Sage Journals: Discover world-class research

Abstract

Dysfunction of the glymphatic system, an intracranial clearance pathway that drains misfolded proteins, has been implicated in the onset of Parkinson’s disease (PD). Recently, the coupling strength of global blood-oxygen-level-dependent (gBOLD) signals and cerebrospinal fluid (CSF) inflow dynamics have been suggested to be an indicator of glymphatic function. Using resting-state functional magnetic resonance imaging (MRI), we quantified gBOLD–CSF coupling strength as the cross-correlation between baseline gBOLD and CSF inflow signals to evaluate glymphatic function and its association with the clinical manifestations of PD. We found that gBOLD–CSF coupling in drug-naïve PD patients was significantly weaker than that in normal controls, but significantly stronger in patients less affected by sleep disturbances than in those more affected by sleep disturbances, based on the PD sleep scale. Furthermore, we collected longitudinal data from patients and found that baseline gBOLD–CSF coupling negatively correlated with the rate of change over time, but positively correlated with the rate of change in UPDRS-III scores. In conclusion, severe gBOLD–CSF decoupling in PD patients may reflect longitudinal motor impairment, thereby providing a potential marker of glymphatic dysfunction in PD.

Keywords

Cerebrospinal fluid global resting-state functional magnetic resonance imaging signal glymphatic system Parkinson’s disease sleep disorder

Introduction

Parkinson’s disease (PD), the second most common neurodegenerative disease, is clinically characterized by motor dysfunction, which is attributed to the progressive loss of dopaminergic (DA) neurons in the substantia nigra (SN) and accompanied by aggregation of misfolded alpha-synuclein (α-syn).^1–4 However, the mechanism of α-syn clearance is not completely understood.

The glymphatic system is a recently discovered continuous fluid-clearance pathway that drains misfolded proteins such as α-syn from the brain along perivascular spaces (PVS), formed by the endfeet of astrocytes that surround the arterioles and venules.^5–8 The glymphatic clearance system promotes the convective flow between interstitial fluid and cerebral fluid, ultimately merging in the downstream extra-cranial lymphatic networks.^9,10 Glymphatic clearance is promoted during sleep and is tightly regulated by circadian rhythms,^10,11 whereby even one night of sleep deprivation resulted in an increased Aβ burden significantly.¹² Moreover, an increasing number of studies have suggested a potential functional role of the glymphatic system in neurodegenerative diseases.^13–18 However, most data have been derived from animal models or tracer-based in vivo imaging. While these studies have provided invaluable insight into glymphatic system physiology, the methods are not translatable for in vivo human studies.¹⁹ Thus, less invasive and more direct imaging strategies are urgently needed to better characterize the role of glymphatic clearance in clinical PD pathophysiology.¹⁹ Diffusion tensor magnetic resonance imaging (MRI) has been used to visualize glymphatic function; but it cannot accurately reflect the dynamic aspects of the glymphatic system.^20,21

Global blood oxygen level-dependent (gBOLD) resting-state functional MRI (rs-fMRI) signals of <0.1 Hz have been associated to cerebrospinal fluid (CSF) dynamics. Therefore, low-frequency oscillations in CSF is a possible resource to non-invasively quantify glymphatic system dynamics.^22–26

In this study, we aimed to compare the strength of gBOLD–CSF coupling between drug-naïve PD patients and healthy individuals. Follow-up visits after 1.94 ± 0.86 years also allowed us to conduct a longitudinal study of the effect of gBOLD–CSF coupling on the progression of motor and non-motor deficits in PD. Furthermore, as all of our PD patients were affected by sleep-related symptoms, we explored the association between gBOLD–CSF coupling and sleep disturbances.^11,24,27,28 Most studies conducted to date have primarily been based on cross-sectional data, and longitudinal evaluation is limited. Therefore, we believe our study provides a novel insight into the role of the glymphatic system in patients with PD.

Materials and methods

Participants

The ethics committee of the Second Affiliated Hospital of Zhejiang University approved this study. All participants provided written informed consent in accordance with the principles of the Declaration of Helsinki. Data were obtained from drug-naïve patients clinically diagnosed with idiopathic PD from the Second Affiliated Hospital, Zhejiang University School of Medicine, from August 2014 to November 2021. PD patients were diagnosed according to the Movement Disorder Society Clinical Diagnostic Criteria. Age- and sex-matched healthy individuals were recruited from the community as normal controls (NC). The following exclusion criteria were applied: (1) parkinsonism attributed to immune disease, metabolic disease, trauma, or other secondary factors; (2) presence of additional neurodegenerative diseases simultaneously such as Alzheimer’s or Huntington’s disease; (3) patients with a medical history of stroke or intracranial arteriovenous malformation; (4) cerebral structural abnormalities, such as those resulting from brain trauma or brain tumor; (5) severe cognitive dysfunction, as determined by Montreal Cognitive Assessment (MoCA) scores less than 10 and/or Minimum Mental State Examination (MMSE) less than 9; (6) left- or bilateral-handed patients; and (7) inability to undergo MRI (Figure 1).

Figure 1.

Flow chart of participant recruitment and exclusion.

Clinical assessments

Clinical and imaging evaluations were performed on all participants in the morning to minimize the impact of the circadian rhythm on glymphatic activity. Baseline demographic data and clinical variables of potential interest were collected. Unified Parkinson’s Disease Rating Scale (UPDRS) scores were also assigned.²⁹ UPDRS part III provided an immediate comprehensive neurological assessment of motor symptoms in participants; UPDRS part I and UPDRS part II were applied to assess the non-motor symptoms and motor symptoms in the daily life of patients, respectively. The stage of functional disability associated with PD among patients were distinguished using the Hoehn-Yahr (H-Y) scale.³⁰ Sleep-related behavioral manifestations of PD were evaluated using the Rapid Eye Movement (REM) Sleep Behavior Disorder Questionnaire-Hong Kong (RBDQ-HK)³¹ and the Parkinson’s Disease Sleep Scale (PDSS).³² Higher RBDQ-HK scores indicate more frequent RBD symptoms and lower PDSS scores indicate more severe sleep symptoms. To identify whether sleep disorders in PD are associated with alterations in gBOLD–CSF coupling, we separated baseline PD patients into PDSS-high and PDSS-low subgroups based on whether their PDSS score was above or below the median. At baseline, PD patients with a RBDQ-HK factor 2 score ≥8 were defined as PD with possible RBD (PD-pRBD).³³ All patients followed the appropriate treatment plan as advised by an experienced neurologist after baseline evaluation. And we underwent follow-up visits for 13 times in NC group and 21 times in PD group. At follow-up, all assessments were conducted at the OFF stage. Total daily levodopa equivalent dose was calculated.³⁴

Imaging data acquisition and preprocessing

Brain images were captured using a 3.0 Tesla MRI scanner (GE Discovery 750), equipped with an 8-channel head coil. A foam pillow was used to immobilize the participant’s head and limit artifacts when scanning. A trained radiologist observed the entire scanning process to ensure image quality. High-resolution T1-weighted structural images were acquired using a fast spoiled gradient recalled sequence: repetition time/echo time [TR/TE] = 7.336/3.036 ms; field of view = 260 × 260 mm²; matrix = 256 × 256; slice thickness = 1.2 mm; and slice number = 196 (sagittal). The rs-fMRI data were acquired using an echo-echo planar image sequence: [TR/TE] = 2000/30 ms; flip angle = 77°; field of view = 240 × 240 mm²; matrix = 64 × 64; slice thickness = 4.0 mm; slice gap = 0 mm; and number of slices = 38 (axial). All rs-fMRI data were preprocessed using the Data Processing & Analysis of Brain Imaging toolbox (version 6.0, http://rfmri.org/dpabi)³⁵ based on the Statistical Parametric Mapping (SPM) 12 (www.fil.ion.ucl.ac.uk/spm).³⁶ The first 10 volumes were removed to allow for the magnetization signal to reach equilibrium. The remaining 195 volumes were corrected to avoid intra-scan differences in acquisition time between slices and realigned for inter-scan head motion correction. The images were spatially smoothed with a 6-mm full width at half maximum Gaussian kernel and the linear temporal trends were removed. Band-pass filtering in the range of 0.01–0.1 Hz was applied. Nuisance regression analysis was not applied to the gBOLD signals, CSF signals, and motion parameters because these variables were the focus of the current study.^25,26

Definition of regions of interest (ROI)

The ROI of the cortical gray matter was defined based on the Harvard-Oxford cortical structural atlases.³⁷ To obtain the mask of the cortical region in the native rs-fMRI data space, the T1-weighted images were linearly registered in the rs-fMRI data, and then co-registered (nonlinear) in the Montreal Neurological Institute (MNI) space in SPM12. The inverse transformation matrix was saved for further analysis. The ROI in the MNI space was transformed back to the native rs-fMRI space. As validated by Fultz et al., this signal reflects CSF inflow and is highly sensitive at the bottom of the cerebellum.²⁴ Therefore, we selected the bottom slices of the fMRI images to acquire the signal. Two neurologists, blinded to clinical data, manually defined the CSF ROIs here based on the rs-fMRI images, then verified using the T1 images.

Quantification of gBOLD–CSF coupling strength

After extracting the rs-fMRI signals from the cortical gray matter and CSF, we calculated the cross-correlation function between gBOLD and CSF signals to quantify their coupling strength over a range of time lags (−10s to 10s) using Pearson’s correlation. As the negative peak at +4s has the same amplitude as the positive peak at −4s, the BOLD–CSF correlation at this negative peak was utilized to quantify BOLD–CSF coupling strength. We also estimated the cross-correlation function between the negative derivative of the gBOLD signal and CSF signal, according to a previous study.²⁶ Using a permutation strategy, we examined the statistical significance of the gBOLD–CSF cross-correlations by randomly matching the gBOLD and CSF signals from different patients. This was repeated 10,000 times to build a null distribution based on their correlations and computed using MATLAB. The value of gBOLD–CSF coupling strength was then calculated by applying the segmented ROIs, respectively.

Statistical analysis

The Kolmogorov-Smirnov test was applied to test data normality. Student’s t test was performed to compare the differences in normally distributed continuous data (age and years in full-time education of participants with follow-up visits) and non-parametric tests to compare the non-normally distributed data (years in full-time education and UPDRS, PDSS, and RBDQ-HK scores). The chi-square test was performed to test the difference in categorical metrics (sex and H-Y stage). Spearman’s correlation was performed to analyze the association between gBOLD–CSF coupling and clinical metrics (UPDRS, PDSS, and RBDQ-HK scores). Logistic regression analysis was performed to adjust for age and sex as confounders. For the follow-up cohort, the annual clinical change rate of each clinical feature was calculated as [(follow-up − baseline)/follow-up interval (year)] × 100%.³⁸ Bonferroni test for multiple-comparison correction was performed to test the correlation between baseline gBOLD and UPDRS-I, II and III scores at baseline and their longitudinal changes. Statistical analysis was performed using Statistical Product and Service Solutions version 26.0 (IBM, New York City, NY, USA). Graphs were generated using GraphPad Prism 9.0 (GraphPad Software, San Diego, California, USA). Normally distributed continuous variables are expressed as mean ± standard deviation and non-normally distributed data as median (third quartile–first quartile). The intraclass correlation coefficient (ICC) of the gBOLD–CSF coupling strength value was used to establish inter-rater reliability in the defined CSF region. And the averaged value is applied in analysis. Statistical significance was considered as a p-value of <0.05.

Results

Participant characteristics

We recruited 159 participants from the Second Affiliated Hospital of Zhejiang University and the surrounding community in Hangzhou, Zhejiang Province. The participants included 82 drug-naïve PD patients and 77 normal controls (NCs). Table 1 summarizes detailed sample characteristics. No significant difference was observed in basic demographic information between the PD and NC groups (age, p = 0.397; sex, p = 0.093). For clinical metrics assessments, all participants received UPDRS evaluation; however, 29 and 17 PD patients did not complete the RBDQ-HK and PDSS assessments, respectively. We found that PD patients at baseline exhibited no differences in PDSS scores compared with NC (p = 0.306) and significantly higher RBDQ-HK scores than NC (p = 0.002). In the follow-up cohort, no significant difference was observed in basic demographic information (age, p = 0.455; sex, p = 0.183), but there was a significant difference in PDSS scores (p < 0.001) and RBDQ-HK scores (p = 0.013).

Table 1.

Characteristic of the cohort.

	HC_BL (n = 77)	HC with F (n = 13)		PD_BL (n = 82)	PD with F (n = 21)		p value (PD vs NC)
	HC_BL (n = 77)	BL	F	PD_BL (n = 82)	BL	F	BL	F
Age (years)	60.75 ± 6.76	60.78 ± 5.86	62.72 ± 5.92	59.74 ± 8.11	58.98 ± 6.87	60.92 ± 7.16	0.397	0.455
Gender (Male/female)	32/45	5/8		45/37	13/8		0.093	0.183
Education (years)	9 (12–8)	10.92 ± 3.55		9 (12–5)	9.67 ± 3.71		0.015*	0.336
Disease duration (years)	N/A	N/A		5 (10–2)	2.14 ± 1.50	4.08 ± 1.35	N/A	N/A
Follow-up time interval (years)	N/A	1.94 ± 0.56		N/A	1.94 ± 1.04		N/A	N/A
RBDQ-HK	8 (11.5–4)	4 (9–2)	9 (11.5–3.5)	16 (26–4)	15.5 (20–4.25)	23 (38–7.5)	0.002**	0.013*
PDSS	134 (142.5–123)	130.5 (138–113.75)	130 (144.25–115.5)	132 (141.25–115)	137 (144.5–125)	133 (137–120)	0.306	0.000**
UPDRS-I	0 (1–0)	0 (1–0)	1 (1–0)	1 (2–0)	1 (2–0)	1 (2.5–0)	0.000**	0.600
UPDRS-II	0 (0–0)	0 (0–0)	0 (1–0)	7 (10.25–5)	6 (9–4.5)	6 (10–0)	0.000**	0.000**
UPDRS-III	0 (0–0)	0 (0–0)	2 (3–0)	21 (31.25–13)	16 (20–10.5)	15 (24.5–10.5)	0.000**	0.000**
UPDRS-total	0 (1–0)	0 (1–0)	3 (5–1)	30 (43.25–20.75)	23 (30.5–17)	22 (35–17.5)	0.000**	0.000**
H-Y stage	N/A	N/A		2 (2–1)	2.5 (3–2)	2 (2.5–2)	N/A	N/A
LED	N/A	N/A		N/A	N/A	450 (550–300)	N/A	N/A

Normally distributed continuous variables represent the mean ± standard deviation while skewed continuous variables represent the median (third quartile – first quartile).

NC: normal control; PD: Parkinson’s disease patient; BL: baseline; F: follow-up; UPDRS: Unified Parkinson’s Disease Rating Scale; RBDQ-HK: Rapid eye movement sleep Behavior Disorder Questionnaire – Hong Kong; PDSS: Parkinson's disease sleep scale; H-Y: Hoehn-Yahr. LED: levodopa equivalent dose; N/A: not applicable.

*Significant result with p < 0.05, **Significant result with p < 0.01.

gBOLD signal couples with CSF signal changes

To examine whether gBOLD signals were strongly coupled to CSF signal changes in all participants, the gBOLD and CSF fMRI signals (Figure 2(a)) were extracted from the whole-brain gray matter regions and CSF region. The gBOLD and CSF signals showed corresponding changes in amplitude. Large CSF peaks often were preceded by a large positive gBOLD peak and followed by a large negative gBOLD peak (Figure 2(b)). The mean gBOLD–CSF cross-correlation function was characterized by a positive peak (B = 0.34, p < 0.001; permutation test with n = 10,000) around the −4-s lag, whereas there was a negative peak (B = −0.28, p < 0.001; permutation test with n = 10,000) at the lag of +4 s (orange dashed line in Figure 2(c), left panel). The cross-correlation function between the CSF signal and the negative first-order derivative of the gBOLD signal displayed a large positive peak at the zero-time lag (−d/dt B = 0.11, p < 0.001; permutation test with n = 10,000). Both cross-correlation functions showed patterns consistent with the data, confirming systematic coupling between the global brain signal and CSF flow.²⁴ The average ICC for the gBOLD–CSF coupling was 0.902, revealing a strong agreement between both researchers.

Figure 2.

Systematic coupling between the gBOLD and CSF signals measured by rs-fMRI. (a) CSF signals from the bottom slice of the fMRI acquisition near the bottom of the cerebellum (red line). This bottom slice is expected to have maximal sensitivity to inflow effects.⁷⁶ CSF inflow from outside the fMRI acquisition volume, which does not experienced repeated radiofrequency pulses, increases the fMRI signal.²⁵ Systematic coupling between the gBOLD and CSF signals was measured by rs-fMRI. The gBOLD and CSF fMRI signals were separately extracted from the whole-brain gray matter regions (yellow mask on a T1-weighted structural MRI in the left panel) and CSF (the red mask) region at the bottom slice of the fMRI acquisition of the bright region in the middle and right panel (b, c). (b) The gBOLD and CSF signals from a healthy control showed corresponding changes of large amplitude. Large positive CSF peaks (black downward arrows) often are preceded by a large positive gBOLD peak and followed by a large negative gBOLD peak (black upward arrows) and (c) The mean gBOLD–CSF cross-correlation function (n = 159 participants) was characterized by a positive peak around the −4-s lag (B = 0.34, p < 0.001; permutation test with n = 10,000), whereas there was a negative peak at the lag of +4 s (B = −0.28, p < 0.001; permutation test with n = 10,000) (orange dashed line). The gray dashed line and the shaded region mark 95% confidence intervals for the mean correlation computed on shuffled signals. Overall, both cross-correlation functions showed patterns consistent with the data confirming systematic coupling between the global brain signal and CSF flow.

Association between BOLD–CSF coupling and PD

A significant correlation was found between age and gBOLD–CSF coupling (B = 0.004, p = 0.009, Figure 3(b)) Increased age has been associated with weaker gBOLD–CSF coupling strength.²⁶ For all following analyses, we adjusted the coupling strength for the effects of age and sex. The gBOLD–CSF coupling exhibited a significant difference, whereby it was lower in the NC group than in the drug-naïve PD groups (p = 0.006, Figure 3(a)). There was a correlation between UPDRS-II and gBOLD–CSF coupling signal (B = −0.006, p = 0.048, Figure 3(d)). however, no significant correlation was found between gBOLD–CSF coupling and UPDRS-III (B = 1.93e-5, p = 0.984, Figure 3(e)) or neurological assessment score (UPDRS-I; B = 0.001, p = 0.945, Figure 3(c)).

Figure 3.

Association of gBOLD–CSF coupling to age, disease condition, and UPDRS scores. (a) Compared with NC, mean gBOLD–CSF coupling strength is significantly weaker (less negative) in PD (adjusted for age and sex; p = 0.006, NC = 77, PD = 82). (b) Age positively correlates with gBOLD–CSF coupling (B = 0.004, p = 0.009, N = 159) and (c, d, e) gBOLD-CSF coupling correlated with various clinical measurements. No significant correlation was found between the gBOLD-CSF coupling and any of the MDS-UPDRS motor (UPDRS-II, B = −0.006, p = 0.048, Bonferroni-corrected p = 0.144, d; UPDRS-III, B = 1.93e-5, p = 0.984, e) or non-motor (UPDRS-I, B = 0.001, p = 0.945, c) subscores of the baseline value.

Association of gBOLD–CSF coupling strength with longitudinal clinical changes

Consistent with comparisons across both drug-naïve PD patients at baseline and follow-up groups, BOLD–CSF coupling significantly correlated with the UPDRS-III score longitudinal changes (B = 36.27, p = 0.008, Bonferroni-corrected p = 0.024, Figure 4(a)). Stronger, that is, more negative gBOLD–CSF coupling at baseline was significantly associated with a higher change rate in gBOLD–CSF coupling, representing a more rapid deterioration of gBOLD–CSF coupling (B = −0.58, p < 0.001, Figure 4(b)). There was no significant correlation found between the gBOLD–CSF coupling at baseline and other UPDRS subscores change rate (UPDRS-I, B = −2.53, p = 0.143; UPDRS-II, B = 12.18, p = 0.039; Bonferroni-corrected p = 0.117; Figure 4(c) and (d)).

Figure 4.

The correlation of gBOLD–CSF coupling at baseline and the rate of clinical changes. (a) Consistent with comparisons across both drug-naïve PD at baseline and follow-up groups. The BOLD–CSF coupling is significantly correlated with the UPDRS-III score changes rate (p = 0.008, Bonferroni-corrected p = 0.024, N = 21). (b) Baseline BOLD–CSF coupling signal in drug-naïve PD patients negatively correlated with gBOLD–CSF (adjusted for age and gender) change rate (B = −0.58, p < 0.001, N = 34) and (c, d) There was no significant correlation found between the gBOLD–CSF coupling at baseline and other MDS-UPDRS subscores change over time (UPDRS-II, B = 12.18, p = 0.039, Bonferroni-corrected p = 0.117, d; UPDRS-I, B = −2.53, p = 0.143, N = 21, c).

gBOLD–CSF coupling and sleep disturbances in PD

The gBOLD–CSF coupling metric was also compared between both the PD and NC groups. In baseline PD, less negative gBOLD–CSF coupling correlated with lower PDSS scores across PDSS subgroups (p = 0.047, Figure 5(a)). We did not find the association of gBOLD–CSF coupling and PDSS scores in PDSS-low group (B = 0.001, p = 0.350, Figure 5(b)) or PDSS-high group (B = 3.04e-4, p = 0.945, Figure 5(c)). No significant difference was found in gBOLD–CSF coupling strength between PD-nRBD and PD-pRBD group (p = 0.291). We found that gBOLD–CSF coupling vs RBDQ-HK associations remained marginally significant in PD-pRBD group (B = 0.004, p = 0.067, N = 25, Figure 5(f)). No association between gBOLD–CSF coupling and RBDQ-HK scores in the PD-nRBD group was found (B = 0.002, p = 0.733, N = 27, Figure 5(e)).

Figure 5.

Associations of gBOLD–CSF coupling and sleep condition in PD patients. (a) Weaker gBOLD–CSF coupling strength was observed in PD with lower PDSS score (more severe sleep disturbance) group across PD with higher PDSS score groups (p = 0.047); (b, c) No association of gBOLD–CSF coupling and PDSS scores in PDSS-low group (B = 0.001, p = 0.350, N = 32, d) or PDSS-high group was found (B = 3.04e-4, p = 0.945, N = 30, c); (d) No significant difference was found in gBOLD–CSF coupling strength between PD-nRBD and PD-pRBD group (p = 0.291) and (e, f) The PD-specific gBOLD–CSF coupling vs RBDQ-HK associations remained significant in PD-pRBD group (B = 0.004, p = 0.067, N = 25, f). No association between gBOLD–CSF coupling and RBDQ-HK scores in the PD-nRBD group was found (B = 0.002, p = 0.733, N = 27, e).

Discussion

To the best of our knowledge, the current study is the first to find that the strength of gBOLD–CSF coupling was significantly lower in drug-naïve PD patients than in healthy individuals. Moreover, gBOLD–CSF decoupling significantly correlated with motor symptom deterioration and sleep disturbance. These results suggest that gBOLD–CSF coupling is a potential non-invasive marker of glymphatic system dysfunction in PD pathogenesis.

gBOLD–CSF coupling was applied as an index to reflect glymphatic function. Intracranial pressure, skull volume, whole CSF volume, cerebral blood flow, and interstitial fluid were all relatively stable.³⁹ An influx of CSF led to a subsequent decrease in cerebral blood flow. According to previous studies, oscillations in gBOLD signal in the 0.01–0.1 Hz range is associated with slow vasomotor pulsation, which is thought to drive glymphatic CSF flow.^40–42 Taken together, low-frequency rs-fMRI signals that indicate gBOLD–CSF coupling likely result from highly coordinated neural and physiological processes that are closely linked to glymphatic clearance.²⁶ Thus, the cross-correlation in gBOLD–CSF signals can be applied to quantify the change in coordination activity between global BOLD signals and CSF inflow signals.

We found that the gBOLD–CSF coupling strength significantly decreased in the drug-naïve PD patients. In animal and autopsy studies, glymphatic dysfunction was also related to pathophysiological mechanisms of neurodegeneration, even before disease onset.^18,43 Previous studies have shown that the glymphatic system is involved in the clearance of α-syn, which is central to the pathogenesis of PD.^13,44–46 For example, the extracellular spreading of α-syn has been proposed to contribute to PD pathology, inducing microenvironmental responses.⁴⁷ In particular, blocking the deep cervical lymph nodes downstream of the glymphatic system aggravated PD pathology in A53T mice.¹⁸ Another study revealed that the emergence of α-syn pathology was followed by delayed meningeal lymphatic drainage in an α-syn preformed fibril-injected PD mouse model.¹⁶ These studies suggested that the glymphatic system is involved in PD pathology. Previous studies have reported glymphatic system deterioration in the early stages of AD and PD.^13,14 The strength of gBOLD–CSF coupling decreased gradually from the healthy controls to AD patients, indicating that the burden of gBOLD–CSF decoupling may increase as the disease progresses from the prodromal phase.²⁵ Given that most PD patients present with significant loss of DA neurons in the SN, it is important to explore the pathophysiological mechanism in the early stages of the disease. Taken together, our results suggest that the gBOLD–CSF coupling may be a sensitive index metric to evaluate glymphatic dysfunction before PD treatment. Further research is required to determine the value of gBOLD–CSF coupling in the identification of prodromal PD.

We reported that baseline gBOLD–CSF coupling positively correlated to changes in UDPRS-III score, while no correlation with baseline UPDRS-III scores was found. This suggests that glymphatic dysfunction may be linked to rapid progression in motor symptoms, regardless of early appearance. UPDRS score is related to dopamine transporter binding.⁴⁸ DA neurons are highly sensitive to changes in the micro-environment, such as iron,⁴⁹ neurotoxicity,⁵⁰ and inflammation.⁵¹ The loss of DA neurons may contribute to the manifestation of motor symptoms. Additionally, we noticed that in the PD follow-up cohort, mean UPDRS-III scores were lower than baseline UPDRS-III scores. Levodopa treatment can improve motor disability even in the OFF state and is invariably less severe than at baseline.⁵² Enhanced glymphatic activity also contributes to the efficiency of drug delivery to the central nervous system.⁵³ We speculated that there may be a time cumulative effect whereby long-term gBOLD–CSF decoupling may lead to inefficient drug delivery or metabolic waste clearance eventually resulting in motor symptom deterioration. Our study found that baseline gBOLD–CSF coupling strength was negatively associated with its rate of change over time. Perhaps, this deterioration might be attributed to compensatory mechanisms, such as altered white matter diffusivity and PVS number and pattern.^54,55

Over 90% of PD patients report various sleep problems, including insomnia, excessive daytime sleepiness, RBD, and circadian abnormalities.⁵⁶ As the disease progresses, disturbances in sleep become increasingly frequent and severe.^57–60 Sleep-related symptoms have a significant impact on PD-related disabilities, compared with motor symptoms.⁶¹ The glymphatic system is thought to exhibit low function during wakefulness and is switched on during slow-wave sleep,^62–64 regulated by circadian rhythms.^{11,12,65–67} Here, we focused on the sleep problems in our PD patients and used PDSS to generally evaluate the 15 most common sleep symptoms in PD.³² We found that gBOLD–CSF coupling strength was weaker in the subgroup with lower PDSS scores,^16,17 indicating that PD patients with better sleep quality achieve more efficient glymphatic transport than those with poorer sleep quality. Furthermore, RBDQ-HK scores were marginally significantly correlated with gBOLD–CSF coupling strength in PD-pRBD patients, but not in PD patients without RBD. Currently, RBD is recognized as an α-synucleinopathy and as a prodromal symptom of PD.^68,69 Consistent with previous studies, PD patients with or without RBD exhibit contradictory clinical manifestations and α-syn pathologies.^70,71 Perhaps, this difference is because PD-RBD represents a more advanced form of the disease compared with PD without RBD.⁷² Taken together, gBOLD–CSF coupling may be a potential tool to assess PD-related disability and study α-syn pathologies and further research is necessary to assess the correlation between sleep-related symptoms in PD and gBOLD–CSF coupling.

This study has a few limitations. First, we used gBOLD–CSF coupling to reflect whole-brain glymphatic function, in the absence of a more accurate method to evaluate glymphatic system function of the effaced region.^73–75 Second, we mostly relied on questionnaires to evaluate clinical metrics. Instead, objective inspection methods, such as polysomnography and dopamine transporter-positron emission tomography, could have better reflected the actual situation in the brain. Third, we recruited our volunteers from a developing region; therefore, a portion of patients may have shown more severe symptoms before drug administration, due to their lack of medical resource accessibility. Lastly, if more patients had been available for follow-up data, the longitudinal study of clinical metrics and the gBOLD–CSF coupling would have higher accuracy.

In conclusion, the current study demonstrated a severe gBOLD–CSF decoupling in patients with drug-naïve PD and may reflect longitudinal motor impairment, thereby providing a potential marker of glymphatic dysfunction in PD. Moreover, gBOLD–CSF decoupling significantly correlated with motor symptom deterioration and sleep disturbance.

Supplemental Material

sj-pdf-1-jcb-10.1177_0271678X231164337 - Supplemental material for Reduced coupling of global brain function and cerebrospinal fluid dynamics in Parkinson’s disease

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X231164337 for Reduced coupling of global brain function and cerebrospinal fluid dynamics in Parkinson’s disease by Zhiyun Wang, Zhe Song, Cheng Zhou, Yi Fang, Luyan Gu, Wenyi Yang, Ting Gao, Xiaoli Si, Yi Liu, Ying Chen, Xiaojun Guan, Tao Guo, Jingjing Wu, Xueqing Bai, Minming Zhang, Baorong Zhang and Jiali Pu in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (grant number: NSFC82001353); and the Key Research and Development Program of Zhejiang Province (grant number: 2020C03020).

Acknowledgements

We would like to thank all the volunteers who participated in our study, as well as their caregivers.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

Z.W and Z.S: study conceptualization, data analysis, and writing the original draft. C.Z and X.G: image evaluation and validation. L.G, W.Y, and T.G: the acquisition of clinical metrics. Y.F and X.S: formal analysis and methods consultant. X.B, T.G, and J.W: acquisition of image data. Y.C and Y.L: program supervision, manuscript reviewing. B.Z: project administration, funding acquisition; M.Z: program supervision, review of the manuscript; J.P: program supervision, review, and editing of the final version of the manuscript. All authors have read and approved the final manuscript.

ORCID iD

Minming Zhang

References

Jellinger

KA.

Neuropathology of sporadic Parkinson’s disease: evaluation and changes of concepts. Mov Disord 2012; 27: 8–30.

Chung

Lee

SB.

Mechanisms of protein toxicity in neurodegenerative diseases. Cell Mol Life Sci 2018; 75: 3159–3180.

Balestrino

Schapira

AHV.

Parkinson disease. Eur J Neurol 2020; 27: 27–42.

Goedert

Masuda-Suzukake

Falcon

Like prions: the propagation of aggregated tau and α-synuclein in neurodegeneration. Brain 2017; 140: 266–278.

Sundaram

Hughes

Peterson

, et al. Establishing a framework for neuropathological correlates and glymphatic system functioning in Parkinson’s disease. Neurosci Biobehav Rev 2019; 103: 305–315.

Valdinocci

Radford

RAW

Goulding

, et al. Extracellular Interactions of Alpha-Synuclein in Multiple System Atrophy. Int J Mol Sci 2018; 19: 4129.

Kress

Iliff

Xia

, et al. Impairment of paravascular clearance pathways in the aging brain. Ann Neurol 2014; 76: 845–861.

Iliff

Wang

Liao

, et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid β. Sci Transl Med 2012; 4: 147ra111.

Louveau

Smirnov

Keyes

, et al. Corrigendum : structural and nervous system lymphatic vessels. Nature 2016; 533: 278.

10.

Xie

Kang

, et al. Sleep drives metabolite clearance from the adult brain. Science 2013; 342: 373–377.

11.

Hablitz

Plá

Giannetto

, et al. Circadian control of brain glymphatic and lymphatic fluid flow. Nat Commun 2020; 11: 4411.

12.

Shokri-Kojori

Wang

Wiers

, et al. β-Amyloid accumulation in the human brain after one night of sleep deprivation. Proc Natl Acad Sci U S A 2018; 115: 4483–4488.

13.

Xiao

Chen

, et al. Deletion of aquaporin-4 in APP/PS1 mice exacerbates brain Aβ accumulation and memory deficits. Mol Neurodegener 2015; 10: 58.

14.

Hoshi

Yamamoto

Shimizu

, et al. Characteristics of aquaporin expression surrounding senile plaques and cerebral amyloid angiopathy in Alzheimer disease. J Neuropathol Exp Neurol 2012; 71: 750–759.

15.

Peng

Achariyar

, et al. Suppression of glymphatic fluid transport in a mouse model of Alzheimer’s disease. Neurobiol Dis 2016; 93: 215–225.

16.

Ding

Wang

Xia

, et al. Impaired meningeal lymphatic drainage in patients with idiopathic Parkinson’s disease. Nat Med 2021; 27: 411–418.

17.

Donahue

Murdos

Jakowec

, et al. Global and Regional Changes in Perivascular Space in Idiopathic and Familial Parkinson’s Disease. Mov Disord 2021; 36: 1126–1136.

18.

Zou

Feng

, et al. Blocking meningeal lymphatic drainage aggravates Parkinson’s disease-like pathology in mice overexpressing mutated α-synuclein. Transl Neurodegener 2019; 8: 7.

19.

Carlstrom

Eltanahy

Perry

, et al. A clinical primer for the glymphatic system. Brain 2022; 145: 843–857.

20.

Absinta

S-K

Nair

, et al. Human and nonhuman primate meninges harbor lymphatic vessels that can be visualized noninvasively by MRI. Elife 2017; 6: 1–15.

21.

Taoka

Masutani

Kawai

, et al. Evaluation of glymphatic system activity with the diffusion MR technique : diffusion tensor image analysis along the perivascular space (DTI ‑ ALPS) in Alzheimer’s disease cases. Jpn J Radiol 2017; 35: 172–178.

22.

Kiviniemi

Wang

Korhonen

Ultra-fast magnetic resonance encephalography of physiological brain activity – Glymphatic pulsation mechanisms?. J Cereb Blood Flow Metab 2016; 36: 1033–1045.

23.

Liu

de Zwart

Schölvinck

, et al. Subcortical evidence for a contribution of arousal to fMRI studies of brain activity. Nat Commun 2018; 9: 395.

24.

Fultz

Bonmassar

Setsompop

, et al. Coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep. Science 2019; 366: 628–631.

25.

Han

Chen

Belkin-Rosen

, et al. Reduced coupling between cerebrospinal fluid flow and global brain activity is linked to Alzheimer disease–related pathology. PLOS Biol 2021; 19: e3001233.

26.

Han

Brown

Zhu

, et al. Decoupling of global brain activity and cerebrospinal fluid flow in Parkinson’s disease cognitive decline. Mov Disord 2021; 36: 2066–2076.

27.

Lysen

Darweesh

SKL

Kamran Ikram

, et al. Sleep and risk of parkinsonism and Parkinson’s disease: a population-based study. Brain 2019; 142: 2013–2022.

28.

Schreiner

Imbach

Werth

, et al. Slow‐wave sleep and motor progression in Parkinson disease. Ann Neurol 2019; 85: 765–770.

29.

Goetz

Tilley

Shaftman

, et al. Movement disorder society-sponsored revision of the unified Parkinson’s disease rating scale (MDS-UPDRS): scale presentation and clinimetric testing results. Mov Disord 2008; 23: 2129–2170.

30.

Goetz

Poewe

Rascol

, et al. Movement disorder society task force report on the Hoehn and Yahr staging scale: status and recommendations the movement disorder society task force on rating scales for Parkinson’s disease. Mov Disord 2004; 19: 1020–1028.

31.

Wing

Lam

, et al. Validation of a new REM sleep behavior disorder questionnaire (RBDQ-HK). Sleep Med 2010; 11: 43–48.

32.

Chaudhuri

Pal

DiMarco

, et al. The Parkinson’s disease sleep scale: a new instrument for assessing sleep and nocturnal disability in Parkinson’s disease. J Neurol Neurosurg Psychiatry 2002; 73: 629–635.

33.

Shen

Xiong

, et al. Validation study of REM sleep behavior disorder questionnaire – Hong Kong (RBDQ-HK) in east China. Sleep Med 2014; 15: 952–958.

34.

Tomlinson

Stowe

Patel

, et al. Systematic review of levodopa dose equivalency reporting in Parkinson’s disease. Mov Disord 2010; 25: 2649–2653.

35.

Yan

C-G

Wang

X-D

Zuo

X-N

, et al. DPABI: data processing & analysis for (resting-state) brain imaging. Neuroinformatics 2016; 14: 339–351.

36.

Friston

Holmes

Worsley

, et al. Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp 1994; 2: 189–210.

37.

Desikan

Ségonne

Fischl

, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 2006; 31: 968–980.

38.

Guo

Zhou

, et al. Longitudinal macro/microstructural alterations of different callosal subsections in Parkinson’s disease using connectivity-based parcellation. 2020; 12: 1–10.

39.

Nakada

Kwee

IL.

Fluid dynamics inside the brain barrier: current concept of interstitial flow, glymphatic flow, and cerebrospinal fluid circulation in the brain. Neuroscientist 2019; 25: 155–166.

40.

Iliff

Wang

Zeppenfeld

, et al. Cerebral arterial pulsation drives paravascular CSF-interstitial fluid exchange in the murine brain. J Neurosci 2013; 33: 18190–18199.

41.

Kiviniemi

Wang

Korhonen

, et al. Ultra-fast magnetic resonance encephalography of physiological brain activity – glymphatic pulsation mechanisms? J Cereb Blood Flow Metab 2016; 36: 1033–1045.

42.

Mestre

Tithof

, et al. Flow of cerebrospinal fluid is driven by arterial pulsations and is reduced in hypertension. Nat Commun 2018; 9: 4878.

43.

Emmanouilidou

Elenis

Papasilekas

, et al. Assessment of α-synuclein secretion in mouse and human brain parenchyma. PLoS One 2011; 6: e22225.

44.

Ohene

Harrison

Nahavandi

, et al. Non-invasive MRI of brain clearance pathways using multiple echo time arterial spin labelling: an aquaporin-4 study. Neuroimage 2019; 188: 515–523.

45.

Lee

Choi

Park

E-J

, et al. Improvement of glymphatic–lymphatic drainage of beta-amyloid by focused ultrasound in Alzheimer’s disease model. Sci Rep 2020; 10: 16144.

46.

Venda

Cragg

Buchman

, et al. α-Synuclein and dopamine at the crossroads of Parkinson’s disease. Trends Neurosci 2010; 33: 559–568.

47.

Lee

Bae

Lee

Extracellular α‑synuclein – a novel and crucial factor in Lewy body diseases. Nat Rev Neurol 2014; 10: 92–98.

48.

Simuni

Siderowf

Lasch

, et al. Longitudinal change of clinical and biological measures in early Parkinson’s disease: Parkinson’s progression markers initiative cohort. Mov Disord 2018; 33: 771–782.

49.

Hare

Double

KL.

Iron and dopamine: a toxic couple. Brain 2016; 139: 1026–1035.

50.

Lee

Song

Huh

, et al. Neuroprotection against 6-OHDA toxicity in PC12 cells and mice through the Nrf2 pathway by a sesquiterpenoid from Tussilago farfara. Redox Biol 2018; 18: 6–15.

51.

Cheng

Liao

Dong

, et al. Microglial autophagy defect causes Parkinson disease-like symptoms by accelerating inflammasome activation in mice. Autophagy 2020; 16: 2193–2205.

52.

Cilia

Cereda

Akpalu

, et al. Natural history of motor symptoms in Parkinson’s disease and the long-duration response to levodopa. Brain 2020; 143: 2490–2501.

53.

Lilius

Blomqvist

Hauglund

, et al. Dexmedetomidine enhances glymphatic brain delivery of intrathecally administered drugs. J Control Release 2019; 304: 29–38.

54.

Chen

Yang

GZX

Getachew

, et al. Focused ultrasound-enhanced intranasal brain delivery of brain-derived neurotrophic factor. Sci Rep 2016; 6: 28599.

55.

Song

, et al. Different Perivascular Space Burdens in Idiopathic Rapid Eye Movement Sleep Behavior Disorder and Parkinson’s Disease. Front Aging Neurosci 2020; 12: 580853.

56.

De Cock

Dodet

Leu-Semenescu

, et al. Safety and efficacy of subcutaneous night-time only apomorphine infusion to treat insomnia in patients with Parkinson’s disease (APOMORPHEE): a multicentre, randomised, controlled, double-blind crossover study. Lancet Neurol 2022; 21: 428–437.

57.

Videnovic

Noble

Reid

, et al. Circadian melatonin rhythm and excessive daytime sleepiness in Parkinson disease. JAMA Neurol 2014; 71: 463–469.

58.

Zhang

Ren

Sanford

, et al. Sleep in Parkinson’s disease: a systematic review and meta-analysis of polysomnographic findings. Sleep Med Rev 2020; 51: 101281.

59.

Anderson

Saffari

, et al. Progression of sleep disturbances in Parkinson’s disease: a 5-year longitudinal study. J Neurol 2021; 268: 312–320.

60.

Diederich

Vaillant

Mancuso

, et al. Progressive sleep ‘destructuring’ in Parkinson’s disease. A polysomnographic study in 46 patients. Sleep Med 2005; 6: 313–318.

61.

Suzuki

Okuma

Uchiyama

, et al. Impact of sleep-related symptoms on clinical motor subtypes and disability in Parkinson’s disease: a multicentre cross-sectional study. J Neurol Neurosurg Psychiatry 2017; 88: 953–959.

62.

Hablitz

Vinitsky

Sun

, et al. Increased glymphatic influx is correlated with high EEG delta power and low heart rate in mice under anesthesia. Sci Adv 2019; 5: eaav5447.

63.

Benveniste

Lee

Ding

, et al. Anesthesia with Dexmedetomidine and Low-dose Isoflurane Increases Solute Transport via the Glymphatic Pathway in Rat Brain When Compared with High-dose Isoflurane. Anesthesiology 2017; 127: 976–988.

64.

Jessen

Munk

ASF

Lundgaard

, et al. The glymphatic system – a beginner’s guide HHS public access. Neurochem Res 2015; 40: 2583–2599.

65.

Y-ES

Finn

Sutphen

, et al. Obstructive sleep apnea decreases central nervous system–derived proteins in the cerebrospinal fluid. Ann Neurol 2016; 80: 154–159.

66.

Cai

Qiao

Kulkarni

, et al. Imaging the effect of the circadian light-dark cycle on the glymphatic system in awake rats. Proc Natl Acad Sci USA 2020; 117: 668–676.

67.

Wang

Liu

, et al. Associations of sleep characteristics with alpha-synuclein in cerebrospinal fluid in older adults. Ann Clin Transl Neurol 2020; 7: 2026–2034.

68.

Postuma

Berg

Advances in markers of prodromal Parkinson disease. Nat Rev Neurol 2016; 12: 622–634.

69.

Jin

Wang

, et al. Apolipoprotein E genotype contributes to motor progression in Parkinson’s disease. Mov Disord 2022; 37: 196–200.

70.

Högl

Stefani

Videnovic

Idiopathic REM sleep behaviour disorder and neurodegeneration – an update. Nat Rev Neurol 2018; 14: 40–55.

71.

Pagano

De Micco

Yousaf

, et al. REM behavior disorder predicts motor progression and cognitive decline in Parkinson disease. Neurology 2018; 91: e894–e905.

72.

Taguchi

Ikuno

Hondo

, et al. α-synuclein BAC transgenic mice exhibit RBD-like behaviour and hyposmia: a prodromal Parkinson’s disease model. Brain 2020; 143: 249–265.

73.

Xue

Liu

Zhang

, et al. Concomitant enlargement of perivascular spaces and decrease in glymphatic transport in an animal model of cerebral small vessel disease. Brain Res Bull 2020; 161: 78–83.

74.

Aribisala

Wiseman

Morris

, et al. Circulating inflammatory markers are associated with magnetic resonance imaging-visible perivascular spaces but not directly with white matter hyperintensities. Stroke 2014; 45: 605–607.

75.

Park

Shin

Chung

, et al. Magnetic Resonance Imaging-Visible Perivascular Spaces in Basal Ganglia Predict Cognitive Decline in Parkinson’s Disease. Mov Disord 2019; 34: 1672–1679.

76.

Gao

Liu

HL.

Inflow effects on functional MRI. Neuroimage 2012; 62: 1035–1039.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.33 MB