Prevalence and Clinical Correlates of Probable REM Sleep Behavior Disorder in Psychiatric Sleep Clinic Outpatients: A Retrospective Study
Article information
Abstract
Objective
This study aimed to investigate the prevalence of probable REM sleep behavior disorder (pRBD) and identify associated factors among psychiatric sleep clinic outpatients.
Methods
We retrospectively reviewed records of first-visit outpatients aged ≥18 years who presented with sleep-related complaints at the psychiatric sleep clinic at a single tertiary university hospital between January 1, 2023, and December 31, 2024. Of 522 patients, 387 (74.1%) completed the REM Sleep Behavior Disorder Screening Questionnaire (RBDSQ), and 356 were included in the analysis. RBDSQ scores ≥5 defined the pRBD group. Demographic and clinical variables, including diagnoses and medication use, were collected. Univariable and multivariable logistic regression analyses identified factors associated with pRBD.
Results
Among participants, 179 (50.3%) were classified as pRBD and 177 (49.7%) as non-RBD. The pRBD group was significantly younger, had a higher body mass index, and showed higher scores for insomnia, depression, anxiety, and trauma exposure, and risk of obstructive sleep apnea (OSA). They were also more likely to be diagnosed with trauma- and stressor-related disorders and selective serotonin reuptake inhibitors (SSRIs). In multivariable analysis, higher anxiety scores (odds ratio [OR]=1.07, 95% confidence interval [CI]=1.02–1.12), sleep-related breathing disorders (OR=7.60, 95% CI=2.05–28.17), and SSRIs use (OR=6.91, 95% CI=1.91–22.11) were independently associated with pRBD.
Conclusion
RBD symptoms were common among psychiatric sleep clinic outpatients. These results suggest that psychiatric patients with marked anxiety, SSRIs use, and sleep-related breathing disorders should be prioritized for screening, particularly before antidepressant initiation, with careful differentiation from OSA.
INTRODUCTION
REM sleep behavior disorder (RBD) is a type of parasomnia characterized by the loss of muscle atonia during REM sleep, which leads to the enactment of dream content [1]. Such dream enactment behaviors (DEB) often manifest as violent or aggressive actions, including shouting, swearing, gesturing, punching, and kicking, which may result in physical injury to the patient or their bed partner [2]. According to a previous study, 31.4% of RBD patients had caused injury to themselves or others during sleep, but only 12.3% had received medical evaluation [3]. Idiopathic RBD (iRBD) has been identified as an early biomarker of α-synucleinopathies, such as Parkinson’s disease, dementia with Lewy bodies, and multiple system atrophy. Longitudinal studies have reported that approximately 40%–65% of patients with iRBD eventually develop a neurodegenerative disorder within 10 years [4], emphasizing the importance of early detection and clinical evaluation.
RBD has also been associated not only with neurodegenerative diseases, but also with psychiatric factors such as anxiety, depression [5,6], and the use of antidepressants [7-9]. Lam et al. [9] reported that among psychiatric outpatients aged 18–65, 5.8% exhibited RBD-like symptoms, 21.0% reported sleep-related injury, and 3.6% experienced sleep-related violence. In particular, patients taking antidepressants have been shown to exhibit increased DEB, likely due to the effects of these medications on REM sleep [5,10].
In a study using video-polysomnography (vPSG) among psychiatric inpatients, approximately 3.6% were diagnosed with RBD, and 8.77% showed REM sleep without atonia [11]. Among patients diagnosed with RBD through vPSG, 46.1% were taking antidepressants [12]. These findings indicate that RBD symptoms may be present among psychiatric outpatients and that they are often unrecognized in clinical assessments. To date, few studies in Korea have examined RBD symptoms and associated risk factors among psychiatric outpatients.
Although polysomnography (PSG) is essential for the definitive diagnosis of RBD, its use as a screening tool in routine clinical practice is often limited. The REM Sleep Behavior Disorder Screening Questionnaire (RBDSQ) is a self-report screening tool designed to systematically assess abnormal behaviors during sleep and is easily applicable in clinical settings [13]. A recent study from outside Korea also employed the RBDSQ to screen for RBD in psychiatric outpatients aged 45 to 80 years [14].
Accordingly, this study aimed to evaluate symptoms of probable RBD (pRBD) and investigate its prevalence and associated factors among psychiatric outpatients at a university hospital in Korea.
METHODS
Participants
This retrospective chart review included patients aged 18 years or older who had their first outpatient visit at the Sleep Clinic of the Department of Psychiatry, Chungnam National University Hospital between January 1, 2023, and December 31, 2024. Out of a total of 522 initial visit charts, 387 patients (74.1%) completed the self-reported questionnaire, whereas 135 patients (25.9%) did not. Of the 387 patients who completed the self-report questionnaire, those with inadequate medical records or a prior diagnosis of parasomnia, including RBD, confirmed by PSG were excluded. After applying these criteria, a total of 356 patient records were included in the final analysis. The flowchart of patient selection is presented in Figure 1. This study was approved by the Institutional Review Board of Chungnam National University Hospital (IRB No. 2024-10-038). Given the retrospective nature of the chart review, the requirement for informed consent was waived.
Flowchart of study population selection. PSG, polysomnography; pRBD, probable REM sleep behavior disorder.
Variables and measures
Demographic and clinical variables
Age, sex, alcohol use, body mass index (BMI), history of central nervous system (CNS) disorders (including neurodegenerative diseases, brain tumors, traumatic brain injury, stroke, and intracranial hemorrhage), psychiatric diagnoses, and current psychiatric medications were collected.
Self-report instruments
REM Sleep Behavior Disorder Screening Questionnaire
The REM Sleep Behavior Disorder Screening Questionnaire (RBDSQ) is a 10-item self-report questionnaire designed to screen for abnormal behaviors during sleep [13]. In a Korean population, a cutoff score of 5 or higher on the RBDSQ demonstrated a sensitivity of 89.4% and a specificity of 98.3% for identifying pRBD [15]. In this study, a total score of 5 on the RBDSQ was used as the cutoff point. Participants with scores of 5 or higher were classified as the pRBD group, while those with scores below 5 were classified as the non-RBD group.
Insomnia Severity Index
The Insomnia Severity Index (ISI) is a self-report scale developed to assess the perceived severity of insomnia [16]. It consists of seven items, each rated on a scale from 0 to 4, with total scores ranging from 0 to 28. Higher scores indicate more severe insomnia, reflecting poorer subjective sleep quality. The reliability of the ISI has been validated in Korean populations, and a cutoff score of 15.5 has been suggested [17].
Beck Depression Inventory and Beck Anxiety Inventory
Depressive and anxiety symptoms were assessed using the Beck Depression Inventory (BDI) and the Beck Anxiety Inventory (BAI), respectively. Both are self-report measures consisting of 21 items, each rated on a scale from 0 to 3, with total scores ranging from 0 to 63 [18,19]. Higher scores indicate greater levels of subjective depression or anxiety. The reliability of both instruments has been validated in Korean populations. A cutoff score of 22 has been suggested for identifying severe depression on the BDI [20] and for severe anxiety on the BAI [21].
Life Events Checklist for DSM-5
To assess exposure to traumatic events, the Korean version of the Life Events Checklist for DSM-5 (LEC-5) was used [22,23]. The LEC-5 includes 16 potentially traumatic events based on the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria, and respondents indicate their level of exposure for each event as “happened to me,” “witnessed it,” “learned about it,” “not sure,” or “does not apply.” In this study, following the classification method used by Pugach et al. [24], the number of events marked as “happened to me” was counted as the total number of experienced traumatic events.
Berlin Questionnaire
The Berlin Questionnaire (BQ) was used to assess obstructive sleep apnea (OSA) risk. It consists of 10 items across three categories: snoring/apnea (6 points), daytime sleepiness/fatigue (3 points), and obesity or hypertension (2 points). A category was considered positive if the score was ≥2 in categories 1 or 2, or ≥1 in category 3 [13]. Participants were classified as high risk for OSA when two or more categories were positive [25].
Psychiatric diagnoses and medication information
All participants were diagnosed and treated by board-certified psychiatrists. Psychiatric disorders were classified based on the DSM-5, Text Revision (DSM-5-TR) into the following categories: mood disorders, anxiety disorders, trauma- and stressor-related disorders, somatic symptom and related disorders, substance-related and addictive disorders, schizophrenia spectrum disorder, and other mental disorders. Sleep disorders were classified according to the International Classification of Sleep Disorders, Third Edition, Text Revision (ICSD-3-TR), as follows: insomnia disorder, sleep-related breathing disorders, central disorders of hypersomnolence, sleep-related movement disorders, and parasomnias.
Psychiatric medications prescribed at the time of the visit were categorized as follows: antipsychotics, mood stabilizers, antidepressants, anxiolytics/hypnotics, and stimulants or wake-enhancing drugs. Antidepressants were further classified into four groups based on their pharmacological mechanisms: selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants, and mixed-mechanism agents.
Statistical analysis
Demographic and clinical characteristics were compared between the pRBD group and the control group. Independent t-tests and chi-square tests were used to compare differences between groups for continuous and categorical variables, respectively. To identify the factors associated with pRBD, multivariable logistic regression analysis was performed. Variables that were statistically significant in the univariate analysis, along with clinically relevant factors previously reported to be associated with pRBD (insomnia, mood disorders [14], and sleep-related breathing disorders [26]), were selected for inclusion in the multivariable logistic regression model. The multivariable logistic regression was performed using the entry method, and all models were adjusted for age and sex. All statistical analyses were conducted using two-tailed tests, with a significance level set at p<0.05. Data were analyzed using SPSS version 25 (IBM Corp.).
RESULTS
The characteristics of the patients visiting the Sleep Clinic of the Department of Psychiatry for the first time are shown in Table 1, divided into those who completed the self-report questionnaires (study participants) and those who did not (non-completers). Compared to non-completers, study participants had a significantly lower prevalence of schizophrenia spectrum disorder (p=0.001) and other mental disorders (p=0.002).
Characteristics of psychiatric sleep clinic outpatients
Table 2 presents a comparison of demographic and clinical characteristics of the study participants, divided into the pRBD group and the non-RBD group. A total of 356 patients were included in the study, of whom 179 (50.3%) were classified as having pRBD, and 177 (49.7%) were classified as the control group (non-RBD). Compared to the non-RBD group, the pRBD group was significantly younger (43.4±15.9 vs. 50.2±13.9 years, p=0.005), had a higher BMI (25.1±6.7 vs. 23.5±3.5 kg/m2, p=0.007). In addition, the pRBD group scored significantly higher on the ISI (17.0±7.0 vs. 12.6±7.5, p<0.001), BDI (25.1±14.6 vs. 19.2±12.3, p<0.001), BAI (21.5±14.6 vs. 11.4±10.4, p<0.001), and LEC-5 (3.4±2.8 vs. 2.3±2.5, p=0.003) than the non-RBD group. The pRBD group had a higher proportion of patients at high risk for OSA according to the BQ compared with the non-RBD group (35.8% vs. 22.0%; p=0.034).
Comparison of demographic and clinical characteristics between pRBD and non-RBD groups
Table 3 shows psychiatric and sleep disorder diagnoses based on the presence or absence of pRBD. Among patients in the pRBD group, 20 patients (11.2%) were diagnosed with trauma- and stressor-related disorders, which was significantly higher than in the non-RBD group (4.5% [8 patients]; p=0.020). No significant differences were observed between the two groups in the other psychiatric or sleep disorder diagnoses.
Comparison of psychiatric and sleep disorder diagnoses between pRBD and non-RBD groups
Table 4 summarizes the use of psychiatric medications in the pRBD group and the non-RBD group. Overall, 52% of participants were prescribed antidepressants. Although the proportion was higher in the pRBD group than in the non-RBD group (57.0% vs. 46.9%), the difference was not statistically significant (p=0.057). However, subgroup analysis by antidepressant class revealed a significantly higher rate of SSRIs use in the pRBD group than in the non-RBD group (44.7% vs. 17.5%, p<0.001). There were no significant between-group differences in the use of other psychiatric medications.
Comparison of psychotropic medication use between pRBD and non-RBD groups
Table 5 shows the factors associated with pRBD. In the multivariable logistic regression analysis, a total of 13 variables including age and sex were entered into the model. As a result, higher BAI scores (odds ratio [OR]=1.068, 95% confidence interval [CI]=1.019–1.120, p=0.006), sleep-related breathing disorders (OR=7.597, 95% CI=2.049–28.167, p=0.002), and the use of SSRIs (OR=6.914, 95% CI=1.909–22.110, p=0.003) were significantly associated with pRBD (Table 5). The model demonstrated good fit (χ2=63.07, df=13, p<0.001), explaining 33.8% and 45.1% of the variance according to the Cox & Snell and Nagelkerke R2 values, respectively.
Multivariable logistic regression analysis of factors associated with probable RBD (n=356)
DISCUSSION
This study investigated the prevalence and associated factors of pRBD among first-visit outpatients at a psychiatric sleep clinic. Among psychiatric outpatients who presented with sleep-related complaints, 74.1% completed the RBDSQ, and 50.3% of these were classified as having pRBD. Factors significantly associated with pRBD were higher BAI scores, a diagnosis of sleep-related breathing disorder, and the use of SSRIs.
The prevalence of RBD has been reported as 1.06% in a largescale study of the general population aged 35 to 75 years [27] and 2.01% among individuals aged 60 and older in a Korean community-based sample [28]. In a psychiatric outpatient clinic-based study of individuals aged 45 and older, the prevalence of pRBD was reported to be 15.9%. Subgroup analyses revealed considerable variation in pRBD prevalence across primary psychiatric diagnoses, ranging from 6.7% to 26.4% [14]. In our study, among patients who visited the Sleep Clinic of the Department of Psychiatry, the prevalence of pRBD was 50.3%, notably higher than that reported in previous studies. Several factors may account for this discrepancy.
First, only patients who voluntarily completed the RBDSQ were included, leading to potential selection bias. Patients who completed the RBDSQ in our study showed a different diagnostic distribution compared with those who did not. Specifically, completers (study participants) had significantly lower proportions of schizophrenia spectrum disorders and other mental disorders. Since previous research has suggested a lower risk of pRBD in schizophrenia spectrum disorders (odds ratio=0.573) [14], their absence from our sample may have contributed to the elevated prevalence. Moreover, the category of other mental disorders in non-completers primarily included patients with mild or subthreshold symptoms or those who visited mainly for routine psychiatric consultation. Such patients may have been less motivated to report their symptoms accurately in the self-report questionnaire. In addition, because this study targeted psychiatric outpatients who primarily presented with sleep-related complaints, even minor sleep problems may have been overreported.
Second, among those who did complete the questionnaires, the pRBD group exhibited more severe psychiatric symptoms, including significantly higher ISI, BDI, and BAI scores, compared with the non-RBD group. This suggests that severity, rather than diagnostic category alone, may have contributed to an elevated prevalence.
Third, pharmacological effects may have influenced RBDSQ responses. SSRIs are known to induce vivid dreams [29] and sleep-related behaviors such as sleep talking [30]. These sleep-related side effects could have been reported as positive responses on the RBDSQ, which may have led to an overestimation of the prevalence of pRBD in our sample.
Fourth, undiagnosed OSA may also have contributed to false positives on the RBDSQ, as severe OSA can mimic RBD-like behaviors. In our study, considering the higher BMI and BQ scores of participants, the possibility of unrecognized OSA increased risk of false-positive findings, thereby inflating the prevalence estimates.
Taken together, these factors indicate that the prevalence observed in this study is likely an overestimation and should be interpreted cautiously. The findings may not be fully generalizable to the broader population of psychiatric outpatients.
In this study, the use of SSRIs was identified as a significant factor associated with RBD symptoms, consistent with findings from numerous previous studies. According to a recent meta-analysis, antidepressants—particularly SSRIs—are among the major risk factors for RBD, with SSRI users exhibiting approximately a 2.4-fold increased risk of developing RBD compared to non-users [8]. Antidepressants interfere with brainstem regulation of REM sleep by disrupting normal muscle atonia during this stage, thereby enabling DEB [31]. While such symptoms may subside after discontinuation of the medication [32,33], in some cases they may represent an early manifestation of iRBD, warranting longitudinal monitoring [34]. Further studies should systematically examine RBD symptoms after antidepressant withdrawal, as their course and underlying mechanisms have not yet been fully clarified.
When comparing the pRBD group to the control group, no diagnostic differences were found for diagnosis of mood disorder or anxiety disorder; however, a significant association was confirmed with higher BAI scores. This suggests that the severity of symptom is more relevant to RBD symptoms than diagnostic category. Unlike the prevalence analysis, where inflated rates may have reflected more severe cases, here the finding indicates that anxiety severity itself may be independently associated with RBD symptoms. In this study, diagnoses were categorized into groups such as mood disorder and anxiety disorder, and the severity of the disorder was not considered. Therefore, the significantly higher BAI scores in the pRBD group suggest that patients with more severe psychiatric disorders may be included in the pRBD group. Previous studies have reported positive correlations between RBDSQ scores and both anxiety (r=0.43, p<0.05) and depression (r=0.56, p<0.05) [35]. In a PSG-confirmed RBD cohort, early-onset psychiatric disorders were common, with depressive disorder found in 49% and anxiety disorder in 43% [36]. Given that depression is generally more prevalent than anxiety disorders, this suggests a high prevalence of anxiety disorders in the RBD patient group. In line with these reports, the significantly higher BAI scores in our pRBD group suggest that anxiety severity plays a critical role. Further research is needed to examine the association between RBD and psychiatric severity, with a particular focus on anxiety.
In our study, higher BMI and a high risk for OSA as determined by the BQ was significantly associated with pRBD in univariate analysis. The multivariable logistic regression analysis demonstrated that a diagnosis of sleep-related breathing disorder was independently associated with pRBD. This association may be attributable to a higher likelihood of comorbid sleep apnea or alterations in REM sleep architecture. Consistent with this, a previous study reported that 107 out of 120 RBD patients (89.2%) were diagnosed with OSA, demonstrating a high prevalence of OSA in RBD patients [37]. Given the high prevalence of obesity and the potential influence of psychiatric medications on metabolism in psychiatric populations, this finding is noteworthy. Further studies are needed to clarify the relationship between RBD and factors related to weight and metabolism.
In this study, both the number of traumatic events (as measured by the LEC-5) and the diagnosis of trauma- and stressor-related disorders were higher in the pRBD group; however, these associations were not statistically significant in multivariable analysis. Prior studies have reported RBD-like symptoms—particularly DEB—among patients with post-traumatic stress disorder, and some have proposed the concept of a distinct parasomnia subtype, termed trauma-associated sleep disorder, to describe this phenomenon [38]. These findings indicate that careful assessment of trauma history and related diagnoses is warranted when evaluating pRBD symptoms in psychiatric outpatients.
The observed pRBD group was younger than the non-RBD group, in contrast to prior studies that identified old age and male sex as risk factors for RBD. To date, most RBD research has focused on middle-aged and older populations, whereas data on younger individuals are limited to studies of early-onset RBD, defined as diagnosis before the age of 50. A previous study reported that early-onset RBD patients are more likely to be female, to have psychiatric comorbidities, and to be prescribed antidepressants [39]. Notably, these characteristics parallel those observed in our pRBD group, which was younger in age, exhibited psychiatric comorbidities including insomnia, depression, anxiety, and frequently reported SSRI use. Given the scarcity of research on RBD in younger individuals, prospective longitudinal studies are needed to clarify the distinct clinical characteristics and symptom trajectories of early-onset RBD patients.
This study has several limitations. First, the possibility of selection bias must be considered. We retrospectively analyzed medical records of psychiatric outpatients whose chief complaint was sleep-related problems and who completed the RBDSQ; only 74.1% of first-visit patients completed the RBDSQ, and their diagnostic distribution differed from that of non-completers. Schizophrenia spectrum disorders were absent among study participants; given previous reports that this group is associated with a lower likelihood of pRBD [14], the prevalence observed here may have been overestimated. Moreover, the non-completer group contained a higher proportion of “other mental disorders,” which likely included mild or subthreshold cases; such patients may have been less motivated to report symptoms accurately, leading to additional bias. Finally, because the study population consisted of psychiatric outpatients with sleep-related complaints, overreporting of minor sleep symptoms is possible. Second, pRBD was assessed solely using the RBDSQ without PSG confirmation. Although the RBDSQ has been standardized in various studies and is widely used as a reliable screening tool for RBD, its specificity is limited. One study reported a sensitivity of 0.96 and a specificity of 0.56 [13], implying that a substantial number of false positives may occur. A meta-analysis reported its accuracy in general population [40], but high false-positive rates have been reported in migraine patients (58.9%) [41] and in OSA, where a higher cutoff (6.5) has been suggested with frequent false positives [42]. Given the higher BMI and BQ scores in our sample, undiagnosed OSA may have contributed to misclassification. Moreover, other sleep-related behavioral disorders (e.g., vivid dreams, sleep talking, sleepwalking, sleep-related eating disorder) can mimic RBD and require PSG for differentiation. Third, this was a single-center retrospective chart review conducted at a specialized psychiatric sleep clinic; therefore, the findings may not be generalizable to other clinical settings or the community population, and causal inferences cannot be made. Fourth, because the RBDSQ relies on self-report, recall or reporting bias may have affected responses, particularly in psychiatric patients with variable insight. Fifth, psychiatric diagnoses were grouped into broad categories such as mood disorder and anxiety disorder, which may not adequately reflect the heterogeneity of subtypes or severity levels. Finally, medication-related variables such as dosage, treatment duration, and adherence were not assessed; these should be considered in future research to more accurately estimate pharmacological effects on RBD-like symptoms.
To our knowledge, this is the first study to investigate RBD in psychiatric outpatients in Korea. Although the prevalence of pRBD observed in this study is likely to have been overestimated due to selection bias and the specific characteristics of our sample, the findings nonetheless provide important clinical implications. Our findings indicate that pRBD symptoms are relatively common among outpatients at a psychiatric sleep clinic and are independently associated with anxiety, SSRI use, and sleep-related breathing disorders. These results suggest that psychiatric patients with these risk factors should be prioritized for screening, particularly before initiating antidepressant therapy, and that careful differentiation from OSA is clinically important. Because RBD may also serve as an early biomarker of neurodegenerative diseases, psychiatric patients reporting such symptoms should be considered for regular clinical monitoring and further diagnostic evaluation, including PSG.
Notes
The authors have no potential conflicts of interest to disclose.
Availability of Data and Material
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Author Contributions
Conceptualization: So-Hyun Ahn, Hyun Seung Chee. Data curation: Jihee Lee. Formal analysis: Jihee Lee, So-Hyun Ahn. Funding acquisition: Hyun Seung Chee. Investigation: all authors. Methodology: Jihee Lee, So-Hyun Ahn. Project administration: So-Hyun Ahn. Supervision: So-Hyun Ahn, Hyun Seung Chee. Validation: all authors. Writing—original draft: Jihee Lee, So-Hyun Ahn. Writing—review & editing: all authors.
Funding Statement
This work was supported by research fund of Chungnam National University.
Acknowledgments
None