In this questionnaire-based study, we found a significantly increased frequency of pRBD in PD patients compared with a control population. The frequency of RBD in our study is comparable with that observed in previous studies. In 469 non-demented Japanese PD patients, the pRBD comorbidity was noted in 31.6% of the individuals (defined using RBDSQ-J
In our study, a cut-off score of 5 or 6 on the RBDSQ-J revealed similar clinical characteristics in the PD patients. The original study of the RBDSQ showed that a cut-off of 5 was useful for differentiating idiopathic RBD patients from healthy subjects, showing a sensitivity of 96% and a specificity of 56% [21
]. Similarly, in our previous validation study of the RBDSQ, an RBDSQ-J cut-off of 5 was useful for differentiating Japanese idiopathic RBD patients from healthy subjects, with a sensitivity of 88.5% and specificity of 96.9% [28
]. Recently, Nomura et al. [29
] reported upon the usefulness of the RBDSQ-J in Japanese PD patients. In that study, a cut-off point of 6 yielded a sensitivity of 84.2% and a specificity of 96.2%. This finding is in contrast with that of a study assessing the sensitivity of clinical interviews (minimal diagnostic criteria of ICSD, revised) for diagnosing PSG-confirmed RBD in PD patients and controls, which revealed that in PD patients, a low sensitivity (33%) and high specificity (90%) was observed whereas in controls, an excellent sensitivity and specificity were noted (sensitivity, 100%; specificity, 99.6%). The authors suggested that these findings were due to common observations of the mild form of RBD and unawareness of sleep problems and other sleep-related disturbances in patients with PD [40
]. The RBDSQ is, therefore, thought to have superior sensitivity for detecting RBD, even in PD patients. In the original version of the RBDSQ, comorbid neurologic diseases such as PD were included in item 10. Patients with PD and RBD frequently exhibit a less severe form of abnormal behavior during REM sleep compared with individuals with idiopathic RBD [5
]. Thus, we believe that applying a cut-off score of 5, which can increase the sensitivity for detecting pRBD, may be suitable for PD patients, and thereafter, PSG may be used upon suspicion of RBD and when the patient is considered suitable for PSG examination.
There were no significant differences in the gender ratio, disease duration, disease severity, UPDRS III scores or PFS scores between the pRBD and non-pRBD groups. Furthermore, the axial-limb ratio was similar between the groups, a finding similar to that of Postuma et al. [6
]; however, they reported that patients with PD who had a symmetric onset showed increased RBD comorbidity and an increased percentage of tonic REM sleep compared with those showing an asymmetric onset [41
]. Several, but not all [6
], studies [13
] have reported that patients with RBD and PD manifest with non-tremor-dominant type PD rather than tremor-dominant type PD. This finding may imply a more extensive brain involvement in those PD patients with RBD than in those without because in non-tremor-dominant type PD, patients show a more rapid motor deterioration whereas tremor-dominant cases show a benign disease course [43
]. Moreover, compared with the tremor-dominant PD type, the akinetic-rigid type (non-tremor-dominant type) showed severe neuronal loss in the medial and lateral substantia nigra and locus coeruleus [44
]. In contrast, we found that disease duration, motor scores and levodopa treatment did not differ between the two PD groups, and we did not find a specific motor subtype that correlated with pRBD, although there was a trend towards a higher tremor score in the pRBD group compared with the non-pRBD group. Similarly, a study of 457 PD patients with sleep disturbances did not find a characteristic clinical subtype for PD with RBD but did report a higher disease severity and longer disease duration in PD patients with RBD than those without [13
]. Because most of our patients were medicated, we cannot exclude that a tremor-dominant type may transition into a non-tremor-dominant type over time and with increased age [45
]. In early PD (disease duration of up to 5 years and HY stage of between 1 and 2.5), the RBD co-morbidity (confirmed by clinical history during the preceding 6 months and a cut-off score
4 on the RBDSQ) was high (55%); however, no differences in the clinical subtype and disease severity have been reported recently [42
Our study demonstrated that the scores for the PDQ-39 cognition and emotional well-being dimensions were higher in patients with pRBD than in those without pRBD. Patients with idiopathic RBD have no cognitive or motor complaints, but they frequently develop characteristic motor and cognitive symptoms, such as impairment of visuospatial abilities and deficits in verbal memory, attention and executive function, which are also observed in patients with PD, PD with dementia, MCI and DLB [46
]. Several risk factors associated with dementia in PD, such as the akinetic-rigid type, hallucinations, longer disease duration and male gender, correlate with RBD in PD [46
]. In our study, hallucinations also correlated significantly with PD-RBD.
The negative impact of RBD on the QOL has been previously reported [7
]. In that report, no changes were found in the PDQ-39 score, but lower QOL scores were noted (these scores were measured using the 36-item short form health survey on emotional and mental functioning) in PD patients with RBD, suggesting that RBD in PD may affect the QOL in a non-disease-specific manner. In contrast, in our study, similar components of the QOL were affected in the PDQ-39, the disease-specific QOL questionnaire for PD. Because the pRBD group showed higher PDSS scores and PDQ-39 cognition dimensions than the group without pRBD, impaired sleep may play a role in cognitive dimensions in this study.
We found that PD patients with pRBD showed significantly higher PDSS-2 scores related to sleep onset insomnia, nocturnal hallucinations and vivid dreams but that the ESS and PSQI scores were similar between the groups. The impairment of REM sleep has been attributed to the occurrence of visual hallucinations and delusions in PD [47
]. RBD is a possible predictor for the development of cognitive impairment and visual hallucinations in PD, and RBD, hallucinations and cognitive impairment may progress together with motor impairment [48
]. Because LED and disease severity were not different between the pRBD and non-pRBD groups, these nocturnal dysfunctions may be related to impairments in non-dopaminergic systems.
The limitations of our study include the absence of PSG examinations, which may have resulted in an underestimation of the frequency of RBD because the frequency of RBD, diagnosed using clinical history alone, ranges from 15% to 46% whereas the frequency of PSG-confirmed RBD varies from 46% to 58% in PD patients [5
]. In addition, the absence of any significant difference in several clinical characteristics and sleep parameters in this study between the PD patients and controls and between the pRBD and non-pRBD groups may be due to insufficient statistical power.