The utility of neuro-QOL in idiopathic REM sleep behavior disorder

Abstract

Introduction

Idiopathic rapid eye movement (REM) sleep behavior disorder (iRBD) can affect quality of life (QOL) for both patient and bed partner; has been less well-studied. Utilizing the Neuro-QOL, we aimed to investigate QOL complaints in subjects with iRBD, and whether QOL changes were associated with phenoconversion to neurodegenerative illness.

Methods

We prospectively enrolled subjects from the “REM Sleep Behavior Disorder Associations with Parkinson’s Disease Study (RAPiDS)” cohort and evaluated them via the NeuroQOL, both at baseline and then at follow-up evaluations. Determination of phenoconversion was ascertained from physical examination and medical chart review.

Results

Of those who completed both evaluations, there were 33 subjects with iRBD, with an average age of 61.9 ± 13.0 years, with 13 women and 26 men. Various QOL changes were found among those who phenoconverted versus those who did not; but following correction, none of these changes were significant.

Conclusions

This is the first time the Neuro-QOL has been studied in iRBD. QOL can be affected in this condition, but other screening tools will likely be needed for future studies.

Introduction

Rapid eye movement (REM) sleep behavior disorder (RBD) consists of abnormally increased muscle tone during REM sleep (noted during polysomnogram [PSG] testing) combined with a history of recurrent nocturnal dream enactment behavior(Schenck et al. 1986). Idiopathic RBD (iRBD) occurs in the absence of conditions known to cause secondary RBD (i.e. autoimmune or inflammatory disorders), and when other causes of possible abnormal nocturnal behavior have been ruled out (i.e. nocturnal seizures)(Iranzo and Santamaria 2005; Peever et al. 2014).

It is well-known that iRBD is linked to Parkinson’s disease (PD) and other alpha-synucleinopathies(Postuma et al. 2019). Previously, we had reported on the psychiatric, autonomic, and sleep impact of iRBD in a cross-sectional fashion(Barone et al. 2020), as well as the neurologic and psychiatric features of impending neurodegeneration in iRBD(Barone et al. 2023). Here, we prospectively focus on quality of life (QOL) aspects of subjects with iRBD and how they change over time. While QOL has been evaluated in those with RBD associated with PD (Barber et al. 2017; Park et al. 2020), to our knowledge this is the first time QOL changes have been monitored over time in iRBD.

We analyzed data from the “REM Sleep Behavior Disorder Associations with Parkinson’s Disease Study (RAPiDS)” cohort (described previously (Barone et al. 2020, 2023)) both at baseline and then at follow-up evaluations utilizing the Neuro-QOL; this is a self-reported QOL measurement system intended to be brief, reliable, valid, responsive, and consistent across various neurologic conditions (Gershon et al. 2012).

The International Parkinson and Movement Disorder Society Task Force has recommended the Neuro-QOL for evaluation of QOL in those with PD; it has been shown to have high internal consistency, test–retest reliability, and good correlation with PD-specific measures (Shulman et al. 2016). As such, the Neuro-QOL has been utilized as an outcome measure in subjects with advanced PD (Fleisher et al. 2020). In another study, evaluating for mild cognitive impairment (MCI) in subjects with PD, the Neuro-QOL was administered at baseline, 1, 2, and 3 years. Increasing subjective cognitive complaints in the Neuro-QOL were associated with development of PD-MCI over 3 years of follow-up (Mills et al. 2020).

Despite these obvious uses for the Neuro-QOL in established PD, the question of its utility in early PD or even in those in the prodromal phase (i.e. iRBD) remains in question. Recently, the Neuro-QOL was evaluated for use as an outcome measure in subjects with early PD; over a 3 year period, there were found to be only modest changes in five of eight Neuro-QOL domains, suggesting that more sensitive outcome measures would likely be required for trials in early PD (Marras et al. 2021).

Our aim was to investigate Neuro-QOL changes over time in those with iRBD. And whether certain Neuro-QOL changes were associated with conversion to PD and other neurodegenerative illness.

Methods

The institutional review board at Weill Cornell Medical College approved this protocol, and informed consent was provided by all participants. Each author completed a Conflict of Interest Disclosure form. This study was funded through a private grant.

All patients underwent polysomnogram (PSG) testing as part of the clinical evaluation of parasomnias and other sleep disorders. Consecutive adult study participants with iRBD were then prospectively recruited from the Weill Cornell Center for Sleep Medicine. Scoring was performed by registered technologists, according to the AASM scoring manual guidelines (Berry et al. 2012) which includes either sustained muscle activity (tonic activity) of the chin EMG or excessive transient muscle activity (phasic activity) of the chin or limb EMG during REM sleep. The EMG technical specifications were as follows: sampling rate of 200 Hz, low frequency filter of 10 Hz, high frequency filter of 100 Hz, maximum electrode impedance of 5 KΩ, and digital resolution was 16 bits per sample. For tonic activity, an epoch of REM sleep with at least 50% of the duration of the epoch having chin EMG amplitude greater than the minimum amplitude demonstrated in non-REM sleep was considered abnormal. For phasic activity, a 30-s epoch of REM sleep was divided into 10 sequential, 3-s mini-epochs, and abnormally elevated REM muscle tone was scored if at least 5 (50%) of the mini-epochs contained bursts of transient muscle activity in the chin or limb EMG. Excessive transient muscle activity bursts were defined as 0.1–5.0 s in duration and at least 4 times as high in amplitude as the background EMG activity (Berry et al. 2012).

Individuals with an existing diagnosis of any neurodegenerative disorder were excluded. Those with other serious neurological disorders including stroke, epilepsy, and a history of brain tumor, hydrocephalus, encephalitis and other disorders were excluded. Additional exclusion criteria were existence of conditions that would confound autonomic and other neurological testing, such as cardiac disease, uncontrolled thyroid disease or diabetes, autoimmune conditions, and specific medications. Following confirmation of iRBD and recruitment into the study, all participants were evaluated with the Neuro-QOL, which was completed during face-to-face interviews, occurring at both baseline and follow-up evaluations. The Neuro-QOL was developed by the National Institute of Neurological Disorders and Stroke and validated in the adult general population, incorporating the assessment of physical, emotional, cognitive and social patient function (Gershon et al. 2012; National Institute of Neurological Disorders and Stroke 2015;). Using a 5-Point Likert Scale it measures 13 domains including: communication, ability to participate in social roles and activities, anxiety, depression, emotional and behavioral dyscontrol, fatigue, lower extremity function, positive affect and well-being, sleep, upper extremity function, stigma, satisfaction with social roles and activities, and cognitive function.

Determination of phenoconversion to neurodegenerative illness was ascertained through physical examination and medical chart review of subjects; terminology including MCI, dementia, Alzheimer’s disease, PD, parkinsonism, multiple system atrophy, and Lewy body dementia were queried for among the iRBD group.

To analyze changes in the measured parameter for each subject over time from baseline to follow up (V01) (Years), we utilized Linear Mixed Effects models that are extensions of simple linear models and incorporate both fixed and random effects (Pinheiro and Bates 2000). Specifically, in this study, we utilized a random intercept model (https://www.statsmodels.org/stable/mixed_linear.html). The dependent variable was the measured parameter of each subject, while the covariate was the variable "time (Years)", accounting for a fixed effect. To consider individual differences, in this model, the random intercept of each subject was used as a random factor accounting for random effects. There were two outputs: coefficient value and p-value. The coefficient value implies the increase (Positive coefficient) or decrease (Negative coefficient) relationship from baseline to V01. The p-value indicates whether the positive or negative correlation is statistically significant (p value < 0.05). During computing, the median value was used to fill missing values. In addition, the p-value (T-test/ Mann–Whitney U test) was provided, which is used to test whether there is the difference between baseline and V01 in any parameter. The p-value was obtained by t-test or Mann–Whitney U test, which is a nonparametric test. Given the volume of testing, a Bonferroni correction was performed to account for Type I error.

Results

For baseline testing, 65 subjects with iRBD were recruited. Of those who completed both baseline and V01, there were 33, with an average age of 63.1 ± 12.8 years, with 9 women and 24 men (see Table 1). During the time frame of this study, various factors played into the large attrition rate, most notable being the COVID-19 pandemic and the resultant lack of subject’s ability and willingness to participate. The follow-up appointments were set to occur approximately 1 year after the initial evaluation, although there were several instances in which the follow-up evaluations took place 2–3 years after the initial evaluation due to the above reasons.

Table 1 Demographics

Based on physical examination and electronic medical record data, 8 subjects developed neurodegenerative illness, which included PD, parkinsonism, Lewy body dementia, and other forms of dementia and/or cognitive impairment; these were labeled the “ND + RBD” (neurodegeneration) group. The 25 iRBD subjects who did not phenoconvert were labeled the “ND-RBD” group (see Table 1).

For the ND + RBD group, there was more fatigue (3 items), less satisfaction with social roles and activities (3 items) and improved cognition (1 item) over time, but no change in any aspect of motor function (see Table 2). In the ND-RBD group, there was a worsening of cognition (2 items) and increased depression (1 item) over time (see Table 3). These changes were initially significant (p < 0.05), but when the Bonferroni correction was applied, all comparisons were found to be non-significant, including a comparison of changes over time between each group (data not shown).

Table 2 Changes in NeuroQOL from baseline (V00) to follow-up (V01) for ND + RBD

Table 3 Changes in NeuroQOL from baseline (V00) to follow-up (V01) for ND-RBD

Discussion

The Neuro-QOL has been validated in patients with multiple sclerosis (Miller et al. 2016) and PD (Nowinski et al. 2016), and has been used in patients with diabetic peripheral neuropathy (Saraf et al. 2022) and epilepsy (Andersson et al. 2022). While QOL changes have been evaluated in those with RBD associated with PD (Barber et al. 2017; Park et al. 2020), to our knowledge this is the first time the Neuro-QOL has been utilized in those with iRBD. While our present study did not demonstrate significant differences across time, there does exist a precedent for QOL changes in those with iRBD, as noted in the seminal work by the International RBD Study Group (Postuma et al. 2019). In that paper, factors predicting phenoconversion included motor abnormalities (both objectively and subjectively), decrease in olfaction, mild cognitive impairment, erectile dysfunction, decrease in color vision, and constipation (Postuma et al. 2019), all of which can negatively impact QOL. Furthermore, an earlier analysis of our cohort corroborates this notion (Barone et al. 2020). But our current findings are in agreement with Marras et al. (Marras et al. 2021), in that, while the Neuro-QOL is very useful in mid- and late-PD, it may not be as useful in iRBD.

To this point, we found somewhat conflicting and contrasting results among those who phenoconverted and those who did not. Whereas it was unsurprising to find that in the ND + RBD group there was more fatigue and reduced social functioning over time, the finding of improved cognition was unexpected, as was the increase in depression and worsening cognition over time in the ND-RBD group. Following correction, none of these changes were statistically significant, which is in agreement with prior studies (Thimm et al. 2021; Carlozzi et al. 2020), suggesting that in certain neurologic illness (i.e. spinal muscular atrophy and Huntington’s disease), the Neuro-QOL may not capture the clinical spectrum of patients’ symptoms, particularly if early in disease or if the monitoring time is not long enough.

It is well-known that RBD may be triggered by medications, especially antidepressants (Biscarini et al. 2024; Lam et al. 2019; Barone 2024); but the implication of whether this is tantamount to impending neurodegeneration is still unclear (Barone and Henchcliffe 2018). Whereas some data suggests that markers of prodromal neurodegeneration may be present in those with antidepressant-associated RBD (Jiang et al. 2017), it also has reported that RBD patients taking antidepressants had less chance of developing neurodegenerative disease than those without antidepressant use (Postuma et al. 2013). More recently, an observational study compared the features of iRBD and antidepressant-related RBD, and it was found that those with antidepressant-related RBD have features suggesting a lower risk of phenoconversion (Biscarini et al. 2024). This question is particularly relevant to our cohort, as more than half of the ND-RBD group had taken antidepressants (n = 14), which begs the question of whether these subjects are suffering with antidepressant-related RBD. If this is the case, then the Neuro-QOL changes noted in this group (i.e. increase in depressed mood and worsening of cognition) might be explained as occurring secondary to a psychiatric disorder instead of neurodegenerative illness. Further studies, including longer-term follow up, will need to be performed to answer such questions.

Ultimately, it seems reasonable to conclude that iRBD is a very heterogenous disorder. A recent meta-analysis by Huang et al. speaks to this point (https://www.statsmodels.org/stable/mixed_linear.html.); the authors noted that different PD-related genes have varying associations with RBD. For example, GBA variants can increase the risk and severity of RBD in both the prodromal and clinical stages of PD, while LRRK2 G2019S is negatively associated with RBD. While not a component of our present study, these variations underscore the heterogeneity that may result in inconclusive findings. Similarly, on the one hand it has been shown that those with iRBD tend to have more sleepiness than controls (predicting a more rapid conversion to parkinsonism and dementia) (Gershon et al. 2012), and on the other hand it has been found that those with iRBD may actually demonstrate less sleepiness than controls (Shulman et al. 2016). These conflicting data, especially when viewed in light of our findings, may be partially explained by the aforementioned heterogeneity and complexity of iRBD, and/or the inherent shortcomings of screening tools.

Limitations

There were some important limitations in our study, including pandemic-related lack of follow-up and possible bias in terms of which subjects presented for evaluation. Small sample size may have affected the impact of our findings, as did the relatively short period from baseline to follow-up, although the follow-up period of 1–3 years is typical for this type of project. Ideally, continued testing of the recruited subjects over a longer timeframe, as well as re-recruitment of those lost to follow-up would garner further important data. The diagnoses of neurodegenerative illness were ascertained from the subjects’ medical records; this was done given the difficulty of in-person examinations during the pandemic. The study enrolled patients from a single sleep center, presenting the possibility of referral bias, and the cohort is derived from an urban setting thus limiting its generalizability. Finally, while there was no control group per se, the Neuro-QOL has been shown to demonstrate evidence of validity insofar that it can differentiate people based upon number of reported health conditions and whether those reported conditions impede normal function (Gershon et al. 2012).

Conclusion

It is clear that much remains to be learned about iRBD, and how this may affect QOL in the short-term, and predict phenoconversion long-term. While the Neuro-QOL questionnaire has been successfully utilized in several neurologic conditions, particularly advanced PD, it seems that its utility in iRBD may not be as profound.

Despite our findings, it is known that QOL decrements are a major component of chronic neurologic illness. Previously, we and others demonstrated that those with iRBD may be subject to mild cognitive changes and other subtle symptoms (Postuma et al. 2019; Barone et al. 2020). Given such prior data, it seems there exists a need for screening patients with iRBD for QOL complaints.

Early detection of iRBD via utilization of QOL screening measures (Beek et al. 2022), or other such tools, may aid in eventually play a role in more accurately stratifying which patient groups are at risk for impending neurodegeneration. Appropriate screening would not only improve care for iRBD patients in the short term, but the identification of potential early indicators of neurodegeneration could facilitate inclusion in trials focused on disease prevention or in slowing progression.