CLINICAL RESEARCH STUDY
Restless Legs Syndrome and Cognitive Function: A Population-based Cross-sectional Study Pamela M. Rist, ScD,a,b Alexis Elbaz, MD, PhD,c,d Carole Dufouil, PhD,e,f Christophe Tzourio, MD, PhD,e,f,1 Tobias Kurth, MD, ScDa,b,e,f,1 a
Division of Preventive Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass; Department of Epidemiology, Harvard TH Chan School of Public Health, Boston, Mass; cSocial and Occupational Determinants of Health, INSERM Centre for Research in Epidemiology and Population Health, Villejuif, France; dUniversité de Versailles St-Quentin, UMRS 1018, Versailles, France; eTeam Neuroepidemiology, INSERM Research Center for Epidemiology and Biostatistics, Bordeaux, France; fCollege of Health Sciences, University of Bordeaux, Bordeaux, France. b
ABSTRACT BACKGROUND: Restless legs syndrome has been speculated to be linked to cognitive impairment through vascular risk factors or through its effect on sleep deprivation. Previous studies on the association between restless legs syndrome and cognitive function have been inconclusive. We performed a cross-sectional analysis of the association between restless legs syndrome and cognitive function using data from a large population-based study of elderly individuals residing in France. METHODS: We used information from 2070 individuals from the Dijon, France center of the Three-City study who had available information on restless legs syndrome and cognitive functioning measures. Restless legs syndrome was assessed using the 4 minimal diagnostic criteria of the International Restless Legs Study Group. During the same wave in which restless legs syndrome status was assessed, cognitive functions also were assessed using 4 tests: Isaacs’ test of verbal/category fluency, the Benton Visual Retention Test, the Trail Making Test B, and the Mini-Mental State Examination. We created a summary global cognitive score by summing the z scores for the 4 tests and used analysis of covariance to explore the association between restless legs syndrome and cognitive function. RESULTS: We did not observe any statistically significant differences in any cognitive z-score between those with restless legs syndrome and those without restless legs syndrome. The mean global z-score after multivariate adjustment was 0.003 (SE 0.173) for those with restless legs syndrome and 0.007 (SE 0.129) for those without restless legs syndrome (P-value ¼ .98). CONCLUSION: Data from this large, population-based study do not suggest that restless legs syndrome is associated with prevalent cognitive deficits in elderly individuals. Ó 2015 Elsevier Inc. All rights reserved. The American Journal of Medicine (2015) -, --KEYWORDS: Cognitive function; Epidemiology; Restless legs syndrome
Restless legs syndrome is a neurological disorder characterized by an urge to move the legs and unpleasant leg sensations, which usually are experienced in the evening or Funding: See last page of article. Conflict of Interest: See last page of article. Authorship: See last page of article. Requests for reprints should be addressed to Tobias Kurth, MD, ScD, INSERM Research Center for Epidemiology and Biostatistics (U897) Team Neuroepidemiology, University of Bordeaux, 146 rue Léo Saignat CS61292, Bordeaux 33076, France. E-mail address: [email protected] 1 These authors contributed equally to this manuscript. 0002-9343/$ -see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjmed.2015.04.011
at night while the individual is at rest. Population-based studies using the minimal diagnostic criteria developed by the International Restless Legs Syndrome Study Group found the prevalence of restless legs syndrome to range from 4% to 29% (average of 14.5%).1-3 The prevalence of the disease is approximately twice as high in women as it is in men, and increases with age.4 Previous studies have shown associations between restless legs syndrome and many cardiovascular risk factors including smoking,5-9 diabetes,6,8,10 hypercholerolemia,8,10 exercise,10 body mass index (BMI)5,10-12 and hypertension.6,13,14 Many of these factors are also risk factors for cognitive impairment or
2
The American Journal of Medicine, Vol -, No -,
-
2015
dementia, which has led to speculation that restless legs Response options were “yes” or “no.” If the participant syndrome may also be associated with cognitive function. In responded “yes,” he or she was asked further: “Do these addition to the potential link between restless legs syndrome unpleasant sensations occur solely or mainly at rest (when and cognition through vascular pathways, the sleep depriyou are sitting or lying down, without moving your legs) vation and poor quality sleep experienced by individuals and do they improve with movement?” and “Are these unwith restless legs syndrome also may be associated with pleasant sensations more intense in the evening or at night poor cognitive performance. than in the morning?” Response Previous studies have examined options for these questions were CLINICAL SIGNIFICANCE the association between restless legs “yes” or “no.” If the participant syndrome and cognitive perforresponded “yes” to all 3 questions, Restless legs syndrome may be linked to mance and have provided conflicthe or she was considered as havcognitive impairment through vascular ing results. Some have observed ing restless legs syndrome. For risk factors or through its effect on sleep deficits in some cognitive dothose respondents who experideprivation. 15-18 mains, but others have found no enced restless legs syndrome Using a population-based cohort of association between restless legs within the past 12 months, we also syndrome and any cognitive funcasked about the frequency of elderly individuals, we examined the tion measure.19,20 Most previous symptoms. Possible response opcross-sectional association between studies were small, with only one tions were: at least once a year but restless legs syndrome prevalence and study having more than 100 particless than once a month, once a cognitive function. ipants with restless legs synmonth, 2 to 4 times per month, 2 This large population-based study does drome.16 Given the discrepant to 3 times per week, 4 to 5 times not support any association between per week, and 6 to 7 times per results and small size of previous week. We collapsed these studies, we performed a crossrestless legs syndrome and existing response options into 2 categories: sectional analysis of the associacognitive deficits across a wide range of 4 or fewer times per month (low tion between restless legs syndrome cognitive domains. frequency) and 2 or more times and cognitive function using data per week (high frequency). from a large population-based study of elderly individuals residing in France.
Cognitive Assessment METHODS The Three-City (3C) Study is a longitudinal cohort study designed to estimate the risk of dementia and cognitive impairment attributable to vascular risk factors.21 The study enrolled (1999-2001) subjects living in 3 French cities (Bordeaux, Dijon, and Montpellier). The present analysis uses data only from subjects living in Dijon, where information about restless legs syndrome status and cognitive function are available. Individuals aged 65 years or older who were living in Dijon, registered on the electoral rolls in 1999, and not institutionalized, were eligible to be included in the study. Each subject provided informed consent and the Ethical Committee of the University Hospital of Kremlin-Bicêtre approved the methods and procedures of the 3C study.
Restless Legs Syndrome Assessment Information on restless legs status was first assessed during the fifth (2008-2009) and sixth (2010-2012) waves of the study. Participants were asked to report the 4 minimal diagnostic criteria of the International Restless Legs Study Group, which have been established and validated in previous studies.22,23 The first question was: “Have you ever felt unpleasant sensation in the legs (restlessness, tingling, tension, annoyances, contractions, twitching, numbness, electricity, etc.) with the irresistible need or want to move?”
During the same study waves in which restless legs syndrome status was assessed, cognitive function also was assessed using a neuropsychological battery that included 4 tests: the Isaacs’ test of verbal/category fluency (Isaac),24 the Benton Visual Retention Test (BVRT),25 the Trail Making Test B (TMTB)26 and the Mini-Mental State Examination (MMSE).27 We calculated TMTB scores by calculating the ratio of time to task completion divided by the number of correct connections. Additionally, we capped time at 300 seconds and total score at 35 (99th percentile at baseline) to reduce the influence of outlying scores. All test results were converted to z-scores using the mean and SD of the original scores of the participants eligible for this study. Because higher times imply worse performance, we multiplied z-scores for TMTB by 1. Additionally, we created a summary global cognitive score by summing the z-scores for the Isaac, BVRT, TMTB, and MMSE tests.
Covariates Trained psychologists collected sociodemographic and medical data on participants during home visits at baseline and during study examination center visits or home visits at the fifth and sixth waves of the study. Information from the baseline and fifth and sixth wave questionnaires were used to determine the participant’s medical history.
Rist et al
Restless Legs Syndrome and Cognitive Function
3
History of cardiovascular disease was defined as a history of myocardial infarction, stroke, angina, percutaneous transluminal coronary angioplasty, or coronary artery bypass surgery. History of diabetes was defined as glycemia 7 mmol/L or use of antidiabetic treatment.28 History of hypertension was defined as measured systolic blood pressure 140 mm Hg or measured diastolic blood pressure 90 mm Hg or antihypertensive treatment. History of high cholesterol was defined as treatment with cholesterollowering medication or having a total cholesterol level of 6.2 mmol/L. History of depression symptoms was defined using the Center for Epidemiological Studies-Depression scale29 with a cutoff of 17 points for men and 23 points for women, as has been done previously.30 For BMI, smoking status, alcohol consumption, and physical activity, we used values from the follow-up wave in which restless legs syndrome was assessed. In the event that this value was missing, we used values from baseline. BMI was calculated from measured height and weight and was categorized as normal weight (12 to 24 grams/day, or drinking >24 grams/day. Physical activity was defined as active vs nonactive because changes in the questionnaires on physical activity precluded more detailed physical activity categorization. Education was assessed at baseline and categorized as low (no education or primary school), medium low (middle school or technical or professional short degree), medium high (secondary level without diploma or technical or professional long degree or baccalaureate), or high (university level).
and high frequency of restless legs syndrome symptoms to those with no history of restless legs syndrome. Most subjects were not missing information on any covariates. Only 2 subjects were missing information on history of high cholesterol and were assigned to having high cholesterol (the most common category). Only one person was missing information on depression status and was classified as not depressed (the most common category). Because over 100 subjects were missing information on migraine status, we created a separate category for missing migraine information. All analyses were performed in SAS
Statistical Analyses We used analysis of covariance (ANCOVA) to explore the association between restless legs syndrome and cognitive function (outcome). Our first model (Model 1) adjusted for age at restless legs syndrome and cognitive assessment and sex. Our second set of models (Model 2) adjusted for age, sex, smoking status, alcohol consumption, physical activity, BMI categories, history of high blood pressure, history of high cholesterol, history of diabetes, history of cardiovascular disease, and education. Because it has been hypothesized that restless legs syndrome may impact cognitive function through its effects on sleep, we performed a sensitivity analysis in which we adjusted for self-reported difficulty sleeping (never, rarely, regularly, often), quality of sleep (good, medium, or mediocre/bad), and intake of sleep medication (yes/no), as well as all variables in Model 2. Restless legs syndrome may also impact cognitive function through its association with depression,31 so we performed another sensitivity analysis in which we adjusted for history of depression (yes/no) in addition to all variables in Model 2. We performed another sensitivity analysis in which we compared those with low
Table 1
Characteristics of Participants by Restless Legs Status
Characteristic Age at restless legs assessment and cognitive interview (mean, SD) Sex (% female) Smoking status (%) Never Past Current Alcohol consumption (%) Nondrinker 0 to 12 g/day 12 to 24 g/day >24 g/day Physically active (%) Body mass index
Restless Legs Syndrome and Cognitive Function: A Population-based Cross-sectional Study Pamela M. Rist, ScD,a,b Alexis Elbaz, MD, PhD,c,d Carole Dufouil, PhD,e,f Christophe Tzourio, MD, PhD,e,f,1 Tobias Kurth, MD, ScDa,b,e,f,1 a
Division of Preventive Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass; Department of Epidemiology, Harvard TH Chan School of Public Health, Boston, Mass; cSocial and Occupational Determinants of Health, INSERM Centre for Research in Epidemiology and Population Health, Villejuif, France; dUniversité de Versailles St-Quentin, UMRS 1018, Versailles, France; eTeam Neuroepidemiology, INSERM Research Center for Epidemiology and Biostatistics, Bordeaux, France; fCollege of Health Sciences, University of Bordeaux, Bordeaux, France. b
ABSTRACT BACKGROUND: Restless legs syndrome has been speculated to be linked to cognitive impairment through vascular risk factors or through its effect on sleep deprivation. Previous studies on the association between restless legs syndrome and cognitive function have been inconclusive. We performed a cross-sectional analysis of the association between restless legs syndrome and cognitive function using data from a large population-based study of elderly individuals residing in France. METHODS: We used information from 2070 individuals from the Dijon, France center of the Three-City study who had available information on restless legs syndrome and cognitive functioning measures. Restless legs syndrome was assessed using the 4 minimal diagnostic criteria of the International Restless Legs Study Group. During the same wave in which restless legs syndrome status was assessed, cognitive functions also were assessed using 4 tests: Isaacs’ test of verbal/category fluency, the Benton Visual Retention Test, the Trail Making Test B, and the Mini-Mental State Examination. We created a summary global cognitive score by summing the z scores for the 4 tests and used analysis of covariance to explore the association between restless legs syndrome and cognitive function. RESULTS: We did not observe any statistically significant differences in any cognitive z-score between those with restless legs syndrome and those without restless legs syndrome. The mean global z-score after multivariate adjustment was 0.003 (SE 0.173) for those with restless legs syndrome and 0.007 (SE 0.129) for those without restless legs syndrome (P-value ¼ .98). CONCLUSION: Data from this large, population-based study do not suggest that restless legs syndrome is associated with prevalent cognitive deficits in elderly individuals. Ó 2015 Elsevier Inc. All rights reserved. The American Journal of Medicine (2015) -, --KEYWORDS: Cognitive function; Epidemiology; Restless legs syndrome
Restless legs syndrome is a neurological disorder characterized by an urge to move the legs and unpleasant leg sensations, which usually are experienced in the evening or Funding: See last page of article. Conflict of Interest: See last page of article. Authorship: See last page of article. Requests for reprints should be addressed to Tobias Kurth, MD, ScD, INSERM Research Center for Epidemiology and Biostatistics (U897) Team Neuroepidemiology, University of Bordeaux, 146 rue Léo Saignat CS61292, Bordeaux 33076, France. E-mail address: [email protected] 1 These authors contributed equally to this manuscript. 0002-9343/$ -see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjmed.2015.04.011
at night while the individual is at rest. Population-based studies using the minimal diagnostic criteria developed by the International Restless Legs Syndrome Study Group found the prevalence of restless legs syndrome to range from 4% to 29% (average of 14.5%).1-3 The prevalence of the disease is approximately twice as high in women as it is in men, and increases with age.4 Previous studies have shown associations between restless legs syndrome and many cardiovascular risk factors including smoking,5-9 diabetes,6,8,10 hypercholerolemia,8,10 exercise,10 body mass index (BMI)5,10-12 and hypertension.6,13,14 Many of these factors are also risk factors for cognitive impairment or
2
The American Journal of Medicine, Vol -, No -,
-
2015
dementia, which has led to speculation that restless legs Response options were “yes” or “no.” If the participant syndrome may also be associated with cognitive function. In responded “yes,” he or she was asked further: “Do these addition to the potential link between restless legs syndrome unpleasant sensations occur solely or mainly at rest (when and cognition through vascular pathways, the sleep depriyou are sitting or lying down, without moving your legs) vation and poor quality sleep experienced by individuals and do they improve with movement?” and “Are these unwith restless legs syndrome also may be associated with pleasant sensations more intense in the evening or at night poor cognitive performance. than in the morning?” Response Previous studies have examined options for these questions were CLINICAL SIGNIFICANCE the association between restless legs “yes” or “no.” If the participant syndrome and cognitive perforresponded “yes” to all 3 questions, Restless legs syndrome may be linked to mance and have provided conflicthe or she was considered as havcognitive impairment through vascular ing results. Some have observed ing restless legs syndrome. For risk factors or through its effect on sleep deficits in some cognitive dothose respondents who experideprivation. 15-18 mains, but others have found no enced restless legs syndrome Using a population-based cohort of association between restless legs within the past 12 months, we also syndrome and any cognitive funcasked about the frequency of elderly individuals, we examined the tion measure.19,20 Most previous symptoms. Possible response opcross-sectional association between studies were small, with only one tions were: at least once a year but restless legs syndrome prevalence and study having more than 100 particless than once a month, once a cognitive function. ipants with restless legs synmonth, 2 to 4 times per month, 2 This large population-based study does drome.16 Given the discrepant to 3 times per week, 4 to 5 times not support any association between per week, and 6 to 7 times per results and small size of previous week. We collapsed these studies, we performed a crossrestless legs syndrome and existing response options into 2 categories: sectional analysis of the associacognitive deficits across a wide range of 4 or fewer times per month (low tion between restless legs syndrome cognitive domains. frequency) and 2 or more times and cognitive function using data per week (high frequency). from a large population-based study of elderly individuals residing in France.
Cognitive Assessment METHODS The Three-City (3C) Study is a longitudinal cohort study designed to estimate the risk of dementia and cognitive impairment attributable to vascular risk factors.21 The study enrolled (1999-2001) subjects living in 3 French cities (Bordeaux, Dijon, and Montpellier). The present analysis uses data only from subjects living in Dijon, where information about restless legs syndrome status and cognitive function are available. Individuals aged 65 years or older who were living in Dijon, registered on the electoral rolls in 1999, and not institutionalized, were eligible to be included in the study. Each subject provided informed consent and the Ethical Committee of the University Hospital of Kremlin-Bicêtre approved the methods and procedures of the 3C study.
Restless Legs Syndrome Assessment Information on restless legs status was first assessed during the fifth (2008-2009) and sixth (2010-2012) waves of the study. Participants were asked to report the 4 minimal diagnostic criteria of the International Restless Legs Study Group, which have been established and validated in previous studies.22,23 The first question was: “Have you ever felt unpleasant sensation in the legs (restlessness, tingling, tension, annoyances, contractions, twitching, numbness, electricity, etc.) with the irresistible need or want to move?”
During the same study waves in which restless legs syndrome status was assessed, cognitive function also was assessed using a neuropsychological battery that included 4 tests: the Isaacs’ test of verbal/category fluency (Isaac),24 the Benton Visual Retention Test (BVRT),25 the Trail Making Test B (TMTB)26 and the Mini-Mental State Examination (MMSE).27 We calculated TMTB scores by calculating the ratio of time to task completion divided by the number of correct connections. Additionally, we capped time at 300 seconds and total score at 35 (99th percentile at baseline) to reduce the influence of outlying scores. All test results were converted to z-scores using the mean and SD of the original scores of the participants eligible for this study. Because higher times imply worse performance, we multiplied z-scores for TMTB by 1. Additionally, we created a summary global cognitive score by summing the z-scores for the Isaac, BVRT, TMTB, and MMSE tests.
Covariates Trained psychologists collected sociodemographic and medical data on participants during home visits at baseline and during study examination center visits or home visits at the fifth and sixth waves of the study. Information from the baseline and fifth and sixth wave questionnaires were used to determine the participant’s medical history.
Rist et al
Restless Legs Syndrome and Cognitive Function
3
History of cardiovascular disease was defined as a history of myocardial infarction, stroke, angina, percutaneous transluminal coronary angioplasty, or coronary artery bypass surgery. History of diabetes was defined as glycemia 7 mmol/L or use of antidiabetic treatment.28 History of hypertension was defined as measured systolic blood pressure 140 mm Hg or measured diastolic blood pressure 90 mm Hg or antihypertensive treatment. History of high cholesterol was defined as treatment with cholesterollowering medication or having a total cholesterol level of 6.2 mmol/L. History of depression symptoms was defined using the Center for Epidemiological Studies-Depression scale29 with a cutoff of 17 points for men and 23 points for women, as has been done previously.30 For BMI, smoking status, alcohol consumption, and physical activity, we used values from the follow-up wave in which restless legs syndrome was assessed. In the event that this value was missing, we used values from baseline. BMI was calculated from measured height and weight and was categorized as normal weight (12 to 24 grams/day, or drinking >24 grams/day. Physical activity was defined as active vs nonactive because changes in the questionnaires on physical activity precluded more detailed physical activity categorization. Education was assessed at baseline and categorized as low (no education or primary school), medium low (middle school or technical or professional short degree), medium high (secondary level without diploma or technical or professional long degree or baccalaureate), or high (university level).
and high frequency of restless legs syndrome symptoms to those with no history of restless legs syndrome. Most subjects were not missing information on any covariates. Only 2 subjects were missing information on history of high cholesterol and were assigned to having high cholesterol (the most common category). Only one person was missing information on depression status and was classified as not depressed (the most common category). Because over 100 subjects were missing information on migraine status, we created a separate category for missing migraine information. All analyses were performed in SAS
Statistical Analyses We used analysis of covariance (ANCOVA) to explore the association between restless legs syndrome and cognitive function (outcome). Our first model (Model 1) adjusted for age at restless legs syndrome and cognitive assessment and sex. Our second set of models (Model 2) adjusted for age, sex, smoking status, alcohol consumption, physical activity, BMI categories, history of high blood pressure, history of high cholesterol, history of diabetes, history of cardiovascular disease, and education. Because it has been hypothesized that restless legs syndrome may impact cognitive function through its effects on sleep, we performed a sensitivity analysis in which we adjusted for self-reported difficulty sleeping (never, rarely, regularly, often), quality of sleep (good, medium, or mediocre/bad), and intake of sleep medication (yes/no), as well as all variables in Model 2. Restless legs syndrome may also impact cognitive function through its association with depression,31 so we performed another sensitivity analysis in which we adjusted for history of depression (yes/no) in addition to all variables in Model 2. We performed another sensitivity analysis in which we compared those with low
Table 1
Characteristics of Participants by Restless Legs Status
Characteristic Age at restless legs assessment and cognitive interview (mean, SD) Sex (% female) Smoking status (%) Never Past Current Alcohol consumption (%) Nondrinker 0 to 12 g/day 12 to 24 g/day >24 g/day Physically active (%) Body mass index
