], which was attenuated after additional adjustment [0.67 (0.28–1.64)]. Odds of electro-clinical RLS in Hispanics did not significantly differ from those observed in whites.
Table 4.
Odds of RLS, PLMs, and electro-clinical RLS by demographics and obesity categories: The Multi-Ethnic Study of Atherosclerosis 2011–2013
| Electro-clinical RLS | RLS, no PLMs | PLMs, no RLS | No RLS /PLMs | Electro-clinical RLS | RLS, no PLMs | PLMs, no RLS | No RLS /PLMs | ||
|---|---|---|---|---|---|---|---|---|---|
| 137 | 328 | 431 | 1142 | 128 | 308 | 413 | 1105 | ||
| n | Model 1 | Model 2 | |||||||
| Race | |||||||||
| White | 247 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| Black | 93 | 0.50 (0.31–0.82) | 1.04 (0.76–1.42) | 0.32 (0.23–0.44) | 0.56 (0.32–0.96) | 1.04 (0.74–1.47) | 0.31 (0.22–0.44) | ||
| Hispanic | 155 | 1.38 (0.90–2.13) | 1.17 (0.84–1.62) | 0.80 (0.60–1.08) | 1.55 (0.91–2.65) | 1.04 (0.70–1.55) | 0.74 (0.52–1.06) | ||
| Chinese- American | 75 | 0.45 (0.20–1.03) | 0.33 (0.18–0.62) | 0.91 (0.61–1.34) | 0.67 (0.28–1.64) | 0.36 (0.19–0.71) | 1.02 (0.66–1.60) | ||
| Age | |||||||||
| ≤67 | 214 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| >67 | 356 | 1.76 (1.23–2.53) | 1.13 (0.88–1.45) | 2.29 (1.82–2.89) | 1.62 (1.09–2.40) | 1.15 (0.87–1.51) | 2.19 (1.71–2.81) | ||
| Sex | |||||||||
| Men | 298 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| Women | 272 | 1.16 (0.81–1.66) | 1.74 (1.35–2.26) | 0.72 (0.57–0.90) | 1.20 (0.81–1.79) | 1.71 (1.30–2.26) | 0.71 (0.56–0.91) | ||
| BMI category | |||||||||
| <25 | 159 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 25 to <30 | 233 | 1.48 (0.91–2.40) | 1.39 (1.00–1.93) | 1.07 (0.81–1.43) | 1.32 (0.80–2.18) | 1.28 (0.91–1.81) | 1.06 (0.79–1.42) | ||
| ≥30 | 186 | 1.55 (0.94–2.55) | 1.27 (0.91–1.79) | 1.13 (0.84–1.54) | 1.23 (0.73–2.10) | 1.13 (0.79–1.62) | 1.14 (0.82–1.56) | ||
Model 1: Race, age, sex, BMI, and site; Model 2: model 1 + physical activity, education, income, smoking status, alcohol consumption, dichotomized CES-D score, and medications. When not the exposure of interest, models were adjusted for the remaining covariates.
Odds of electro-clinical RLS tended to be higher among participants who were older than 67 years, female, overweight, and/or obese when compared to participants younger than 67 years, males, or of normal BMI; however, results were only statistically significant for age. Compared to participants ≤67 years old, those aged >67 years had odds of electro-clinical RLS of 1.76 (1.23–2.53) in minimally adjusted models and 1.62 (1.09–2.40) in fully adjusted models.
A sensitivity analysis was conducted restricted to the 1638 participants without diabetes, a possible source of secondary RLS (Table 5) [3, 7, 15, 23]. Most estimates remained unchanged, though the electro-clinical RLS odds ratio estimate among blacks was no longer significant in either the fully adjusted models. This may have been due to poorer precision in this analysis.
Table 5.
Odds of RLS, PLMS, and electro-clinical RLS by demographics and obesity categories, among those without diabetes (n = 1638): The Multi-Ethnic Study of Atherosclerosis 2011–2013
| Electro-clinical RLS | RLS, no PLMs | PLMs, no RLS | No RLS /PLMs | Electro-clinical RLS | RLS, no PLMs | PLMs, no RLS | No RLS /PLMs | ||
|---|---|---|---|---|---|---|---|---|---|
| 104 | 256 | 348 | 927 | 99 | 239 | 334 | 896 | ||
| n | Model 1 | Model 2 | |||||||
| Race | |||||||||
| White | 656 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| Black | 421 | 0.53 (0.30–0.94) | 1.14 (0.80–1.61) | 0.41 (0.29–0.59) | 0.62 (0.33–1.15) | 1.21 (0.83–1.76) | 0.40 (0.28–0.59) | ||
| Hispanic | 355 | 1.34 (0.82–2.20) | 1.16 (0.80–1.68) | 0.83 (0.59–1.16) | 1.66 (0.90–3.04) | 1.06 (0.68–1.66) | 0.77 (0.52–1.16) | ||
| Chinese- American | 206 | 0.49 (0.20–1.19) | 0.32 (0.16–0.62) | 0.94 (0.62–1.45) | 0.76 (0.29–1.99) | 0.36 (0.17–0.73) | 1.01 (0.62–1.64) | ||
| Age | |||||||||
| ≤67 | 892 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| >67 | 746 | 1.57 (1.04–2.36) | 1.03 (0.77–1.36) | 2.18 (1.69–2.82) | 1.57 (1.00–2.45) | 1.06 (0.78–1.44) | 2.10 (1.60–2.77) | ||
| Sex | |||||||||
| Men | 892 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| Women | 746 | 1.18 (0.78–1.78) | 1.72 (1.28–2.30) | 0.66 (0.51–0.85) | 1.33 (0.85–2.08) | 1.70 (1.24–2.32) | 0.67 (0.51–0.88) | ||
| BMI category | |||||||||
| <25 | 501 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 25 to <30 | 642 | 1.36 (0.82–2.26) | 1.23 (0.87–1.75) | 0.95 (0.70–1.28) | 1.24 (0.73–2.10) | 1.13 (0.78–1.63) | 0.95 (0.69–1.30) | ||
| ≥30 | 492 | 1.24 (0.71–2.18) | 1.19 (0.82–1.73) | 1.10 (0.78–1.53) | 1.09 (0.61–1.97) | 1.06 (0.72–1.58) | 1.10 (0.78–1.57) | ||
Model 1: Race, age, sex, BMI, and site; Model 2: model 1 + physical activity, education, income, smoking status, alcohol consumption, dichotomized CES-D score, and medications. When not the exposure of interest, models were adjusted for the remaining covariates.
Sensitivity analyses were also conducted to explore the impact of dopaminergic medications on our findings. Only 10 participants in our sample (0.5% of our total sample) were taking these medications. Results were virtually identical in analyses where we (1) classified these individuals as having RLS and PLMS and (2) excluded them (data not shown).
Associations of demographics and BMI with RLS, PLMI, and PLMAI
Secondary analyses presented in Supplementary Tables provide prevalence ratio estimates for associations of demographic factors and BMI with RLS (Supplementary Table 2), PLMI > 15 (Supplementary Table 3), and PLMAI > 5 (Supplementary Table 4). In contrast to the electro-clinical RLS models where the prevalence was similar, RLS prevalence ratios (PRs) were qualitatively higher among Hispanics compared to whites (PR [95% CI = 1.25 [0.98–1.58]). For Chinese-Americans, the prevalence of RLS was significantly lower compared to whites (0.50 [0.32–0.80]). Women had a 51% higher prevalence of RLS when compared to men (1.51 [1.25–1.81]), which was stronger than the association for electro-clinical RLS. Prevalence ratios for PLMI > 15 were significantly lower for blacks compared to whites (0.46 [0.37–0.58]), higher for individuals >67 years compared to those younger (1.60 [1.37–1.87]), and lower for women (0.79 [0.68–0.91]). Estimates for PLMAI > 5 were 56% lower for blacks compared to whites (0.44 [0.29–0.69]) and 141% higher for participants >67 years (2.41 [1.74–3.35]).
Discussion
Within the context of the community-based MESA cohort, which includes a multi-racial/ethnic sample of older adults, 6.7% of the cohort met criteria for electro-clinical RLS, defined as having both RLS by IRLSSG criteria and PLMS (PLMI > 15). The odds of having electro-clinical RLS were significantly higher among older individuals, but lower for blacks relative to whites. The odds of electro-clinical RLS also tended to be greater among those who were obese, but this relationship was not statistically significant. The prevalence of RLS without PLMS was 16.1%, while the prevalence of PLMS without RLS was 21.1%. A total of 29.5% of those with RLS also had PLMS. Notably, electro-clinical RLS was not associated with sex while RLS alone was more common in women. While RLS-PLMs was no different in Chinese compared to whites, RLS alone was less prevalent in Chinese.
Prevalence of electro-clinical RLS
Of the MESA Sleep participants, 6.7% were classified as having electro-clinical RLS. Comparatively, under the 2003 diagnostic criteria for RLS, prevalence in the study population was 22.8%. Prior studies, which have typically defined RLS without considering PLMS or PLMA, have estimated the prevalence of RLS to be 1%–15% in North American populations [15, 16]. The wide range of estimates generated by these studies may be due, in part, to challenges in consistently eliciting symptoms of RLS, the potential lack of specificity of these symptoms, and the lack of objective measures for RLS. One of the current issues in the estimation of RLS prevalence lies with the subjective nature of using questionnaire data as the diagnostic method [7, 9, 23]. It has been suggested that the inclusion of PLMS to the essential diagnostic criteria for RLS may improve the specificity of RLS diagnoses [9, 10]. The difference seen between the prevalences of electro-clinical RLS (6.7%) and questionnaire-based RLS (22.8%) in the present study supports this notion. Our results also suggest that RLS and/or PLMS phenotypes differ by age, sex, and race/ethnicity, and use of both RLS and PLMS criteria could better characterize individuals with different subphenotypes. Further research is needed to better understand whether the entities of electro-clinical RLS, RLS alone and PLMS alone are distinct, and if so, the optimal measurement approaches for identifying these often overlapping conditions.
Association of electro-clinical RLS with race, sex, age, and BMI
Our results expand upon findings from prior studies that have reported associations between race/ethnicity and RLS [2, 24, 25]. Specifically, in our sample, there was evidence that black participants had lower odds of electro-clinical RLS than white participants, while Hispanic participants had odds similar to those observed in whites. The association in black participants is supported by prior studies done in the United States which suggest that the prevalence of RLS is higher in European-Americans than in African-Americans and Latino Americans [16, 26]. We observed a tendency for lower odds of electro-clinical RLS and a significantly lower odds of RLS in the Chinese-American participants compared to whites, consistent with prior work suggesting that Asian populations may have a significantly lower prevalence of RLS when compared to other racial/ethnic groups [3, 27, 28]. Relatively few studies of RLS have been conducted in East Asian populations; in a Singaporean population, the prevalence of RLS was estimated to be 0.1%, while in a Korean population the estimate ranged from 0.9% to 12.1% [3, 7]. Notably, these studies did not use uniform definitions of RLS.
As in other studies [7, 16, 27], older age was associated with higher odds of RLS in the MESA population. Using the electro-clinical RLS definition, participants aged 67 years and older were found to have 62% greater odds of having RLS, compared to younger study participants. Age is considered a risk factor for RLS with a number of studies suggesting a higher prevalence of RLS in the elderly population [7, 16, 27]. The prevalence of PLMS is also known to increase with advanced age [29].
Sex was not associated with odds of electro-clinical RLS in the MESA sample. Similarly, the prevalences of PLMI and PLMAI did not differ by sex in the MESA sample. Prior studies have reported that females are more likely to experience RLS than are males [3, 7, 16, 27, 28]. When defining RLS by questionnaire only, our data also show that the odds of RLS was higher in females than in males. Yeh et al. [16] have previously warned that sex differences in RLS assessed by questionnaire may reflect differences in thresholds for reporting symptom, with men tending not to report symptoms until they are severe. Alternatively, women may preferentially be affected by pathologies resulting in RLS symptoms but not to the nocturnal manifestations of PLMS.
Contrary to our expectations, obese participants did not have higher odds of electro-clinical RLS. Several recent studies have implicated overweight and obesity as risk factors for the development of RLS and PLMS [1, 16, 17]. For instance, a cross-sectional study by Gao et al. [17] reported a positive correlation between RLS and increasing BMI, with odds of RLS being 42% higher for participants with BMI > 30 compared to those with BMI < 23 kg/m2. A prospective study by De Vito et al. [1] evaluating the association of obesity with RLS also suggested that obese individuals had as much as 57% greater odds of developing RLS. While estimates in our study were somewhat elevated among the overweight (BMI 25 to < 30) and obese (BMI ≥ 30) groups as seen in other studies, these figures were not statistically significant. Notably, our population was aged 54–93 years, which is older than most prior studies that have evaluated this research question. It is possible that unintentional weight loss, which is common among the elderly [30], may have influenced our findings. It is also possible that these findings were influenced by the potential confounding influences of obesity on sleep-disordered breathing. Limb movements are not scored when occurring concurrently with sleep disordered breathing events, which could result in under-estimation of PLMs in individuals with sleep-disordered breathing who tend to be obese.
Physiologic underpinnings of observed associations
Existing literature suggests that the biological mechanism for RLS and PLMS involves dopaminergic dysfunction and iron deficiency. Genetic factors have also been implicated as contributing to RLS [31]. Epidemiological data suggest that levels of iron stored in the body differ by race [32, 33]; which, along with genetic differences, could help explain the differences in both RLS and PLMS prevalence between racial/ethnic groups [5, 34, 35].
There are also plausible biologic mechanisms to explain our finding that the prevalence of electro-clinical RLS was higher in older as opposed to younger individuals. Evidence suggests that the typical age-related increase in levels of brain iron is inhibited in individuals suffering from RLS [3, 16, 36]. Also, recent research has suggested that aging is associated with a loss of dopaminergic function [37]. It is possible that age-related inhibition of brain iron stores may be compounded by a simultaneous reduction in dopaminergic function, leading to a higher prevalence of electro-clinical RLS in older individuals.
Strengths and limitations
This study has several important limitations. The cross-sectional nature of this study prevented us from evaluating incident electro-clinical RLS. Also, in 2012 the IRLSSG adopted new RLS diagnostic criteria [5], which required that conditions that can mimic RLS be ruled out, such as edema, arthritis, leg cramps, or myalgias. Unfortunately, as our data collection began in 2011, we utilized the older IRLSSG criteria [19], and this may have contributed to the high prevalence of RLS without electrical evidence in the present study. Also, the present analysis did not account for severity of RLS. It is also important to consider that only one night of objective PLMS measurement was conducted, which may have led to misclassification given the known night-to-night variability of PLMS [38, 39]. In assessing other factors of interest, this study was limited by an inability to fully evaluate age as a risk factor for electro-clinical RLS, given the older age of our study population (range 54 to 93 years). Additionally, small sample sizes within racial/ethnic groups resulted in low precision for some analyses. This was most noticeable for analyses comparing Chinese Americans to other racial/ethnic groups. Despite these limitations, this study has several strengths. The greatest strengths of our study are the overall large sample size in a racially/ethnically diverse cohort (allowing comparisons of multiple racial/ethnic groups while controlling for location thus limiting the effect of environment), and assessments for both RLS and PLMS. PLMs were recorded using piezo sensors-which, however, have shown to provide reliable estimates of leg movements. By using objective measures to detect PLMS and by incorporating PLMS into the current clinical diagnostic criteria for RLS, a more specific definition of RLS can be utilized. However, individuals experiencing symptoms of RLS without PLMS remain clinically important, and further research is needed to understand this subgroup.
Conclusions
Of the MESA study participants, 6.7% were classified as meeting study criteria for electro-clinical RLS. Odds of electro-clinical RLS were elevated for older compared to younger participants, tended to be lower among participants who were black and of Chinese ancestry, relative to whites, and no different in men compared to women. This study provides some of the first information about the prevalence of electro-clinical RLS in the general population, and how it varies by key demographic characteristics.
Supplementary material
Supplementary material is available at SLEEP online.
Funding
This research was supported by National Heart, Lung, and Blood Institute grants (R01HL098433 [MESA Sleep] and T32-HL069764) and contracts (HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169). Additional support was from grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Acknowledgments
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
MESA data was collected at all MESA sites: St. Paul, Minnesota; Baltimore, Maryland; Chicago, Illinois; Forsyth County, North Carolina; Los Angeles, California; Manhattan, New York. Statistical analysis took place at the University of Minnesota.
Notes
Conflict of interest statement. None declared.
References
- 1. De Vito K, et al. Prospective study of obesity, hypertension, high cholesterol, and risk of restless legs syndrome. Mov Disord. 2014;29(8):1044–1052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ferini-Strambi L, et al. The relationship among restless legs syndrome (Willis-Ekbom Disease), hypertension, cardiovascular disease, and cerebrovascular disease. J Neurol. 2014;261(6):1051–1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Nagandla K, et al. Restless legs syndrome: pathophysiology and modern management. Postgrad Med J. 2013;89(1053):402–410. [DOI] [PubMed] [Google Scholar]
- 4. Walters AS, et al. Review of the relationship of restless legs syndrome and periodic limb movements in sleep to hypertension, heart disease, and stroke. Sleep. 2009;32(5):589–597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Allen RP, et al. ; International Restless Legs Syndrome Study Group 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(8):860–873. [DOI] [PubMed] [Google Scholar]
- 6. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-5. Washington, DC: American Psychiatric Association; 2013. [Google Scholar]
- 7. Rye DB, et al. Restless legs syndrome and periodic leg movements of sleep. Neurol Clin. 2012;30(4):1137–1166. [DOI] [PubMed] [Google Scholar]
- 8. Montplaisir J, et al. Clinical, polysomnographic, and genetic characteristics of restless legs syndrome: a study of 133 patients diagnosed with new standard criteria. Mov Disord. 1997;12(1):61–65. [DOI] [PubMed] [Google Scholar]
- 9. Koo BB. Restless legs syndrome: would you like that with movements or without?Tremor Other Hyperkinet Mov (N Y). 2015;5:316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Benes H, et al. Empirical evaluation of the accuracy of diagnostic criteria for Restless Legs Syndrome. Sleep Med. 2009;10(5):524–530. [DOI] [PubMed] [Google Scholar]
- 11. Stefansson H, et al. A genetic risk factor for periodic limb movements in sleep. N Engl J Med. 2007;357(7):639–647. [DOI] [PubMed] [Google Scholar]
- 12. Winkelmann J, et al. Genome-wide association study of restless legs syndrome identifies common variants in three genomic regions. Nat Genet. 2007;39(8):1000–1006. [DOI] [PubMed] [Google Scholar]
- 13. Winkelman JW. Periodic limb movements in sleep–endophenotype for restless legs syndrome?N Engl J Med. 2007;357(7):703–705. [DOI] [PubMed] [Google Scholar]
- 14. Khan FH, et al. Iron, dopamine, genetics, and hormones in the pathophysiology of restless legs syndrome. J Neurol. 2017;264(8):1634–1641. [DOI] [PubMed] [Google Scholar]
- 15. Venkateshiah SB, et al. Restless legs syndrome. Crit Care Clin. 2015;31(3):459–472. [DOI] [PubMed] [Google Scholar]
- 16. Yeh P, et al. Restless legs syndrome: a comprehensive overview on its epidemiology, risk factors, and treatment. Sleep Breath. 2012;16(4):987–1007. [DOI] [PubMed] [Google Scholar]
- 17. Gao X, et al. Obesity and restless legs syndrome in men and women. Neurology. 2009;72(14):1255–1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Bild DE, et al. Multi-Ethnic Study of Atherosclerosis: objectives and design. Am J Epidemiol. 2002;156(9):871–881. [DOI] [PubMed] [Google Scholar]
- 19. Allen RP, et al. ; Restless Legs Syndrome Diagnosis and Epidemiology workshop at the National Institutes of Health; International Restless Legs Syndrome Study Group 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(2):101–119. [DOI] [PubMed] [Google Scholar]
- 20. Zucconi M, et al. ; International Restless Legs Syndrome Study Group (IRLSSG) The official World Association of Sleep Medicine (WASM) standards for recording and scoring periodic leg movements in sleep (PLMS) and wakefulness (PLMW) developed in collaboration with a task force from the International Restless Legs Syndrome Study Group (IRLSSG). Sleep Med. 2006;7(2):175–183. [DOI] [PubMed] [Google Scholar]
- 21. Claman DM, et al. ; Study of Osteoporotic Fractures Research Group Periodic leg movements are associated with reduced sleep quality in older men: the MrOS Sleep Study. J Clin Sleep Med. 2013;9(11):1109–1117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Radloff LS. The CES-D Scale. A self-report depression for research in the general population. Appl Psychol Measurement 1977;1:385–401. [Google Scholar]
- 23. Salas RE, et al. All the wrong moves: a clinical review of restless legs syndrome, periodic limb movements of sleep and wake, and periodic limb movement disorder. Clin Chest Med. 2010;31(2):383–395. [DOI] [PubMed] [Google Scholar]
- 24. Innes KE, et al. Prevalence of restless legs syndrome in North American and Western European populations: a systematic review. Sleep Med. 2011;12(7):623–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Koo BB, et al. ; Osteoporotic Fractures in Men (MrOS) Study Group Association of incident cardiovascular disease with periodic limb movements during sleep in older men: outcomes of sleep disorders in older men (MrOS) study. Circulation. 2011;124(11):1223–1231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Sawanyawisuth K, et al. Ethnic differences in the prevalence and predictors of restless legs syndrome between Hispanics of Mexican descent and non-Hispanic Whites in San Diego County: a population-based study. J Clin Sleep Med. 2013;9(1):47–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Ekbom K, et al. Restless legs syndrome. J Intern Med. 2009;266(5):419–431. [DOI] [PubMed] [Google Scholar]
- 28. Ohayon MM, et al. Epidemiology of restless legs syndrome: a synthesis of the literature. Sleep Med Rev. 2012;16(4):283–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Ancoli-Israel S, et al. Periodic limb movements in sleep in community-dwelling elderly. Sleep. 1991;14(6):496–500. [DOI] [PubMed] [Google Scholar]
- 30. Stajkovic S, et al. Unintentional weight loss in older adults. CMAJ. 2011;183(4):443–449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Freeman AA, et al. The molecular basis of restless legs syndrome. Curr Opin Neurobiol. 2013;23(5):895–900. [DOI] [PubMed] [Google Scholar]
- 32. McLaren CE, et al. Relationship between transferrin saturation and iron stores in the African American and US Caucasian populations: analysis of data from the third National health and nutrition examination survey. Blood. 2001;98(8):2345–2351. [DOI] [PubMed] [Google Scholar]
- 33. Zacharski LR, et al. Association of age, sex, and race with body iron stores in adults: analysis of NHANES III data. Am Heart J. 2000;140(1):98–104. [DOI] [PubMed] [Google Scholar]
- 34. O’Brien LM, et al. Ethnic difference in periodic limb movements in children. Sleep Med. 2007;8(3):240–246. [DOI] [PubMed] [Google Scholar]
- 35. Scofield H, et al. Periodic limb movements during sleep: population prevalence, clinical correlates, and racial differences. Sleep. 2008;31(9):1221–1227. [PMC free article] [PubMed] [Google Scholar]
- 36. Rizzo G, et al. Imaging brain functional and metabolic changes in restless legs syndrome. Curr Neurol Neurosci Rep. 2013;13(9):372. [DOI] [PubMed] [Google Scholar]
- 37. Villar-Cheda B, et al. Aging-related dysregulation of dopamine and angiotensin receptor interaction. Neurobiol Aging. 2014;35(7):1726–1738. [DOI] [PubMed] [Google Scholar]
- 38. Ferri R, et al. Night-to-night variability of periodic leg movements during sleep in restless legs syndrome and periodic limb movement disorder: comparison between the periodicity index and the PLMS index. Sleep Med. 2013;14(3):293–296. [DOI] [PubMed] [Google Scholar]
- 39. Sforza E, et al. Night-to-night variability in periodic leg movements in patients with restless legs syndrome. Sleep Med. 2005;6(3):259–267. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.