Cortical thickness alterations in Parkinson's disease related anxiety: A cross-sectional 7 T MRI study

Abstract

Background

Anxiety is a common non-motor symptom in Parkinson's disease (PD) and has previously been associated with changes in cortical thickness of various brain regions.

Objective

To identify changes in cortical thickness in PD-related anxiety.

Methods

148 patients from an ongoing cohort study were included: 30 PD patients with anxiety, 73 PD patients without anxiety and 45 healthy controls. Anxiety was measured with the Parkinson Anxiety Scale (PAS). A 7 T structural MRI scan was performed and used to compare cortical thickness between these groups. Region of interest (ROI) as well as whole-brain analyses were performed to identify differences.

Results

ROI analyses revealed a strong negative association between the cortical thickness of the left lingual gyrus and the severity of anxiety in PD patients (R = −0.71; p = 0.006). Additional significant strong negative associations with the severity of anxiety in PD patients were observed in the frontal and cingulate regions (R between −0.56 and −0.65). Whole-brain analysis revealed a significant cluster of cortical thinning in the left anterior cortex and the dorsolateral prefrontal cortex weakly associated with PAS total score across all groups (R = −0.25, p = 0.00201).

Conclusions

This study is the first to report a strong negative association between left lingual gyrus thickness and anxiety severity in PD. Additionally, anxiety in early PD is associated with cortical thinning in the fronto-cingulate region, mainly affecting left sided structures. Future studies should examine whether these cortical changes can predict the anxiety progression patterns or the treatment response in PD patients.

Plain language summary

Anxiety is a common symptom in Parkinson's disease (PD) and has been linked to changes in thickness of certain cortical areas in the brain. Our study examines whether specific cortical areas show an altered thickness in early diagnosed PD patients that suffer from anxiety compared with those not suffering from anxiety. 148 participants were included: 30 PD patients with anxiety, 73 PD patients without anxiety, and 45 people without PD. Each participant underwent an ultra-high-field 7 T MRI-scan to determine regional cortical thickness. By comparing the thickness of different regions between the groups we found that people with both PD and anxiety displayed thinning of the ‘lingual gyrus’, a specific area in the back part of the brain, compared to the other two groups. Secondly, the people in the PD group with more severe anxiety showed thinning of the ‘fronto-cingulate region’, an area located more toward the front of the brain, compared to the other two groups. Finally, in all three groups, people with a higher anxiety score showed thinning of the ‘prefrontal cortex’, an area in the front of the brain, compared to the participants with a low score for anxiety. The exact role of these areas in the brain of anxious people with PD is not clear yet. Still, the results are interesting and should be further investigated.

Keywords

Parkinson's disease anxiety cortical thickness lingual gyrus fronto-cingulate cortex

Introduction

Parkinson's disease (PD) is a neurodegenerative disease characterized by both motor and non-motor symptoms, including neuropsychiatric symptoms. Among these neuropsychiatric symptoms, anxiety is one of the most frequent with approximately 31% of PD patients experiencing significant anxiety symptoms.¹ The underlying mechanisms of PD-related anxiety are still poorly understood. In PD, increased anxiety levels have been linked to structural and functional changes in anxiety-related circuits, consisting of the ‘fear circuit’ and the ‘limbic anxiety circuit’.² The fear circuit is a neuronal circuit involved in fear processing and includes the amygdala, anterior cingulate cortex (ACC), the medial prefrontal cortex (mPFC), the insular cortex (IC), the hippocampus and the striatum. The limbic loop of the cortico-striato-thalamo-cortical circuit, also called the “limbic anxiety circuit”, is involved in emotional control. While altered functional connectivity has been observed between structures within these circuits, little is known about alterations of the cortical grey matter structures.²

In the general population, most studies on fear and anxiety disorders reported a decrease in cortical thickness of the ventromedial (vmPFC) and dorsolateral (dlPFC) prefrontal cortex, as well as cortical thinning in several other cortical regions, such as the lateral orbitofrontal cortex (OFC), medial OFC, ACC and caudal middle frontal gyrus.^3–6 In PD patients, a reduced cortical thickness in the bilateral ACC and left parietal regions is reported in patients with anxiety compared to patients without anxiety.⁷ Additionally, higher levels of anxiety correlate with a reduction in grey matter volume in the ACC and precuneus.⁸ These findings suggest that cingulate and parietal areas may be involved in the underlying mechanisms of PD-related anxiety through disrupted emotion regulation, reflecting dysfunction in the fear circuit. However, studies focusing specifically on cortical thickness in patients with PD-related anxiety are limited, and so far, none was based on ultra-high field 7-Tesla MRI, which can reveal subtle cortical alterations that may be undetectable at 3 T resolution. Identifying changes in cortical thickness in these patients could help to decipher the underlying mechanisms of PD-related anxiety, which is essential for development of more effective treatment options.

The aim of this study was to identify changes of brain cortical thickness in PD patients suffering from anxiety using ultra-high field 7-Tesla MRI scans. For PD patients with anxiety symptoms, we hypothesize a reduction of cortical thickness in areas involved in the anxiety circuits, such as the ACC and mPFC, as well as in the parietal regions, which were previously associated with anxiety.^7,8

Methods

Study design

We analyzed data from the TRACK-PD study.⁹ This is an on-going longitudinal observational study that includes recently diagnosed PD patients without dementia (≤ 3 years after diagnosis), and healthy controls (HC), with 4-year of follow-up assessments. All data were collected at Maastricht University Medical Centre. At each visit (baseline, 2- and 4-year follow-up) participants were extensively assessed with (non)motor-scores, questionnaires neuropsychological battery and underwent 7-Tesla brain MRI scanning. In this cross-sectional study, we only used the baseline dataset. The study design is described in more detail by Wolters et al.⁹ Ethical approval was provided by the local medical ethical committee, and the study has been registered (Dutch Trial Register, NL7558). Written informed consent was obtained from all participants.

Study population

Patients were recruited locally and nationwide. Local recruitment targeted the PD population of the movement disorder clinic in Maastricht University Medical Centre, Maastricht, the Netherlands. Nationwide recruitment was conducted through online advertising. Patients included in this study had received a diagnosis of idiopathic PD in the past three years and were between 35 and 80 years old. The diagnosis of PD was made by a neurologist based on the MDS clinical diagnostic criteria.¹⁰ Patients with other neurological disorders, dementia or severe cognitive impairment (Montreal Cognitive Assessment (MoCA) score <24), contra-indications for MRI, major depressive disorder (MDD) or abuse of alcohol, drugs or benzodiazepines were excluded.

At baseline, demographic and clinical variables were collected including sex, age, handedness, and disease duration.

The non-motor, motor symptoms and disease severity were assessed by the Movement Disorders Society - Unified Parkinson's Disease Rating Scale (MDS-UPDRS) part I and III and the Hoehn & Yahr scale (H&Y), respectively. The ‘Beck Depression Inventory’ (BDI) was used to assess depression.⁹

Assessment of anxiety

The Parkinson Anxiety Scale (PAS) was used for the assessment of anxiety symptoms. It has subsections for persistent anxiety (PAS-A), episodic/situational anxiety (PAS-B), and avoidance behavior (PAS-C).¹¹ A PAS-total score was calculated by summing the three sub-scores. PD-patients were considered to have significant PD-related anxiety if they had a score above the cut-off in at least one of the three subparts of the scale: PAS-A > 9, PAS-B > 3, or PAS-C > 3. PD participants were divided into two groups: 1) PD-patients with anxiety (Anx-PD), 2) PD-patients without anxiety (noAnx-PD), alongside a third group of healthy controls (HC).

Image acquisition

MRI-data was acquired using a 7-Tesla MRI-scanner (Magnetom, Siemens, Erlangen, Germany) equipped with a Nova Medical 32-channel head coil. Dielectric pads were applied to enhance the signal in the temporal brain regions. Cardiac and respiratory physiological signals were measured synchronized with the scan start. A whole-brain MP2RAGE (Magnetization Prepared 2 Rapid Acquisition Gradient Echoes) acquisition was used with an acquisition time of 10:57 min, resulting in a T1-weighted image and a quantitative T1 map. These settings were used: TE = 2.51 ms, TR = 5000 ms, TI = 900 ms and 2750 ms, flip angle = 5° and 3°, FoV = 208 mm, resolution (x-y-z) = 0.65 × 0.65 × 0.65mm³, slices = 240, orientation = sagittal (more details in the published protocol) (9).

Image processing and cortical thickness extraction

After dicom to nifti conversion using dicomx, structural brain MRI scans were processed using FreeSurfer (version 6.0.0) (https://surfer.nmr.mgh.harvard.edu/) for extraction and segmentation of the cortical thickness.¹² Pial and white surfaces were automatically segmented. The “Desikan-Killiany” cortical atlas, a gyral based atlas including 34 cortical areas bilaterally, was used to define the cortical regions of interest (ROI) in MNI-space.¹³

Quality control was performed by trained researchers (ML, GC) and each surface was visually inspected and corrected manually as needed. Quality control steps recommended for FreeSurfer were performed.

The average cortical thickness of each cortical region of the atlas was extracted and provided in mm for each participant.

Cortical thickness analysis at vertex level

The open-access multimodal processing and data co-registration pipeline micapipe (V.0.2.0) was used to process anatomical images.¹⁴ T1w images were deobliqued, reoriented to standard neuroscience orientation (LPI), corrected for intensity nonuniformity, intensity normalized, and skull stripped. Cortical features were projected from the native space to the fsaverage5 midsurface (with approximately 10,000 surface vertices per hemisphere) and smoothed at 5 mm full width at half maximum (FWHM), using nilearn python library and Workbench tools (v. 1.4.2). Finally, all cortical surface maps were individually z-scored.

Statistical analysis

To investigate the relationship between anxiety and cortical thickness, we employed two separate analyses: a whole-brain analysis and a region of interest (ROI) analysis.

The ROI analysis was employed to assess mean cortical thickness within pre-defined brain regions. Using the mean cortical thickness of each region instead of focusing on individual voxels makes this analysis suitable for the detection of regional alterations in cortical thickness. Additionally, it facilitates comparison to existing literature on cortical thickness as it uses pre-defined cortical areas.

The whole-brain analysis served as an exploratory analysis to observe differences in cortical thickness on a vertex level without a priori assumptions about regions that could possibly be affected. Conducting this analysis allowed the detection of diffuse alterations in cortical thickness that might not be identified in a more targeted analysis.

The significance threshold was set at p-value < 0.05 and corrected for multiple comparisons –using the False Discovery Rate (FDR) when necessary. The numerical variables were described as means and standard deviations, the ordinal and the categorical variables as frequencies and percentages. The normality of distribution was assessed using a Kolmogorov-Smirnov test.

Descriptive analyses

Group differences were assessed using Chi2 tests for categorical variables and ANOVA tests for continuous variables, using SPSS, version 29 (SPSS, Chicago, 2023). A Mann-Whitney U test was used to compare the H&Y stages.

Region of interest analysis

Cortical thickness was analyzed across all 34 bilaterally defined regions from the Desikan-Killiany atlas.¹³ The mean cortical thickness of each region was normalized using Z-scores.

First, group comparison analyses of those z-scores between the three groups were performed using MANCOVA tests with age, sex and BDI-score as co-variables. No other covariables were added.

Secondly, hierarchical multiple regression analyses were performed to examine the relationship between the PAS-total score and the mean cortical thickness (Z-scores) of the cortical regions. This regression-based approach was chosen over simple correlational analyses to control for potential confounders and to quantify the effect of PAS-total score on cortical thickness alterations.

In this analysis, age, sex and depressive symptoms (BDI-score) were set as nuisance regressors in the first block (model 1) of all regression models, whereas PAS-score (independent variable) was separately added to the second block of the model (model 2). The mean cortical thickness (Z-score) of each cortical region was set as a dependent variable.

We ensured that all models met the assumptions for multiple regression analyses, including normality of the residuals, multicollinearity, and homoscedasticity.

Whole-brain analysis

Surface-based linear models (SLM in python brainstat toolbox) were performed to investigate the relationship between two key variables: BDI score and PAS score. Both variables were included as fixed effects in a single model, allowing assessment of their individual contributions to cortical thickness while controlling for the other variable. Age and sex were added as nuisance covariates. No other covariates were added.

This model was then fitted to our neuroimaging data, specifically the cortical thickness measurements. We applied a cluster-wise correction for multiple comparisons using both FDR and Random Field Theory (RFT) methods, with a cluster-defining threshold of p < 0.01.

Results

Demographic and clinical characteristics

Of the 151 participants originally included in the study, one was excluded at baseline due to withdrawal of consent. Therefore, 150 participants were included in this analysis. After MRI image quality control, 2 additional participants were excluded because of too poor quality of the MRI data despite correction and reprocessing steps. The 148 participants consisted of 30 PD patients with anxiety (Anx-PD), 73 PD patients without anxiety (NoAnx-PD), and 45 healthy controls. Baseline characteristics of these 148 subjects are reported in Table 1.

Table 1.

Demographic, cognitive and clinical characteristics in Parkinson patients with anxiety, without anxiety and healthy controls.

Variables	PD with anxiety	PD without anxiety	HC	p
Demographic variables
Effective (n = 148)	30 (20.3%)	73 (49.3%)	45 (30,4%)
Sex (women)	10 (33.3%)	22 (30.1%)	14 (31.1%)	0.95
Age^$	62.5 (± 9.1)	62.0 (± 8.13)	60.7 (± 8.1)	0.59
Dominant hand (right)	27 (90%)	61 (83.6%)	42 (93.3%)	0.27
Disease duration^$	1.63 (± 0.85)	1.61 (± 0.76)	NA	0.94
Side onset (right)	12 (40%)	38 (52.1%)	NA	0.54
LEDD (mg)^$	467.6 (± 260.6)	374.2 (± 211.2)	NA	0.15
Cognitive variables
MoCA (/30)^$	27.6 (± 1.59)	28.0 (± 1.67)	27.7 (± 1.56)	0.60
Clinical variables
MDS-UPDRS 1^$
1–1 Cognitive disorders	1 (0–2)	0 (0–2)	NA	0.02
1–3 Depressive mood	0 (0–3)	0 (0–2)	NA	0.02
1–4 Anxious state	1 (0–2)	0 (0–2)	NA	<0.0001
1–5 Apathy	0 (0–2)	0 (0–1)	NA	<0.001
MDS-UPDRS 3 (/132)^$	19.3 (± 6.3)	18.7 (± 8.0)	NA	0.45
Hoehn & Yahr stage^$	2 (1–2)	2 (1–3)	NA	0.60
Depression score (BDI total)^$	11.2 (± 5.7)	5.7 (± 3.8)	3.51 (±3.60)	<0.001
Parkinson Anxiety Scale (PAS)
Total (/48)	15.7 (± 4.0)	5.7 (± 3.4)	8.1 (± 2.9)	<0.0001
A subscale (/20)	8.1 (± 2.8)	3.9 (± 2.5)	3.7 (± 2.6)	<0.0001
B subscale (/16)	4.2 (± 1.8)	1.0 (± 1.0)	1.3 (± 1.6)	<0.0001
C subscale (/12)	3.4 (± 2.0)	0.9 (±	0.9 (± 1.3)	<0.0001

$: described as median and range; LEDD: Levodopa Equivalent Daily Dosages; MoCA: Montreal Cognitive Assessment; MDS-UPDRS: Movement Disorder Society Unified Parkinson Disease Rating Scale; BDI: Beck's Depression Inventory; PAS: Parkinson Anxiety Scale. Bold: significant differences only relevant items to this study are shown

ROI-based analyses

There was no significant difference in the mean cortical thickness of each ROI using Z-scores between the three groups (Supplemental Table 1).

In the group of PD patients (both anxious and non-anxious), there were significant weak negative associations between the PAS-total score and the cortical thickness of the left temporal cortex (fusiform gyrus, lingual gyrus, middle and superior temporal gyrus), the left medial orbitofrontal cortex, and the left rostral middle frontal gyrus (R between 0.27 and 0.39; Supplemental Table 2) There was no association of anxiety with the ROIs of the right hemisphere (Supplemental Table 2).

There were significant weak negative associations between the PAS-A sub-score (persistent anxiety) and the cortical thickness in the superior frontal cortex, the rostral middle frontal cortex, the middle and inferior temporal cortex and the lingual cortex, in the left hemisphere (R between 0.34 and 0.40; Supplemental Table 3).

In the group of PD patients with anxiety, there was a significant strong negative association (R = −0.71, p = 0.001) between the cortical thickness of the left lingual gyrus and the severity of anxiety according to the PAS-total score (Table 2 and Figure 1). Additionally, there were significant strong negative associations between the PAS-B sub-score (episodic anxiety) and the cortical thickness in the superior and inferior frontal cortex, pars orbitalis and pars triangularis, the paracentral cortex, the isthmus and posterior cingulate cortex and the inferior temporal cortex in the left hemisphere; and the pericalcarine cortex in the right hemisphere (R ranging from −0.57 to - 0.65; Table 2 and Figure 2).

Figure 1.

Regression analysis between cortical thickness and severity of anxiety symptoms according to PAS-total score in Parkinson's disease (PD) patients and healthy controls: Visualization of the negative association between the cortical thickness (Z-score) of the left lingual gyrus and PAS-total score in anxious PD patients. Adjusted by sex, age, and BDI-score.

Figure 2.

Representation of the multiple linear regression analysis in PD-patients with anxiety (upper image) and in PD-patients both with and without anxiety groups (lower image).

Table 2.

Regression analysis between the cortical thickness in brain region of interest (ROI) and PAS-total score (A) and PAS-B subscore (B) in Parkinson's disease PD patients with anxiety subgroup.

A. PD-patients with anxiety group (PAS-total score)
Left Hemisphere (adjusted by age, sex & BDI)
ROI	R	p model	p PAS	Beta PAS
Lingual gyrus	−0.71	0.001	0.006	−0.47
B. PD-patients with anxiety group (PAS-B score)
Left Hemisphere (adjusted by age, sex & BDI)
ROI	R	p model	p PAS	Beta PAS
Superior frontal cortex	−0.57	0.038	0.022	−0.47
Inferior frontal cortex pars-orbitalis	−0.59	0.027	0.008	−0.56
Inferior frontal cortex pars-triangularis	−0.60	0.021	0.025	−0.45
Paracentral cortex	−0.62	0.014	0.046	−0.39
Posterior cingulate cortex	−0.59	0.025	0.042	−0.41
Isthmus cingulate cortex	−0.65	0.007	0.046	−0.38
Inferior temporal cortex	−0.59	0.028	0.039	−0.42
Right Hemisphere (adjusted by age, sex & BDI)
ROI	R	p model	p PAS	Beta PAS
Pericalcarine cortex	−0.56	0.043	0.018	−0.50

BDI: Beck depression inventory; PAS: Parkinson anxiety scale.

p-values have been corrected for multiple comparisons using the False Discovery Rate (FDR) method. Only strong significant associations are provided.

There was no significant association between the cortical thickness and the PAS (sub)scores, in the HC group.

Whole-brain vertex-wise cortical thickness analysis

In the three groups, there was a significant cluster in PAS-total score, as in PAS-A subscale, which shows a weak negative association with cortical thickness in the left dorsolateral prefrontal cortex (DLPFC), mostly in the anterior part of the middle frontal gyrus (R = −0.25, p = 0.00201) (Figure 3). There were no significant associations with PAS-B and PAS-C sub-scores.

Figure 3.

Whole-brain vertex-wise cortical thickness analysis. PAS-total score shows a significant negative association with cortical thickness in the middle frontal gyrus in Parkinson patients with anxiety (anxPD), compared to Parkinson's Disease patients without anxiety (noAnxPD) and healthy controls (HC). Adjusted for age, sex and BDI-score. PAS: Parkinson Anxiety Scale.

Discussion

In this study, we examined the differences in cortical thickness between PD patients with anxiety, PD patients without clinically relevant anxiety and healthy controls. We conducted both ROI-based and whole-brain analyses. The ROI-based analyses revealed no between-group differences but did show strong negative associations between the PAS-total score and the cortical thickness of the left lingual gyrus in both anxious and non-anxious PD patients and between the PAS-B score and the cortical thickness of the left fronto-cingulate, left temporal and right pericalcarine cortex in the group of PD patients with anxiety. Furthermore, there were weak negative associations between PAS (sub)scores and parts of the left frontal and temporal cortex in the group of PD patients (with and without anxiety). In the whole-brain analyses, there was a weak association between the cortical thickness of the left DLPFC and the PAS-total and PAS-A scores. Those results are partly in line with previous findings showing reduced thickness of the left fronto-cingulate cortex in PD-related anxiety, but they also show that underlying mechanisms could differ depending on the subtype of anxiety.^7,8,15 These results also suggest a potential implication of the lingual cortex.

PD related anxiety is associated with cortical thinning in visual and prefrontal regions

While most recent studies on cortical thickness in anxiety used 3 T MRI,^3,4,6,8 the application of ultra-high field 7 T MRI in an early PD population enabled higher spatial resolution and potentially more accurate cortical thickness measurements.¹⁶ The ROI-based analyses revealed a strong negative association between the cortical thickness of the lingual gyrus in the temporo-occipital lobe and anxiety severity in PD patients. This is a novel finding that may reflect the potential enhanced sensitivity of 7 T MRI, which could explain why this area was not reported more often with regards to PD anxiety in 3 T MRI studies. The whole brain analyses show a weak negative association between anxiety severity and cortical thickness of the dlPFC, which is mentioned more often with regards to anxiety in literature.^5,7

The lingual gyrus is part of a large network that encodes, analyzes, and gives emotional meaning to visual information and is connected to the amygdala through the ventral visual stream.^17–19 Existing literature on anxiety mentions that the lingual gyrus was thicker in patients with social anxiety disorder.²⁰ While this contrasts with our finding of cortical thinning, it may suggest that anxiety patients are particularly sensitive to interactions with the environment. Structural thinning of the lingual gyrus could disrupt the integration of visual and emotional information, potentially exacerbating anxiety symptoms in PD patients.²¹ Future research should investigate the potential of the lingual gyrus being a potential imaging marker for anxiety progression in PD.

The PFC is a key structure of both the fear and limbic circuits and thinning in this area has previously been observed in anxious individuals and anxious PD patients.^7,22 The observed negative association between anxiety severity and cortical thickness in the dlPFC may reflect the vulnerability of structures within these circuits for cortical thinning in anxiety.^23,24 The present finding aligns with prefrontal regions responsible for cognitive control over fear and emotion showing cortical thinning in anxiety disorders.^5,7

Episodic anxiety in PD patients is associated with cortical thickness of the fronto-cingulate regions

The ROI-based analyses showed a strong negative association between episodic anxiety severity and cortical thickness of structures within the left fronto-cingulate region. The fronto-cingulate regions play a key role in mediating top-down regulation of emotions through the inhibition of limbic structures such as the amygdala, which are involved in fear signaling.^7,25 Through this inhibitory regulation, the structures within this region can reduce negative emotional activity and intrusive thoughts.

In non-PD patients suffering from anxiety, the strength of the functional connectivity between the posterior cingulate cortex and amygdala showed a negative association with the severity of anxiety.²⁶ Cortical thinning in the frontal and cingulate regions within the fear circuit could alter the activation of these structures, reducing top-down regulation. Consequently, this could lead to overactivation of the amygdala, resulting in an excessive response to perceived threats.^2,7

In terms of cortical thickness, our findings in the fronto-cingulate region are in line with previous findings in anxious PD patients.^2,8,15 While cortical thickness in this region is associated with episodic anxiety, rather than persistent anxiety, it highlights the vulnerability of these fear-processing areas in PD patients. Possible therapeutic options, such as cognitive behavioral therapy (CBT), have shown to induce functional connectivity changes between brain structures in PD.²³ Similarly, neurofeedback therapy proved to improve symptoms in anxiety.²⁷ Future studies could investigate the effect of such therapies on the longitudinal changes in cortical thickness in these areas.

Subtypes of anxiety in PD are associated with different patterns of cortical thinning

Our findings reveal that the patterns of cortical thickness can vary significantly between PAS-total and PAS sub-scores. Specifically, PAS-A (persistent anxiety) and PAS-B (episodic anxiety) demonstrate significant negative associations with cortical thickness, while PAS-C (avoidance behavior) shows no significant results.

This pattern of alterations per anxiety subtype suggests that different forms of anxiety exhibit unique patterns of cortical thinning.^5,28 While the results support our hypothesis of cortical alterations in an overactive fear circuit and underactive limbic circuit in anxiety, the underlying mechanisms that cause these alterations could be more extensive than those usually associated with PD related anxiety. In this population, the severity of anxiety is likely due to a combination of factors and is not exclusively due to the neurobiological changes associated with PD. For instance, trait anxiety presents before the onset and diagnosis of the disease as well as the individual coping strategies in response to the emotional stress caused by PD and its symptoms. These findings highlight that research into anxiety must consider the various clinical subtypes, which may have different underlying mechanisms and result in different clinical trajectories.

Additionally, both the ROI and whole brain analyses suggest that cortical thinning is confined to the left side of the brain. While hemispheric lateralization is well-known for language and problem-solving, its role in emotional processing is less clear.²⁹ The observed pattern of cortical thinning lateralizing to the left hemisphere may indicate that anxiety could be hemisphere specific, possibly aligning with some functions being hemisphere specific.

Strengths and limitations

This study has several strengths as we included a large and well-characterized sample of PD patients and healthy controls. The PAS-score was used as a reliable measure of anxiety symptoms in PD. Imaging was performed with an ultra-high field 7-Tesla MRI scanner. However, this study also has some limitations. First, TRACK-PD is a large cohort study designed to assess the structural characteristics of PD patients.⁹ While anxiety assessment was performed for patient characterization, the study design was not focused specifically on anxiety. Secondly, since alterations in cortical thickness were the prime focus in this analysis, we did not analyze subcortical structures such as the basal ganglia and amygdala. Therefore, further analyses including subcortical structures and functional connectivity are needed. The use of a 7-Tesla MRI scanner will help to study these structures with high precision in further analyses. Thirdly, we excluded patients that were diagnosed with MDD. As there is a high comorbidity between anxiety and depression in PD,³⁰ the generalizability of this study could be limited. However, with this approach we aimed to reduce the confounding effects of depression to allow clearer interpretation of anxiety-related findings. Finally, we focused on the baseline data of an on-going cohort. PD patients were then at a relatively early stage of the disease. Using the follow-up data once they are available is necessary to have a more comprehensive understanding of the mechanisms of PD-related anxiety according to the patients’ clinical trajectory.

Conclusion

The main finding of our study is that anxiety in early PD is associated with cortical thinning of left-hemisphere regions, with the fronto-cingulate region and the left lingual gyrus in particular. Most notably, the strong negative association between PAS-total scores and the left lingual gyrus thickness, identified with 7 T MRI, highlights a novel finding in PD anxiety. While the exact implication of both results in PD anxiety remains unclear, the observed association with anxiety symptoms is strong and significant and warrants further investigation. Our findings aid in the understanding of the underlying pathophysiology of anxiety symptoms in PD.

Supplemental Material

sj-docx-1-pkn-10.1177_1877718X251367302 - Supplemental material for Cortical thickness alterations in Parkinson's disease related anxiety: A cross-sectional 7 T MRI study

Supplemental material, sj-docx-1-pkn-10.1177_1877718X251367302 for Cortical thickness alterations in Parkinson's disease related anxiety: A cross-sectional 7 T MRI study by Michiel van Lier, Guillaume Carey, Romain Viard, Mark L Kuijf, Amée F Wolters, Kathy Dujardin and Albert FG Leentjens in Journal of Parkinson's Disease

Footnotes

Acknowledgements

The authors thank all participants of the study for their patience and cooperation.

ORCID iDs

Michiel van Lier

Guillaume Carey

Romain Viard

Mark L Kuijf

Amée F Wolters

Kathy Dujardin

Albert FG Leentjens

Ethical considerations

Ethical approval was provided by the local medical ethical committee, and the study has been registered (Dutch Trial Register, NL7558).

Consent to participate

Written informed consent was obtained from all participants.

Consent for publication

Not applicable

Author contributions

Michiel van Lier: Research project: Organization; Execution; Statistical Analysis: Execution; Manuscript: Writing of the first draft. Guillaume Carey: Research project: Conception; Organization; Execution; Statistical Analysis: Design; Execution; Review & Critique; Manuscript: Review & Critique. Romain Viard: Statistical Analysis: Design; Execution; Manuscript: Review & Critique. Mark L Kuijf: Statistical Analysis: Review and Critique; Manuscript: Review & Critique. Amée F. Wolters: Research project: Organization; Execution; Manuscript: Review & Critique. Kathy Dujardin: Statistical Analysis: Review and Critique; Manuscript: Review & Critique. Albert FG Leentjens: Research project: Conception; Organization; Statistical Analysis: Design; Execution; Manuscript: Review & Critique.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: MRI data included in this study were acquired at Scannexus (Maastricht, The Netherlands). This study is funded by a grant from the Weijerhorst Foundation (Stichting de Weijerhorst). This Dutch foundation supports and promotes activities in the public interest, in particular scientific research. The sponsor has no role in the design of the study and collection, analysis, and interpretation of data. Also, the sponsor was not involved in writing this manuscript.

Declaration of conflicting interest

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: A.F.G. Leentjens received consultancy fees from Lilly. G. Carey participated to advisory boards for ORKYN’ and Abbvie and gave consultancies to Abbvie. He was employed by University of Reims (French Government) and earned a grant from France Parkinson.

Mark L. Kuijf, Stijn Michielse and Albert F.G. Leentjens were employed by Maastricht University (Dutch government). Amée Wolters was employed by Catharina Hospital Eindhoven. Kathy Dujardin was employed by University of Lille (French Government).

Data availability statement

This study uses data from the dataset TRACK-PD. The data that supports the findings of this study are available and can be requested from M.L. Kuijf. The data are not publicly available due to privacy considerations.

Supplemental material

Supplemental material for this article is available online.

References

Broen

MPG

Narayen

Kuijf

, et al. Prevalence of anxiety in Parkinson’s disease: a systematic review and meta-analysis. Mov Disord 2016; 31: 1125–1133.

Carey

Görmezoğlu

Jong

, et al. Neuroimaging of anxiety in Parkinson’s disease: a systematic review. Mov Disord 2021; 36: 327–339.

Brühl

Hänggi

Baur

, et al. Increased cortical thickness in a frontoparietal network in social anxiety disorder. Hum Brain Mapp 2014; 35: 2966–2977.

Molent

Maggioni

Cecchetto

, et al. Reduced cortical thickness and increased gyrification in generalized anxiety disorder: a 3T MRI study. Psychol Med 2018; 48: 2001–2010.

Sobral

Morgado

Moreira

, et al. Association between cortical thickness and anxiety measures: a scoping review. Psychiatry Res Neuroimaging 2022; 319: 111423.

Zhao

Chen

Zhang

, et al. Gray matter abnormalities in non-comorbid medication-naive patients with major depressive disorder or social anxiety disorder. EBioMedicine 2017; 21: 228–235.

Carey

Lopes

Viard

, et al. Anxiety in Parkinson’s disease is associated with changes in the brain fear circuit. Parkinsonism Relat Disord 2020; 80: 89–97.

Wee

Wen

M-C

Kandiah

, et al. Neural correlates of anxiety symptoms in mild Parkinson’s disease: a prospective longitudinal voxel-based morphometry study. J Neurol Sci 2016; 371: 131–136.

Wolters

Heijmans

Michielse

, et al. The TRACK-PD study: protocol of a longitudinal ultra-high field imaging study in Parkinson’s disease. BMC Neurol 2020; 20: 292.

10.

Postuma

Berg

Stern

, et al. MDS Clinical diagnostic criteria for Parkinson’s disease. Mov Disord 2015; 30: 1591–1601.

11.

Leentjens

AFG

Dujardin

Pontone

, et al. The Parkinson Anxiety Scale (PAS): development and validation of a new anxiety scale. Movement Disord 2014; 29: 1035–1043.

12.

Fischl

Kouwe

Avd

Destrieux

, et al. Automatically parcellating the human cerebral cortex. Cereb Cortex 2004; 14: 11–22.

13.

Desikan

Ségonne

Fischl

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

14.

Cruces

Royer

Herholz

, et al. Micapipe: a pipeline for multimodal neuroimaging and connectome analysis. NeuroImage 2022; 263: 119612.

15.

Betrouni

Alazard

Bayot

, et al. Anxiety in Parkinson’s disease: a resting-state high density EEG study. Neurophysiol Clin 2022; 52: 202–211.

16.

Lüsebrink

Wollrab

Speck

. Cortical thickness determination of the human brain using high resolution 3T and 7T MRI data. Neuroimage 2013; 70: 122–131.

17.

Kitada

Johnsrude

Kochiyama

, et al. Brain networks involved in haptic and visual identification of facial expressions of emotion: an fMRI study. Neuroimage 2010; 49: 1677–1689.

18.

Ousdal

Andreassen

Server

, et al. Increased amygdala and visual cortex activity and functional connectivity towards stimulus novelty is associated with state anxiety. PLoS One 2014; 9: e96146.

19.

Isenberg

Silbersweig

Engelien

, et al. Linguistic threat activates the human amygdala. Proc Natl Acad Sci U S A 1999; 96: 10456–10459.

20.

Frick

Howner

Fischer

, et al. Cortical thickness alterations in social anxiety disorder. Neurosci Lett 2013; 536: 52–55.

21.

Tinaz

Courtney

Stern

. Focal cortical and subcortical atrophy in early Parkinson’s disease. Mov Disord 2011; 26: 436–441.

22.

Hilbert

Pine

Muehlhan

, et al. Gray and white matter volume abnormalities in generalized anxiety disorder by categorical and dimensional characterization. Psychiatry Res Neuroimaging 2015; 234: 314–320.

23.

Carey

Lopes

Moonen

AJH

, et al. Cognitive behavioral therapy for anxiety in Parkinson’s disease induces functional brain changes. J Parkinsons Dis 2023; 13: 93–103.

24.

Kenwood

Kalin

Barbas

. The prefrontal cortex, pathological anxiety, and anxiety disorders. Neuropsychopharmacology 2022; 47: 260–275.

25.

Etkin

Egner

Kalisch

. Emotional processing in anterior cingulate and medial prefrontal cortex. Trends Cogn Sci 2011; 15: 85–93.

26.

Hahn

Stein

Windischberger

, et al. Reduced resting-state functional connectivity between amygdala and orbitofrontal cortex in social anxiety disorder. Neuroimage 2011; 56: 881–889.

27.

Chen

Xiao

Belkacem

, et al. Efficacy evaluation of neurofeedback-based anxiety relief. Front Neurosci 2021; 15: 758068.

28.

Saviola

Pappaianni

Monti

, et al. Trait and state anxiety are mapped differently in the human brain. Sci Rep 2020; 10: 11112.

29.

Duboc

Dufourcq

Blader

, et al. Asymmetry of the brain: development and implications. Annu Rev Genet 2015; 49: 1–26.

30.

Menza

Robertson-Hoffman

Bonapace

. Parkinson’s disease and anxiety: comorbidity with depression. Biol Psychiatry 1993; 34: 465–470.

Supplementary Material

Please find the following supplemental material available below.

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

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

0.03 MB

0.00 MB