Cerebral microstructural alteration patterns in motor subtypes of Parkinson’s disease: a neurite orientation dispersion and density imaging study

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disease characterized by a range of motor symptoms such as bradykinesia, resting tremor, and postural instability, as well as various non-motor manifestations (1). The clinical presentations of PD are highly heterogeneous, and patients can be classified into the tremor-dominant (TD) subtype or postural instability and gait difficulty (PIGD) subtype based on the predominant motor symptoms. It has been reported that TD exhibits a less severe motor impairment compared to PIGD (2). Besides, patients with PIGD have a higher risk of developing cognitive impairments and neuropsychiatric abnormalities, whereas those of the TD subtype tend to have relatively mild non-motor symptoms (3,4). As such, it is vital to explore the underlying neuropathological mechanisms that account for differences in clinical manifestations between these two subtypes. A deeper understanding of these subtypes may provide further insights into the heterogeneity of PD symptoms.

Previous studies have explored the origins of differences between the two subtypes by identifying changes in white matter (WM) and gray matter (GM). Diffusion tensor imaging (DTI) provides parameters which reflect the integrity of WM in a noninvasive way, which is the most common approach employed in diffusion studies and plays a vital role in uncovering neural changes in motor subtypes (5). TD patients have been shown to exhibit higher fractional anisotropy (FA) compared to PIGD patients in projection, association, and commissural tracts, which may contribute to the more severe symptoms and poorer prognosis observed in PIGD (6). For GM, research using voxel-based morphometry (VBM) has found widespread GM reduction in PIGD compared to TD, particularly in areas involving motor, cognitive, limbic, and associative functions (7). Additionally, studies have shown distinct cortical gyrification patterns, cortical surface area, and sulcal depth in motor subtypes (8,9). However, the results of previous studies have been inconsistent, which might be attributed to distinct sample sizes or methodological limitations. DTI assumes that the diffusion of water in the brain is unrestricted and follows a Gaussian distribution, which fails to reflect the microstructure specifically. It has been shown to be poorly sensitive to alterations such as axonal density, axonal direction dispersion, or myelination (10); the conventional structural magnetic resonance imaging (MRI) demonstrates relatively low sensitivity to early or micro impairments in GM, which may lead to some insignificant macrostructural differences in previous studies.

To reflect the microstructural changes specifically, neurite orientation dispersion and density imaging (NODDI) (11), an advanced multi-compartment diffusion model, has been applied in studies of some diseases, such as multiple sclerosis and Alzheimer’s disease (12,13). In NODDI, the neurite density index (NDI) represents the density of neurites, including both axons and dendrites; the orientation dispersion index (ODI) characterizes the bending and fanning of axons in WM as well as the extent of dendritic branching in GM; and the volume fraction of isotropic diffusion (fiso) represents the fraction of cerebrospinal fluid (CSF). At present, few studies using NODDI have detected the alterations in WM and GM in motor subtypes of PD.

We hypothesized that microstructural alterations in WM and GM detected by NODDI could comprehensively reveal neurostructural differences of diverse clinical presentations between TD and PIGD. Our goal was to investigate distinct cerebral microstructural patterns in TD and PIGD and identify clinically relevant biomarkers. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-34/rc).

Methods

Participants

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the institutional ethical committee of The First Affiliated Hospital of Sun Yat-sen University (No. [2023]134) and informed consent was provided by all participants. Seventy-four PD patients were included from The First Affiliated Hospital of Sun Yat-sen University between March 2021 and March 2024. All of the PD patients were diagnosed according to the Movement Disorder Society (MDS) clinical diagnostic criteria for PD (14), and were classified according to their MDS-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) (15). Based on the MDS-UPDRS-III subitems (MDS-UPDRS TD/PIGD ratio), patients with PD were categorized into TD, PIGD, or intermediate groups (where the TD/PIGD ratio was ≥1.15 for TD, ≤0.9 for PIGD, and >0.9 and <1.15 for the intermediate group). The exclusion criteria in this study were as follows: (I) patients classified into the intermediate group; (II) cases with a history of other neurologic or psychiatric disorders, brain trauma, or severe cerebrovascular diseases; and (III) a Mini-Mental State Examination (MMSE) score <24. Furthermore, 47 healthy controls (HCs) were enrolled during the same period, matched to the patients in both subgroups for age and sex.

All patients took part in motor and nonmotor assessments in the “drug-off” state, and the relevant clinical data were collected, including duration of disease, MDS-UPDRS, Hoehn & Yahr stage, levodopa equivalent daily dose (LEDD), MMSE, Montreal Cognitive Assessment (MoCA), Hamilton Anxiety Scale (HAMA), and Hamilton Depression Scale (HAMD). Besides, age, sex, and years of education were recorded for all participants.

MRI acquisition

All cases were scanned on a 3.0 Tesla MR imaging system (MAGNETOM Prisma, Siemens Healthineers, Erlangen, Germany) equipped with a 64-channel head coil. Foam pads and ear plugs were used to reduce head motion and noise. The echo planar imaging (EPI) diffusion-weighted data were acquired along 64 gradient directions, with three b values for each direction (b value =0, 1,000, 2,000 s/mm²). The sequence parameters were as follows: repetition time (TR) =2,400 msec; echo time (TE) =83 msec; field of view (FOV) =220×220 mm²; matrix =88×88; slice thickness =2.5 mm; slice gap =0 mm; and acquisition time, 7 min 59 s.

Diffusion MRI preprocessing

All data were visually checked in all three orthogonal (axial, sagittal, and coronal) views and were converted to Neuroimaging Informatics Technology Initiative (NIfTI) format. The diffusion images were preprocessed using FMRIB Software Library (FSL 6.0.5; https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/), including removing the Gibbs artifacts, brain extraction, and the correction with regard to eddy current distortion and head motion (16). Then, we utilized the accelerated microstructure imaging via convex optimization (AMICO; https://github.com/daducci/AMICO.wiki.git) technique with improved processing time to obtain NODDI-derived metrics (NDI, ODI, and fiso) for all participants (17).

Tract-based spatial statistics (TBSS)

We used the TBSS (18) to analyze WM. Firstly, the FMRIB non-linear registration tool was used to align all FA maps to the standard Montreal Neurological Institute (MNI) space, and all FA maps in the standard space were averaged. Next, we created the mean FA skeleton using a threshold of 0.2 to exclude peripheral tracts and GM, and include major WM pathways. Finally, the FA maps and the NODDI-derived maps of each participant were projected onto the mean FA skeleton.

GM-based spatial statistics (GBSS)

GBSS (19) is a statistical pipeline analog of the TBSS framework, and it can perform voxel-wise analysis on GM. Firstly, we obtained the CSF fraction from the fiso maps and WM fraction estimated by two-tissue class segmentation of FA maps using the Atropos segmentation tool in Advanced Normalization Tools (ANTs) (20). Then, GM fraction maps were estimated in the native diffusion space by subtracting CSF fraction and WM fraction. Each tissue segmentation map was then multiplied by their respective tissue weighting (CSF =0, GM =1, WM =2) to increase tissue contrast and summed together to generate “pseudo T1-weighted” images. Next, the “pseudo T1-weighted” images were used to build a study-specific template using ANTs. Native diffusion space NODDI parameter maps (NDI and ODI) and GM fraction maps were warped to the template space by applying the corresponding warp fields from the previous step. The average GM fraction maps were skeletonized, and the final skeleton was kept to only include voxels with GM fraction >0.65 in >70% of participants. For each case, all NODDI parameter maps and GM fraction maps were projected onto the GM skeleton. The remaining voxels on the participants’ skeletons with non-satisfactory GM fraction (≤0.65) were filled with the average of the surrounding satisfactory voxels on the skeleton (GM fraction >0.65) weighted by their closeness with a Gaussian kernel (default: σ=2 mm).

Statistical analysis

Statistical analyses for demographic and clinical data were performed using the software SPSS 26.0 (IBM Corp., Armonk, NY, USA). The Shapiro-Wilk test was used for analysis of normal distribution. Demographic data were analyzed by one-way analysis of variance (ANOVA) or Kruskal-Wallis test among three groups. Difference in sex distribution was compared using Chi-squared test. Clinical variables were compared between the two subtypes using independent-samples Student’s t-test or Mann-Whitney U test. The level of statistical significance was set to P<0.05.

For both TBSS and GBSS analyses, inter-group comparisons were conducted (TD vs. HCs, PIGD vs. HCs, and TD vs. PIGD), adjusted for age, sex, and years of education. The general linear model (GLM) was fitted for each voxel in all the skeletonized NODDI parameter maps. The FSL’s randomize tool with 5,000 permutations was conducted for groupwise statistical comparisons. To correct the multiple comparisons, we used threshold-free cluster enhancement (TFCE), which controlled family-wise error (FWE) rate, and voxels with a TFCE-corrected P_FWE<0.05 were considered statistically significant. Furthermore, in each motor subtype, voxel-wise correlational analyses were conducted to investigate the relationships between NODDI metrics and relevant clinical data (including MDS-UPDRS III, Hoehn & Yahr stage, MMSE, MoCA, HAMA, and HAMD) adjusted for age, sex, and years of education. The same multiple comparisons correction method and statistical significance threshold used in group comparisons were consistently employed in correlation analyses.

Results

Demographic and clinical characteristics

According to MDS-UPDRS III subitems, 49 PD patients were classified as PIGD, and 25 PD patients were classified as TD. The demographic and clinical characteristics of all participants for TBSS and GBSS are presented in Table 1. There was no significant difference in age, sex distribution, and years of education among each subtype and the HCs group. For patients with PD, individuals with TD showed a significantly higher tremor score, and inversely, those with PIGD showed a significantly higher PIGD score. Compared with TD, PIGD had higher Hoehn & Yahr stage. MDS-UPDRS III score, disease duration, LEDD, MMSE, MoCA, and HAMD score did not significantly differ between the motor subtypes. Additionally, the participants included for TBSS exhibited that the PIGD group showed higher HAMA scores than did the TD group.

TBSS analysis

After visually checking by an experienced neuro-radiologist, we excluded one TD patient and 13 PIGD patients as their WM in brainstem and cerebellum were not fully covered. Ultimately, 24 TD patients, 36 PIGD patients, and 47 HCs were included for TBSS analysis (Figure 1).

Figure 1 A flowchart of the subject inclusion process. fiso, the volume fraction of isotropic diffusion; GBSS, gray matter-based spatial statistics; HC, healthy control; MDS-UPDRS, Movement Disorder Society-Unified Parkinson’s Disease Rating Scale; NDI, neurite density index; NODDI, neurite orientation dispersion and density imaging; ODI, orientation dispersion index; PD, Parkinson’s disease; PIGD, postural instability and gait difficulty; TBSS, tract-based spatial statistics; TD, tremor-dominant.

In the TBSS analysis, patients with PIGD showed increased ODI in right superior and posterior corona radiata compared to those with TD (P_FWE<0.05) (Figure 2A). Besides, the PIGD group showed lower FA mainly in left posterior thalamic radiation and right superior corona radiata (Figure 2A).

Figure 2 Microstructural differences in WM from TBSS analysis. PIGD demonstrated increased ODI of corona radiata than TD, with decreased NDI in several tracts than HCs. (A) WM displaying significantly lower FA (colored purple) and higher ODI (colored green) in PIGD compared to TD. (B) Significantly reduced FA and NDI (colored blue) in PIGD compared to HCs (P_FWE<0.05). FA, fractional anisotropy; FWE, family-wise error; HC, healthy control; NDI, neurite density index; ODI, orientation dispersion index; PIGD, postural instability and gait difficulty; TBSS, tract-based spatial statistics; TD, tremor-dominant; WM, white matter.

Compared with HCs, the PIGD group showed widespread decreased FA mainly in bilateral corticospinal tracts, internal capsule, corona radiata, corpus callosum, left external capsule, bilateral superior longitudinal fasciculus, fornix, and bilateral superior, inferior, and middle cerebellar peduncles (P_FWE<0.05) (Figure 2B). Furthermore, the PIGD group showed reduced NDI in regions mainly covering bilateral corona radiata, corpus callosum, posterior thalamic radiation, and bilateral superior longitudinal fasciculus (P_FWE<0.05) (Figure 2B) (specific information in Table S1). No significant differences in WM were found between TD and HCs.

GBSS analysis

According to the result of visual assessment, two additional TD patients and one PIGD patient were excluded based on the samples in TBSS for incomplete coverage of the cerebral cortex; one additional PIGD patient and one HC were excluded based on the samples in TBSS for “pseudo T1-weighted” images with poor quality. Ultimately, 22 TD patients, 34 PIGD patients, and 46 HCs were included in the GBSS analysis (Figure 1).

There was no significant difference found in GBSS between these two PD subtypes.

Compared with HCs, patients with PIGD exhibited widespread reduced NDI in bilateral temporal lobes (including bilateral superior, middle, and inferior temporal gyrus, hippocampus, parahippocampal gyrus, fusiform gyrus, and amygdala) and bilateral middle and inferior occipital gyrus. Besides, PIGD showed a decreased ODI in bilateral lingual gyrus, right occipital gyrus, hippocampus, parahippocampal, and fusiform gyrus (P_FWE<0.05) (Figure 3A). Patients with TD exhibited reduced NDI, mainly in bilateral temporal lobes (including bilateral inferior temporal gyrus, left middle and superior temporal gyrus, bilateral hippocampus, and parahippocampal gyrus), left insula, left inferior frontal gyrus, and right inferior occipital gyrus when compared to HCs (Figure 3B) (specific information in Table S2).

Figure 3 Microstructural differences in GM from GBSS analysis. Both subtypes showed distinct microstructural changes relative to HCs. (A) GM demonstrating significantly decreased NDI and ODI (colored blue) in PIGD than HCs. (B) Significantly decreased NDI (colored blue) in TD than HCs (P_FWE<0.05). FWE, family-wise error; GBSS, gray matter-based spatial statistics; GM, gray matter; HC, healthy control; NDI, neurite density index; ODI, orientation dispersion index; PIGD, postural instability and gait difficulty; TD, tremor-dominant.

Correlations between alterations and clinical data

For patients with the PIGD subtype, we found negative correlations between the ODI of several WM tracts and MMSE (r=−0.66, P_FWE<0.001). These were clustered mainly in right hemisphere, involving corpus callosum, internal capsule, corona radiata, posterior thalamic radiation, and superior longitudinal fasciculus (Figure 4A). In those with the TD subtype, multiple WM tracts showed positive correlations between NDI and MoCA (r=0.71, P_FWE<0.001). These were clustered mainly bilaterally involved corpus callosum, bilateral internal capsule, corona radiata, posterior thalamic radiation, and superior longitudinal fasciculus (Figure 4B) (specific information in Table S1). Moreover, we also found a positive correlation between the NDI of left putamen and MoCA (r=0.71, P_FWE=0.001) (Figure 5) (specific information in Table S2).

Figure 4 Correlations between microstructural alterations of WM and clinical data in two subtypes. (A) In PIGD, the significant negative correlation was identified between ODI of several WM fiber tracts and MMSE. (B) In TD, the significant positive correlation was identified between NDI of multiple WM fiber tracts and MoCA. MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; NDI, neurite density index; ODI, orientation dispersion index; PIGD, postural instability and gait difficulty; TD, tremor-dominant; WM, white matter.

Figure 5 A correlation between microstructural alteration of putamen and MoCA in TD. In TD, a significant positive correlation was identified between NDI of putamen and MoCA. MoCA, Montreal Cognitive Assessment; NDI, neurite density index; TD, tremor-dominant.

Discussion

Our study systematically evaluated the cerebral microstructural differences between the two subtypes of TD and PIGD using NODDI. In the TBSS analysis, the right superior and posterior corona radiata of patients with PIGD showed a higher ODI than those with TD. Furthermore, the FA and NDI of multiple WM fiber tracts in the PIGD group were lower than those in HCs. The GBSS analysis indicated that compared to HCs, the patterns of alterations differed between the two subtypes. However, the NDI and ODI in bilateral temporal lobes of both motor subtypes were decreased. The application of NODDI enables the identification of more specific microstructural differences, which can exhibit a wider range of alterations in both WM and GM, especially in PIGD, supporting the clinical observations that the PIGD subtype tends to bear a higher risk of developing non-motor abnormalities. Additionally, we observed significant correlations between NODDI metrics of multiple WM tracts and cognitive score in both subtypes and the NDI of putamen also exhibited significant positive correlations with MoCA in the TD group.

WM microstructural alterations and clinical correlations

This study revealed lower FA of several WM tracts but higher ODI of corona radiata in the PIGD group compared to the TD group. FA is a commonly used indicator to depict the integrity of WM, which can be affected by myelination, axonal density, and fiber organization, and the ODI characterizes the extent of dispersion of WM fiber orientation. A reduction in FA can be accompanied by an increase of ODI (21,22). Our study adopted a more specific approach for exploring the microstructural damage, instead of using the more general description of “impairment of WM integrity”. A higher ODI of corona radiata reflected more severe fiber bundle impairment in PIGD patients. The corona radiata contains ascending and descending projection fibers from subcortical nuclei to the cerebral cortex, as well as part of the association and callosal fibers (23). This structure serves as the neuroanatomical backbone for perceptual and motor functions, as well as higher cognitive functions (24). The impairment of corona radiata is related to gait abnormality, which is probably due to disruption of the projecting fibers connecting to the motor-related cortex and sensory deficits (6,25). The current results supported the theory that the corona radiata may play a role in the heterogeneity of symptoms between the two subtypes.

Compared to HCs, the PIGD group demonstrated a significant reduction of neurite density in multiple WM tracts reflected by NDI. The TD group did not show any difference when compared with HCs. Studies detecting WM lesions on fluid-attenuated inversion recovery (FLAIR) images (26) and those using DTI (5,27) have found that PIGD has more significant WM lesions than TD. Our study revealed that the impairment of WM tracts may manifest as a decrease in fiber bundle density and may explain the poorer prognosis and the presence of non-motor symptoms in PIGD. Besides, the absence of difference between TD and HCs also aligned with the previous conclusion that the TD subtype had less impairment in neural circuits and had potential neural compensation, which probably made the course of disease more benign relative to PIGD (6,28,29). Additionally, several studies have linked PD heterogeneity to genetic mutations. For instance, G2385R LRRK2 patients tend to show faster motor progression and cognitive decline, whereas G2019S LRRK2 patients often demonstrate milder symptom progression, with their PIGD subtype not correlating with higher UPDRS scores (30,31). Furthermore, a previous DTI study indicated preserved corticospinal tract compensation in monogenic PD patients (32). Our observed WM microstructural differences may reflect genetic influences in motor subtypes, although our study did not include genetic analysis. Future studies of genotype-specific variations between subtypes may yield novel insights.

Additionally, the ODI of several WM tracts was negatively correlated with MMSE in the PIGD subtype, and the NDI of multiple WM tracts was positively correlated with MoCA in the TD subtype. The relevant regions involve multiple projection, association, and callosal fibers. As the largest fiber bundle in cerebrum, corpus callosum is implicated in cognition, executive function, and attention (33,34). The superior longitudinal fasciculus constitutes the WM fiber tract connecting multiple lobes and is capable of transmitting information via the cortico-cortical circuit, which exerts a crucial role in function of cognition, pre-motor, motor, visual-spatial, and auditory (35,36). The pattern of alterations in projection and association fibers has been reported in previous studies about cognition in PD (37,38). Interestingly, we found that the WM regions showing significant correlations with MoCA and MMSE scores differed in our analysis. The MoCA assesses multiple cognitive domains (e.g., executive function, visuospatial skills, and attention) involving broad regions including but not limited to frontal and occipital lobes, consistent with the widely distributed WM correlates observed in the TD group of our study. Compared to the MMSE, the MoCA is likely to be more suitable as a screening instrument for cognitive impairment due to its higher specificity and lack of ceiling effects (39), which accounts for the observed differences in correlational analyses between the two assessment methods. Moreover, the area of NODDI metrics related to cognition was broader in TD. The more severe loss of fiber bundle density in PIGD likely resulted in a weaker cognitive correlation, and this association may be better reflected in the dispersion of the fiber bundles indicated by ODI. In contrast, the damage to fibers in TD was relatively minor, leading to a more evident correlation with cognition. Overall, our findings suggested that NODDI may be a promising biomarker in WM for assessing cognitive status across different subtypes.

GM microstructural alterations and clinical correlations

In GBSS analysis, we did not discover any microstructural distinction between these two subtypes, but each subtype had alterations of GM when respectively compared to HCs. During the progression of the PD, neurons or subcortical fiber bundles may be affected by abnormal pathological processes (such as aberrant α-synuclein aggregation, dysfunction of mitochondria, lysosomes or vesicle transport, synaptic transport issues, and neuroinflammation) (1) and lead to microstructural damage in GM. Several previous studies have reported macrostructural differences in GM among different motor subtypes, but few studies have investigated microstructural alteration in GM (7,9,40). The NDI represents the density of neurites including axons and dendrites, and it can reflect the dendritic density in cortex and subcortical nuclei. The extensive NDI reduction we found in the cortex of both subtypes might reflect decreased dendritic density due to the effects of abnormal pathological processes in PD.

Both subtypes showed lower NDI of the temporal (including hippocampus) and occipital lobes, with PIGD having a more extensive involvement. The temporal lobe is associated with cognitive performance, motor skills, attention, and information processing speed (41,42), among which the hippocampus and parahippocampal gyrus play key roles in cognition and psychiatric symptoms (43,44). The temporal lobe has been found to atrophy in PD patients in previous studies, and some have reported that the volume of the temporal lobe is correlated with the PIGD score (40). Also, the reduced NDI we found in the occipital lobe probably explained the visual impairments (including color vision, contrast sensitivity, and difficulty completing complex visual tasks) (45,46), cognitive impairments, and frailty in PD patients (47). Increased cortical lesions may help to explain why patients with the PIGD subtype experience more severe non-motor symptoms compared to those with the TD subtype. Interestingly, we found abnormalities of the inferior frontal gyrus and insula in TD. The inferior frontal gyrus triangular part is involved in the regulation of unconscious perceptual processing and information processing (48), and the frontal regions are predominantly associated with the insula (49). Although we did not find abnormalities in the frontal and insular cortices in the PIGD group, we queried whether these areas, as well as the temporal lobes, contribute to non-motor symptoms in PD, and we hypothesized that the distribution of areas influencing non-motor symptoms may differ among the various subtypes.

In TD, we observed a positive correlation between the NDI of the left putamen and MoCA. Patients with PD exhibit relatively substantial iron deposition and degeneration in the basal ganglia nucleus (50,51), and previous studies have reported that the lesion of the putamen contributes to the deficits of neurocognition (52,53). Such impairments may be manifested as the loss of neuronal synapses prior to visible alteration of macrostructure, which may only become apparent upon further progression of the underlying pathology. The changes of NDI in basal ganglia could be reflective of underlying neuroanatomical impairment, specifically loss of neurite and neuronal fiber projections of this brain region. The observed correlation substantiated that the NDI of putamen might serve as a potential marker for the cognitive impairments in TD subtype.

Some limitations of our study should be noted. First, this was a cross-sectional study. The dynamic changes in WM and GM of the two subtypes during the course of PD remained unknown. Second, the presence of an unbalanced sample size between these two subtypes might have contributed to non-significant results in the current study, necessitating larger longitudinal studies with more evenly distributed samples. Third, most of the patients in the current study were defined to be in Hoehn & Yahr stage 2 or above, and the average duration of disease was about 8 years. Further studies including patients with earlier stages of PD will be better placed to interpret the microstructure distinction in de novo, untreated, or preclinical stages of PD.

Conclusions

Both motor subtypes showed abnormal microstructure in WM and GM, with a more extensive range of impairment in PIGD. Our findings supported a more benign course of non-motor symptoms and prognosis in TD than PIGD. Furthermore, the microstructural changes were associated with cognitive function, suggesting the potential of NODDI-derived metrics as biomarkers for cognitive symptoms in subtypes of PD. Overall, understanding these microstructural patterns can enhance our comprehension of the heterogeneous symptomatology observed in PD.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-34/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-34/dss

Funding: This study was supported by the National Natural Science Foundation of China (No. NSFC 82172015) and the Guangdong Basic and Applied Basic Research Foundation (No. 2022A1515011264).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-34/coif). C.Z. is an employee from Siemens Healthcare, and in the current study, he provided technical support, like post-processing, etc. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the institutional ethical committee of The First Affiliated Hospital of Sun Yat-sen University (No. [2023]134), and informed consent was obtained for all study subjects.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bloem BR, Okun MS, Klein C. Parkinson's disease. Lancet 2021;397:2284-303. [Crossref] [PubMed]
2. Jankovic J, McDermott M, Carter J, Gauthier S, Goetz C, Golbe L, Huber S, Koller W, Olanow C, Shoulson I, Stern M, Tanner C, Weiner WParkinson Study Group. Variable expression of Parkinson's disease. Neurology 1990;40:1529. [Crossref] [PubMed]
3. Alves G, Larsen JP, Emre M, Wentzel-Larsen T, Aarsland D. Changes in motor subtype and risk for incident dementia in Parkinson's disease. Mov Disord 2006;21:1123-30. [Crossref] [PubMed]
4. Burn DJ, Rowan EN, Allan LM, Molloy S, O'Brien JT, McKeith IG. Motor subtype and cognitive decline in Parkinson's disease, Parkinson's disease with dementia, and dementia with Lewy bodies. J Neurol Neurosurg Psychiatry 2006;77:585-9. [Crossref] [PubMed]
5. Maetzler W, Vervoort G, Leunissen I, Firbank M, Heremans E, Nackaerts E, Vandenberghe W, Nieuwboer A. Structural Brain Alterations in Motor Subtypes of Parkinson's Disease: Evidence from Probabilistic Tractography and Shape Analysis. PLoS One 2016;11:e0157743. [Crossref] [PubMed]
6. Wen MC, Heng HSE, Lu Z, Xu Z, Chan LL, Tan EK, Tan LCS. Differential White Matter Regional Alterations in Motor Subtypes of Early Drug-Naive Parkinson's Disease Patients. Neurorehabil Neural Repair 2018;32:129-41. [Crossref] [PubMed]
7. Rosenberg-Katz K, Herman T, Jacob Y, Giladi N, Hendler T, Hausdorff JM. Gray matter atrophy distinguishes between Parkinson disease motor subtypes. Neurology 2013;80:1476-84. [Crossref] [PubMed]
8. Li J, Zhang Y, Huang Z, Jiang Y, Ren Z, Liu D, Zhang J, La Piana R, Chen Y. Cortical and subcortical morphological alterations in motor subtypes of Parkinson's disease. NPJ Parkinsons Dis 2022;8:167. [Crossref] [PubMed]
9. Tang X, Zhang Y, Liu D, Hu Y, Jiang L, Zhang J. Association of Gyrification Pattern, White Matter Changes, and Phenotypic Profile in Patients With Parkinson Disease. Neurology 2021;96:e2387-94. [Crossref] [PubMed]
10. Qiu A, Mori S, Miller MI. Diffusion tensor imaging for understanding brain development in early life. Annu Rev Psychol 2015;66:853-76. [Crossref] [PubMed]
11. Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. Neuroimage 2012;61:1000-16. [Crossref] [PubMed]
12. Parker TD, Slattery CF, Zhang J, Nicholas JM, Paterson RW, Foulkes AJM, Malone IB, Thomas DL, Modat M, Cash DM, Crutch SJ, Alexander DC, Ourselin S, Fox NC, Zhang H, Schott JM. Cortical microstructure in young onset Alzheimer's disease using neurite orientation dispersion and density imaging. Hum Brain Mapp 2018;39:3005-17. [Crossref] [PubMed]
13. Sacco S, Caverzasi E, Papinutto N, Cordano C, Bischof A, Gundel T, et al. Neurite Orientation Dispersion and Density Imaging for Assessing Acute Inflammation and Lesion Evolution in MS. AJNR Am J Neuroradiol 2020;41:2219-26. [Crossref] [PubMed]
14. Postuma RB, Berg D, Stern M, Poewe W, Olanow CW, Oertel W, Obeso J, Marek K, Litvan I, Lang AE, Halliday G, Goetz CG, Gasser T, Dubois B, Chan P, Bloem BR, Adler CH, Deuschl G. MDS clinical diagnostic criteria for Parkinson's disease. Mov Disord 2015;30:1591-601. [Crossref] [PubMed]
15. Stebbins GT, Goetz CG, Burn DJ, Jankovic J, Khoo TK, Tilley BC. How to identify tremor dominant and postural instability/gait difficulty groups with the movement disorder society unified Parkinson's disease rating scale: comparison with the unified Parkinson's disease rating scale. Mov Disord 2013;28:668-70. [Crossref] [PubMed]
16. Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage 2012;62:782-90. [Crossref] [PubMed]
17. Daducci A, Canales-Rodríguez EJ, Zhang H, Dyrby TB, Alexander DC, Thiran JP. Accelerated Microstructure Imaging via Convex Optimization (AMICO) from diffusion MRI data. Neuroimage 2015;105:32-44. [Crossref] [PubMed]
18. Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, Watkins KE, Ciccarelli O, Cader MZ, Matthews PM, Behrens TE. Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 2006;31:1487-505. [Crossref] [PubMed]
19. Nazeri A, Mulsant BH, Rajji TK, Levesque ML, Pipitone J, Stefanik L, Shahab S, Roostaei T, Wheeler AL, Chavez S, Voineskos AN. Gray Matter Neuritic Microstructure Deficits in Schizophrenia and Bipolar Disorder. Biol Psychiatry 2017;82:726-36. [Crossref] [PubMed]
20. Avants BB, Tustison NJ, Wu J, Cook PA, Gee JC. An open source multivariate framework for n-tissue segmentation with evaluation on public data. Neuroinformatics 2011;9:381-400. [Crossref] [PubMed]
21. Yang L, Li J, Huang C, Du Y, Li C, Huang B, Hou F, Zhao L, Guo H, Hu J, Ouyang X, Liu J. Altered orientation dispersion index of white matter in individuals with insomnia during the COVID-19 pandemic: A study combining neuroimaging technique and Mendelian randomization. Sleep Med 2024;114:167-77. [Crossref] [PubMed]
22. Zhao X, Shi J, Dai F, Wei L, Zhang B, Yu X, Wang C, Zhu W, Wang H. Brain Development From Newborn to Adolescence: Evaluation by Neurite Orientation Dispersion and Density Imaging. Front Hum Neurosci 2021;15:616132. [Crossref] [PubMed]
23. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology 2004;230:77-87. [Crossref] [PubMed]
24. Schmahmann JD, Pandya DN. Disconnection syndromes of basal ganglia, thalamus, and cerebrocerebellar systems. Cortex 2008;44:1037-66. [Crossref] [PubMed]
25. Wang M, Jiang S, Yuan Y, Zhang L, Ding J, Wang J, Zhang J, Zhang K, Wang J. Alterations of functional and structural connectivity of freezing of gait in Parkinson’s disease. J Neurol 2016;263:1583-92. [Crossref] [PubMed]
26. Bohnen NI, Müller ML, Zarzhevsky N, Koeppe RA, Bogan CW, Kilbourn MR, Frey KA, Albin RL. Leucoaraiosis, nigrostriatal denervation and motor symptoms in Parkinson's disease. Brain 2011;134:2358-65. [Crossref] [PubMed]
27. Gu Q, Huang P, Xuan M, Xu X, Li D, Sun J, Yu H, Wang C, Luo W, Zhang M. Greater loss of white matter integrity in postural instability and gait difficulty subtype of Parkinson's disease. Can J Neurol Sci 2014;41:763-8. [Crossref] [PubMed]
28. Helmich RC, Hallett M, Deuschl G, Toni I, Bloem BR. Cerebral causes and consequences of parkinsonian resting tremor: a tale of two circuits? Brain 2012;135:3206-26. [Crossref] [PubMed]
29. Paulus W, Jellinger K. The neuropathologic basis of different clinical subgroups of Parkinson's disease. J Neuropathol Exp Neurol 1991;50:743-55. [Crossref] [PubMed]
30. Marras C, Alcalay RN, Caspell‐Garcia C, Coffey C, Chan P, Duda JE, Facheris MF, Fernández‐Santiago R, Ruíz‐Martínez J, Mestre T, Saunders‐Pullman R, Pont‐Sunyer C, Tolosa E, Waro BLRRK2 Cohort Consortium. Motor and nonmotor heterogeneity of LRRK2‐related and idiopathic Parkinson's disease. Mov Disord 2016;31:1192-202. [Crossref] [PubMed]
31. Nabli F, Ben Sassi S, Amouri R, Duda JE, Farrer MJ, Hentati F. Motor phenotype of LRRK2‐associated Parkinson's disease: A tunisian longitudinal study. Mov Disord 2015;30:253-8. [Crossref] [PubMed]
32. Dayan R, Bick AS, Weill C, Bauer M, Erdman HB, Israel Z, Bergman H, Levin N, Arkadir D. Atrophy-related corticospinal changes in advanced Parkinson's disease are associated with the genetic etiology of the disease. J Parkinsons Dis 2024;14:1584-93. [Crossref] [PubMed]
33. Di Paola M, Luders E, Di Iulio F, Cherubini A, Passafiume D, Thompson PM, Caltagirone C, Toga AW, Spalletta G. Callosal atrophy in mild cognitive impairment and Alzheimer's disease: different effects in different stages. Neuroimage 2010;49:141-9. [Crossref] [PubMed]
34. Zheng Z, Shemmassian S, Wijekoon C, Kim W, Bookheimer SY, Pouratian N. DTI correlates of distinct cognitive impairments in Parkinson's disease. Hum Brain Mapp 2014;35:1325-33. [Crossref] [PubMed]
35. Nagano-Saito A, Washimi Y, Arahata Y, Kachi T, Lerch JP, Evans AC, Dagher A, Ito K. Cerebral atrophy and its relation to cognitive impairment in Parkinson disease. Neurology 2005;64:224-9. [Crossref] [PubMed]
36. Martino J, De Lucas EM. Subcortical anatomy of the lateral association fascicles of the brain: A review. Clin Anat 2014;27:563-9. [Crossref] [PubMed]
37. Liao TW, Wang JJ, Tsai CC, Wang PN, Chen YL, Wu YM, Wu YR. A fixel-based analysis of white matter reductions early detects Parkinson disease with mild cognitive impairment. Biomed J 2024;47:100678. [Crossref] [PubMed]
38. Duncan GW, Firbank MJ, Yarnall AJ, Khoo TK, Brooks DJ, Barker RA, Burn DJ, O'Brien JT. Gray and white matter imaging: A biomarker for cognitive impairment in early Parkinson's disease? Mov Disord 2016;31:103-10. [Crossref] [PubMed]
39. Hoops S, Nazem S, Siderowf AD, Duda JE, Xie SX, Stern MB, Weintraub D. Validity of the MoCA and MMSE in the detection of MCI and dementia in Parkinson disease. Neurology 2009;73:1738-45. [Crossref] [PubMed]
40. Yin L, Zhu Z, Fu J, Zhou C, Liu Z, Li Y, Luo Z, Zhu Y, Xu Z, Yang X. Differences in gray matter atrophy and functional connectivity between motor subtypes of Parkinson's disease. Acta Neurol Belg 2024;124:1899-911. [Crossref] [PubMed]
41. Vikene K, Skeie GO, Specht K. Compensatory task-specific hypersensitivity in bilateral planum temporale and right superior temporal gyrus during auditory rhythm and omission processing in Parkinson's disease. Sci Rep 2019;9:12623. [Crossref] [PubMed]
42. Achiron A, Chapman J, Tal S, Bercovich E, Gil H, Achiron A. Superior temporal gyrus thickness correlates with cognitive performance in multiple sclerosis. Brain Struct Funct 2013;218:943-50. [Crossref] [PubMed]
43. Chung SJ, Park YH, Yoo HS, Lee YH, Ye BS, Sohn YH, Lee JM, Lee PH. Mild cognitive impairment reverters have a favorable cognitive prognosis and cortical integrity in Parkinson's disease. Neurobiol Aging 2019;78:168-77. [Crossref] [PubMed]
44. Christopher L, Duff-Canning S, Koshimori Y, Segura B, Boileau I, Chen R, Lang AE, Houle S, Rusjan P, Strafella AP. Salience network and parahippocampal dopamine dysfunction in memory-impaired Parkinson disease. Ann Neurol 2015;77:269-80. [Crossref] [PubMed]
45. Weil RS, Schrag AE, Warren JD, Crutch SJ, Lees AJ, Morris HR. Visual dysfunction in Parkinson's disease. Brain 2016;139:2827-43. [Crossref] [PubMed]
46. Zarkali A, McColgan P, Leyland LA, Lees AJ, Weil RS. Visual Dysfunction Predicts Cognitive Impairment and White Matter Degeneration in Parkinson's Disease. Mov Disord 2021;36:1191-202. [Crossref] [PubMed]
47. Chen YS, Chen HL, Lu CH, Chen MH, Chou KH, Tsai NW, Yu CC, Chiang PL, Lin WC. Reduced lateral occipital gray matter volume is associated with physical frailty and cognitive impairment in Parkinson's disease. Eur Radiol 2019;29:2659-68. [Crossref] [PubMed]
48. Shi J, Huang H, Jiang R, Mao X, Huang Q, Li A. The Right Inferior Frontal Gyrus Plays an Important Role in Unconscious Information Processing: Activation Likelihood Estimation Analysis Based on Functional Magnetic Resonance Imaging. Front Neurosci 2022;16:781099. [Crossref] [PubMed]
49. Kann S, Zhang S, Manza P, Leung HC, Li CR. Hemispheric Lateralization of Resting-State Functional Connectivity of the Anterior Insula: Association with Age, Gender, and a Novelty-Seeking Trait. Brain Connect 2016;6:724-34. [Crossref] [PubMed]
50. Ye FQ, Allen PS, Martin WR. Basal ganglia iron content in Parkinson's disease measured with magnetic resonance. Mov Disord 1996;11:243-9. [Crossref] [PubMed]
51. Thomas GEC, Leyland LA, Schrag AE, Lees AJ, Acosta-Cabronero J, Weil RS. Brain iron deposition is linked with cognitive severity in Parkinson's disease. J Neurol Neurosurg Psychiatry 2020;91:418-25. [Crossref] [PubMed]
52. Shen B, Pan Y, Jiang X, Wu Z, Zhu J, Dong J, Zhang W, Xu P, Dai Y, Gao Y, Xiao C, Zhang L. Altered putamen and cerebellum connectivity among different subtypes of Parkinson's disease. CNS Neurosci Ther 2020;26:207-14. [Crossref] [PubMed]
53. Madden DJ, Merenstein JL. Quantitative susceptibility mapping of brain iron in healthy aging and cognition. Neuroimage 2023;282:120401. [Crossref] [PubMed]