Abstract
Keywords
INTRODUCTION
Although nonmotor symptom (NMS) fluctuations related to wearing-off and On/Off phenomena have been described to be present in up to 100% of patients with motor fluctuations [1], NMS fluctuations have only been recognized as an important aspect of Parkinson’s disease (PD) quite recently. NMS fluctuations can be more troublesome and disabling than motor disturbances [1–3], and fluctuations of some NMS such as anxiety/depression and pain have negative effects on health-related quality of life (hr-QoL) [4, 5]. Patterns of NMS fluctuations are heterogeneous and complex with psychiatric NMS fluctuating more frequently and severely compared to autonomic and sensory NMS [4, 5]. Demographic parameters and motor function do not correlate with NMS or their fluctuation frequency or severity. Since previous reports aimed to study NMS fluctuations in direct conjunction with motor oscillations meaning that NMS were examined within On and/or Off motor states and subsequently correlated with motor function, there are no data available on the temporal context of NMS fluctuations and motor oscillations. The present study used a NMS diary asking subjects to hourly judge the presence of nine key NMS in parallel to the standard PD motor home diary to dissect the temporal connection between NMS fluctuations and motor oscillations in advanced PD patients. We chose key psychiatric NMS, which are likely to fluctuate during the day, and – in contrast – autonomic NMS, which are less likely to fluctuate [4], to obtain first information on this novel topic of NMSresearch.
SUBJECTS AND METHODS
Subjects
Subjects across all ages and disease severities fulfilling UK PD Brain Bank criteria [6] with or without documented motor fluctuations were enrolled as convenient cohorts at two inpatient Movement Disorder centers between January 2013 and March 2014. PD patients were classified as non-fluctuators if they had no reports of motor fluctuations in their medical records and/or history as well as no points in Unified PD Rating Scale (UPDRS) part IV questions 31 through 38 [7], while fluctuators had documented motor fluctuations (medical history/records and UPDRS part IV). All PD subjects were inpatients and admitted to the Movement Disorders centers for optimizing PD treatment and multidisciplinary intensive rehabilitation treatment. Healthy control subjects were recruited within the same time period and investigated at the adjoined outpatient clinics. PD subjects were excluded if they had an identifiable cause of parkinsonism or signs for atypical parkinsonism, psychosis, or dementia (Montreal Cognitive Assessment [MoCA] [8]≤26 points) or other relevant conditions interfering with the study protocol. All subjects provided written informed consent and the study was approved by institutional review boards at participatingsites.
Assessments
We assessed basic demographic data including type of motor complication, modified Hoehn-Yahr score [9], UPDRS [7], NMS Scale (NMSS) [10, 11], BDI-1A [12], and PDQ-39 [13]. All subjects underwent a diary training session for a pair of diaries: (a) the modified diary for motor function introduced by Hauser and colleagues with four different motor states (asleep, motor Off, On without dyskinesia, and On with dyskinesia) [14] and (b) a novel NMS diary asking to rate nine key NMS (psychiatric NMS: Anxiety, depressive mood, inner restlessness, difficulties with concentration; fatigue; autonomic NMS: Excessive sweating, sialorrhea, bladder urgency and dizziness) as present or absent during awake time using the same questions as the NMSQuest [15]. During the training session, patients were instructed as to how the different functional states are defined and how the diary should be completed (placing a tick mark on a daily diary card every 60 minutes reflecting their predominant status over the prior hour period; for time asleep, the diary was completed upon awakening). Patients then completed both diaries on 5 consecutive days. All diary sets were included for analysis. Individual time periods were excluded from analysis if there was no response or more than one response on either diary, or if the patient indicated they were asleep on one diary but not the other. For all analyses except for the 24-hours distribution analyses, hours marked as asleep or switches from or towards asleep were omitted from the calculations. Of note, if NMS was rated as “present” the NMS state in that hour was interpreted herein as NMS Off state and if NMS was rated as “absent” the state was interpreted as NMS On state to allow for an easier comparison with motor states.
Statistics
Statistical comparisons of data between subject groups and motor/NMS patterns were calculated using χ2 test, Fisher’s exact test, Wilcoxon rang order test, Kruskal-Wallis test with post-hoc Bonferroni-adjusted Man-Whitney U-test, one-way ANOVA with post-hoc unpaired two-sided Bonferroni-adjusted t-test, Mann-Whitney U-test, paired or unpaired two-sided t-test as appropriate. Pearson’s correlation test was used to examine correlations (|r| < 0.3 was considered a weak, |r| = 0.3–0.59 a moderate, |r|≥0.6 a strong correlation). Data were analyzed using SPSS 21.0 (SPSS Inc., Chicago, IL). If not mentioned otherwise, all data are displayed as means ± SD [range], median [interquartile range; IQR] or numbers (%), significance level was set at P < 0.05 (two-sided test).
RESULTS
Subject cohort
A total of 32 PD subjects (21 [66%] women) with motor fluctuations (n = 15, 10 [67%] women) or without motor fluctuations (n = 17, 11 [65%] women) and healthy 15 controls (9 [60%] women) were enrolled into this study (for demographics and clinical data, see Table 1). Within the subject group with motor fluctuations (duration: 50 ± 51 [2–153] months), we observed differences in Hoehn-Yahr stage between On (2.2 ± 0.5 [1–3]) and Off state (2.8 ± 0.9 [2–5]; P < 0.001 [Wilcoxon test]), in UPDRSIII motor score between On (14.8 ± 5.7 [5–28]) and Off state (27.3 ± 8.3 [12–44]; P < 0.001 [paired two-sidedt-test]). All subjects of this group displayed at least one type of motor complication (median: 4 [IQR: 2/5]) with On-Off phenomenon (n = 12 [80%]), end-of dose akinesia (n = 10 [67%]), Off period dystonia (n = 9 [60%]) and peak-dose dyskinesia (n = 9 [60%]) as the most frequent complications.
Diary usage
PD home diary usage time periods with respect to motor and NMS states are summarized in Table 1. Of the total time awake (healthy controls: 1,140; non-fluctuators: 1,320 hour; fluctuators: 1,198 hours), healthy controls indicated that they spent no single hour in motor Off state. Although we used strict selection criteria for non-fluctuating patients (see Materials & Methods section), 3 non-fluctuating patients (17%) rated together 2.0% hours as motor Off state (27 hours) and the remaining time in On state. Fluctuating patients rated 27.6% as motor Off, 52.1% in motor On without dyskinesia, and 20.3% in On with dyskinesia (72.4% total On time; P < 0.001; χ2 test). We detected no motor state switches in healthy controls, only 1.1% of maximal possible motor state switches in non-fluctuators, but 22.4% in fluctuators (P < 0.001; χ2 test).
Motor state and NMS state periods
If not stated otherwise, for further analyses we used the total motor On time (combined On without and On with dyskinesia time) for correlations with NMS data for clarity. As summarized in Table 2, motor Off state time and Off time for psychiatric NMS were more frequent in fluctuating patients compared to non-fluctuators and controls, while hours in Off state for autonomic symptoms were equally distributed between the groups. The most frequently indicated NMS in fluctuators was fatigue with 20.9% of total time awake, followed by inner restlessness (16.7%) and depressed mood (13.2%). Occurrence patterns of the four possible distributions of NMS occurrence with respect to the motor state ratings on the single hour time period level are displayed in Fig. 1. In fluctuating patients, NMS Off time periods were similarly distributed between motor On and Off periods for most NMS (ratios of NMS Off/motor Off divided by NMS Off/motor On: 0.8–1.5), with anxiety (ratio: 4.5) and the autonomic symptoms sialorrhea, bladder urgency and dizziness (ratios: 0.3–0.5) as the major exceptions.
Comparing the frequencies of NMS Off state time between the two motor states on the hour level, all NMS except sialorrhea and bladder urgency were more frequently rated as present in motor Off compared to On state hours (Fig. 1C). Differentiating motor On state without and On state with dyskinesia revealed a continuous reduction of the occurrence of depressive mood from motor Off state through On state without to On state with dyskinesia (post hocχ2 test with Bonferroni adjustment; Supplementary Table S1). In contrast, inner restlessness and dizziness are less frequent in motor On without dyskinesia, but equally frequent in motor Off and dyskinetic state hours (post hocχ2 tests with Bonferroni adjustment; Supplementary Table S1).
Correlation analyses between motor Off and NMS Off state times (relative to total time recorded per patient) from patient’s diary data using Pearson’s correlations test revealed moderate correlations (r = 0.451–0.577; P < 0.05) only for the psychiatric NMS anxiety, depression, inner restlessness and concentration/attention and for fatigue, but no or only weak correlations for the autonomic NMS.
Circadian patterns of NMS
We next plotted the 24-hours profiles of motor symptoms and NMS for all three subject groups to elucidate the circadian patterns of NMS in conjunction with motor fluctuations. In healthy controls (Supplementary Figure S1), there is a randomized distribution of NMS over the day for most NMS, but a peak of NMS present in the evening for fatigue and in the early morning for excessive sweating with frequencies of > 10% or more of total hours. In non-fluctuating PD patients (Supplementary Figure S2) there is a small peak of motor Off state in the early afternoon in conjunction with fatigue (but rated by only 2 individual patients), but an irregular inverse U-shaped distribution for the other NMS. In fluctuating patients (Fig. 2), motor symptoms and NMS showed 3 small peaks for motor Off state and anxiety, which were less visible for depressive mood and inner restlessness (morning, early afternoon and early evening). The other NMS showed an increase in the morning with relative stable frequencies for the rest of the day time.
Motor and non-motor fluctuations
We next analyzed the quantity of switches between motor On and Off state and NMS states and their interdependency (Table 2, Fig. 3). Changes from or towards “asleep” were excluded from the analyses. Switches between motor states and NMS states of anxiety, depression, inner restlessness and concentration were more frequently indicated in fluctuators compared to controls and non-fluctuators, while for autonomic NMS (except dizziness) no differences between the groups were detected. For depression, switch rate was also higher in non-fluctuators than in controls. Interestingly, for fatigue and dizziness we detected differences between both PD groups and the control group, but not between the two PD groups. Changes of motor state and fatigue state were most frequently indicated by fluctuating PD patients (23.6% and 25.6% of all possible changes), followed by inner restlessness (12.5%) and concentrationproblems (8.1%). To estimate the interdependency of motor and NMS fluctuations, we analyzed the switch patterns on the single hour time period level (Fig. 3). In non-fluctuating patients, no single NMS state switch in conjunction with a motor state switch could be detected. In fluctuating patients, most changes of NMS states were independent of motor state switches or in discordance with motor state switches (NMS and motor state switched in opposite direction) with a concordance rate of only 25.9–42.9% of all NMS switches for psychiatric and 6.5–16.5% for autonomic NMS (Fig. 3B). Similar results were obtained when On with and without dyskinesia were analyzed separately.
Correlation analyses between the number of motor and NMS state switches (relative to maximal possible switches per patient) from patient’s diaries using Pearson’s correlations test revealed moderate correlations (r = 0.356–0.593; P < 0.05) only for the psychiatric NMS anxiety, depression, inner restlessness, concentration/attention, but not for fatigue (r = 0.063; P = 0.733) or autonomic NMS.
Correlations of NMS diary data with other measures
We next correlated the NMS diary data with demographic and clinical data in PD patients (combined cohorts of non-fluctuators and fluctuators; Supplementary Table S2). In general, we found moderate correlations between hours in NMS Off state (relative to total time recorded) as well as numbers of NMS state switches (relative to maximal possible switches per patient) and disease duration and levodopa equivalent dose only for psychiatric NMS, but not for fatigue and autonomic NMS. NMS Off state hours and switches correlated moderately with hr-Qol measure (PDQ-39 sum score) for depression, inner restlessness, and concentration/attention as well as fatigue (r = 0.341–0.594; P < 0.05). Correlation analysis of only mood (anxiety and depression) NMS Off state diary data with independent measures of related constructs revealed moderate correlations with the corresponding NMSS domains/items, BDI score and PDQ-39 domains (r = 0.335–0.392; P < 0.05). No other relevant correlations were detected between NMS diary data and demographic and other PD-related clinical measures (sex, age, disease duration, duration of fluctuations, Hoehn-Yahr stage, UPDRSIII motor score in On and Off state). Correlations analyses with the cohort of fluctuating patients alone revealed very similar results.
DISCUSSION
Although NMS fluctuations in advanced PD have received increasing research attention during the recent years, the circadian occurrence patterns of NMS and the temporal connection between NMS fluctuations and motor oscillations remain largely enigmatic. Since previous studies used the motor state as the basis for investigating frequency and severity of NMS fluctuations [1, 16–19], these studies drew conclusions neither on circadian occurrence patterns of NMS (as shown for the motor symptoms using the PD home diary since more than 10 years) [14, 21] nor on temporal dependency of motor and NMS fluctuations. Here we used here a NMS diary in conjunction with the standard motor diary [14] to provide data on circadian occurrence and temporal interrelations between motor and major psychiatric and autonomic NMS state fluctuations.
For all investigated NMS, Off state (hours with NMS rated as present) were less frequent compared to motor Off state and NMS On-Off state switches as a measure of NMS fluctuations were less prevalent in comparison to motor state oscillations. In close agreement with previous reports showing more frequent and severe fluctuations of psychiatric NMS [4, 5],Off state time and On-Off switches of psychiatric NMS were moderately correlated with motor Off state time while autonomic NMS Off state time and switches showed only weak or no association with motor Off state time/switches. However, changes in NMS state occurred largely independent of changes in motor states with 57–74% discordant state changes for psychiatric and 84–94% discordant changes for autonomic NMS changes.
The 24-hours circadian profiles of NMS in healthy control subjects revealed a randomized distribution of NMS with low frequencies over the day for most NMS, but a peak with frequencies above 10% of total ratings per individual hour for fatigue in the evening and for excessive sweating in the early morning. Interestingly, in general there is not such a clear pattern of NMS occurrence over the day in non-fluctuating PD patients, but an irregular inverse U-shaped distribution for NMS occurrence (the small peak of fatigue Off state in the early afternoon came off by ratings from only 2 individual patients). In fluctuating patients, we observed 3 small peaks for motor Off state and anxiety in the morning, early afternoon and evening, which were less pronounced for depressive mood and inner restlessness. The other NMS showed relative stable frequencies for the rest of the day time.
This is the first use of a NMS diary using questions of the NMSQuest [15] to estimate NMS circadian occurrence and their temporal connection to motor oscillations. The higher proportion of missing data from the NMS diary compared to the motor diary (for details of diary usage, refer to Table 1) suggests that patients had more difficulties in rating their NMS than their motor state on an hourly basis. There is however – to our knowledge – no other tool available to assess NMS with high frequency (several times per day) and validation of these data will be challenging due to the lack of suitable external validation criteria. However, the comparison of the PD home diary data from fluctuating patients with healthy controls and non-fluctuators showed a much higher frequency of hours with NMS marked as present (NMS Off state) and much more NMS state switches strongly suggesting that we – at least in part – measured NMS fluctuations specifically occurring in PD patients with motor fluctuations. Indeed, convergent validities of diary data on the fluctuating symptoms (particularly mood symptoms) [4, 5] with independent measures of related constructs were in general moderate. We suggest future studies using a short standard interview to assess both motor function and key NMS simultaneously at several time points per day to confirm our diary data. Mobile devices with reminder functions might further improve practicability and adherence to diary entries.
Our study has several limitations: Firstly, we recruited a rather heterogeneous and small sized patient cohort, which is however similar to other larger cohorts investigating NMS in advanced PD [4, 22]. Although we used strict criteria defining non-fluctuating patients, three patients of this group (17%) reported short (1 hour) motor Off periods in their diaries. The reasons for these ratings are unclear, but these patients might experience very early fluctuations not detected during study recruitment. Secondly, as a consequence of the relatively small sample size, the occurrence of autonomic NMS (numbers of hours with NMS rated as present and numbers of On-Off-state switches) was low in both PD groups. The statistical analyses of the autonomic NMS have thus little authority and presumably limited clinical relevance. Thirdly, we selected only nine NMS without severity quantification to limit data quality as little as possible due to bothering the patients too much which might lead to early diary fatigue [23, 24]. We chose NMS with major impact on hr-QoL that are likely to fluctuate during the day (psychiatric NMS and fatigue) and – in comparison – that are likely not to fluctuate (autonomic NMS) [4, 25]. This limitation together with the heterogeneity of NMS fluctuation patterns [4, 5] restricts the interpretation of the results to the nine investigated NMS.
Together, we provide first data on the temporal context of NMS fluctuations during the day showing a very low concordance rate of NMS fluctuations with motor oscillations for all investigated NMS. However, we observed similar patterns of motor Off state and NMS Off state distributions over the day formost psychiatric NMS with morning, early afternoon and early evening Off periods. The other NMS, particularly fatigue, showed relatively stable frequencies over the day with no association with motor Off periods. These data together with the low concordance rates of NMS and motor On-Off switches implicate similar circadian fluctuations patterns of psychiatric NMS compared to those of motor fluctuations, but without close timing and/or with different kinetics of NMS and motor fluctuations. We found no evidence for NMS fluctuations in PD patients without motor oscillations. Our results are of particular interest, since they might explain – at least in part – the weak or even absent correlations of NMS fluctuation severities with motor function measures observed in previous studies, which examined NMS fluctuations only in close conjunction with the functional motor state.
CONFLICT OF INTEREST
AS has received unrestricted research grants and consultancy fees from Global Kinetics Corporation, Melbourne, Australia. The authors have no other conflict of interest to report related to this manuscript.
Footnotes
ACKNOWLEDGMENTS
The authors thank all patients for their participation and the study center personnel, namely Cecile Bosredon, Dorena Galle, Rowena Karl and Antonia Maass. This study was supported in part by an unrestricted research grant from Global Kinetics Corporation (GKC), Melbourne, Australia. The financial sponsors of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had the final responsibility for the decision to submit for publication. We also acknowledge the international Parkinson’s disease nonmotor group who has contributed to the concept of this work.