Background
Sleep disturbance caused by obstructive sleep apnea is recognized as a contributing factor to adverse cardiovascular outcomes. However, the effect of restless legs syndrome, another common cause of fragmented sleep, on cardiac structure, function, and long-term outcomes is not known. The aim of this study was to assess the effect of frequent leg movement during sleep on cardiac structure and outcomes in patients with restless legs syndrome.
Methods
In our retrospective study, patients with restless legs syndrome referred for polysomnography were divided into those with frequent (periodic movement index > 35/hour) and infrequent (≤35/hour) leg movement during sleep. Long-term outcomes were determined using Kaplan-Meier and logistic regression models.
Results
Of 584 patients, 47% had a periodic movement index > 35/hour. Despite similarly preserved left ventricular ejection fraction, the group with periodic movement index > 35/hour had significantly higher left ventricular mass and mass index, reflective of left ventricular hypertrophy (LVH). There were no significant baseline differences in the proportion of patients with hypertension, diabetes, hyperlipidemia, prior myocardial infarction, stroke or heart failure, or the use of antihypertensive medications between the groups. Patients with frequent periodic movement index were older, predominantly male, and had more prevalent coronary artery disease and atrial fibrillation. However, on multivariate analysis, periodic movement index > 35/hour remained the strongest predictor of LVH (odds ratio, 2.45; 95% confidence interval, 1.67–3.59; P < .001). Advanced age, female sex, and apnea-hypopnea index were other predictors of LVH. Patients with periodic movement index > 35/hour had significantly higher rates of heart failure and mortality over median 33-month follow-up.
Conclusions
Frequent periodic leg movement during sleep is an independent predictor of severe LVH and is associated with increased cardiovascular morbidity and mortality.
Sleep disturbance is increasingly recognized as a predisposing factor for cardiovascular dysfunction contributing to increased morbidity and mortality, particularly in the elderly. The relationship between obstructive sleep apnea and cardiovascular disease has been well recognized. However, the effect of restless legs syndrome (RLS), another common cause of fragmented sleep that affects >12 million Americans, on cardiac structure and function is not known. More than 80% of patients with RLS have periodic leg movement, particularly during sleep. This movement sometimes occurs throughout the night, resulting in repeated arousals, sleep fragmentation, and increased sympathetic activity, contributing to mental and physical exhaustion with adverse effects on body functions. Fluctuations in blood pressure, heart rate, and heart rate variability have been well described as occurring in association with periodic leg movement during sleep (PLMS). Although higher prevalence rates of coronary artery disease, heart failure, and stroke were reported in patients with RLS with severe symptoms determined by a self-administered questionnaire, the effect of severity of PLMS on cardiac performance and long-term outcomes has not been systematically assessed. The objective of this study was to define the impact of high PLMS frequency, quantified as periodic movement index during polysomnography, on cardiac structure and function and its ability to independently predict adverse cardiovascular outcomes.
Methods
Patient Population
All patients with indications for polysomnography because of sleep disorders and clinical suspicion of RLS as assessed from their medical records (International Classification of Diseases, Ninth Revision, code 333.94) between January 2000 and August 2007 were included in the study. Because clinical diagnosis of RLS requires strict criteria and cannot be assessed from a retrospective review of medical records, the study focused on objective documentation of the severity of periodic leg movements during sleep on overnight polysomnography. Those with baseline echocardiography performed at the Mayo Clinic (Rochester, MN) <6 months before polysomnography were included in the study. Both polysomnographic and echocardiographic studies were performed in the state-of-the-art laboratories at the Mayo Clinic in Rochester, and the data was interpreted by physicians with expertise in sleep medicine and cardiac echocardiography. Patients with end-stage renal disease, severe neuropathy, Parkinson’s disease, and class III or IV heart failure were excluded. Particular attention was paid to confirm diagnosis of hypertension because of its impact on cardiac structure and function. Individual blood pressure recordings documenting hypertension could not be obtained from every patient, but all patient records were manually reviewed, and the diagnosis of hypertension was accepted when confirmed by the primary care physician or cardiologist’s note at three separate occasions and established by the use of medications directed primarily at blood pressure control. The study was approved by the Mayo Clinic Institutional Review Board.
Polysomnography
All participants underwent sleep studies at the sleep laboratory using a computerized polysomnographic device with simultaneous noninvasive acquisition of electrocardiogram, electroencephalogram, electro-oculogram, submental and leg electromyogram, nasal and oral airflow, chest and abdominal respiratory movement, snoring microphone, and body position. Presence of PLMS was defined according to American Sleep Disorders Association criteria as leg movement that is otherwise not explained by insomnia or hypersomnia and after excluding other causes of sleep disturbances, including breathing-related sleep disturbance. PLMS was scored as bursts of muscle activity on the anterior tibialis electromyogram of 0.5 to 5.0 sec in duration, >25% of the amplitude of a voluntary movement, and as part of a series of four or more movements separated by 5 to 90 sec, only if unrelated to voluntary or breathing-related disturbance. PLMS frequency per hour of total sleep time was quantified as the periodic movement index. PLMS was classified as frequent if the periodic movement index was >35/hour, a value we obtained from a receiver operating characteristic curve plotting periodic movement index against the echocardiographic finding of left ventricular (LV) hypertrophy (LVH) (area under the curve, 0.703; 95% confidence interval [CI], 0.664–0.740; P < .001). PLMS followed by an arousal from sleep was considered relevant in the generation of sleep disturbances and classified as movement-related arousal index, calculated as the number of PLMS-related arousals per hour of total sleep time. Obstructive and central apnea were defined by a reduction in airflow to <20% of baseline lasting ≥10 sec, with or without respiratory effort, respectively, and hypopnea as a reduction in airflow by 50% for ≥10 sec accompanied by ≥4% oxygen desaturation or arousal or both. Apnea-hypopnea index was defined as the sum of apnea and hypopnea episodes per hour of total sleep time. Arousals due to breathing-related incidents were quantified per hour of total sleep time as breathing-related arousal index. The overall arousal index per hour of total sleep time was calculated as the sum of movement-related arousal index and breathing-related arousal index.
Echocardiography
M-mode and two-dimensional echocardiographic studies were performed from standard transthoracic windows using Acuson Sequoia (Siemens Medical Solutions, Malvern, PA) and Vivid 7 (GE Healthcare, Waukesha, WI) equipment, according to the recommendations of the American Society of Echocardiography (ASE). Echocardiographic studies were performed by experienced sonographers and interpreted by cardiologists with expertise in echocardiography at the Mayo Clinic in Rochester. LV dimensions, including diameters and the end-diastolic thicknesses of the interventricular septum and posterior wall, were measured from M-mode images in the standard parasternal long-axis view. LV mass was calculated from LV linear dimensions using the Devereux formula and indexed to body surface area. Relative wall thickness was obtained using the formula (2 × posterior wall thickness in diastole)/LV diastolic dimensions. Relative wall thickness was categorized as concentric if >0.42 or eccentric if ≤0.42. LVH was defined according to criteria outlined by the ASE, with LV mass index (i.e., LV mass adjusted for body surface area) >88 g/m 2 in women and >102 g/m 2 in men by two-dimensional echocardiography and LV mass >150 g in women and >200 g in men. Moderate to severe LVH was determined in a sex-specific manner per ASE guidelines (LV mass index >100 g/m 2 in women and >116 g/m 2 in men). LV ejection fraction was calculated using Simpson’s rule, and LV diastolic function was evaluated using early (E-wave) and late (A-wave) diastolic transmitral velocities, the E/A ratio, and the E-wave deceleration time obtained from the spectral pulsed-wave Doppler recordings at the tip of mitral valve leaflets in the apical four-chamber view. Apical four-chamber and two-chamber views were used to determine left atrial area and length. Left atrial volume was calculated per ASE criteria using the algorithm (0.85 × four-chamber left atrial area × two-chamber or apical long-axis left atrial area)/average of the two lengths obtained from orthogonal views. Volume was then indexed to body surface area.
Follow-Up
Patients were divided into two groups on the basis of the frequency of PLMS: periodic movement index >35/hour was classified as frequent and periodic movement index ≤35/hour as infrequent. Follow-up information was obtained from a comprehensive medical record system. Baseline clinical, echocardiographic, and polysomnographic parameters were compared between the two groups to correlate the severity of PLMS with cardiac structural and functional alterations. Differences in cardiovascular end points, including incident heart failure, recurrent (two or more times) cardiac hospitalization, and death, were compared between the two PLMS groups. Cardiac hospitalization was defined as hospital admission for any of the following primary diagnoses: heart failure, dysrhythmia, and myocardial ischemia or infarction. Predictors of LVH and long-term outcomes were also determined.
Statistical Analysis
The differences between patient groups were tested using Wilcoxon’s rank-sum test for continuous variables and χ 2 tests or Fisher’s exact tests for categorical variables. Continuous variables are summarized as medians with interquartile ranges and discrete variables as numbers and percentages. Predictors of outcomes were identified using logistic regression analysis to calculate odds ratios (ORs) with 95% CIs and to examine the effect of each variable on adverse cardiovascular events. All values were two tailed, and P values < .05 were considered statistically significant. Survival of patients free of heart failure, death, and rehospitalization was estimated using the Kaplan-Meier method, and differences in survival were compared using log-rank tests. All statistical analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC).
Results
Patient Population
Of 584 patients with RLS referred for polysomnography who also had baseline echocardiograms, 274 (47%) had periodic movement index >35/hour. Table 1 summarizes baseline characteristics of the study population. No significant differences were present between the groups in the proportions of patients with prior diagnoses of hypertension, use of antihypertensive medications, diabetes mellitus, hyperlipidemia, prior myocardial infarctions, heart failure, stroke, and abnormal LV ejection fractions at the time of polysomnography. Patients with higher periodic movement index were older ( P < .001), predominantly male ( P < .001), had lower body mass index ( P = .02), and had higher rates of coronary artery disease ( P < .01) and atrial fibrillation ( P < .01).
| Characteristic | Overall | Periodic movement index ≤35/h | Periodic movement index >35/h | P |
|---|---|---|---|---|
| n (%) | 584 | 310 (53.1%) | 274 (46.9%) | |
| Age (y) | 65 (55–72) | 61 (52–70) | 68 (61–75) | <.001 |
| Women | 298 (49.0%) | 183 (59.0%) | 103 (37.6%) | <.001 |
| Body mass index (kg/m 2 ) | 33.8 ± 7.9 | 34.5 ± 8.2 | 33 ± 7.4 | .02 |
| Body surface area (m 2 ) | 2.07 ± 0.25 | 2.06 ± 0.26 | 2.07 ± 0.24 | .65 |
| Hypertension | 399 (68.3%) | 209 (67.4%) | 190 (69.3%) | .62 |
| Diabetes | 222 (38.0%) | 114 (36.8%) | 108 (39.4%) | .51 |
| Hyperlipidemia | 343 (58.7%) | 175 (56.5%) | 168 (61.3%) | .23 |
| Coronary artery disease | 216 (37.0%) | 98 (31.6%) | 118 (43.1%) | <.01 |
| Myocardial infarction | 125 (21.4%) | 68 (21.9%) | 57 (20.8%) | .74 |
| Heart failure | 93 (15.9%) | 46 (14.8%) | 47 (17.2%) | .45 |
| Atrial fibrillation | 143 (24.5%) | 61 (19.7%) | 82 (29.9%) | <.01 |
| Aspirin | 276 (47%) | 138 (44%) | 138 (51%) | .12 |
| ACE inhibitors/ARBs | 212 (36%) | 104 (33%) | 108 (40%) | .27 |
| β-blockers | 297 (51%) | 149 (58%) | 148 (54%) | .11 |
| Calcium channel blockers | 134 (23%) | 71 (23%) | 63 (23%) | .91 |
| Diuretics | 239 (41%) | 124 (40%) | 115 (42%) | .54 |
| Nitrates | 68 (12%) | 35 (11%) | 33 (12%) | .73 |
| Periodic movement index (/h) | 32 (3–72) | 3 (0–14) | 76 (52–107) | <.001 |
| Total sleep time < 240 min | 419 (71.8%) | 205 (66.1%) | 214 (78.1%) | <.01 |
| Arousal index (/h) | 32 (21–51) | 28 (17–46) | 38 (25–56) | <.001 |
| Movement related | 14% (0%–37%) | 2% (0%–14%) | 32% (17%–51%) | <.001 |
| Breathing related | 54% (27%–78%) | 57% (29%–84%) | 51% (23%–70%) | <.001 |
| Apnea-hypopnea index (/h) | 9 (3–25) | 6 (2–19) | 12 (5–29) | <.001 |
Differences in Polysomnographic Variables
Results of the overnight polysomnographic parameters are summarized in Table 1 . A higher proportion of patients with frequent PLMS had shorter sleep times (≤4 hours) compared with those with infrequent PLMS ( P = .001), with a marked increase in the number of arousals related to movement ( P = .001). The mean apnea-hypopnea index was also higher in patients with periodic movement index >35/hour, but breathing-related arousals were more common in those with infrequent periodic movement ( P = .001).
Cardiac Structural and Functional Differences
The effect of frequent PLMS on cardiac structure and function was determined by echocardiography. Despite similarly preserved LV ejection fractions between those with frequent and infrequent PLMS ( P = .12), significant differences existed in LV mass ( P < .001) and LV mass index ( P < .001) as well as interventricular septal wall and posterior wall thicknesses ( P < .001), as summarized in Table 2 . The proportion of patients with LVH and moderate to severe LVH was significantly higher in those with frequent PLMS ( P < .001). Internal LV end-diastolic and end-systolic dimensions were within the normal range of the population but significantly higher in the group with periodic movement index >35/hour ( P < .001). The overall pattern of ventricular geometry in patients with hypertrophy indicated a greater percentage of eccentric LVH in the group with periodic movement index >35/hour ( P < .001), whereas the group with periodic movement index ≤35/hour had a predominantly concentric form of hypertrophy ( P < .001) ( Figure 1 ). A trend toward a higher prevalence of patients with concentric remodeling also was observed in the group with periodic movement index ≤35/hour group (27% vs 20% in the frequent movement group, P = .08). In patients with frequent PLMS, the mean relative wall thickness was 0.43 ± 0.08, compared with 0.44 ± 0.09 in patients with infrequent PLMS. Both groups had increased left atrial volume index, but the magnitude of atrial enlargement was significantly higher in the group with frequent periodic movement ( P < .001), which also had a higher prevalence of atrial fibrillation ( Table 1 ). There were no significant differences between both groups in the number of patients with permanent atrial fibrillation (five in the frequent group vs four in the infrequent group) at the time of echocardiography.
| Characteristic | Overall | Periodic movement index ≤35/h | Periodic movement index >35/h | P |
|---|---|---|---|---|
| n (%) | 584 | 310 (53.1%) | 274 (46.9%) | |
| LV ejection fraction (%) | 63 (58–67) | 64 (58–67) | 62 (56–66) | .12 |
| LV mass (g) | 194 (163–241) | 185 (153–219) | 213 (174–278) | <.001 |
| LV mass index (g/m 2 ) | 94 (79–117) | 89 (76–105) | 104 (86–131) | <.001 |
| Septal thickness (mm) | 11 (10–12) | 10 (10–12) | 11 (10–12) | <.001 |
| Posterior wall thickness (mm) | 10 (9–12) | 10 (9–11) | 11 (10–12) | <.001 |
| LV diastolic dimension (mm) | 49 (46–54) | 49 (45–53) | 50 (47–55) | <.001 |
| LV systolic dimension (mm) | 31 (28–36) | 30 (27–34) | 32 (29–38) | <.001 |
| Left atrial volume index (cm 3 /m 2 ) | 32 (28–39) | 32 (27–36) | 34 (29–44) | <.001 |
| LVH (gender specific) ∗ | 292 (50%) | 127 (41%) | 165 (60%) | <.001 |
| Moderate to severe LVH (gender specific) † | 198 (33.9%) | 74 (23.9%) | 124 (45.3%) | <.001 |
∗ LV mass index >88 g/m 2 in women and >102 g/m 2 in men.
Predictors of LVH
Univariate and multivariate predictors of moderate to severe LVH in patients with RLS as well as ORs, 95% CIs, and P values are summarized in Table 3 . Significant predictors of moderate to severe LVH on univariate analysis were periodic movement index >35/hour, total sleep time <4 hours, atrial fibrillation, age, and apnea-hypopnea index. After correction for age, sex, and other risk factors for hypertrophy by logistic regression, periodic movement index >35/hour remained the strongest independent predictor of LVH severity (OR, 2.45; 95% CI, 1.67–3.58; P < .001). Age, female sex, and apnea-hypopnea index were also independently predictive of severe LVH.
| Characteristic | Univariate | Multivariate | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P | OR | 95% CI | P | |
| Age | 1.04 | 1.02–1.06 | <.001 | 1.03 | 1.01–1.04 | .003 |
| Female gender | 1.06 | 0.76–1.50 | .722 | 1.64 | 1.11–2.43 | .013 |
| Hypertension | 1.37 | 0.94–2.00 | .102 | 1.11 | 0.73–1.69 | .605 |
| Diabetes | 0.99 | 0.70–1.41 | .991 | 0.98 | 0.67–1.45 | .937 |
| Myocardial infarction | 1.08 | 0.71–1.63 | .730 | 1.12 | 0.72–1.75 | .617 |
| Atrial fibrillation | 1.53 | 1.03–2.25 | .033 | 1.38 | 0.89–2.15 | .149 |
| Total sleep time <240 min | 1.60 | 1.07–2.38 | .021 | 1.03 | 0.66–1.61 | .896 |
| Periodic movement index >35/hr | 2.64 | 1.85–3.75 | <.001 | 2.45 | 1.67–3.59 | <.001 |
| Apnea-hypopnea index | 1.01 | 1.01–1.02 | <.001 | 1.01 | 1.01–1.02 | .004 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
