In Parkinson’s disease (PD), peripheral immune alterations and chronic low-grade inflammation are present, and the thyroid plays a role in immune regulation. However, how immune status influences thyroid function, and at what level such effects emerge, remains unclear. This retrospective study included 396 patients with PD and 699 age- and sex-matched controls without thyroid disease. We examined serum thyroid indices, including thyroid-stimulating hormone (TSH), free thyroxine (FT4), free triiodothyronine (FT3), and the FT3/FT4 ratio. We also assessed inflammatory markers, including C-reactive protein (CRP), lymphocyte percentage, neutrophil percentage, and the neutrophil-to-lymphocyte ratio (NLR). We performed group comparisons, partial correlation analyses, interaction models, and segmented regression to assess thyroid–immune relationships. To assess generalizability, key associations were examined in an independent population-based cohort from the United States (NHANES). We found that inflammatory markers differed between groups: PD patients showed higher CRP, increased lymphocyte percentage, and reduced neutrophil percentage and NLR. Thyroid indices showed selective changes, with lower total T3, total T4, FT3, and the FT3/FT4 ratio in PD, whereas TSH and FT4 remained within their reference ranges. We further observed that the associations between thyroid indices and inflammatory markers differed by disease status. In controls, the FT3/FT4 ratio showed consistent associations with immune markers. In PD, these associations were attenuated and became dependent on immune status. These findings indicate that thyroid regulation in PD is context-dependent, characterized by preserved central axis activity but immune-linked modulation of peripheral hormone conversion. This dissociation provides a framework for understanding heterogeneous thyroid findings in PD.
Circadian syndrome (CircS), a cluster of conditions stemming from circadian rhythm disruption, is a potential contributor to Parkinson’s disease (PD), but its longitudinal association with PD remains unclear. This study aimed to assess the longitudinal association between CircS and PD risk, examine interaction with genetic susceptibility, and evaluate whether CircS provides additional risk information beyond metabolic syndrome. We included 303,613 participants from the UK Biobank cohort who were free of PD at baseline. Associations were estimated using Cox proportional hazards regression, with results presented as hazard ratios (HRs) and 95% confidence intervals (CIs). Following adjustment for sociodemographic, lifestyle, and genetic factors, participants with CircS exhibited a 22% increased risk of PD (HR 1.22, 95% CI: 1.11–1.34) compared to those without, while metabolic syndrome showed no independent association. The dose-response relationship between CircS component burden and PD risk was significantly nonlinear (Pnonlinearity < 0.0001). The adjusted HRs were 1.51 (95% CI 1.21–1.88), 1.88 (95% CI 1.36–2.61), and 5.74 (95% CI 2.90–11.37) for CircS scores of 5, 6, and 7, respectively. A notable interaction was found between CircS and PD polygenic risk score, with CircS conferring the greatest increase in PD risk in individuals exhibiting low genetic risk. These findings suggest that the prespecified CircS phenotype may capture risk-related information beyond that reflected by MetS and highlight circadian health as a potentially modifiable domain that may inform future PD prevention strategies.
The α-synuclein seed amplification assay (synSAA) is a biomarker test that identifies underlying synuclein pathology in living patients. Cerebrospinal fluid (CSF) synSAA distinguishes Type 2 synuclein-seeds (syn-seeds) found in multiple system atrophy (MSA) from Type 1 syn-seeds found in other synucleinopathies. Skin has been proposed as an alternative biospecimen for synSAA, but syn-seeds have not been consistently detected and identified in MSA. Here, we report a skin_synSAA for the detection of syn-seeds across synucleinopathies. The 126-participant cohort was enriched in Parkinson’s disease (PD, n = 64) and MSA (n = 28) cases. Skin_synSAA reached 69% and 75% sensitivities for PD and MSA. Agreement with CSF_synSAA reached 84.3%. Specificity was 100%. Amplification pattern of MSA skin samples allowed clear identification of Type 2 syn-seeds. Remarkably, there was 97.8% agreement in syn-seed type between skin and CSF. Syn-seeds detected in skin correlated with those in CSF and allowed detection and differentiation of underlying synuclein pathology by means of skin_synSAA.
There is an unclear relationship between socioeconomic status (SES)—a social exposure—and PD risk, where those with a higher SES may be more at risk, or the inverse may be true. The risk of PD may not be directly related to SES, but exposures like outdoor air pollution may be involved, particularly as associations between SES and air pollution exposure, and between air pollution exposure and neurodegenerative outcomes have been noted. Scopus, PubMed, and PsycINFO were searched to identify studies analysing degree of exposure to chronic ambient outdoor air pollution and incidence of PD, or the association between individual measures of SES and degree of exposure to outdoor air pollution. 25,416 studies were screened, with 37 fitting inclusion criteria. Meta-analyses were conducted by air pollutant, with narrative syntheses when this was not possible. There was evidence of a positive association between exposure to NO2 and PM2.5, and PD incidence. Higher and lower SES were associated with greater air pollution exposure, varying by location. The review had a low certainty of evidence, according to GRADE. Exposure to air pollution may be associated with greater risk of PD, and SES may play a role in the degree of air pollution exposure.
Bradykinesia is a hallmark sign of Parkinson’s disease (PD). Current clinical assessment is susceptible to inter-rater variability and only provides a single, ordinal severity score. Here, we propose a computer vision-based framework to quantify distinct motor characteristics from video recordings of the leg agility test, including slowness (bradykinesia), reduced amplitude (hypokinesia), progressive decrement (sequence effect) and irregularity (hesitation-halts). We validate our approach using a large-scale dataset of 3097 video recordings captured from 443 participants in the Personalized Parkinson Project. We demonstrated that features reflecting bradykinesia, hypokinesia and hesitation-halts differed significantly across clinical severity ratings, whereas sequence effect features showed no consistent differences. Furthermore, data-driven analysis using principal component analysis with varimax rotation revealed that two additional feature dimensions (i.e., beyond the four conventional domains) may be necessary to fully characterize motor abnormalities during the leg agility test. We also explored the relationship between finger-tapping (a distal upper limb task) and leg agility (a proximal lower extremity task). The analyses showed that upper and lower limb tasks provide differential insights into PD motor impairment, and that asymmetry between the least and most affected side was more pronounced in the upper extremities. This work highlights the ability of video-based assessment to provide objective characterization of motor impairment in PD, with the potential to support both in-clinic evaluations and at-home remote monitoring. Future work will investigate responsiveness to medication and longitudinal disease progression.
Synucleinopathies are defined by pathological α-synuclein states, including conformation, aggregation, modification, and seeding competence, that abundance alone cannot capture. Extracellular vesicles may preserve these states during transit from brain to blood, but mechanistic plausibility, diagnostic discrimination, and brain provenance are distinct claims. We argue that proteomics can interrogate vesicle identity, cargo and proteoforms, while unresolved provenance, pre-analytical variability, and incomplete clinical validation leave the biochemical biopsy concept plausible but unvalidated.
Predicting new-onset motor fluctuations and levodopa-induced dyskinesias (LID) is crucial for optimizing Parkinson’s disease management. To establish a transparent prognostic framework, we applied explainable machine learning to real-world, multicentric clinical data from 247 patients to forecast the 3-year onset of these complications. Evaluated strictly on complication-free patients, the models achieved moderate predictive power (LID MCC = 0.28; fluctuations MCC = 0.32). SHAP-based interpretability confirmed predictions aligned accurately with established clinical knowledge, driven primarily by levodopa duration and Levodopa Equivalent Daily Dose, with risk increasing significantly above a 300–400 mg threshold. Crucially, an ablation study revealed that excluding patients with pre-existing complications from training caused model sensitivity to collapse, demonstrating that the full spectrum of disease severity is essential for robust risk stratification. Ultimately, this rigorous methodological stress-test demonstrates that baseline clinical features alone yield limited absolute sensitivity, highlighting the necessity of integrating dynamic, longitudinal data to achieve clinical-grade individualized prediction.
Parkinson's disease (PD) is a progressive neurodegenerative disorder with a prolonged prodromal phase and complex motor symptoms. Despite improved clinical criteria, early diagnosis and longitudinal monitoring remain challenging. While cerebrospinal fluid (CSF) and plasma metabolites and proteins show biomarker potential, their utility in predictive models is insufficiently characterized. We employed a secondary computational approach to integrate proteometabolomic profiles from CSF and plasma samples of >1100 Parkinson's Progression Markers Initiative (PPMI) participants. Using multi-omics machine learning, we identified biofluid-specific signatures and evaluated predictive performance. Twenty-one biomarker candidates were validated across three models (SVM, GLMNET, RF); SVM and GLMNET achieved the highest recall (83-86%) and AUCs of 0.84-0.89. Longitudinal mixed-effects modeling revealed eight candidates associated with progression across diagnostic stages. We identified a three-part molecular framework characterizing neurodegeneration: a diagnostic subpanel reflecting early microbiome dysregulation (secretory granins and metabolites) and synaptic breakdown; a second subpanel monitoring phenoconversion via neurogenesis precursors and extracellular matrix proteins; and a third subpanel tracking progression through chronic neuroinflammation and immune activation. This integrated multi-omics approach provides a robust framework for stage-specific PD monitoring and potential clinical deployment.
The biological heterogeneity of Parkinson’s disease impedes disease-modifying trials and demands stratified approaches. We present a framework applying pharmacogenetics across three domains: mechanistic stratification, drug-host interactions, and prognostic enrichment. Genetic heterogeneity operates from polygenic scores to single variants, and mechanistic and prognostic effects may conflict with direct consequences for trial power. Realising this potential requires larger stratified trials, sensitive endpoints and collaborative infrastructure to detect genotype-by-treatment interactions with adequate power.
Freezing of Gait (FOG) is a disabling feature of Parkinson’s disease (PD). However, the neural mechanisms underlying voluntary stopping impairments remain unclear. This study used a validated Virtual Reality (VR) gait paradigm to investigate the neural correlates of cued voluntary stopping in 15 PD participants with FOG, 10 PD participants without FOG, and 12 healthy controls. PD participants were assessed on and off dopaminergic medication. Blood-oxygen-level-dependent (BOLD) signals during voluntary stopping were compared with walking and rest periods. In the OFF state, PD participants showed impaired voluntary stopping relative to healthy controls, relying predominantly on the Supplementary Motor Area (SMA). Dopaminergic medication restored stopping performance to levels comparable to healthy controls and was associated with broader recruitment of occipital, cerebellar, and inferior frontal regions. Freezers exhibited greater stopping deficits and relied on the SMA in both medication states. They also demonstrated reduced functional connectivity across occipital, temporal, and cerebellar regions in the ON state and recruited fewer cortical regions compared to non-freezers, particularly in the OFF state. The findings suggest that PD participants with FOG fail to engage a disseminated voluntary stopping network. Future work should explore interactions between the stopping network and brainstem gait and balance systems.
Quantitative assessment of motor impairment in Parkinson’s disease (PD) remains limited, particularly in tracking how deficits evolve over time. Current bedside scoring systems, including the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS), capture clinically important observations but condense them into categorical, semi-subjective ratings. This manual scoring process is imprecise, limiting diagnostic and prognostic accuracy and hindering our understanding of PD impairments and their progression. While there have been recent efforts to improve scoring via sensor-based measurements and machine learning, these approaches have limitations impeding their utility and clinical adoption. In principle, video-based methods enable comprehensive quantification and monitoring of motor function, but previous approaches largely have used imprecise 2D movement features, analyzed only individual motor tasks, and not addressed disease progression. Here, we introduce an interpretable, quantitative framework built on a synchronized, markerless 3D pose tracking system. We recorded three routinely performed MDS-UPDRS motor tasks from a large cohort of patients with PD and healthy subjects, and extracted task-level, clinically explainable kinematic features. These activities probed complementary motor subsystems—fine hand, forearm rotation, and whole-body locomotion—offering a concise yet comprehensive view of PD motor function. From 3D poses, we developed how movement patterns differed between PD and healthy subjects and how these patterns shifted as PD progressed to form distinct movement characteristics. Using machine learning models combining 3D motor features from multiple activities across multiple spatiotemporal scales, we also automatically classified diagnostic status, duration-defined PD severity, and inferred time since diagnosis. These results establish a transparent digital-biomarker framework that may support future longitudinal monitoring of PD in clinics and trials.
α-Synuclein (αSyn) aggregation is a central driver for Parkinson’s disease (PD), leading to the formation of toxic oligomeric species that promote microglial reactivity and establish a self-sustaining cycle of neuroinflammation and neurodegeneration. This process accelerates dopaminergic neuron loss in the substantia nigra and contributes to motor and cognitive deficits. Targeting both αSyn aggregation and inflammation represents a promising therapeutic strategy. This study evaluated 3-monothiopomalidomide (3MP), a novel immunomodulatory imide drug with improved safety over Pomalidomide. In vitro, 3MP reduced αSyn–induced neuronal death, microglial activation, and inflammation. Thioflavin T assay and Thioflavin S staining in SH-SY5Y cells showed significant inhibition of αSyn aggregation. In vivo, chronic administration in a αSyn-based PD rat model preserved dopaminergic neurons, improved motor and cognitive function, reduced αSyn aggregates and neuroinflammation in the midbrain and anterior cingulate cortex. Overall, 3MP shows strong dual-action potential by modulating both αSyn aggregation and neuroinflammation, supporting its development as a disease-modifying therapy for PD.
Parkinson’s disease (PD) pathogenesis elicits diverse cellular dysregulations, including mislocalized transcription factors and nuclear envelope malformations in affected dopaminergic neurons. Environmental risk factors, particularly mitochondrial toxicants and neurotoxicant exposures, are increasingly recognized as important contributors to PD susceptibility and progression by promoting oxidative stress and mitochondrial impairment in vulnerable nigrostriatal dopaminergic neurons. However, the contribution of channel-forming nucleoporins (Nups) to environmental neurotoxicant- and mitochondrial dysfunction-related neurodegenerative processes remains unexplored. Herein, we identified pathological abnormalities in both the levels and cellular distribution of specific Nups, which are key components of the nuclear pore complex, in mitochondria-impaired dopaminergic neurons in PD. We observed that mitochondrial dysfunction reduces the expression of nuclear basket Nups153 and 50, as well as the scaffold Nup107, and disrupts the localization of Ran GTPase in in vitro and in vivo dopaminergic neuron models of mitochondrial dysfunction in PD. Importantly, the nuclear pore central channel component Nup62 mislocalizes and accumulates in the cytoplasm of mitochondria-stressed dopaminergic neurons. Mitochondrial stress also interferes with bidirectional nucleocytoplasmic transport of proteins mediated by Nups in dopaminergic neuronal cells. We observed Nup pathology and Ran gradient loss in nigral dopaminergic neurons of PD patient brains, which highlights the clinical relevance of nuclear pore dysfunction. Collectively, these findings provide direct evidence that environmental neurotoxicant-linked mitochondrial dysfunction can impair the nuclear pore complex and nucleocytoplasmic transport, mechanistically linking Nup-related abnormalities to dopaminergic neurodegeneration and PD pathogenesis.
Coffee consumption has been associated with Parkinson’s disease (PD), but findings remain inconsistent. We examined whether this association differs by CYP1A2 genotype and sex in 435,551 UK Biobank participants free of PD at baseline. During a median follow-up of 15.7 years, 3319 incident PD cases were identified. In overall analyses, coffee intake was not associated with PD risk. However, coffee intake significantly interacted with CYP1A2 genotype. Among AA carriers, coffee intake below 5 cups/day was associated with lower PD risk (HR 0.83, 95% CI 0.74–0.93), and spline analyses indicated a nonlinear association with the lowest risk at around 3 cups/day. In contrast, high coffee intake (≥5 cups/day) was associated with higher PD risk in AC carriers (HR 1.43, 95% CI 1.11–1.85) and CC carriers (HR 1.84, 95% CI 1.08–3.11). Sex-stratified analyses showed that the inverse association in AA carriers was more evident in women, whereas increased risk at high intake among AC/CC carriers was mainly observed in men. These findings support heterogeneity in the association between coffee intake and PD risk according to CYP1A2 genotype.
Multiple system atrophy (MSA) is a severe neurodegenerative disorder with various underlying pathophysiological features. Mitochondrial dysfunction has been implied as a viable treatment target in patients with MSA. Yet, there is a lack of in-vivo studies examining regional metabolic differences between the Parkinsonian (MSAp) and the cerebellar (MSAc) subtype of MSA. Twenty-four patients with MSA (12 patients with MSAp and 12 patients with MSAc), 24 patients with Parkinson's disease (PD), and 24 age- and sex-matched healthy controls (HCs) underwent clinical evaluations and multimodal neuroimaging, including 31P-MRSI targeting the basal ganglia and the cerebellum. Ratios of high-energy phosphorus-containing metabolites (HEPs) were compared between groups. Only patients with MSAc showed decreased HEP levels in the cerebellum. Conversely, no differences in basal ganglia HEP levels appeared between both MSA subtypes, patients with PD, or HCs. Our findings provide preliminary in-vivo evidence for regionally detectable bioenergetic alterations in the cerebellum of patients with MSAc, while basal ganglia results, particularly in MSAp, require cautious interpretation because of the spatial-resolution limits of the present 31P-MRSI approach.
Cognitive and motor impairments are common in Parkinson’s disease (PD), but rapid decline trajectories remain difficult to predict at the individual level. Identifying reliable early-stage prognostic markers could define disease-modification windows and improve trial enrichment. Classification models were developed to predict rapid decline, defined as a decrease of ≥5 points on the Montreal Cognitive Assessment (MoCA) and an increase of ≥10 points on the MDS-UPDRS3 from baseline to any timepoint 3–5 years post-baseline. Models were trained using longitudinal MRI-derived regional atrophy rates and baseline clinical features. Model performance was evaluated using AUROC and complementary classification metrics; calibration, performance ceilings, and clinical utility were further assessed using calibration curves and decision curve analysis (DCA). Feature analysis was performed to identify clinically informative predictors. Structural MRI (e.g., annualized atrophy rate) demonstrated limited prognostic utility for individual-level prediction; this remained true with finer-grained atlases. In contrast, baseline clinical features yielded substantially stronger discrimination for both outcomes. Notably, cognitive prediction remained robust after removal of baseline MoCA, whereas motor prediction was strongly dependent on baseline UPDRS3. External validation preserved high NPVs (cognitive: 0.908 [0.872–0.940]; motor: 0.863 [0.818–0.901]). Incorporating first-year trajectory slopes improved AUROC by approximately 5%, supporting a single follow-up visit as a practical refinement timepoint. Performance gains rapidly plateaued after inclusion of a small number of high-value features, indicating an early performance ceiling. Decision curve analysis suggested potential net benefit across selected threshold probabilities, but prospective evaluation in broader and more heterogeneous populations is required before clinical implementation.
Despite global efforts to develop disease-modifying therapies for Parkinson’s disease (PD), recurrent trial failures persist, likely driven by profound and unstratified molecular heterogeneity. The advent of α-syn seed amplification assays (SAA) has enabled biologically pure cohorts, yet underlying disease trajectories remain highly variable. In this study, we leveraged deep baseline dual-compartment proteomics (4,785 CSF and 5,400 plasma proteins) from a deeply phenotyped, strictly SAA-positive PD cohort (N = 114) with a 5-year longitudinal clinical follow-up. Unsupervised machine learning revealed two robust biological subtypes driven by diametrically opposed micro-pathological crises: a synaptic/neuronal failure versus a lysosomal/glial collapse. Strikingly, despite this massive molecular divergence, these subtypes exhibited indistinguishable macroscopic clinical phenotypes, uniform striatal dopaminergic denervation (DaTscan) profiles, and strictly parallel disease progression over 5 years. Furthermore, this dichotomy occurred entirely independently of canonical genetic mutations. This pervasive “clinical masking effect” explains how targeting unstratified cohorts inadvertently dilutes subtype-specific therapeutic signals into statistical noise. To overcome this macroscopic disguise, we translated our central findings into a non-invasive, 5-protein machine-learning plasma panel (FAM3B, ACTA2, ATL3, PEBP4, and BTNL10P), achieving a cross-validated Area Under the Curve (AUC) of 0.867. Using a large-scale multi-center replication dataset, we definitively demonstrated that critical peripheral biomarker signals are completely submerged in unstratified clinical cohorts, yet they are robustly restored when stratified by our mechanism-driven framework. Ultimately, this signature provides an actionable non-invasive roadmap for precision patient stratification in future clinical trials.
Abstract Parkinson’s disease (PD) is a heterogeneous neurodegenerative disorder with variable long-term outcomes. Blood-based biomarkers for prognostic prediction remain underdeveloped. We aimed to develop a peripheral blood transcriptomic signature for predicting long-term complications in PD using unsupervised data-driven methods. Using RNA sequencing data from 541 PD patients and 180 healthy controls from the Parkinson’s Progression Markers Initiative (PPMI), we employed weighted gene correlation network analysis (WGCNA) to identify disease-associated gene modules. Contrastive principal component analysis was applied to derive a pseudo-temporal (PT) trajectory score reflecting molecular disease progression. The prognostic value of PT scores was assessed through Cox regression for cognitive impairment, freezing of gait (FOG), wearing-off, and levodopa-induced dyskinesia. WGCNA identified two PD-associated co-expression modules enriched for immune/inflammatory pathways. PT scores derived from these modules showed significant correlations with cognitive, autonomic, and axial motor symptoms. In multivariable Cox regression, higher PT scores independently predicted cognitive impairment (HR = 7.31, P = 0.002) and FOG (HR = 3.43, P < 0.001), but not predominantly dopaminergic motor complications. These findings demonstrate that peripheral blood transcriptomic signatures capture aspects of PD pathophysiology that are not fully explained by nigrostriatal dopaminergic degeneration alone, serving as potential prognostic biomarkers for cognitive impairment and FOG.
Patients with Parkinson’s disease (PD) can experience cognitive impairments linked to low frequency “theta” 4 Hz cortical activity and cognitive control. Our study investigated effects of 4 Hz subthalamic nucleus (STN) deep brain stimulation (DBS) on cognitive performance in PD patients with cognitive impairments. We recruited 17 PD participants with (n = 10) and without (n = 7) cognitive impairment. We compared motor and cognitive performance during 4 Hz STN DBS, typical DBS for motor symptoms of PD (~130 Hz) and DBS OFF. Motor performance was tested by Part III of the Movement Disorders Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS-III). Cognitive performance was tested by the Multi-Source Interference Task (MSIT), which requires conflict resolution to respond accurately. Motor function improved with 4 Hz STN DBS and further improved with ~130 Hz STN DBS. Compared to DBS OFF, reaction times were decreased during 4 Hz STN DBS and were further decreased at ~130 Hz. Strikingly, 4 Hz DBS alone improved accuracy compared to DBS OFF and ~130 Hz STN DBS. These data suggest that 4 Hz STN stimulation can improve performance in PD patients with cognitive impairments. Our findings will guide therapies targeted at improving cognition in PD and could broaden low-frequency stimulation interventions.
