|Home | About | Journals | Submit | Contact Us | Français|
To examine risk factors for REM sleep behavior disorder (RBD) in a large-scale community-based study.
This community-based study included 12,784 Chinese adults (10,556 men and 2,228 women, aged 24 years or older) who were free of Parkinson disease and dementia in 2012. Probable RBD (pRBD) status was determined by a validated questionnaire (Chinese RBD questionnaire–Hong Kong) in 2012. Potential risk factors—including age, sex, smoking, socioeconomic status, physical activity, obesity, consumption of tea (surrogate for caffeine intake) and alcohol, serum concentrations of lipids and glucose, and chronic disease status—were assessed in 2006. Logistic regression was used to calculate odds ratios and 95% confidence intervals and to test differences in prevalence of pRBD across exposures.
Prevalence of pRBD was 5.9% in men and 4.1% in women. In the fully adjusted model, risk factors that were significantly associated with a higher risk of having pRBD included lower education level, coal mining and other blue collar occupation, lower physical activity level, diabetes or prediabetes, lower body mass index, head injury, higher low-density lipoprotein level, and chronic olfactory and taste dysfunction. In sensitivity analyses, restricting to pRBD cases with symptom onset within 1 year or excluding coal miners or those with history of head injury generated similar results.
We found several potential risk factors for pRBD, including socioeconomic status, head injury, olfactory and taste dysfunction, and various cardiovascular risk factors. Future prospective studies to establish the temporal relationship between these potential risk factors and RBD are warranted.
REM sleep behavior disorder (RBD) is characterized by symptoms of dream enactment and loss of muscle atonia during REM sleep as confirmed by polysomnography. RBD may precede the emergence of neurodegenerative disorders by more than a decade, and may be a risk factor for Parkinson disease (PD), dementia with Lewy bodies (DLB), and multiple system atrophy.1,–3 For example, in a study of patients with RBD without prior neurodegenerative diseases, about one-third developed parkinsonism or dementia over 5 years.4 Similarly, in another study of cognitively normal adults, those with probable RBD (pRBD) at baseline, i.e., positive replies to screening questions, were at higher risk for subsequent development of mild cognitive impairment or PD.5 Given the association between RBD and neurodegenerative disorders, understanding RBD risk factors may have wide implications: RBD risk factors may be useful screening tools to identify those at risk for development of neurodegenerative disorders. RBD risk factors, including head injury, smoking, farming occupation, and ischemic heart disease, were identified in a hospital-based case-control study of patients with idiopathic RBD; about one-third of these patients with RBD in this case-control study developed neurodegenerative diseases in a 4-year follow-up.6,–8 To further enhance our understanding of RBD risk factors, we conducted a community-based study in approximately 13,000 Chinese adults in the outpatient setting to investigate risk factors for pRBD—assessed by a validated RBD questionnaire, though not diagnostically confirmed—within 6 years of exposure.
Participants of this study were a subset of the Kailuan cohort. As previously described, the Kailuan cohort is a multicenter study that includes 101,510 Chinese adults (81,110 men and 20,400 women) aged 18–98 years in 2006–2007 (baseline of the cohort) living in Tangshan, an industrial city in China.9 All the participants underwent questionnaire assessments and clinical and laboratory examinations, which were conducted in the 11 hospitals responsible for health care of the community. Information on lifestyle and health status was updated via questionnaires every 2 years.
Participants in the current study were included from 12,990 adults (10,725 men and 2,265 women) who were evaluated at the Kailuan hospital, one of the 11 participating hospitals, and were free of PD and dementia in 2012. We excluded 216 participants (1.7%) who did not provide information on studied exposures or RBD, leaving 12,784 participants (10,556 men and 2,228 women) in the current analysis.
The study was approved jointly by the Ethics Committee of the Kailuan General Hospital and the Human Subjects Committee at Brigham and Women's Hospital/Harvard Medical School.
Probable RBD status was determined based on completion of the validated RBD questionnaire–Hong Kong (RBDQ-HK) in 2012. The RBDQ-HK is a Chinese screening instrument for diagnosis of RBD, comprising 13 questions related to RBD clinical features, which are rated on scales of lifetime occurrence and recent 1-year frequency.10 Based on a validation study including 107 Chinese patients with RBD and 107 controls, the RBDQ-HK demonstrated good sensitivity (82.2%), specificity (86.9%), internal consistency, and test-retest reliability.10 More recently, the RBDQ-HK was validated in East China, including in patients with PD or obstructive sleep apnea, with similar sensitivity and specificity.11 Questions in the RBDQ-HK screened for frequency of dreams and nightmares, content of dreams, disturbance of sleep or the bed partner, and vocalizations or behaviors during sleep.10 Of note, RBD status in the current study was determined by questionnaire answers, and was not diagnostically confirmed by physician assessment and polysomnography.
To reduce the possibility of reverse causality (i.e., pRBD that led to change in status of potential risk factors), we used cohort baseline data for exposures (2006), except for chronic olfactory and taste function, which were not collected until 2012.
Exposures that were ascertained from questionnaires in 2006 included educational level, income level, marital status, occupation, physical activity, smoking status, alcohol consumption, tea consumption (surrogate for caffeine consumption), use of medications (e.g., antihypertensive medications), and history of cancer, myocardial infarction, and stroke. Additional information about head injury was collected by reviewing medical records. In 2012, we also collected information on chronic olfactory and taste dysfunction (with minimum duration of 3 months), using the questions previously used in the National Health Interview Survey.12
Weight and height were measured by trained field workers during the 2006 interview. Body mass index (BMI) was calculated as weight (kg)/height (m2). Participants were grouped into 3 categories based on their BMI: normal weight (<24 kg/m2), overweight (24–27.9 kg/m2), or obese (≥28 kg/m2). Systolic and diastolic blood pressures were measured twice from the seated position using a mercury sphygmomanometer. The average of the 2 readings was used for analysis. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg or use of antihypertensive medications in the last 2 weeks irrespective of blood pressure status. Participants with systolic blood pressure between 120 and 129 mm Hg or diastolic blood pressure 80–89 mm Hg were categorized as prehypertension.
Fasting blood samples were collected in 2006 and concentrations of blood glucose, triglyceride, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol were determined using an autoanalyzer (Hitachi 747; Hitachi, Tokyo, Japan) at the central laboratory of the Kailuan General Hospital. Diabetes was defined by a fasting blood glucose concentration ≥7 mmol/L, or active treatment with insulin or any oral hypoglycemic agent. Impaired fasting glucose was defined by a fasting blood glucose concentration between 5.6 and 6.9 mmol/L.
Statistical analyses were completed with SAS version 9.1 (SAS Institute, Cary, NC). Logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) and to test differences in prevalence of pRBD across exposures, including age, sex, educational level, income, marital status, occupation, physical activity, smoking, alcohol and tea consumption, obesity, diabetes, hypertension, history of myocardial infarction, stroke, cancer, and head injury, chronic olfactory and taste dysfunction, and serum concentrations of triglycerides, LDL, and HDL. We fit 2 multivariate models: model 1 was adjusted for age and sex, while model 2 was adjusted for all aforementioned exposures. Linear trends were tested for significance by treating categories of each studied exposure as an ordinal variable.
Since approximately 38% of the study participants worked underground in the coal mines, we conducted a sensitivity analysis after excluding these participants, due to concern of generalizability. Although we used the 2006 data as the exposure, we still could not eliminate the possibility of reverse causality, so we conducted a sensitivity analysis by restricting to the cases who had pRBD occurring only during the recent 1 year (i.e., symptom onset in 2011–2012). A sensitivity analysis was also performed excluding those with history of head injury.
Additionally, we examined potential interactions of the studied exposures with sex and age (as pRBD is reported to be more common in men and older adults13) in relation to pRBD risk, by including multiplicative terms between sex/age and each of these factors in the logistic regression models, with adjustment for other potential confounders.
Baseline characteristics of the cohort in 2006 are shown in tables 1 and and2.2. Prevalence of pRBD was 5.9% in men and 4.1% in women (p difference = 0.02 after adjustment for age). Risk factors that were significantly associated with higher risk of having pRBD in the fully adjusted model included lower level of education, coal mining or other blue collar occupation (compared to white collar occupation), lower level of physical activity, diagnosis of diabetes or prediabetes, lower BMI, history of head injury, higher LDL level, and chronic olfactory and taste dysfunction (tables 3 and and44 and table e-1 on the Neurology® Web site at Neurology.org). In the sensitivity analyses, excluding of coal miners, or those with history of head injury, or restricting to pRBD cases with symptom onset within 1 year, generated similar results (table e-2).
We did not find significant interactions between sex and studied exposures in relation to pRBD risk (p interaction ≥ 0.05 for all), except for triglyceride level (table e-3). We also observed that age might modify the association of income, marital status, occupation, and alcohol drinking with pRBD status (p interaction <0.05 for all, table e-4). However, it remained unclear whether observed sex and age differences were due to chance.
We provide data from a large-scale community-based study on risk factors in pRBD. Our results show that higher risk of pRBD is associated with lower socioeconomic status, history of head injury, chronic olfactory and taste dysfunction, and cardiovascular risk factors.
Understanding risk factors for RBD is particularly important in the context of its strong association with neurodegenerative diseases.14 If RBD and other neurodegenerative diseases share similar pathophysiology, then it is expected that they share some similar risk factors. Head injury is associated with pRBD in our study, as well as in the other RBD epidemiologic study6; likewise, epidemiologic studies—though sometimes conflicting—have shown significant associations between head injury and PD.15 However, in our study, smoking and caffeine intake are not significantly associated with pRBD, confirming one of the major findings of the other RBD epidemiologic study that caffeine and alcohol intake did not appear protective for RBD.6 This finding is surprising considering the epidemiologic evidence supporting associations between reduced risk PD and caffeine consumption16,17 or smoking.18,–21
Olfactory dysfunction is a risk factor shared by both RBD and PD. Poor olfactory function is also often associated with PD, and can predate development of PD.22 Our finding that chronic olfactory dysfunction is associated with higher risk of pRBD complements existing evidence available from smaller studies. In past studies, patients with idiopathic RBD performed worse on olfactory function tests compared to controls.4,23,24 Poor olfactory discrimination in idiopathic RBD was related to lower cerebral blood flow in bilateral anterior hippocampal gyri.25 Consistently, patients with RBD who developed neurodegenerative disease over 5 years had worse baseline olfaction compared to those who did not develop neurodegenerative disease.4 Two recent studies on patients with idiopathic RBD confirmed that olfactory function helped predict early development of PD or DLB.26,27
We report the novel result of a significant association between chronic taste dysfunction and pRBD, even after adjustment for olfactory dysfunction. While the relationship between taste dysfunction and pRBD is not well-understood, several past studies have reported that individuals with PD or parkinsonism had significantly worse taste function relative to healthy controls.28,–31 Given the connection between RBD and PD, it is possible that taste alterations can be features or risk factors of both.
We demonstrate that vascular risk factors such as diabetes, LDL level, and physical inactivity—though not smoking—are risk factors for pRBD. Similarly, in a previously published case-control study on RBD risk factors, ischemic heart disease was associated with a higher risk of idiopathic RBD, even after adjusting for cardiovascular risk factors.7 This study also reported associations between smoking and inhaled glucocorticoids and risk of idiopathic RBD.7 The pathophysiologic link between RBD and cardiovascular risk factors is currently unclear. RBD is associated with autonomic dysfunction,4,32 which can increase cardiovascular risk. Alternatively, it is possible that cardiovascular risk factors may play a role in RBD pathogenesis, especially as they may increase the risk of cerebrovascular disease, including vascular parkinsonism. Furthermore, physical activity is not simply an indicator of cardiovascular health, but has also been shown to be independently beneficial in PD.33 In PD, reverse causality (i.e., early prodromal PD preventing exercise) has been a concern in interpreting the relationship between exercise and PD, but our study assesses the prodromal state of RBD that is not defined by motor dysfunction, providing evidence against the potential reverse causality.
Our study reveals environmental and socioeconomic risk factors for RBD. In our cohort, pRBD is associated with the coal mining occupation. In a previous case-control study of RBD, mining occupation was likewise associated with 52% higher RBD risk, although the association was not significant.6 Interestingly, the parkinsonogenic neurotoxin N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) can be produced during coal preparation.34 There is a case report of a coal preparation plant worker who developed parkinsonism.35 pRBD is also associated with low level of occupation and education in our study. As a possible pathophysiologic explanation, RBD is associated with problems with cognitive function,5 which may lead to lower levels of achievement in education and career.
In past clinical and epidemiologic studies, the majority of patients with RBD were male, though male sex is not an independent risk factor for pRBD in our study.6,36 However, the demographics of our cohort do not necessarily reflect the true demographics of RBD, as RBD status in our cohort is questionnaire-based and not diagnostically confirmed by physician assessment and polysomnography. Some RBD mimics, including insomnia, are more common in women,37 and the high prevalence of pRBD in our study likely reflects inclusion of RBD mimics. In comparison, in an elderly Korean study that underwent polysomnography to confirm diagnoses, the age- and sex-adjusted estimate of RBD prevalence was 2.01%.38
Our study has several limitations. First, as discussed, RBD status was determined per self-report, and not diagnostically confirmed by physician evaluation and polysomnography. We did not collect information on several mimics of RBD, such as posttraumatic stress disorder or nightmares. Questionnaires were subject to recall bias and inaccuracies. Based on RBD prevalence, which was shown in the aforementioned Korean study,38 and sensitivity/specificity of the RBDQ-HK,10 only an expected ~12% of participants who screened positive for pRBD likely had RBD. Given the increased risk of misclassification of RBD status with this methodology, we used the term pRBD in this report to indicate that RBD diagnoses were not clinically confirmed. Notably, previous studies reported association between pRBD and increased risk of neurodegenerative diseases.5 Furthermore, misclassification of RBD tended to neutralize differences between groups, so any findings were likely of larger amplitude than were observed. Second, for studied exposures, we had information on medications for hypertension, diabetes, and hyperlipidemia, but not on other medications that could lead to RBD symptoms, such as antidepressants.7 Third, there was a lack of generalizability to other settings and populations, as this study focused on a specific group of employees—including coal miners—in China. Fourth, this study was limited by its cross-sectional study design, so the temporal relationship between studied exposures and pRBD was unclear. However, the use of exposures from 2006 to predict pRBD status in 2012 introduced a prospective component, and sensitivity analysis with restriction to those with symptom onset in the past year generated similar results.
The advantages of our study include the large, systematically evaluated study population and the high quality of data obtained, with 98% of the participants providing complete responses. Overall, in this large-scale community-based study, we identified multiple risk factors for RBD symptoms, including clinical and socioeconomic risk factors. Future prospective studies to establish the temporal relationship between these potential risk factors and RBD will be important. Identifying RBD symptoms in patients is important not only because of the RBD symptoms themselves, but also because of the association between RBD and neurodegenerative disorders. Knowledge of RBD risk factors can help raise clinical suspicion of RBD and guide interventions to prevent or attenuate RBD symptoms. Furthermore, as neuroprotective therapies become available in the future, identifying RBD early will be of particular importance for early intervention.
Supplemental data at Neurology.org
Drafting/revising the manuscript (J.C.W., J.L., M.P., S.C., A.W., S.W., and X.G.), study concept/design (X.G.), analysis or interpretation of data (J.C.W., M.P., S.W., and X.G.), acquisition of data (S.W. and X.G.), study supervision (S.W. and X.G.), and obtaining funding (X.G.). Dr. Gao, who received an MS in Epidemiology and Biostatistics in 2001, completed data analyses.
This study was supported by a grant (R21 NS087235-01A1) from the NIH. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
J. Wong and J. Li report no disclosures relevant to the manuscript. M. Pavlova served on a committee of the American Academy of Sleep Medicine and received research study funding from Lundbeck Inc. S. Chen and A. Wu report no disclosures relevant to the manuscript. S. Wu received funding from the Chinese Science Foundation. X. Gao has served on committees of the Sleep Research Society, American Academy of Sleep Medicine, and Parkinson Study Group and received funding from the NIH/NINDS. Go to Neurology.org for full disclosures.