Analysis of AQP4 SNPs and Their Association With REM Sleep Behavior Disorder and Hallucinations in Parkinson’s Disease Authors: Cristina Agliardi, Franca Rosa Guerini, Milena Zanzottera, Elisabetta Bolognesi, Mario Meloni, Francesca Lea Saibene, Roberta Zangaglia, Andrea Sturchio, Carlo Casali, Cherubino Di Lorenzo, Jorge Navarro, Brigida Minafra, Mario Clerici Published: 2025 ABSTRACT AbstractThe glymphatic system is a glial-dependent network responsible for the clearance of waste products from the brain through cerebrospinal fluid (CSF) circulation. This process, which involves astrocytes and the water channel AQP4, facilitates the removal of harmful proteins like β-amyloid and tau, making it crucial in neurodegenerative diseases such as Alzheimer's (AD) and Parkinson's disease (PD). AQP4 dysfunction has been linked to the accumulation of these proteins and related pathologies. This study aimed to investigate the association between AQP4 gene polymorphisms (rs2075575, rs162009, and rs335929) and PD in an Italian cohort, consisting of 380 individuals with PD enrolled in a rehabilitation protocol and 461 healthy controls. The analysis found no significant correlation between the selected AQP4 single nucleotide polymorphisms (SNPs) and PD risk. However, the rs162009 AA genotype was associated with a lower risk of developing REM sleep behavior disorder (RBD) symptoms in PD patients, while rs2075575 was linked to hallucinations in these individuals. These findings suggest a potential role for AQP4 polymorphisms in sleep disturbances and psychotic symptoms in PD, but further research is needed to confirm these results and understand the complex interactions between the glymphatic system and PD pathophysiology.Keywords: Parkinson’s disease, AQP4, Polymorphism, Single nucleotide polymorphism (SNP), Sleep, REM sleep behavior disorder (RBD) Introduction The glymphatic system is defined as the glial-dependent flow of cerebrospinal fluid (CSF) through the brain parenchyma, which enables the osmotic clearance of waste products from the brain. This is achieved by creating a convective flow of CSF and brain interstitial fluid (ISF) [1]. Specifically, CSF enters the deep brain via periarterial spaces and exits along the perivascular spaces of the draining veins, creating a periarterial CSF inflow and a perivenous or perineuronal ISF outflow pathway [2]. The glymphatic fluid eventually drains through the meningeal lymph vessels to the cervical lymph nodes. Astrocytes, key components of the blood–brain barrier (BBB), play a crucial role in this system. On their end-feet, which surround blood vessels, astrocytes express the water channel AQP4, which is abundant in the human brain, spinal cord, and optic nerve. AQP4, like other aquaporins (AQPs), assembles into tetramers to form a water-selective channel [3]. AQP4 has several critical functions, including the regulation of extracellular volume, modulation of neuro-excitation, synaptic plasticity, BBB permeability, and, importantly, waste disposal within the functions of the glymphatic system [4]. This last function is particularly significant in facilitating the clearance of waste proteins such as β-amyloid, tau, and α-synuclein from the brain parenchyma. Consequently, much research has focused on the relationship between glymphatic system dysfunction and the pathogenesis of neurodegenerative diseases, such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), which are characterized by the accumulation of amyloidogenic proteins. A post-mortem study found that AQP4 expression was associated with aging in both healthy individuals and AD patients. Moreover, impaired AQP4 perivascular localization was linked to AD and an increased β-amyloid burden, suggesting a role for AQP4 in AD development [5]. Regarding PD, a large body of experimental evidence points to a potential role for AQP4 in the disease’s pathogenesis and progression. Clinical studies have shown that AQP4 mRNA levels are reduced in the blood samples of people with PD (pwPD) [6], and there is a negative correlation between α-synuclein and AQP4 levels in the brain neocortex of pwPD [7]. In a mouse model of PD, where recombinant human α-synuclein was injected into the substantia nigra and AQP4 was knocked out, α-synuclein clearance was reduced, dopaminergic neurons were lost, tyrosine hydroxylase (TH)-positive neurons were depleted, and PD-like symptoms were accelerated [8]. Several studies in recent years have explored the potential involvement of AQP4 gene single-nucleotide polymorphisms (SNPs) in PD risk and progression. In particular, Sun et al. reported that rs2075575 GA genotype was associated with an increased risk of PD while the G allele was linked to REM sleep behavior disorder (RBD) in pwPD. Notably, the rs335929 A allele was also observed to correlate with memory impairment in the same individuals [9]. Two other studies indicated that the rs162009 polymorphism influences spontaneous brain activity and cognitive decline in pwPD [10, 11]. All these results stem from studies conducted in Chinese population. Given the lack of research on AQP4 gene polymorphisms in the Italian population, this study aimed to genotype Italian pwPD and age- and sex-matched healthy controls for AQP4 SNPs rs2075575, rs162009, and rs335929 to evaluate their possible role in PD pathogenesis. Material and Methods Study Population A total of 841 Caucasian Italians, including 380 individuals with idiopathic Parkinson’s disease and 461 age- and sex-matched healthy controls (HC) were enrolled in the study. Individuals of Sardinian ancestry were excluded due to their known genetic homogeneity. PwPD were diagnosed by expert neurologists based on the Movement Disorder Society (MDS) clinical diagnostic criteria for PD [12]. Exclusion criteria included vascular parkinsonism, atypical parkinsonism, familial forms of PD, drug-induced parkinsonism, and brain tumors. All healthy controls underwent neurologic examination to rule out any movement disorders, particularly PD. All pwPD were receiving antiparkinsonian treatment, while control group participants exhibited no signs or symptoms of neurological disease. Subjects were recruited from three different institutions: the Parkinson’s Disease and Movement Disorders Unit at the IRCCS Mondino Foundation, Pavia, Italy; Ospedale di Circolo and Fondazione Macchi, Varese, Italy; and the Neurology Unit of IRCCS Fondazione Don Carlo Gnocchi, Milan, Italy, where PD patients were enrolled in a rehabilitation protocol. Ethical approval for the study was obtained from local Ethics Committee (02_21/06/2005) and all participants provided informed consent. The recorded demographic and clinical variables for patients included age, sex, age at onset and levodopa daily dose (LEDD). Disease stage and motor severity were assessed using the modified Hoehn and Yahr (mH&Y) scale and the MDS-Unified Parkinson’s Disease Rating Scale Part III (MDS-UPDRS-III) [13], respectively. RBD was clinically diagnosed using the RBD Screening Questionnaire (RBDSQ) [14], with a cut-off score of ≥ 6 for PD patients used to define RBD [15]. Additionally, a sleep-focused interview was conducted by a neurologist specialized in sleep disorders to rule out other sleep-related issues. The presence of psychiatric symptoms, including the presence of hallucinations was reported as being or not being present as a result of in-depth clinical interviews with patients and/or their partners and families. Cognitive performance was assessed using the Mini-Mental State Examination (MMSE) [16], with scores corrected for age and education. Sample Collection and DNA Extraction Whole blood samples were collected in EDTA-containing Vacutainer tubes (Becton Dickinson Co., Rutherford, NJ), and genomic DNA was extracted from peripheral blood mononuclear cells (PBMCs) using the standard phenol/chloroform method. The DNA concentration for each sample was determined by measuring optical density at 260 nm using a spectrophotometer (SmartSpec Plus, Bio-Rad, Irvine, CA, USA). DNA samples were stored at −20 °C until further use. SNP Selection, Description, and Genotyping Three AQP4 gene single nucleotide polymorphisms (SNPs) were selected for analysis based on previous studies reporting associations with PD. These SNPs are rs2075575, rs162009, and rs335929. Rs2075575 is located in intron 1 of the AQP4 gene, rs162009 is in the promoter region of the M23 variant, and rs335929 is found in the 3' UTR region of the gene. Genotyping of these SNPs was performed using allelic discrimination real-time PCR with pre-designed TaqMan assays (C__15863033_20 for rs2075575, C___1303573_30 for rs162009, and C___1303566_10 for rs335929) (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The PCR protocol included a hot start at 95 °C for 10 min, followed by 40 cycles at 94 °C for 15 s and 60 °C for 1 min. Fluorescence detection was performed at 60 °C. Reactions were conducted in 10 μL volumes using TaqMan Genotyping Master Mix on 96-well plates with a CFX96 instrument (Bio-Rad, Hercules, CA, USA). Control samples representing all possible genotypes, as well as a negative control, were included in each run. Statistical Analysis The statistical power of the sample size was calculated using the University of Michigan's online calculator, derived from the CaTS power calculator for two-stage association studies [17] (https://csg.sph.umich.edu/abecasis/cats/gas_power_calculator/). Chi-square or Fisher’s exact tests were employed to assess deviations in SNP genotype distribution from Hardy–Weinberg equilibrium. Chi-square statistics were applied to 2xN tables where appropriate to compare case–control differences in SNP genotype, allele distributions and the associations of SNPs genotypes with RBD status and the presence of hallucinations. Associations were evaluated using odds ratios (OR) with 95% confidence intervals (CI). Two-sided p-values were considered significant when < 0.05, Bonferroni correction for multiple comparisons was applied when necessary (pc). Haplotype analyses were performed using the SNPStat online software available at https://www.snpstats.net/start.htm. One-way ANOVA test was used to analyse the possible association between SNPs genotypes and quantitative variables such as age at onset, clinical scales and MMSE scores. Binary logistic regression was conducted with the presence of hallucinations as the dependent variable and the AQP4 rs2075575 genotypes and LEDD as covariates. All statistical analyses were performed using SPSS software (v.29.0.1.0, IBM, USA) and Jamovi (https://www.jamovi.org/). Results The AQP4 rs2075575, rs162009, and rs335929 SNPs were selected both because they have been already reported to be involved in PD pathology and because they are not in linkage disequilibrium and cover a large part of the gene. The SNPs were analyzed in 380 Italian pwPD and 461 age- and sex-matched healthy controls. Demographic and clinical data for the study population are shown in Table 1 (Table 1). Table 1Study population demographic and clinical dataPopulation characteristicspwPDHCpN380461males/females, (%/%)164/216, (43.2/56.8)196/265, (42.5/57.5)0.43Age at enrollment (years), mean ± SD69.79 ± 8.6869.95 ± 11.770.83Age at onset (years), mean ± SD61.73 ± 10.07n.amH&Y, mean (SD)2.21 ± 0.73n.aMDS-UPDRS-Part III, mean (SD)20.10 ± 11.10n.aLEDD, mean (SD)574,31 ± 338,41n.aRBD (yes/no)90/212n.avHallucinations (yes/no)44/222n.avMMSE, mean (SD)26.79 ± 3.2528.23 ± 1.63 < 0.001pwPD = people with PD; HC = healthy control; n.a. = not applicable; n.av. = not available; SD = standard deviation; mH&Y = modified Hoehn and Yahr; MDS-UPDRS-Part III = Unified Parkinson’s disease rating scale; LEDD = Levodopa equivalent daily dose; RBD = REM sleep behavior disorder; MMSE = Mini-Mental State Examination Study population demographic and clinical data pwPD = people with PD; HC = healthy control; n.a. = not applicable; n.av. = not available; SD = standard deviation; mH&Y = modified Hoehn and Yahr; MDS-UPDRS-Part III = Unified Parkinson’s disease rating scale; LEDD = Levodopa equivalent daily dose; RBD = REM sleep behavior disorder; MMSE = Mini-Mental State Examination The Mini-Mental State Examination (MMSE) score was lower in pwPD compared to healthy controls (p < 0.001). The sample size was sufficient to detect an odds ratio (OR) of 1.5 with a 90% power, assuming an additive model of inheritance, a minor allele frequency of 0.192 (for rs335929, based on data from the 1000 Genomes project for the Toscani in Italy population), and a significance level of 0.01. No deviation from Hardy–Weinberg equilibrium was observed in either group (all p-values > 0.05). Genotype and allele distributions for the AQP4 SNPs showed no statistically significant differences between pwPD and healthy controls, even when stratified by sex (Table 2). Table 2The AQP4 rs2075575, rs162009 and rs335929 SNPs genotypes and alleles distributions in pwPD and healthy controls (HC)rs2075575pwPD (M + F)HC (M + F)p valueOR (95% CI)pwPD (M)HC (M)p valueOR (95% CI)pwPD (F)HC (F)p valueOR (95% CI)GenotypeN(%)N(%)N(%)N(%)N(%)N(%)AA100(26.3)116(25.2)0.701.06 (0.78–1.45)35(21.3)51(26.0)0.300.77 (0.47–1.26)65(30.1)65(24.5)0.171.32 (0.88–1.99)GA196(51.6)226(49.0)0.461.11 (0.84–1.45)88(53.7)88(44.9)0.091.42 (0.94–2.16)108(50.0)138(52.1)0.650.92 (0.64–1.32)GG84(22.1)119(25.8)0.210.82 (0.59–1.12)41(25.0)57(29.1)0.390.81 (0.51–1.30)43(19.9)62(23.4)0.360.81 (0.52–1.26)Total3804610.461641960.252162650.35AlleleA396(52.1)458(49.7)158(48.2)190(48.5)238(55.1)268(50.6)G364(47.9)464(50.3)170(51.8)202(51.5)194(44.9)262(49.4)Total7609220.341.10 (0.91–1.34)3283920.940.99 (0.74–1.33)4325300.161.20 (0.93–1.55)rs162009pwPD(M + F)HC(M + F)p valueOR (95% CI)pwPD(M)HC(M)p valueOR (95% CI)pwPD(F)HC(F)p valueOR (95% CI)GenotypeN(%)N(%)N(%)N(%)N(%)N(%)AA37(9.7)47(10.2)0.820.95 (0.60–1.50)16(9.8)21(10.7)0.770.90 (0.45–1.80)21(9.7)26(9.8)0.980.99 (0.53–1.82)GA170(44.7)203(44.0)0.841.03 (0.78–1.35)73(44.5)90(45.9)0.790.95 (0.62–1.44)97(44.9)113(42.6)0.621.10(0.76–1.58)GG173(45.5)211(45.8)0.940.99 (0.75–1.30)75(45.7)85(43.4)0.651.10 (0.72–1.67)98(45.4)126(47.5)0.640.92 (0.64–1.32)Total3804610.971641960.892162650.88AlleleA244(32.1)297(32.2)105(32.0)132(33.7)139(32.2)165(31.1)G516(67.9)625(67.8)223(68.0)260(66.3)293(67.8)365(68.9)Total7609220.960.99 (0.81–1.22)3283920.640.93 (0.68–1.27)4325300.731.05 (0.80–1.38)rs335929pwPD(M + F)HC(M + F)p valueOR (95% CI)pwPD(M)HC(M)p valueOR (95% CI)pwPD(F)HC(F)p valueOR (95% CI)GenotypeN(%)N(%)N(%)N(%)N(%)N(%)AA237(62.4)284(61.6)0.821.03 (0.78–1.37)100(61.0)121(61.7)0.880.97 (0.63–1.49)137(63.4)163(61.5)0.661.09 (0.75–1.58)CA125(32.9)158(34.3)0.680.94 (0.70–1.25)60(36.6)66(33.7)0.571.14 (0.73–1.76)65(30.1)92(34.7)0.280.81 (0.55–1.19)CC18(4.7)19(4.1)0.671.16 (0.59–2.26)4(2.4)9(4.6)0.290.52 (0.14–1.70)14(6.5)10(3.8)0.181.77 (0.79–4.20)Total3804610.851641960.52162650.27AlleleA599(78.8)726(78.7)260(79.3)308(78.6)339(78.5)418(78.9)C161(21.2)196(21.3)68(20.7)84(21.4)93(21.5)112(21.1)Total7609220.971.00 (0.79–1.27)3283920.821.04 (0.73–1.50)4325300.880.98 (0.72–1.33)pwPD = people with PD; HC = healthy control; OR = Odd Ratio; CI = confidence interval; N = number; M = male; F = female The AQP4 rs2075575, rs162009 and rs335929 SNPs genotypes and alleles distributions in pwPD and healthy controls (HC) pwPD (M + F) HC (M + F) pwPD (M) HC (M) pwPD (F) HC (F) 1.10 (0.76–1.58) pwPD (M + F) HC (M + F) pwPD (M) HC (M) pwPD (F) HC (F) pwPD = people with PD; HC = healthy control; OR = Odd Ratio; CI = confidence interval; N = number; M = male; F = female Haplotype analysis also revealed no association between the co-segregation of these SNPs and PD risk (Table 3). Table 3Haplotype analysis of the AQP4 rs2075575, rs162009 and rs335929 SNPsHaplotypePwPD (n, %)HC (n, %)Chi2pOR (95%CI)AAC47 (6.1)49 (5.3)0.590.441.17 (0.78–1.77)AGA340 (44.7)394 (42.7)0.680.411.08 (0.89–1.32)GAA91 (11.9)106 (11.4)0.090.761.05 (0.78–1.41)GAC101 (13.2)133 (14.4)0.450.500.91 (0.69–1.20)GGA163 (21.4)217 (23.5)1.040.310.89 (0.70–1.12)SNP = single nucleotide polymorphism; pwPD = people with PD; HC = healthy control; p = p value; OR = odds ratio; CI = confidence interval Haplotype analysis of the AQP4 rs2075575, rs162009 and rs335929 SNPs SNP = single nucleotide polymorphism; pwPD = people with PD; HC = healthy control; p = p value; OR = odds ratio; CI = confidence interval Linkage Disequilibrium (LD) test plot (D and r2) are reported in Fig. 1 Linkage disequilubrium (LD) plots for rs2075575, rs162009 and rs335929 SNPs The potential influence of the AQP4 SNPs on clinical parameters was further investigated. The MMSE score was not influenced by any of the three SNPs, whether considering PD patients and controls together or pwPD alone (data not shown). Among the 302 pwPD tested for REM behavior disorder (RBD), 90 were found to have RBD. The AQP4 rs162009 polymorphism was associated with RBD (p = 0.046), with the AA genotype being “protective” against RBD. This genotype was present in 12.3% of RBD-negative patients and only 3.3% of RBD-positive patients (p = 0.01, pc = 0.02, OR = 0.25, 95% CI = 0.06–0.77) (Table 4). Logistic regression was performed with RBD positivity as the dependent variable, and rs162009 genotypes, age, sex, and LEDD as covariates. The overall model fit was statistically significant (p = 0.03). The rs162009 AA genotype compared to the GA genotype was statistically significant (p = 0.02), but not when compared to the GG genotype (p = 0.06). None of the other covariates were found to be significant. Table 4The AQP4 SNPs rs162009 and rs2075575 genotypes and alleles distribution in pwPD with/without RBD and with/without hallucinationsrs162009pwPD-RBDpwPD-noRBDp valueOR (95% CI)GenotypeN(%)N(%)AA3(3.3)26(12.3)0.01, 0.02*0.25 (0.06–0.77)GA45(50.0)89(42.0)0.200.72 (0.44–1.91)GG42(46.7)97(45.8)0.881.04 (0.63–1.70)Total902120.046AlleleA51(28.3)141(33.3)G129(71.7)283(66.7)Total1804240.240.79 (0.54–1.16)rs2075575pwPD-hallucinationspwPD-no hallucinationsp valueOR (95% CI)GenotypeN(%)N(%)AA6(13.6)63(28.4)0.02, 0.04*0.40 (0.15–0.95)GA32(72.7)106(47.7)0.001, 0.002*2.91 (1.44–6.14)GG6(13.6)53(23.9)0.130.50 (0.18–1.21)Total44222 < 0.01AlleleA44(50.0)232(52.3)G44(50.0)212(47.7)Total884240.700.91 (0.58–1.45)pwPD = people with PD; RBD: REM sleep behavior disorder; OR = Odds ratio; CI = confidence interval; * p value with Bonferroni correction for 2 degrees of freedom The AQP4 SNPs rs162009 and rs2075575 genotypes and alleles distribution in pwPD with/without RBD and with/without hallucinations pwPD = people with PD; RBD: REM sleep behavior disorder; OR = Odds ratio; CI = confidence interval; * p value with Bonferroni correction for 2 degrees of freedom The AQP4 rs2075575 and rs335929 SNPs were not associated with RBD. Data on hallucinations were available for 266 pwPD. The AQP4 rs2075575 SNP was found to be associated with hallucinations (p < 0.01) The GA genotype was identified as a "risk" factor for hallucinations, being present in 47.7% of pwPD who did not experience hallucinations, compared to 72.7% of those who did (p = 0.001, pc = 0.002, OR = 2.91, 95% CI = 1.44–6.14). On the other hand, the AA genotype emerged as a "protective" factor (p = 0.02, pc = 0.04, OR = 0.40, 95% CI = 0.15–0.95), as shown in Table 3. The results were confirmed through binary logistic regression, with the presence of hallucinations as the dependent variable and the AQP4 rs2075575 genotypes and LEDD as covariates. LEDD was found to be associated with hallucinations (p = 0.01, Z score = 2.56, OR = 1.001, 95% CI = 1.0003–1.002). Additionally, the GA genotype (p = 0.02, Z score = 2.39, OR = 3.17, 95% CI = 1.23–8.13) and the AA genotype (p = 0.02, Z score = −2.39, OR = 0.32, 95% CI = 0.12–0.81) were also significant. The age at PD onset was not influenced by any of the AQP4 SNPs analyzed (data not shown). Discussion Together, our results suggest that the AQP4 rs2075575, rs162009, and rs335929 SNPs are not associated with PD risk in the Italian population. However, among pwPD, carriers of the rs162009 AA genotype appear to have a lower risk of developing RBD symptoms compared to those carrying the rs162009 GA + GG genotypes, indicating a potential role for this SNP in sleep disorders. It has been proposed that rs162009, located in a promoter region of the AQP4 gene, may increase the transcription of the AQP4-M23 isoform, which is primarily localized at astrocytic end-feet and facilitates glymphatic fluid circulation [2, 18]. We also found that the AQP4 rs2075575 genotype correlates with the presence of hallucinations in PD patients, marking the first report of an association between AQP4 SNPs and this symptom. These are preliminary results that warrants further investigation. A number of studies have previously examined the same SNPs. A case–control study in non-Hispanic Caucasians suggested that the rs162009 SNP is associated with a slower rate of cognitive decline in pwPD, with AA and AG genotype carriers exhibiting slower progression to dementia, less Aβ deposition in several brain regions, and better performance on neuropsychological tests. Healthy controls carrying the A allele also performed better on semantic fluency tests [11]. Furthermore, Jiang et al. found that in Chinese populations, the rs162009 SNP was associated with brain activity measured by clinical and MRI assessments, suggesting its involvement in the neuro-pathophysiology of PD, particularly with protective effects on the right caudate head function and cognitive performance on the Montreal Cognitive Assessment (MoCA) [10]. Another study by Sun et al. identified an association between the rs2075575 SNP and PD susceptibility and RBD in Chinese individuals, while rs335929 was associated with memory impairment but not disease risk [9]. The absence of replication of previously reported associations may reflect population-specific genetic architecture, differences in environmental exposures, or variation in study design, including sample size and selection criteria. Little is known about the potential biological significance of the AQP4 gene SNPs rs2075575 and rs335929. Rs2075575 is located in the promoter region, about 1 kb upstream of the AQP4 gene, in an area that lacks known transcription factor binding sites [19]. However, this SNP could be in linkage disequilibrium with one or more other polymorphisms, either in the promoter or coding regions of the gene, which can affect AQP4 expression and function. On the other hand, rs162009 is located in the promoter region of the M23 AQP4 isoform, which is crucial for the assembly of orthogonal arrays of particles (OAPs) at astrocytic end-feet. These structures are essential for maintaining AQP4 polarization and supporting glymphatic clearance. Genetic variation at this locus may affect promoter activity, thereby influencing AQP4 expression levels or the relative abundance of AQP4 isoforms (M1 vs. M23). Such alterations could impair AQP4 localization or function, potentially leading to glymphatic dysfunction. It is worth noting that sleep disturbances are a strong risk factor for PD [20], and α-synuclein deposition is crucial for both PD and idiopathic RBD. Post-mortem studies have shown that RBD patients have caudal brain degeneration due to synucleinopathy, and the majority of those with idiopathic RBD eventually develop a synucleinopathy such as PD, multiple system atrophy (MSA), or dementia with Lewy bodies (DLB) [21]. Reduced AQP4 expression and depolarization exacerbate α-synuclein deposition, leading to neuroinflammation. It is reasonable to assume that RBD could impair glymphatic capacity [13], and reduced AQP4 expression exacerbates α-synuclein deposition. Thus, bidirectional relationships likely exist between the glymphatic system, sleep quality, α-synuclein deposition, and PD onset [22]. The current findings of an association between AQP4 SNPs and RBD as and psychotic symptoms such as hallucinations in pwPD, are intriguing and warrant further investigation. However, our study does have some limitations: the cohort size is relatively small, and RBD was diagnosed clinically rather than through gold standard instrumental diagnosis methods such as videopolysomnography (vPSG), which would offer greater accuracy. Future replication studies involving larger cohorts and more accurately diagnosed patients for both RBD and psychiatric symptoms along with functional studies, are necessary to confirm the results and to obtain a deeper understanding of the complex interactions between the components of the glymphatic system, BBB integrity, α-synuclein load, sleep disturbances, and the onset of PD. Author Contributions Conceptualization, C.A. and F.R.G.; methodology: C.A., M.Z. and E.B.; data curation, C.A., F.R.G., E.B, and M.Z.; formal analysis, C.A., F.R.G. and M.C.; funding acquisition: M.C, F.R.G.; investigation, C.A., F.R.G., E.B. and M.Z.; resources: M.M., F.L.S., R.Z., C.C., C.D.L., B.M., A.S., J.N., project administration, C.A., F.G.; supervision, M.C.; writing—original draft: C.A. and M.C.; writing—review and editing, C.A., F.R.G., and M.C. All authors have read and agreed to the published version of the manuscript. Funding This study was funded by the Italian Ministry of health [Ricerca Corrente 2025, RC2025] CUP: H43C25000380001 project code:RRC-2025-3686061, also supported by #NEXTGENERATIONEU (NGEU) and funded by the Ministry of University and Research (MUR), National Recovery and Resilience Plan (NRRP), project MNESYS (PE0000006) – A Multiscale integrated approach to the study of the nervous system in health and disease (DN. 1553 11.10.2022). The work was also partially supported by a grant from Fondazione Romeo ed Enrica Invernizzi. Data Availability Raw data are published in Zenodo repository, 10.5281/zenodo.14843944. Declarations Competing interest The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement Raw data are published in Zenodo repository, 10.5281/zenodo.14843944.