Data-Driven Neurobiological Subtyping of Parkinson’s Disease Using Diffusion MRI-Derived Isotropic Diffusion

Data-Driven Neurobiological Subtyping of Parkinson’s Disease Using Diffusion MRI-Derived Isotropic Diffusion | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Data-Driven Neurobiological Subtyping of Parkinson’s Disease Using Diffusion MRI-Derived Isotropic Diffusion Anupa A Vijayakumari, Daniel Teixeira-Dos-Santos, Ken E Sakaie, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7880492/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 09 Feb, 2026 Read the published version in Neuroradiology → Version 1 posted You are reading this latest preprint version Abstract Purpose Parkinson’s disease (PD) exhibits marked clinical and biological heterogeneity. This study aimed to identify neurobiologically defined PD subtypes using isotropic diffusion (ISO), a diffusion MRI metric sensitive to extracellular water, and to determine whether these subtypes differ in baseline motor profiles and longitudinal progression. Methods Baseline ISO values were extracted from 12 subcortical motor regions in 156 de novo PD patients from the Parkinson’s Progression Markers Initiative. Hierarchical clustering was applied to ISO values to derive data-driven subtypes. Group differences in baseline motor severity were evaluated using t-tests, and linear mixed-effects models assessed longitudinal changes in ISO and motor scores over four years. Results Two subtypes emerged: subtype 1 (n = 62) with lower ISO values and subtype 2 (n = 94) with higher ISO across all regions. Subtype 2 showed greater baseline rigidity and bradykinesia. Longitudinally, subtype 1 exhibited significant ISO increases, whereas subtype 2 remained stable at elevated levels. However, motor progression rates did not differ significantly between subtypes. Conclusion Diffusion MRI-derived ISO identified distinct neurobiological subtypes of PD with divergent trajectories of extracellular pathology but similar clinical progression. ISO may serve as a sensitive biomarker for PD heterogeneity, warranting further validation in larger, long-term cohorts. Machine learning Parkinson’s disease Diffusion MRI Cluster analysis Extracellular water Isotropic diffusion Figures Figure 1 Figure 2 Figure 3 Introduction Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized by the selective loss of dopaminergic neurons in the substantia nigra and the accumulation of α-synuclein pathology throughout the brain [ 7 , 15 ]. While PD is traditionally diagnosed based on cardinal motor symptoms including bradykinesia, rigidity, tremor, and postural instability and gait difficulty (PIGD), clinical presentation and disease progression vary considerably among individuals [ 10 , 29 ]. This heterogeneity has significant implications for prognosis, treatment response, and clinical trial design, underscoring the need for biological markers that can better define PD subtypes. Diffusion MRI provides a powerful tool to investigate microstructural changes associated with PD [ 1 ]. Prior work using bi-compartment diffusion models has shown that extracellular water–sensitive metrics, such as free water, are elevated in basal ganglia and motor circuits, reflecting neuroinflammatory and neurodegenerative processes [ 2 , 19 , 21 , 22 , 27 ]. Notably, Bower et al. demonstrated that patients classified clinically as PIGD exhibited greater free-water increases compared to tremor-dominant patients [ 2 ]. While this finding highlights the biological sensitivity of diffusion-derived metrics, the subtypes were defined a priori by clinical phenotype rather than emerging directly from imaging data. Thus, the identification of subtypes grounded directly in neurobiological markers, including those sensitive to extracellular pathology, remains an unmet need. Isotropic diffusion (ISO), a model-free diffusion metric that is derived within the generalized Q-sampling Imaging (GQI) framework [ 34 ], quantifies the isotropic component of water diffusion arising from cerebrospinal fluid, edema, or tissue loss. Higher ISO values have been linked to demyelination and edema [ 25 ], suggesting that ISO can capture disease-related alterations. Yet, whether ISO can stratify PD into neurobiologically distinct subgroups that exhibit different trajectories of neurodegenerative brain changes as well as motor progression has not been investigated. To address this gap, the present study aimed to apply unsupervised clustering to baseline ISO values extracted from subcortical motor regions in a cohort of de novo PD patients from the Parkinson’s Progression Markers Initiative (PPMI) [ 16 , 17 ]. Our objectives were to (1) identify distinct PD subtypes based on ISO patterns, (2) characterize their baseline motor phenotypes, and (3) evaluate whether these subtypes exhibit varying trajectories of ISO changes and motor progression over a 4-year follow-up. Methods Participants We included 156 de novo Parkinson’s disease (PD) patients from the Parkinson’s Progression Markers Initiative (PPMI) database ( https://www.ppmi-info.org/access-data-specimens/download-data ; RRID: SCR_006431), accessed through the standard application process. Inclusion criteria were: diagnosis of PD within the past two years; a positive DaTscan confirming diagnosis; no dopaminergic treatment within six months of enrollment; availability of diffusion MRI data at baseline and at the 4-year follow-up; and MDS-UPDRS-III motor scores obtained at baseline (in the drug-naïve state) and at the 4-year follow-up (in the OFF-medication state). Patients were excluded if they had dementia or atypical parkinsonian syndromes, significant neurological or psychiatric conditions, or with structural brain abnormalities, poor-quality imaging data (e.g., motion artifacts, or susceptibility distortions). Ethics Approval The PPMI study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines and is registered on ClinicalTrials.gov (NCT01141023). All participants provided written informed consent under protocols approved by the local ethics committees of participating sites, listed at https://www.ppmi-info.org/about-ppmi/ppmi-clinical-sites . One of the authors (AAV) obtained permission to access tier 3 PPMI data, which were provided in a fully de-identified format. The PPMI Data and Publications Committee reviewed and administratively approved this manuscript in accordance with PPMI data use policies. As all data were de-identified and previously collected under approved protocols, no additional ethics committee approvals were required for this analysis. Image Acquisition Diffusion MRI data was acquired using a Siemens 3T TrioTim MRI scanner equipped with a 12-channel Matrix head coil. Data from baseline and 4-year timepoints were downloaded for analysis. The diffusion sequence employed a two-dimensional echo-planar imaging (EPI) protocol with the following acquisition parameters: repetition time (TR) = 900 ms; echo time (TE) = 88 ms; flip angle = 90°; voxel dimensions = 2 × 2 × 2 mm³; 72 axial slices; and 64 diffusion-encoding directions with a b-value of 1000 s/mm². Additionally, a single non-diffusion-weighted (b = 0 s/mm²) volume was included. Further details of the PPMI study design and imaging protocols are available in Marek et al. [ 17 ] and the PPMI MRI Operations Manual [ https://www.ppmi-info.org/wp content/uploads/2017/06/PPMI-MRI-Operations-Manual-V7.pdf]. Image processing Diffusion MRI data were processed using DSI Studio ( http://dsi-studio.labsolver.org ; version 2024 release), a widely used software platform for diffusion model reconstruction and extraction of microstructural metrics [ 28 , 35 ]. Preprocessing included correction for motion and eddy current distortions to enhance data quality and mitigate artifacts. Raw images were converted into DSI Studio’s native .src format. Reconstruction was performed using GQI, a model-free framework that estimates spin distribution functions (SDFs) and nonlinearly registers them to the standard ICBM-152 space [ 32 ]. Reconstruction quality was assessed using the goodness-of-fit coefficient (R²) between each subject’s anisotropy map and the standard space [ 30 ]. Only datasets meeting a threshold of R² >0.70 were included. Each reconstructed dataset was saved as a .fib file, which encodes voxel-wise diffusion metrics, including the isotropic diffusion (ISO) value. The ISO metric used here is model-free and derived from GQI, potentially offering increased robustness to partial volume effects in small subcortical structures [ 33 ]. ISO quantifies the direction-independent (isotropic) component of the diffusion signal, serving as a proxy for extracellular water and capturing diffusion related to cerebrospinal fluid (CSF), edema, or neurodegenerative tissue loss. We extracted ISO values using the connectometry module in DSI Studio [ 31 ] for twelve PD–relevant subcortical brain regions of interest (ROIs) defined by the ATAG atlas [ 12 ], including bilateral red nucleus (RN), substantia nigra (SN), subthalamic nucleus (STN), striatum, globus pallidus externus (GPe) and internus (GPi). These subcortical regions were selected because of their involvement in PD pathologic changes that occur early in the disease [ 4 – 6 , 14 ]. For each subject and time point (baseline and 4-year), mean ISO values were computed for each ROI. These regional ISO metrics were used in subsequent clustering and statistical analyses. Motor Outcome Measures Motor symptoms were assessed using Part III of the Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale (MDS-UPDRS-III) [ 9 ]. Total scores and subdomain scores were extracted, including rigidity (item 3.3), bradykinesia (items 3.4–3.8), postural instability and gait difficulty (PIGD; items 3.10–3.13), and tremor (items 3.15–3.17), at two time points: baseline and 4 years. To minimize the influence of dopaminergic medication and capture disease-related motor progression, only OFF-medication scores were included. After filtering for OFF-state availability, 78 patients had usable motor data at the 4-year visit. Data-Driven Clustering Hierarchical agglomerative clustering was applied to identify subtypes based on ISO values extracted from 12 PD-relevant brain regions (bilateral SN, STN, RN, striatum, GPi and GPe). No clinical variables were used in the clustering. All ISO features were z-score standardized across participants to ensure comparability. Clustering was performed using Ward’s linkage and Euclidean distance, implemented via the scipy.cluster.hierarchy library in Python. A dendrogram was generated to visualize hierarchical distances, and the optimal number of clusters (or subtypes) was selected by identifying the largest vertical linkage distance, resulting in a two-cluster solution. Final cluster assignments were derived using the fcluster() function with maxclust = 2 (Fig. 1 ). These neurobiologically defined subtypes were used for subsequent group comparisons and longitudinal modeling. Statistical Analysis Baseline group differences in demographic, clinical, and ISO variables were assessed. Independent two-sample t-tests were used to compare ISO values between subtypes across 12 PD-relevant brain regions defined by the ATAG atlas, thereby verifying that the clustering procedure produced neurobiologically distinct groups. Age differences were evaluated using independent two-sample t-tests, while sex distribution was compared using chi-square tests. We also determined whether these ISO-based subtypes differed in motor severity at baseline using independent two-sample t-tests. We employed linear mixed-effects models (LMEMs) to evaluate whether subtypes exhibited different progression trajectories over time in both motor scores and ISO values. Separate models were constructed for each outcome domain, with motor scores (total and subdomain) and ISO values from each brain region modeled independently as outcome variables. The models included fixed effects for time, subtype, and their interaction (time × subtype), with subject-specific intercepts modeled as random effects. A significant main effect of time indicated overall progression across the cohort, while a significant main effect of subtype reflected baseline differences between groups. The time × subtype interaction term was of particular interest, as it tested whether the rate of change in motor severity and regional ISO burden differed between the two subtypes, thus assessing whether these ISO-defined groups exhibited different disease progression. Age and sex were included as covariates. To correct for multiple comparisons, false discovery rate (FDR) correction was applied using the Benjamini-Hochberg procedure (p < 0.05). All statistical analyses were conducted using Python (v3.11.13), with core packages including pandas, scipy, scikit-learn, statsmodels, and matplotlib. Results Baseline Demographic Characteristics Subtype 1 included 62 patients (mean age 60.7 ± 9.6 years; 39 males and 23 females), and subtype 2 included 94 patients (mean age 61.0 ± 9.8 years; 61 males and 33 females). Age did not differ significantly between subtypes ( t = -0.17, p = 0.87). Sex distribution was also not significantly different ( χ² = 0.01, p = 0.93), indicating that ISO-based clustering was not confounded by demographic variables. ISO-Based Clustering Identified Two Distinct PD Subtypes Hierarchical clustering of baseline ISO values from 12 PD-relevant subcortical brain regions, extracted using the ATAG atlas, revealed two distinct subtypes (Fig. 1 ). At baseline, subtype 1 exhibited lower ISO levels across all regions, while subtype 2 showed elevated ISO values, suggesting a higher neurodegenerative burden. Independent two-sample t -tests confirmed significant baseline differences between subtype 1 and subtype 2 in all 12 regions (all brain regions p < 0.05, FDR-corrected), verifying that the clustering produced neurobiologically distinct groups (Table 1 ). Table 1 Baseline Isotropic Levels Across ATAG Regions in ISO-Defined Parkinson’s Disease Subtypes ATAG regions Subtype 1 (n = 62) (mean ± SD) Subtype 2 (n = 94) (mean ± SD) t p RN left 0.31 ± 0.07 0.47 ± 0.10 -10.18 < 0.001 RN right 0.27 ± 0.09 0.42 ± 0.09 -9.86 < 0.001 SN left 0.26 ± 0.08 0.41 ± 0.08 -10.93 < 0.001 SN right 0.21 ± 0.07 0.32 ± 0.06 -10.67 < 0.001 STN left 0.38 ± 0.13 0.60 ± 0.13 -9.59 < 0.001 STN right 0.26 ± 0.10 0.41 ± 0.10 -8.31 < 0.001 Striatum left 0.42 ± 0.11 0.64 ± 0.14 -10.48 < 0.001 Striatum right 0.43 ± 0.11 0.66 ± 0.14 -10.75 < 0.001 GPe left 0.22 ± 0.06 0.35 ± 0.06 -12.01 < 0.001 GPe right 0.22 ± 0.06 0.35 ± 0.06 -11.93 < 0.001 GPi left 0.25 ± 0.07 0.39 ± 0.07 -11.95 < 0.001 GPi right 0.25 ± 0.07 0.39 ± 0.07 -12.22 < 0.001 SD : standard deviation. ATAG : Atlas of the Basal Ganglia. RN : Red Nucleus, SN : Substantia Nigra, STN : Subthalamic Nucleus, GPe : Globus Pallidus Externa, GPi : Globus Pallidus Interna. t : t-statistic from independent samples t-test comparing Subtype 1 and Subtype 2. p : p-value; all comparisons are significant after correction for multiple comparisons (p < 0.05, FDR corrected) Motor Severity Differences at Baseline At baseline, subtype 2 exhibited significantly greater rigidity (p = 0.004) and bradykinesia (p = 0.01) compared to subtype 1. Total UPDRS-III scores were also higher in subtype 2 (p = 0.05), although this difference did not remain significant after correction for multiple comparisons. No significant differences were observed in baseline tremor or PIGD scores (both p > 0.05; Table 2 ). Table 2 MDS-UPDRS-III Total and Subdomain at Baseline by ISO-Defined PD Subtypes Motor scores Subtype 1 (n = 62) (mean ± SD) Subtype 2 (n = 94) (mean ± SD) t p Total 18.71 ± 8.69 21.59 ± 9.14 -1.95 0.05 Rigidity 3.52 ± 2.35 4.83 ± 2.96 -2.93 0.004 Bradykinesia 7.34 ± 4.36 9.24 ± 5.17 -2.39 0.01 PIGD 0.73 ± 0.91 0.65 ± 0.68 0.60 0.54 Tremor 2.84 ± 2.51 2.49 ± 2.07 0.94 0.34 SD : standard deviation. PIGD : Postural Instability and Gait Difficulty. t : t-statistic from independent samples t-test comparing Subtype 1 and Subtype 2. p : p-value; values < 0.05 indicate statistical significance. Longitudinal Modeling of Isotropic Diffusion Trajectories Across all ATAG regions, LMEMs revealed significant main effects of time and subtype, as well as significant time × subtype interactions (all p < 0.05; Table 3 ). Main effects of time indicated progressive increases in ISO over the 4-year follow-up, regardless of subtype. A robust subtype effect was also observed, with subtype 2 consistently showing higher ISO than subtype 1 across all regions. Importantly, significant negative time × subtype interactions emerged in every region, reflecting divergent longitudinal trajectories: subtype 1 began with lower ISO values but exhibited progressive increases over time, whereas subtype 2 started with elevated ISO levels and remained comparatively stable across follow-up. These findings suggest that subtype 1 patients “catch up” in ISO burden over time. Figure 2 illustrates the longitudinal changes in mean ISO values averaged across the 12 ATAG regions, with region-specific trajectories shown in Supplementary Fig. 1 (Online Resource 1). Table 3 Longitudinal Effects of Time, ISO-Based Subtype, and Their Interaction on Isotropic Diffusion Levels in ATAG Brain Regions ATAG Time Subtype Time × Subtype β (95% CI) p β (95% CI) p β (95% CI) p RN left 0.03 (0.02–0.04) < 0.001 0.17 (0.13–0.20) < 0.001 -0.03 (-0.04 – -0.02) < 0.001 RN right 0.03 (0.02–0.03) < 0.001 0.15 (0.12–0.18) < 0.001 -0.03 (-0.04 – -0.01) < 0.001 SN left 0.03 (0.02–0.04) < 0.001 0.15 (0.12–0.18) < 0.001 -0.03 (-0.04 – -0.01) < 0.001 SN right 0.02 (0.01–0.02) < 0.001 0.12 (0.09–0.14) < 0.001 -0.02 (-0.03 – -0.01) < 0.001 STN left 0.03 (0.02–0.05) < 0.001 0.21 (0.17–0.26) < 0.001 -0.04 (-0.05 – -0.02) < 0.001 STN right 0.03 (0.02–0.04) < 0.001 0.14 (0.11–0.18) < 0.001 -0.02 (-0.03 – -0.01) 0.002 Striatum left 0.04 (0.03–0.05) < 0.001 0.23 (0.18–0.27) < 0.001 -0.04 (-0.05 – -0.02) < 0.001 Striatum right 0.04 (0.03–0.05) < 0.001 0.23 (0.19–0.28) < 0.001 -0.04 (-0.06 – -0.02) < 0.001 GPe left 0.02 (0.01–0.03) < 0.001 0.13 (0.11–0.15) < 0.001 -0.02 (-0.03 – -0.01) < 0.001 GPe right 0.02 (0.01–0.03) < 0.001 0.13 (0.11–0.15) < 0.001 -0.02 (-0.03 – -0.01) < 0.001 GPi left 0.02 (0.02–0.03) < 0.001 0.14 (0.12–0.17) < 0.001 -0.02 (-0.03 – -0.01) < 0.001 GPi right 0.03 (0.02–0.03) < 0.001 0.14 (0.12–0.17) < 0.001 -0.02 (-0.03 – -0.02) < 0.001 β indicates the estimated fixed-effect slope from linear mixed-effects models. The “Time” term reflects the average annual rate of change in isotropic diffusion (ISO) across participants. The “Subtype” term indicates baseline differences in ISO between ISO-defined subtypes. The “Time × Subtype” interaction term assesses whether the rate of ISO change over time differs by subtype. RN = Red Nucleus; SN = Substantia Nigra; STN = Subthalamic Nucleus; GPe = Globus Pallidus Externa; GPi = Globus Pallidus Interna; CI = confidence interval. Bold p-values indicate statistical significance after FDR correction (p < 0.05). Longitudinal Modeling of Motor Progression LMEMs assessed the effects of time, subtype, and their interaction on motor symptom progression over 4 years (Table 4 ). Significant main effects of time were observed for total MDS-UPDRS-III scores, rigidity, bradykinesia, and PIGD (all p < 0.05), indicating overall worsening of motor symptoms across patients. A significant main effect of subtype was not detected for motor scores. However, time × subtype interactions were not significant for the total score or any motor subdomain scores (all p > 0.05), suggesting that progression rates did not differ between subtypes. Figure 3 illustrates longitudinal changes in MDS-UPDRS-III total and motor subdomain scores from baseline to 4-year follow-up in ISO-defined PD subtypes. Table 4 Longitudinal Effects of Time, ISO-Based Subtypes, and Their Interaction on Motor Progression in Parkinson’s Disease MDS-UPDRS-III Time Subtype Time × Subtype β (95% CI) p β (95% CI) p β (95% CI) p Total 2.25 (1.29–3.20) < 0.001 2.79 (-0.56–6.15) 0.22 0.08 (-1.18–1.34) 0.90 Rigidity 0.64 (0.37–0.92) < 0.001 1.30 (0.32–2.28) 0.05 -0.27 (-0.64–0.10) 0.24 Bradykinesia 0.86 (0.39–1.34) < 0.001 1.90 (0.16–3.64) 0.08 -0.05 (-0.67–0.58) 0.91 PIGD 0.16 (0.04–0.28) 0.03 -0.09 (-0.47–0.28) 0.72 0.12 (-0.04–0.27) 0.24 Tremor 0.08 (-0.14–0.31) 0.57 -0.36 (-1.14–0.43) 0.51 0.18 (-0.12–0.47) 0.37 β indicates the estimated fixed-effect slope from linear mixed-effects models. The "Time" term reflects the average annual change in motor scores across all participants. The "Subtype" term represents baseline differences between ISO-defined subtypes. The "Time × Subtype" interaction tests whether the rate of motor progression differs between subtypes. CI = confidence interval; PIGD = Postural Instability and Gait Difficulty. Bold p-values indicate statistical significance after FDR correction (p < 0.05). Discussion In this study, we applied a data-driven clustering approach to ISO values derived from diffusion MRI to identify neurobiologically defined subtypes of PD. Two distinct subtypes emerged: one with elevated ISO across subcortical motor regions at baseline (subtype 2) and another with relatively lower ISO levels (subtype 1). These groups differed not only in baseline imaging profiles but also in motor severity, with subtype 2 exhibiting greater rigidity and bradykinesia. Longitudinal analyses further revealed different trajectories of ISO changes: subtype 1 displayed progressive increases in ISO burden over four years, whereas subtype 2 remained relatively stable at elevated levels. While ISO-defined subtypes differed neurobiologically, these differences did not translate into divergent clinical trajectories, as both groups demonstrated comparable rates of motor progression. Shippey et al. highlighted that extracellular processes play a central role in PD pathogenesis, with extracellular vesicles serving as vehicles for the spread of misfolded α-synuclein aggregates that disrupt mitochondrial function, impair axonal transport, and ultimately drive neuronal death [ 26 ]. ISO provides an in vivo marker of these extracellular processes, as it quantifies the direction-independent component of water diffusion [ 33 ]. Elevated ISO has been linked to extracellular water burden arising from microstructural alterations such as demyelination, astrogliosis, or neurodegenerative tissue loss [ 25 ]. Importantly, ISO is not specific to a single pathological mechanism but reflects the net effect of these processes. In our study, we found that subtype 2 patients exhibited high baseline ISO values that remained stable over four years, whereas subtype 1 patients began with lower ISO but showed progressive increases, suggesting distinct courses of extracellular pathology across subgroups. Together, these findings demonstrate that although Shippey’s review emphasizes the fundamental role of extracellular pathology in PD, our study shows that their burden and temporal dynamics are not uniform across patients. However, caution is warranted, as ISO is not specific to a single pathological mechanism. Future studies integrating imaging with histopathological data will be essential to disentangle the underlying contributors to ISO changes. It is also important to note that another diffusion-derived metric, free water, has been widely investigated as an indicator of extracellular pathology in PD. It is derived from a bi-compartment model that separates the isotropic contribution of extracellular water from the tissue signal [ 20 ]. In PD, free water is consistently elevated across subcortical regions, and longitudinal studies indicate that it increases with disease progression and predicts decline in motor function [ 2 , 11 , 18 , 19 , 22 ]. By contrast, our study applied ISO, an alternative approach to quantifying isotropic diffusion. While ISO and free water both reflect extracellular processes, they are derived differently and can be viewed as methodologically distinct approaches. Future work comparing ISO- and free water-defined subtypes will be needed to clarify their ability in distinguishing PD subtypes and to establish their clinical relevance. In addition to the imaging differences, we also observed motor distinctions between ISO-defined subtypes. At baseline, subtype 2 patients exhibited greater rigidity and bradykinesia, whereas tremor and PIGD scores were comparable across groups. These findings suggest that elevated ISO burden may align with greater motor severity early in the disease course. However, longitudinal analyses revealed that both subtypes progressed at similar rates across total and domain-specific motor scores over four years, despite different ISO trajectories. This dissociation between imaging and clinical progression suggests that while ISO-defined subtypes capture underlying neurobiological variability, this does not manifest as divergent motor progression when measured with MDS-UPDRS-III. However, it should also be noted that attrition reduced the available sample for 4-year OFF-medication data to 78 participants, which may have limited statistical power to detect subtle interaction effects. Thus, caution is warranted in interpreting the motor findings, and future studies with larger cohorts, longer follow-up, and more sensitive motor measures will be needed to clarify the clinical significance of neurobiologically defined subtypes. Our findings highlight a key distinction between cross-sectional and longitudinal perspectives on PD heterogeneity. At baseline, ISO-defined subtypes demonstrated both neurobiological and clinical differences, with subtype 2 showing greater isotropic diffusion burden alongside more severe rigidity and bradykinesia. These results suggest that neurobiological subtyping can capture meaningful clinical variation early in the disease course, supporting its utility in cross-sectional settings. However, longitudinal analyses revealed that, despite divergent trajectories of ISO changes, motor progression did not differ significantly between subtypes. This dissociation emphasizes that while neurobiological markers can stratify patients at baseline, their long-term clinical impact may not be straightforward. Future studies should therefore account for both cross-sectional and longitudinal dimensions when evaluating the clinical relevance of biologically defined subtypes, as reliance on baseline differences alone may overestimate their prognostic utility. Several limitations should be acknowledged. First, the clinical measures analyzed were restricted to MDS-UPDRS-III in the OFF state, which is subjective and prone to inter-subject variability [ 8 , 23 ], and the use of more objective motor assessments will be important in future work [ 3 , 13 , 24 ]. Second, PD heterogeneity includes a wide spectrum of non-motor symptoms such as cognitive impairment, rapid eye movement sleep behavior disorder, and autonomic dysfunction, which were not addressed in this analysis [ 36 ]. Future studies that incorporate these outcomes may reveal stronger clinical correlates of ISO-defined subtypes. Finally, our analysis was confined to subcortical motor regions, which are known to be centrally involved in PD pathophysiology, and future work should also investigate cortical and cerebellar circuits to provide a more complete characterization of PD heterogeneity. Conclusion This study shows that diffusion MRI-derived ISO revealed two neurobiologically distinct subtypes of PD. Patients with higher baseline ISO values remained stable over four years, whereas those with lower baseline ISO values showed progressive increases, indicating divergent trajectories of extracellular pathology. Despite these imaging differences, both subtypes demonstrated similar rates of motor progression, underscoring a dissociation between neurobiological and clinical change. These findings provide initial evidence that ISO is sensitive to heterogeneity in underlying pathology, although its prognostic and clinical significance remain uncertain. Larger studies with longer follow-up will be needed to clarify whether ISO-based subtyping contributes to understanding disease progression. Declarations Ethics Approval and Consent to Participate: This study analyzed data obtained from a publicly available, fully de-identified research database that had received prior ethics approval from the originating study’s institutional review boards. All participants in the original study provided written informed consent. As only anonymized data were used, and no new data were collected or identifiable information accessed, additional institutional ethics approval was not required. All analyses were conducted in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Competing Interests: The authors declare no competing interests. Code Availability Analysis code available on reasonable request. Author Contribution AAV contributed to the study conception and design, data extraction, data analysis and interpretation, and drafting and revision of the manuscript. DTS contributed to study interpretation and revision of the manuscript. KS and HHF contributed to revision of the manuscript. BLW contributed to the study conception and design and revision of the manuscript. All authors read and approved the final manuscript. Data Availability Data were obtained from the Parkinson’s Progression Markers Initiative (PPMI; [https://www.ppmi-info.org](https:/www.ppmi-info.org) ) and are available to qualified investigators upon application per PPMI policies. References Bergamino M, Keeling EG, Mishra VR, Stokes AM, Walsh RR (2020) Assessing White Matter Pathology in Early-Stage Parkinson Disease Using Diffusion MRI: A Systematic Review. Front Neurol 11:314 Bower AE, Crisomia SJ, Chung JW, Martello JP, Burciu RG (2023) Free water imaging unravels unique patterns of longitudinal structural brain changes in Parkinson's disease subtypes. Front Neurol 14:1278065 Chen L, Cai G, Weng H, Yu J, Yang Y, Huang X, Chen X, Ye Q (2020) More Sensitive Identification for Bradykinesia Compared to Tremors in Parkinson's Disease Based on Parkinson's KinetiGraph (PKG). Front Aging Neurosci 12:594701 Colpan ME, Slavin KV (2010) Subthalamic and red nucleus volumes in patients with Parkinson’s disease: Do they change with disease progression? Parkinsonism Relat Disord 16:398–403 Damier P, Hirsch EC, Agid Y, Graybiel AM (1999) The substantia nigra of the human brain: II. Patterns of loss of dopamine-containing neurons in Parkinson's disease. Brain 122:1437–1448 DeLong MR, Wichmann T (2007) Circuits and Circuit Disorders of the Basal Ganglia. Arch Neurol 64:20–24 Eriksen JL, Wszolek Z, Petrucelli L (2005) Molecular Pathogenesis of Parkinson Disease. Arch Neurol 62:353–357 Evers LJW, Krijthe JH, Meinders MJ, Bloem BR, Heskes TM (2019) Measuring Parkinson's disease over time: The real-world within-subject reliability of the MDS-UPDRS. Mov Disord 34:1480–1487 Goetz CG, Fahn S, Martinez-Martin P, Poewe W, Sampaio C, Stebbins GT, Stern MB, Tilley BC, Dodel R, Dubois B, Holloway R, Jankovic J, Kulisevsky J, Lang AE, Lees A, Leurgans S, LeWitt PA, Nyenhuis D, Olanow CW, Rascol O, Schrag A, Teresi JA, Van Hilten JJ, LaPelle N (2007) Movement Disorder Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS): Process, format, and clinimetric testing plan. Mov Disord 22:41–47 Greenland JC, Williams-Gray CH, Barker RA (2019) The clinical heterogeneity of Parkinson's disease and its therapeutic implications. Eur J Neurosci 49:328–338 Guttuso T Jr., Bergsland N, Hagemeier J, Lichter DG, Pasternak O, Zivadinov R (2018) Substantia Nigra Free Water Increases Longitudinally in Parkinson Disease. AJNR Am J Neuroradiol 39:479–484 Keuken MC, Forstmann BU (2015) A probabilistic atlas of the basal ganglia using 7 T MRI. Data Brief 4:577–582 King LA, Priest KC, Salarian A, Pierce D, Horak FB (2012) Comparing the Mini-BESTest with the Berg Balance Scale to Evaluate Balance Disorders in Parkinson's Disease. Parkinsons Dis 2012:375419 Kish SJ, Shannak K, Hornykiewicz O (1988) Uneven pattern of dopamine loss in the striatum of patients with idiopathic Parkinson's disease. Pathophysiologic and clinical implications. N Engl J Med 318:876–880 Lo Bianco C, Ridet JL, Schneider BL, Deglon N, Aebischer P (2002) alpha -Synucleinopathy and selective dopaminergic neuron loss in a rat lentiviral-based model of Parkinson's disease. Proc Natl Acad Sci U S A 99:10813–10818 Marek K, Chowdhury S, Siderowf A, Lasch S, Coffey CS, Caspell-Garcia C, Simuni T, Jennings D, Tanner CM, Trojanowski JQ, Shaw LM, Seibyl J, Schuff N, Singleton A, Kieburtz K, Toga AW, Mollenhauer B, Galasko D, Chahine LM, Weintraub D, Foroud T, Tosun-Turgut D, Poston K, Arnedo V, Frasier M, Sherer T (2018) The Parkinson's progression markers initiative (PPMI) - establishing a PD biomarker cohort. Ann Clin Transl Neurol 5:1460–1477 Marek K, Jennings D, Lasch S, Siderowf A, Tanner C, Simuni T, Coffey C, Kieburtz K, Flagg E, Chowdhury S (2011) The Parkinson progression marker initiative (PPMI). Prog Neurobiol 95:629–635 Ofori E, Pasternak O, Planetta PJ, Burciu R, Snyder A, Febo M, Golde TE, Okun MS, Vaillancourt DE (2015) Increased free water in the substantia nigra of Parkinson's disease: a single-site and multi-site study. Neurobiol Aging 36:1097–1104 Ofori E, Pasternak O, Planetta PJ, Li H, Burciu RG, Snyder AF, Lai S, Okun MS, Vaillancourt DE (2015) Longitudinal changes in free-water within the substantia nigra of Parkinson's disease. Brain 138:2322–2331 Pasternak O, Sochen N, Gur Y, Intrator N, Assaf Y (2009) Free water elimination and mapping from diffusion MRI. Magn Reson Med 62:717–730 Pasternak O, Westin CF, Bouix S, Seidman LJ, Goldstein JM, Woo TU, Petryshen TL, Mesholam-Gately RI, McCarley RW, Kikinis R, Shenton ME, Kubicki M (2012) Excessive extracellular volume reveals a neurodegenerative pattern in schizophrenia onset. J Neurosci 32:17365–17372 Planetta PJ, Ofori E, Pasternak O, Burciu RG, Shukla P, DeSimone JC, Okun MS, McFarland NR, Vaillancourt DE (2015) Free-water imaging in Parkinson’s disease and atypical parkinsonism. Brain 139:495–508 Regnault A, Boroojerdi B, Meunier J, Bani M, Morel T, Cano S (2019) Does the MDS-UPDRS provide the precision to assess progression in early Parkinson's disease? Learnings from the Parkinson's progression marker initiative cohort. J Neurol 266:1927–1936 Schlachetzki JCM, Barth J, Marxreiter F, Gossler J, Kohl Z, Reinfelder S, Gassner H, Aminian K, Eskofier BM, Winkler J, Klucken J (2017) Wearable sensors objectively measure gait parameters in Parkinson's disease. PLoS ONE 12:e0183989 Shen CY, Tyan YS, Kuo LW, Wu CW, Weng JC (2015) Quantitative Evaluation of Rabbit Brain Injury after Cerebral Hemisphere Radiation Exposure Using Generalized q-Sampling Imaging. PLoS ONE 10:e0133001 Shippey LE, Campbell SG, Hill AF, Smith DP (2022) Propagation of Parkinson's disease by extracellular vesicle production and secretion. Biochem Soc Trans 50:1303–1314 Vijayakumari AA, Parker D, Osmanlioglu Y, Alappatt JA, Whyte J, Diaz-Arrastia R, Kim JJ, Verma R (2021) Free Water Volume Fraction: An Imaging Biomarker to Characterize Moderate-to-Severe Traumatic Brain Injury. J Neurotrauma 38:2698–2705 Vijayakumari AA, Sakaie KE, Fernandez HH, Walter BL (2025) Parkinson's disease subtypes and their association with probable rapid eye movement sleep behavior disorder severity: a brainstem tractography and machine learning investigation. Brain Imaging Behav 19:189–194 Wüllner U, Borghammer P, Choe CU, Csoti I, Falkenburger B, Gasser T, Lingor P, Riederer P (2023) The heterogeneity of Parkinson's disease. J Neural Transm (Vienna) 130:827–838 Yeh F-C, Tang P-F, Tseng W-YI (2013) Diffusion MRI connectometry automatically reveals affected fiber pathways in individuals with chronic stroke. NeuroImage: Clin 2:912–921 Yeh FC, Badre D, Verstynen T (2016) Connectometry: A statistical approach harnessing the analytical potential of the local connectome. NeuroImage 125:162–171 Yeh FC, Tseng WY (2011) NTU-90: a high angular resolution brain atlas constructed by q-space diffeomorphic reconstruction. NeuroImage 58:91–99 Yeh FC, Verstynen TD, Wang Y, Fernández-Miranda JC, Tseng WY (2013) Deterministic diffusion fiber tracking improved by quantitative anisotropy. PLoS ONE 8:e80713 Yeh FC, Wedeen VJ, Tseng WY (2010) Generalized q-sampling imaging. IEEE Trans Med Imaging 29:1626–1635 Zhang H, He WJ, Liang LH, Zhang HW, Zhang XJ, Zeng L, Luo SP, Lin F, Lei Y (2021) Diffusion Spectrum Imaging of Corticospinal Tracts in Idiopathic Normal Pressure Hydrocephalus. Front Neurol 12:636518 Zhang N, Liu W, Ye M, Cohen AD, Zhang Y (2015) The heterogeneity of non-motor symptoms of Parkinson's disease. Neurol Sci 36:577–584 Additional Declarations No competing interests reported. Supplementary Files OnlineResource1.pdf Cite Share Download PDF Status: Published Journal Publication published 09 Feb, 2026 Read the published version in Neuroradiology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7880492","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":532326864,"identity":"8598cf01-3805-4489-b5b7-4d23dde28744","order_by":0,"name":"Anupa A Vijayakumari","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAs0lEQVRIie3PrQ7CMBSG4W+pXTI7td1CFzzXUjIxwwieXzeJI7sMLgBxlibMwBXMVKGLR1AyDKqtI6Gvf/KdA4RCP1kMUhwZECkPIjgmAOPuBAKY7Z1Jcrx1JJbT6tBI4HHe2Ek6LIQ5rKzba4movfcOM8OcG8LqU5oQi+liF/lIdhVPGdwIH4kUH7K2k2L8pS/ev3QtkZ1kQy21fq7ypJGR0rS1k6/MhPQkJt+VUCgU+ode7EQ9AJgXtlAAAAAASUVORK5CYII=","orcid":"","institution":"Cleveland Clinic","correspondingAuthor":true,"prefix":"","firstName":"Anupa","middleName":"A","lastName":"Vijayakumari","suffix":""},{"id":532326865,"identity":"e73fdf16-3e9e-4da4-aec6-e0073ca82e01","order_by":1,"name":"Daniel Teixeira-Dos-Santos","email":"","orcid":"","institution":"Cleveland Clinic","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Teixeira-Dos-Santos","suffix":""},{"id":532326866,"identity":"c84b5285-8b9f-4224-a925-811b15299f5c","order_by":2,"name":"Ken E Sakaie","email":"","orcid":"","institution":"Cleveland Clinic","correspondingAuthor":false,"prefix":"","firstName":"Ken","middleName":"E","lastName":"Sakaie","suffix":""},{"id":532326867,"identity":"8ee568b3-975f-48f8-8619-79caa8e3619e","order_by":3,"name":"Hubert H Fernandez","email":"","orcid":"","institution":"Cleveland Clinic","correspondingAuthor":false,"prefix":"","firstName":"Hubert","middleName":"H","lastName":"Fernandez","suffix":""},{"id":532326868,"identity":"1141fcea-c4c3-4a98-a6c2-a38e0b1a7e88","order_by":4,"name":"Benjamin L Walter","email":"","orcid":"","institution":"Cleveland Clinic","correspondingAuthor":false,"prefix":"","firstName":"Benjamin","middleName":"L","lastName":"Walter","suffix":""}],"badges":[],"createdAt":"2025-10-16 19:08:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7880492/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7880492/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00234-026-03939-4","type":"published","date":"2026-02-09T15:58:28+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":94051083,"identity":"f1ca67d0-99d8-4c4b-afa7-d06c2142a03d","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":241887,"visible":true,"origin":"","legend":"","description":"","filename":"ISOmanuscript10202025.docx","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/62066fe1caa8cd5e4b0a92b1.docx"},{"id":94051390,"identity":"e5af40fd-7d37-42fb-bdf7-6139e3868e7a","added_by":"auto","created_at":"2025-10-22 00:04:39","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6880,"visible":true,"origin":"","legend":"","description":"","filename":"5d54565d249a4e3ca0810ed9e55b5de3.json","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/7afcc14dfb13a2ba3402e67f.json"},{"id":94051087,"identity":"e99dba1e-5a32-4822-a76e-f32523b1298b","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":232338,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/f5e5d2dafeeb02a4196414e8.pdf"},{"id":94051094,"identity":"68f09d22-6a37-499c-890a-316dde098078","added_by":"auto","created_at":"2025-10-21 23:56:40","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":120043,"visible":true,"origin":"","legend":"","description":"","filename":"5d54565d249a4e3ca0810ed9e55b5de31enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/8b93b95724be8d04d129ff4d.xml"},{"id":94051391,"identity":"e9901eab-a302-4d57-8df3-61d2b4c27a1e","added_by":"auto","created_at":"2025-10-22 00:04:39","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13236,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/02c2a973dd36685bd3bce548.png"},{"id":94051089,"identity":"07e135a5-c91f-4230-8499-a15270cf1025","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8065,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/fbe37591bb0b3b310a88020f.png"},{"id":94051092,"identity":"a2af953b-c00e-4103-8557-e4eae9dad4a5","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20408,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/3b9730004d032e17b17fd21c.png"},{"id":94051093,"identity":"814f7473-4875-4654-a6ea-73d03ec044f1","added_by":"auto","created_at":"2025-10-21 23:56:40","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117256,"visible":true,"origin":"","legend":"","description":"","filename":"5d54565d249a4e3ca0810ed9e55b5de31structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/731e3458b9075dd0bfaf9559.xml"},{"id":94051095,"identity":"c2bd50da-5866-490c-ab8c-4f5904f806d5","added_by":"auto","created_at":"2025-10-21 23:56:40","extension":"html","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":125594,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/8d0dee328a8aac66c5a62671.html"},{"id":94051085,"identity":"43822eec-9dff-4a75-afe9-183420a5e1b3","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":32047,"visible":true,"origin":"","legend":"\u003cp\u003eHierarchical clustering dendrogram based on baseline isotropic diffusion (ISO) values. ISO values from 12 PD-relevant subcortical brain regions (bilateral red nucleus, substantia nigra, subthalamic nucleus, striatum, globus pallidus externus, and globus pallidus internus) were used as input features. Clustering was performed using Ward’s linkage. The red dashed horizontal line denotes the cluster threshold at linkage distance = 20, yielding a two-cluster solution.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/b39913bbf1335c80f34ef6ed.png"},{"id":94051082,"identity":"011e2762-c48d-4221-9b67-3f1ec1de77ec","added_by":"auto","created_at":"2025-10-21 23:56:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":31202,"visible":true,"origin":"","legend":"\u003cp\u003eLongitudinal trajectories of mean isotropic diffusion (ISO) values averaged across all 12 ATAG regions in Parkinson’s disease subtypes. Subtype 1 is shown in blue and subtype 2 in orange, with error bars indicating standard errors at baseline and 4-year follow-up.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/8f6bca80dbd02259f06070fe.png"},{"id":94051392,"identity":"34397879-6633-42b6-89ee-754983d1737d","added_by":"auto","created_at":"2025-10-22 00:04:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":76103,"visible":true,"origin":"","legend":"\u003cp\u003eLongitudinal changes in MDS-UPDRS-III total and motor subdomain scores over 4 years in Parkinson’s disease subtypes. Subtype 1 is shown in blue and Subtype 2 in orange, with error bars representing standard errors. Panels show (a) Total score, (b) Rigidity, (c) Bradykinesia, (d) Postural Instability and Gait Difficulty (PIGD), and (e) Tremor.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/7d4f67a0b996aa09a2fc6543.png"},{"id":102785589,"identity":"4380bcf2-f63d-458f-a870-2e0dbed2186a","added_by":"auto","created_at":"2026-02-16 16:08:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1074925,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/0b9f6815-2874-49af-bc76-d5562b182edc.pdf"},{"id":94051393,"identity":"40cf0e3f-8918-436c-a38d-af108ceff0f6","added_by":"auto","created_at":"2025-10-22 00:04:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":232338,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7880492/v1/753f8d9430a578b2575870fd.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Data-Driven Neurobiological Subtyping of Parkinson’s Disease Using Diffusion MRI-Derived Isotropic Diffusion","fulltext":[{"header":"Introduction","content":"\u003cp\u003eParkinson's disease (PD) is a progressive neurodegenerative disorder characterized by the selective loss of dopaminergic neurons in the substantia nigra and the accumulation of α-synuclein pathology throughout the brain [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. While PD is traditionally diagnosed based on cardinal motor symptoms including bradykinesia, rigidity, tremor, and postural instability and gait difficulty (PIGD), clinical presentation and disease progression vary considerably among individuals [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. This heterogeneity has significant implications for prognosis, treatment response, and clinical trial design, underscoring the need for biological markers that can better define PD subtypes.\u003c/p\u003e\u003cp\u003eDiffusion MRI provides a powerful tool to investigate microstructural changes associated with PD [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Prior work using bi-compartment diffusion models has shown that extracellular water\u0026ndash;sensitive metrics, such as free water, are elevated in basal ganglia and motor circuits, reflecting neuroinflammatory and neurodegenerative processes [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Notably, Bower et al. demonstrated that patients classified clinically as PIGD exhibited greater free-water increases compared to tremor-dominant patients [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. While this finding highlights the biological sensitivity of diffusion-derived metrics, the subtypes were defined \u003cem\u003ea priori\u003c/em\u003e by clinical phenotype rather than emerging directly from imaging data. Thus, the identification of subtypes grounded directly in neurobiological markers, including those sensitive to extracellular pathology, remains an unmet need.\u003c/p\u003e\u003cp\u003eIsotropic diffusion (ISO), a model-free diffusion metric that is derived within the generalized Q-sampling Imaging (GQI) framework [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], quantifies the isotropic component of water diffusion arising from cerebrospinal fluid, edema, or tissue loss. Higher ISO values have been linked to demyelination and edema [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], suggesting that ISO can capture disease-related alterations. Yet, whether ISO can stratify PD into neurobiologically distinct subgroups that exhibit different trajectories of neurodegenerative brain changes as well as motor progression has not been investigated.\u003c/p\u003e\u003cp\u003eTo address this gap, the present study aimed to apply unsupervised clustering to baseline ISO values extracted from subcortical motor regions in a cohort of de novo PD patients from the Parkinson\u0026rsquo;s Progression Markers Initiative (PPMI) [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Our objectives were to (1) identify distinct PD subtypes based on ISO patterns, (2) characterize their baseline motor phenotypes, and (3) evaluate whether these subtypes exhibit varying trajectories of ISO changes and motor progression over a 4-year follow-up.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eParticipants\u003c/h2\u003e\u003cp\u003eWe included 156 de novo Parkinson\u0026rsquo;s disease (PD) patients from the Parkinson\u0026rsquo;s Progression Markers Initiative (PPMI) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ppmi-info.org/access-data-specimens/download-data\u003c/span\u003e\u003cspan address=\"https://www.ppmi-info.org/access-data-specimens/download-data\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; RRID: SCR_006431), accessed through the standard application process. Inclusion criteria were: diagnosis of PD within the past two years; a positive DaTscan confirming diagnosis; no dopaminergic treatment within six months of enrollment; availability of diffusion MRI data at baseline and at the 4-year follow-up; and MDS-UPDRS-III motor scores obtained at baseline (in the drug-na\u0026iuml;ve state) and at the 4-year follow-up (in the OFF-medication state). Patients were excluded if they had dementia or atypical parkinsonian syndromes, significant neurological or psychiatric conditions, or with structural brain abnormalities, poor-quality imaging data (e.g., motion artifacts, or susceptibility distortions).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eEthics Approval\u003c/h3\u003e\n\u003cp\u003e The PPMI study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines and is registered on ClinicalTrials.gov (NCT01141023). All participants provided written informed consent under protocols approved by the local ethics committees of participating sites, listed at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ppmi-info.org/about-ppmi/ppmi-clinical-sites\u003c/span\u003e\u003cspan address=\"https://www.ppmi-info.org/about-ppmi/ppmi-clinical-sites\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. One of the authors (AAV) obtained permission to access tier 3 PPMI data, which were provided in a fully de-identified format. The PPMI Data and Publications Committee reviewed and administratively approved this manuscript in accordance with PPMI data use policies. As all data were de-identified and previously collected under approved protocols, no additional ethics committee approvals were required for this analysis.\u003c/p\u003e\n\u003ch3\u003eImage Acquisition\u003c/h3\u003e\n\u003cp\u003eDiffusion MRI data was acquired using a Siemens 3T TrioTim MRI scanner equipped with a 12-channel Matrix head coil. Data from baseline and 4-year timepoints were downloaded for analysis. The diffusion sequence employed a two-dimensional echo-planar imaging (EPI) protocol with the following acquisition parameters: repetition time (TR)\u0026thinsp;=\u0026thinsp;900 ms; echo time (TE)\u0026thinsp;=\u0026thinsp;88 ms; flip angle\u0026thinsp;=\u0026thinsp;90\u0026deg;; voxel dimensions\u0026thinsp;=\u0026thinsp;2 \u0026times; 2 \u0026times; 2 mm\u0026sup3;; 72 axial slices; and 64 diffusion-encoding directions with a b-value of 1000 s/mm\u0026sup2;. Additionally, a single non-diffusion-weighted (b\u0026thinsp;=\u0026thinsp;0 s/mm\u0026sup2;) volume was included. Further details of the PPMI study design and imaging protocols are available in Marek et al. [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and the PPMI MRI Operations Manual [\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ppmi-info.org/wp\u003c/span\u003e\u003cspan address=\"https://www.ppmi-info.org/wp\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e content/uploads/2017/06/PPMI-MRI-Operations-Manual-V7.pdf].\u003c/p\u003e\n\u003ch3\u003eImage processing\u003c/h3\u003e\n\u003cp\u003eDiffusion MRI data were processed using DSI Studio (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dsi-studio.labsolver.org\u003c/span\u003e\u003cspan address=\"http://dsi-studio.labsolver.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; version 2024 release), a widely used software platform for diffusion model reconstruction and extraction of microstructural metrics [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Preprocessing included correction for motion and eddy current distortions to enhance data quality and mitigate artifacts. Raw images were converted into DSI Studio\u0026rsquo;s native .src format. Reconstruction was performed using GQI, a model-free framework that estimates spin distribution functions (SDFs) and nonlinearly registers them to the standard ICBM-152 space [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Reconstruction quality was assessed using the goodness-of-fit coefficient (R\u0026sup2;) between each subject\u0026rsquo;s anisotropy map and the standard space [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Only datasets meeting a threshold of R\u0026sup2; \u0026gt;0.70 were included. Each reconstructed dataset was saved as a .fib file, which encodes voxel-wise diffusion metrics, including the isotropic diffusion (ISO) value. The ISO metric used here is model-free and derived from GQI, potentially offering increased robustness to partial volume effects in small subcortical structures [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. ISO quantifies the direction-independent (isotropic) component of the diffusion signal, serving as a proxy for extracellular water and capturing diffusion related to cerebrospinal fluid (CSF), edema, or neurodegenerative tissue loss.\u003c/p\u003e\u003cp\u003eWe extracted ISO values using the connectometry module in DSI Studio [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e] for twelve PD\u0026ndash;relevant subcortical brain regions of interest (ROIs) defined by the ATAG atlas [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], including bilateral red nucleus (RN), substantia nigra (SN), subthalamic nucleus (STN), striatum, globus pallidus externus (GPe) and internus (GPi). These subcortical regions were selected because of their involvement in PD pathologic changes that occur early in the disease [\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. For each subject and time point (baseline and 4-year), mean ISO values were computed for each ROI. These regional ISO metrics were used in subsequent clustering and statistical analyses.\u003c/p\u003e\n\u003ch3\u003eMotor Outcome Measures\u003c/h3\u003e\n\u003cp\u003eMotor symptoms were assessed using Part III of the Movement Disorder Society-sponsored revision of the Unified Parkinson\u0026rsquo;s Disease Rating Scale (MDS-UPDRS-III) [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Total scores and subdomain scores were extracted, including rigidity (item 3.3), bradykinesia (items 3.4\u0026ndash;3.8), postural instability and gait difficulty (PIGD; items 3.10\u0026ndash;3.13), and tremor (items 3.15\u0026ndash;3.17), at two time points: baseline and 4 years. To minimize the influence of dopaminergic medication and capture disease-related motor progression, only OFF-medication scores were included. After filtering for OFF-state availability, 78 patients had usable motor data at the 4-year visit.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eData-Driven Clustering\u003c/h2\u003e\u003cp\u003eHierarchical agglomerative clustering was applied to identify subtypes based on ISO values extracted from 12 PD-relevant brain regions (bilateral SN, STN, RN, striatum, GPi and GPe). No clinical variables were used in the clustering. All ISO features were z-score standardized across participants to ensure comparability. Clustering was performed using Ward\u0026rsquo;s linkage and Euclidean distance, implemented via the scipy.cluster.hierarchy library in Python. A dendrogram was generated to visualize hierarchical distances, and the optimal number of clusters (or subtypes) was selected by identifying the largest vertical linkage distance, resulting in a two-cluster solution. Final cluster assignments were derived using the fcluster() function with maxclust\u0026thinsp;=\u0026thinsp;2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These neurobiologically defined subtypes were used for subsequent group comparisons and longitudinal modeling.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eBaseline group differences in demographic, clinical, and ISO variables were assessed. Independent two-sample t-tests were used to compare ISO values between subtypes across 12 PD-relevant brain regions defined by the ATAG atlas, thereby verifying that the clustering procedure produced neurobiologically distinct groups. Age differences were evaluated using independent two-sample t-tests, while sex distribution was compared using chi-square tests. We also determined whether these ISO-based subtypes differed in motor severity at baseline using independent two-sample t-tests. We employed linear mixed-effects models (LMEMs) to evaluate whether subtypes exhibited different progression trajectories over time in both motor scores and ISO values. Separate models were constructed for each outcome domain, with motor scores (total and subdomain) and ISO values from each brain region modeled independently as outcome variables. The models included fixed effects for time, subtype, and their interaction (time \u0026times; subtype), with subject-specific intercepts modeled as random effects. A significant main effect of time indicated overall progression across the cohort, while a significant main effect of subtype reflected baseline differences between groups. The time \u0026times; subtype interaction term was of particular interest, as it tested whether the rate of change in motor severity and regional ISO burden differed between the two subtypes, thus assessing whether these ISO-defined groups exhibited different disease progression. Age and sex were included as covariates. To correct for multiple comparisons, false discovery rate (FDR) correction was applied using the Benjamini-Hochberg procedure (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). All statistical analyses were conducted using Python (v3.11.13), with core packages including pandas, scipy, scikit-learn, statsmodels, and matplotlib.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eBaseline Demographic Characteristics\u003c/h2\u003e\u003cp\u003eSubtype 1 included 62 patients (mean age 60.7\u0026thinsp;\u0026plusmn;\u0026thinsp;9.6 years; 39 males and 23 females), and subtype 2 included 94 patients (mean age 61.0\u0026thinsp;\u0026plusmn;\u0026thinsp;9.8 years; 61 males and 33 females). Age did not differ significantly between subtypes (\u003cem\u003et\u003c/em\u003e = -0.17, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.87). Sex distribution was also not significantly different (\u003cem\u003eχ\u0026sup2;\u003c/em\u003e = 0.01, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.93), indicating that ISO-based clustering was not confounded by demographic variables.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eISO-Based Clustering Identified Two Distinct PD Subtypes\u003c/h2\u003e\u003cp\u003eHierarchical clustering of baseline ISO values from 12 PD-relevant subcortical brain regions, extracted using the ATAG atlas, revealed two distinct subtypes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). At baseline, subtype 1 exhibited lower ISO levels across all regions, while subtype 2 showed elevated ISO values, suggesting a higher neurodegenerative burden. Independent two-sample \u003cem\u003et\u003c/em\u003e-tests confirmed significant baseline differences between subtype 1 and subtype 2 in all 12 regions (all brain regions \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR-corrected), verifying that the clustering produced neurobiologically distinct groups (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBaseline Isotropic Levels Across ATAG Regions in ISO-Defined Parkinson\u0026rsquo;s Disease Subtypes\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eATAG regions\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSubtype 1 (n\u0026thinsp;=\u0026thinsp;62)\u003c/p\u003e\u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSubtype 2 (n\u0026thinsp;=\u0026thinsp;94)\u003c/p\u003e\u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003et\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.31\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.47\u0026thinsp;\u0026plusmn;\u0026thinsp;0.10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-10.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.27\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.42\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-9.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.41\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-10.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.21\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-10.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.38\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.60\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-9.59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.41\u0026thinsp;\u0026plusmn;\u0026thinsp;0.10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-8.31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStriatum left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.42\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.64\u0026thinsp;\u0026plusmn;\u0026thinsp;0.14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-10.48\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStriatum right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.43\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.66\u0026thinsp;\u0026plusmn;\u0026thinsp;0.14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-10.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPe left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.35\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-12.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPe right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.22\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.35\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-11.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPi left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-11.95\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPi right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-12.22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003e\u003cem\u003eSD\u003c/em\u003e: standard deviation. \u003cem\u003eATAG\u003c/em\u003e: Atlas of the Basal Ganglia. \u003cem\u003eRN\u003c/em\u003e: Red Nucleus, \u003cem\u003eSN\u003c/em\u003e: Substantia Nigra, \u003cem\u003eSTN\u003c/em\u003e: Subthalamic Nucleus, \u003cem\u003eGPe\u003c/em\u003e: Globus Pallidus Externa, \u003cem\u003eGPi\u003c/em\u003e: Globus Pallidus Interna. \u003cem\u003et\u003c/em\u003e: t-statistic from independent samples t-test comparing Subtype 1 and Subtype 2. \u003cem\u003ep\u003c/em\u003e: p-value; all comparisons are significant after correction for multiple comparisons (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR corrected)\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eMotor Severity Differences at Baseline\u003c/h2\u003e\u003cp\u003eAt baseline, subtype 2 exhibited significantly greater rigidity (p\u0026thinsp;=\u0026thinsp;0.004) and bradykinesia (p\u0026thinsp;=\u0026thinsp;0.01) compared to subtype 1. Total UPDRS-III scores were also higher in subtype 2 (p\u0026thinsp;=\u0026thinsp;0.05), although this difference did not remain significant after correction for multiple comparisons. No significant differences were observed in baseline tremor or PIGD scores (both \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05; Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMDS-UPDRS-III Total and Subdomain at Baseline by ISO-Defined PD Subtypes\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMotor scores\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSubtype 1 (n\u0026thinsp;=\u0026thinsp;62)\u003c/p\u003e\u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSubtype 2 (n\u0026thinsp;=\u0026thinsp;94)\u003c/p\u003e\u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003et\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e18.71\u0026thinsp;\u0026plusmn;\u0026thinsp;8.69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e21.59\u0026thinsp;\u0026plusmn;\u0026thinsp;9.14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-1.95\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003e0.05\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRigidity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e3.52\u0026thinsp;\u0026plusmn;\u0026thinsp;2.35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e4.83\u0026thinsp;\u0026plusmn;\u0026thinsp;2.96\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-2.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e0.004\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBradykinesia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e7.34\u0026thinsp;\u0026plusmn;\u0026thinsp;4.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e9.24\u0026thinsp;\u0026plusmn;\u0026thinsp;5.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-2.39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e0.01\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePIGD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e0.73\u0026thinsp;\u0026plusmn;\u0026thinsp;0.91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e0.65\u0026thinsp;\u0026plusmn;\u0026thinsp;0.68\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.54\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTremor\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e\u003cp\u003e2.84\u0026thinsp;\u0026plusmn;\u0026thinsp;2.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e\u003cp\u003e2.49\u0026thinsp;\u0026plusmn;\u0026thinsp;2.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.94\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.34\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eSD\u003c/em\u003e: standard deviation. \u003cem\u003ePIGD\u003c/em\u003e: Postural Instability and Gait Difficulty. \u003cem\u003et\u003c/em\u003e: t-statistic from independent samples t-test comparing Subtype 1 and Subtype 2. \u003cem\u003ep\u003c/em\u003e: p-value; values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicate statistical significance.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eLongitudinal Modeling of Isotropic Diffusion Trajectories\u003c/h2\u003e\u003cp\u003eAcross all ATAG regions, LMEMs revealed significant main effects of time and subtype, as well as significant time \u0026times; subtype interactions (all p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Main effects of time indicated progressive increases in ISO over the 4-year follow-up, regardless of subtype. A robust subtype effect was also observed, with subtype 2 consistently showing higher ISO than subtype 1 across all regions. Importantly, significant negative time \u0026times; subtype interactions emerged in every region, reflecting divergent longitudinal trajectories: subtype 1 began with lower ISO values but exhibited progressive increases over time, whereas subtype 2 started with elevated ISO levels and remained comparatively stable across follow-up. These findings suggest that subtype 1 patients \u0026ldquo;catch up\u0026rdquo; in ISO burden over time. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e illustrates the longitudinal changes in mean ISO values averaged across the 12 ATAG regions, with region-specific trajectories shown in Supplementary Fig.\u0026nbsp;1 (Online Resource 1).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eLongitudinal Effects of Time, ISO-Based Subtype, and Their Interaction on Isotropic Diffusion Levels in ATAG Brain Regions\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026minus;\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eATAG\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eTime\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eSubtype\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eTime \u0026times; Subtype\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.17 (0.13\u0026ndash;0.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.03 (-0.04 \u0026ndash; -0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.15 (0.12\u0026ndash;0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.03 (-0.04 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.15 (0.12\u0026ndash;0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.03 (-0.04 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.02 (0.01\u0026ndash;0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.12 (0.09\u0026ndash;0.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTN left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.05)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.21 (0.17\u0026ndash;0.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.04 (-0.05 \u0026ndash; -0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTN right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.14 (0.11\u0026ndash;0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e0.002\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStriatum left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.04 (0.03\u0026ndash;0.05)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.23 (0.18\u0026ndash;0.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.04 (-0.05 \u0026ndash; -0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStriatum right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.04 (0.03\u0026ndash;0.05)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.23 (0.19\u0026ndash;0.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.04 (-0.06 \u0026ndash; -0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPe left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.02 (0.01\u0026ndash;0.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.13 (0.11\u0026ndash;0.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPe right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.02 (0.01\u0026ndash;0.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.13 (0.11\u0026ndash;0.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPi left\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.02 (0.02\u0026ndash;0.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.14 (0.12\u0026ndash;0.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGPi right\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.03 (0.02\u0026ndash;0.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.14 (0.12\u0026ndash;0.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.02 (-0.03 \u0026ndash; -0.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eβ indicates the estimated fixed-effect slope from linear mixed-effects models. The \u0026ldquo;Time\u0026rdquo; term reflects the average annual rate of change in isotropic diffusion (ISO) across participants. The \u0026ldquo;Subtype\u0026rdquo; term indicates baseline differences in ISO between ISO-defined subtypes. The \u0026ldquo;Time \u0026times; Subtype\u0026rdquo; interaction term assesses whether the rate of ISO change over time differs by subtype. RN\u0026thinsp;=\u0026thinsp;Red Nucleus; SN\u0026thinsp;=\u0026thinsp;Substantia Nigra; STN\u0026thinsp;=\u0026thinsp;Subthalamic Nucleus; GPe\u0026thinsp;=\u0026thinsp;Globus Pallidus Externa; GPi\u0026thinsp;=\u0026thinsp;Globus Pallidus Interna; CI\u0026thinsp;=\u0026thinsp;confidence interval. Bold p-values indicate statistical significance after FDR correction (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eLongitudinal Modeling of Motor Progression\u003c/h2\u003e\u003cp\u003eLMEMs assessed the effects of time, subtype, and their interaction on motor symptom progression over 4 years (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Significant main effects of time were observed for total MDS-UPDRS-III scores, rigidity, bradykinesia, and PIGD (all p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), indicating overall worsening of motor symptoms across patients. A significant main effect of subtype was not detected for motor scores. However, time \u0026times; subtype interactions were not significant for the total score or any motor subdomain scores (all p\u0026thinsp;\u0026gt;\u0026thinsp;0.05), suggesting that progression rates did not differ between subtypes. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates longitudinal changes in MDS-UPDRS-III total and motor subdomain scores from baseline to 4-year follow-up in ISO-defined PD subtypes.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eLongitudinal Effects of Time, ISO-Based Subtypes, and Their Interaction on Motor Progression in Parkinson\u0026rsquo;s Disease\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026minus;\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eMDS-UPDRS-III\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eTime\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eSubtype\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eTime \u0026times; Subtype\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eβ (95% CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.25 (1.29\u0026ndash;3.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.79 (-0.56\u0026ndash;6.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e0.08 (-1.18\u0026ndash;1.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.90\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRigidity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.64 (0.37\u0026ndash;0.92)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.30 (0.32\u0026ndash;2.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.05\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.27 (-0.64\u0026ndash;0.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.24\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBradykinesia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.86 (0.39\u0026ndash;1.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.90 (0.16\u0026ndash;3.64)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.08\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e-0.05 (-0.67\u0026ndash;0.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.91\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePIGD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.16 (0.04\u0026ndash;0.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e0.03\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.09 (-0.47\u0026ndash;0.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e0.12 (-0.04\u0026ndash;0.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.24\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTremor\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.08 (-0.14\u0026ndash;0.31)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.36 (-1.14\u0026ndash;0.43)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e\u003cp\u003e0.18 (-0.12\u0026ndash;0.47)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.37\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eβ indicates the estimated fixed-effect slope from linear mixed-effects models. The \"Time\" term reflects the average annual change in motor scores across all participants. The \"Subtype\" term represents baseline differences between ISO-defined subtypes. The \"Time \u0026times; Subtype\" interaction tests whether the rate of motor progression differs between subtypes. CI\u0026thinsp;=\u0026thinsp;confidence interval; PIGD\u0026thinsp;=\u0026thinsp;Postural Instability and Gait Difficulty. Bold p-values indicate statistical significance after FDR correction (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we applied a data-driven clustering approach to ISO values derived from diffusion MRI to identify neurobiologically defined subtypes of PD. Two distinct subtypes emerged: one with elevated ISO across subcortical motor regions at baseline (subtype 2) and another with relatively lower ISO levels (subtype 1). These groups differed not only in baseline imaging profiles but also in motor severity, with subtype 2 exhibiting greater rigidity and bradykinesia. Longitudinal analyses further revealed different trajectories of ISO changes: subtype 1 displayed progressive increases in ISO burden over four years, whereas subtype 2 remained relatively stable at elevated levels. While ISO-defined subtypes differed neurobiologically, these differences did not translate into divergent clinical trajectories, as both groups demonstrated comparable rates of motor progression.\u003c/p\u003e\u003cp\u003eShippey et al. highlighted that extracellular processes play a central role in PD pathogenesis, with extracellular vesicles serving as vehicles for the spread of misfolded α-synuclein aggregates that disrupt mitochondrial function, impair axonal transport, and ultimately drive neuronal death [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. ISO provides an in vivo marker of these extracellular processes, as it quantifies the direction-independent component of water diffusion [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Elevated ISO has been linked to extracellular water burden arising from microstructural alterations such as demyelination, astrogliosis, or neurodegenerative tissue loss [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Importantly, ISO is not specific to a single pathological mechanism but reflects the net effect of these processes. In our study, we found that subtype 2 patients exhibited high baseline ISO values that remained stable over four years, whereas subtype 1 patients began with lower ISO but showed progressive increases, suggesting distinct courses of extracellular pathology across subgroups. Together, these findings demonstrate that although Shippey\u0026rsquo;s review emphasizes the fundamental role of extracellular pathology in PD, our study shows that their burden and temporal dynamics are not uniform across patients. However, caution is warranted, as ISO is not specific to a single pathological mechanism. Future studies integrating imaging with histopathological data will be essential to disentangle the underlying contributors to ISO changes.\u003c/p\u003e\u003cp\u003eIt is also important to note that another diffusion-derived metric, free water, has been widely investigated as an indicator of extracellular pathology in PD. It is derived from a bi-compartment model that separates the isotropic contribution of extracellular water from the tissue signal [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. In PD, free water is consistently elevated across subcortical regions, and longitudinal studies indicate that it increases with disease progression and predicts decline in motor function [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. By contrast, our study applied ISO, an alternative approach to quantifying isotropic diffusion. While ISO and free water both reflect extracellular processes, they are derived differently and can be viewed as methodologically distinct approaches. Future work comparing ISO- and free water-defined subtypes will be needed to clarify their ability in distinguishing PD subtypes and to establish their clinical relevance.\u003c/p\u003e\u003cp\u003eIn addition to the imaging differences, we also observed motor distinctions between ISO-defined subtypes. At baseline, subtype 2 patients exhibited greater rigidity and bradykinesia, whereas tremor and PIGD scores were comparable across groups. These findings suggest that elevated ISO burden may align with greater motor severity early in the disease course. However, longitudinal analyses revealed that both subtypes progressed at similar rates across total and domain-specific motor scores over four years, despite different ISO trajectories. This dissociation between imaging and clinical progression suggests that while ISO-defined subtypes capture underlying neurobiological variability, this does not manifest as divergent motor progression when measured with MDS-UPDRS-III. However, it should also be noted that attrition reduced the available sample for 4-year OFF-medication data to 78 participants, which may have limited statistical power to detect subtle interaction effects. Thus, caution is warranted in interpreting the motor findings, and future studies with larger cohorts, longer follow-up, and more sensitive motor measures will be needed to clarify the clinical significance of neurobiologically defined subtypes.\u003c/p\u003e\u003cp\u003eOur findings highlight a key distinction between cross-sectional and longitudinal perspectives on PD heterogeneity. At baseline, ISO-defined subtypes demonstrated both neurobiological and clinical differences, with subtype 2 showing greater isotropic diffusion burden alongside more severe rigidity and bradykinesia. These results suggest that neurobiological subtyping can capture meaningful clinical variation early in the disease course, supporting its utility in cross-sectional settings. However, longitudinal analyses revealed that, despite divergent trajectories of ISO changes, motor progression did not differ significantly between subtypes. This dissociation emphasizes that while neurobiological markers can stratify patients at baseline, their long-term clinical impact may not be straightforward. Future studies should therefore account for both cross-sectional and longitudinal dimensions when evaluating the clinical relevance of biologically defined subtypes, as reliance on baseline differences alone may overestimate their prognostic utility.\u003c/p\u003e\u003cp\u003eSeveral limitations should be acknowledged. First, the clinical measures analyzed were restricted to MDS-UPDRS-III in the OFF state, which is subjective and prone to inter-subject variability [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], and the use of more objective motor assessments will be important in future work [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Second, PD heterogeneity includes a wide spectrum of non-motor symptoms such as cognitive impairment, rapid eye movement sleep behavior disorder, and autonomic dysfunction, which were not addressed in this analysis [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Future studies that incorporate these outcomes may reveal stronger clinical correlates of ISO-defined subtypes. Finally, our analysis was confined to subcortical motor regions, which are known to be centrally involved in PD pathophysiology, and future work should also investigate cortical and cerebellar circuits to provide a more complete characterization of PD heterogeneity.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study shows that diffusion MRI-derived ISO revealed two neurobiologically distinct subtypes of PD. Patients with higher baseline ISO values remained stable over four years, whereas those with lower baseline ISO values showed progressive increases, indicating divergent trajectories of extracellular pathology. Despite these imaging differences, both subtypes demonstrated similar rates of motor progression, underscoring a dissociation between neurobiological and clinical change. These findings provide initial evidence that ISO is sensitive to heterogeneity in underlying pathology, although its prognostic and clinical significance remain uncertain. Larger studies with longer follow-up will be needed to clarify whether ISO-based subtyping contributes to understanding disease progression.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study analyzed data obtained from a publicly available, fully de-identified research database that had received prior ethics approval from the originating study\u0026rsquo;s institutional review boards. All participants in the original study provided written informed consent. As only anonymized data were used, and no new data were collected or identifiable information accessed, additional institutional ethics approval was not required. All analyses were conducted in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eCode Availability\u003c/h2\u003e\n\u003cp\u003eAnalysis code available on reasonable request.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eAAV contributed to the study conception and design, data extraction, data analysis and interpretation, and drafting and revision of the manuscript. DTS contributed to study interpretation and revision of the manuscript. KS and HHF contributed to revision of the manuscript. BLW contributed to the study conception and design and revision of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eData were obtained from the Parkinson\u0026rsquo;s Progression Markers Initiative (PPMI; [https://www.ppmi-info.org](https:/www.ppmi-info.org) ) and are available to qualified investigators upon application per PPMI policies.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBergamino M, Keeling EG, Mishra VR, Stokes AM, Walsh RR (2020) Assessing White Matter Pathology in Early-Stage Parkinson Disease Using Diffusion MRI: A Systematic Review. Front Neurol 11:314\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBower AE, Crisomia SJ, Chung JW, Martello JP, Burciu RG (2023) Free water imaging unravels unique patterns of longitudinal structural brain changes in Parkinson's disease subtypes. Front Neurol 14:1278065\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen L, Cai G, Weng H, Yu J, Yang Y, Huang X, Chen X, Ye Q (2020) More Sensitive Identification for Bradykinesia Compared to Tremors in Parkinson's Disease Based on Parkinson's KinetiGraph (PKG). Front Aging Neurosci 12:594701\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eColpan ME, Slavin KV (2010) Subthalamic and red nucleus volumes in patients with Parkinson\u0026rsquo;s disease: Do they change with disease progression? Parkinsonism Relat Disord 16:398\u0026ndash;403\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDamier P, Hirsch EC, Agid Y, Graybiel AM (1999) The substantia nigra of the human brain: II. Patterns of loss of dopamine-containing neurons in Parkinson's disease. Brain 122:1437\u0026ndash;1448\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDeLong MR, Wichmann T (2007) Circuits and Circuit Disorders of the Basal Ganglia. Arch Neurol 64:20\u0026ndash;24\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEriksen JL, Wszolek Z, Petrucelli L (2005) Molecular Pathogenesis of Parkinson Disease. Arch Neurol 62:353\u0026ndash;357\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEvers LJW, Krijthe JH, Meinders MJ, Bloem BR, Heskes TM (2019) Measuring Parkinson's disease over time: The real-world within-subject reliability of the MDS-UPDRS. Mov Disord 34:1480\u0026ndash;1487\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGoetz CG, Fahn S, Martinez-Martin P, Poewe W, Sampaio C, Stebbins GT, Stern MB, Tilley BC, Dodel R, Dubois B, Holloway R, Jankovic J, Kulisevsky J, Lang AE, Lees A, Leurgans S, LeWitt PA, Nyenhuis D, Olanow CW, Rascol O, Schrag A, Teresi JA, Van Hilten JJ, LaPelle N (2007) Movement Disorder Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS): Process, format, and clinimetric testing plan. Mov Disord 22:41\u0026ndash;47\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGreenland JC, Williams-Gray CH, Barker RA (2019) The clinical heterogeneity of Parkinson's disease and its therapeutic implications. Eur J Neurosci 49:328\u0026ndash;338\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuttuso T Jr., Bergsland N, Hagemeier J, Lichter DG, Pasternak O, Zivadinov R (2018) Substantia Nigra Free Water Increases Longitudinally in Parkinson Disease. AJNR Am J Neuroradiol 39:479\u0026ndash;484\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeuken MC, Forstmann BU (2015) A probabilistic atlas of the basal ganglia using 7 T MRI. Data Brief 4:577\u0026ndash;582\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKing LA, Priest KC, Salarian A, Pierce D, Horak FB (2012) Comparing the Mini-BESTest with the Berg Balance Scale to Evaluate Balance Disorders in Parkinson's Disease. Parkinsons Dis 2012:375419\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKish SJ, Shannak K, Hornykiewicz O (1988) Uneven pattern of dopamine loss in the striatum of patients with idiopathic Parkinson's disease. Pathophysiologic and clinical implications. N Engl J Med 318:876\u0026ndash;880\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLo Bianco C, Ridet JL, Schneider BL, Deglon N, Aebischer P (2002) alpha -Synucleinopathy and selective dopaminergic neuron loss in a rat lentiviral-based model of Parkinson's disease. Proc Natl Acad Sci U S A 99:10813\u0026ndash;10818\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMarek K, Chowdhury S, Siderowf A, Lasch S, Coffey CS, Caspell-Garcia C, Simuni T, Jennings D, Tanner CM, Trojanowski JQ, Shaw LM, Seibyl J, Schuff N, Singleton A, Kieburtz K, Toga AW, Mollenhauer B, Galasko D, Chahine LM, Weintraub D, Foroud T, Tosun-Turgut D, Poston K, Arnedo V, Frasier M, Sherer T (2018) The Parkinson's progression markers initiative (PPMI) - establishing a PD biomarker cohort. Ann Clin Transl Neurol 5:1460\u0026ndash;1477\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMarek K, Jennings D, Lasch S, Siderowf A, Tanner C, Simuni T, Coffey C, Kieburtz K, Flagg E, Chowdhury S (2011) The Parkinson progression marker initiative (PPMI). Prog Neurobiol 95:629\u0026ndash;635\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOfori E, Pasternak O, Planetta PJ, Burciu R, Snyder A, Febo M, Golde TE, Okun MS, Vaillancourt DE (2015) Increased free water in the substantia nigra of Parkinson's disease: a single-site and multi-site study. Neurobiol Aging 36:1097\u0026ndash;1104\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOfori E, Pasternak O, Planetta PJ, Li H, Burciu RG, Snyder AF, Lai S, Okun MS, Vaillancourt DE (2015) Longitudinal changes in free-water within the substantia nigra of Parkinson's disease. Brain 138:2322\u0026ndash;2331\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePasternak O, Sochen N, Gur Y, Intrator N, Assaf Y (2009) Free water elimination and mapping from diffusion MRI. Magn Reson Med 62:717\u0026ndash;730\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePasternak O, Westin CF, Bouix S, Seidman LJ, Goldstein JM, Woo TU, Petryshen TL, Mesholam-Gately RI, McCarley RW, Kikinis R, Shenton ME, Kubicki M (2012) Excessive extracellular volume reveals a neurodegenerative pattern in schizophrenia onset. J Neurosci 32:17365\u0026ndash;17372\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePlanetta PJ, Ofori E, Pasternak O, Burciu RG, Shukla P, DeSimone JC, Okun MS, McFarland NR, Vaillancourt DE (2015) Free-water imaging in Parkinson\u0026rsquo;s disease and atypical parkinsonism. Brain 139:495\u0026ndash;508\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRegnault A, Boroojerdi B, Meunier J, Bani M, Morel T, Cano S (2019) Does the MDS-UPDRS provide the precision to assess progression in early Parkinson's disease? Learnings from the Parkinson's progression marker initiative cohort. J Neurol 266:1927\u0026ndash;1936\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchlachetzki JCM, Barth J, Marxreiter F, Gossler J, Kohl Z, Reinfelder S, Gassner H, Aminian K, Eskofier BM, Winkler J, Klucken J (2017) Wearable sensors objectively measure gait parameters in Parkinson's disease. PLoS ONE 12:e0183989\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShen CY, Tyan YS, Kuo LW, Wu CW, Weng JC (2015) Quantitative Evaluation of Rabbit Brain Injury after Cerebral Hemisphere Radiation Exposure Using Generalized q-Sampling Imaging. PLoS ONE 10:e0133001\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShippey LE, Campbell SG, Hill AF, Smith DP (2022) Propagation of Parkinson's disease by extracellular vesicle production and secretion. Biochem Soc Trans 50:1303\u0026ndash;1314\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVijayakumari AA, Parker D, Osmanlioglu Y, Alappatt JA, Whyte J, Diaz-Arrastia R, Kim JJ, Verma R (2021) Free Water Volume Fraction: An Imaging Biomarker to Characterize Moderate-to-Severe Traumatic Brain Injury. J Neurotrauma 38:2698\u0026ndash;2705\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVijayakumari AA, Sakaie KE, Fernandez HH, Walter BL (2025) Parkinson's disease subtypes and their association with probable rapid eye movement sleep behavior disorder severity: a brainstem tractography and machine learning investigation. Brain Imaging Behav 19:189\u0026ndash;194\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eW\u0026uuml;llner U, Borghammer P, Choe CU, Csoti I, Falkenburger B, Gasser T, Lingor P, Riederer P (2023) The heterogeneity of Parkinson's disease. J Neural Transm (Vienna) 130:827\u0026ndash;838\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeh F-C, Tang P-F, Tseng W-YI (2013) Diffusion MRI connectometry automatically reveals affected fiber pathways in individuals with chronic stroke. NeuroImage: Clin 2:912\u0026ndash;921\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeh FC, Badre D, Verstynen T (2016) Connectometry: A statistical approach harnessing the analytical potential of the local connectome. NeuroImage 125:162\u0026ndash;171\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeh FC, Tseng WY (2011) NTU-90: a high angular resolution brain atlas constructed by q-space diffeomorphic reconstruction. NeuroImage 58:91\u0026ndash;99\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeh FC, Verstynen TD, Wang Y, Fern\u0026aacute;ndez-Miranda JC, Tseng WY (2013) Deterministic diffusion fiber tracking improved by quantitative anisotropy. PLoS ONE 8:e80713\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeh FC, Wedeen VJ, Tseng WY (2010) Generalized q-sampling imaging. IEEE Trans Med Imaging 29:1626\u0026ndash;1635\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang H, He WJ, Liang LH, Zhang HW, Zhang XJ, Zeng L, Luo SP, Lin F, Lei Y (2021) Diffusion Spectrum Imaging of Corticospinal Tracts in Idiopathic Normal Pressure Hydrocephalus. Front Neurol 12:636518\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang N, Liu W, Ye M, Cohen AD, Zhang Y (2015) The heterogeneity of non-motor symptoms of Parkinson's disease. Neurol Sci 36:577\u0026ndash;584\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Machine learning, Parkinson’s disease, Diffusion MRI, Cluster analysis, Extracellular water, Isotropic diffusion","lastPublishedDoi":"10.21203/rs.3.rs-7880492/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7880492/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e\u003cp\u003eParkinson\u0026rsquo;s disease (PD) exhibits marked clinical and biological heterogeneity. This study aimed to identify neurobiologically defined PD subtypes using isotropic diffusion (ISO), a diffusion MRI metric sensitive to extracellular water, and to determine whether these subtypes differ in baseline motor profiles and longitudinal progression.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eBaseline ISO values were extracted from 12 subcortical motor regions in 156 de novo PD patients from the Parkinson\u0026rsquo;s Progression Markers Initiative. Hierarchical clustering was applied to ISO values to derive data-driven subtypes. Group differences in baseline motor severity were evaluated using t-tests, and linear mixed-effects models assessed longitudinal changes in ISO and motor scores over four years.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eTwo subtypes emerged: subtype 1 (n\u0026thinsp;=\u0026thinsp;62) with lower ISO values and subtype 2 (n\u0026thinsp;=\u0026thinsp;94) with higher ISO across all regions. Subtype 2 showed greater baseline rigidity and bradykinesia. Longitudinally, subtype 1 exhibited significant ISO increases, whereas subtype 2 remained stable at elevated levels. However, motor progression rates did not differ significantly between subtypes.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eDiffusion MRI-derived ISO identified distinct neurobiological subtypes of PD with divergent trajectories of extracellular pathology but similar clinical progression. ISO may serve as a sensitive biomarker for PD heterogeneity, warranting further validation in larger, long-term cohorts.\u003c/p\u003e","manuscriptTitle":"Data-Driven Neurobiological Subtyping of Parkinson’s Disease Using Diffusion MRI-Derived Isotropic Diffusion","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-21 23:56:35","doi":"10.21203/rs.3.rs-7880492/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7c45d34c-672f-4668-9764-25e1d79859b3","owner":[],"postedDate":"October 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-16T16:05:38+00:00","versionOfRecord":{"articleIdentity":"rs-7880492","link":"https://doi.org/10.1007/s00234-026-03939-4","journal":{"identity":"neuroradiology","isVorOnly":false,"title":"Neuroradiology"},"publishedOn":"2026-02-09 15:58:28","publishedOnDateReadable":"February 9th, 2026"},"versionCreatedAt":"2025-10-21 23:56:35","video":"","vorDoi":"10.1007/s00234-026-03939-4","vorDoiUrl":"https://doi.org/10.1007/s00234-026-03939-4","workflowStages":[]},"version":"v1","identity":"rs-7880492","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7880492","identity":"rs-7880492","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}