Understanding Anxiety in Cervical Dystonia: An Imaging Study

Abstract

Background

Anxiety may precede motor symptoms in cervical dystonia (CD) and is associated with an earlier onset of dystonia. Our understanding of anxiety in CD is inadequate.

Objective

To investigate brain networks associated with anxiety in CD.

Methods

Twenty‐six subjects with idiopathic CD underwent MRI Brain without contrast. Correlational tractography was derived using Diffusion MRI connectometry. Quantitative Anisotropy (QA) was used in deterministic diffusion fiber tracking. Correlational tractography was then used to correlate QA with State–Trait Anxiety Inventory (STAI) state (STAI‐S) and trait (STAI‐T) subscales.

Results

Connectometry analysis showed direct correlation between state anxiety and QA in tracts from amygdala to thalamus/ pulvinar bilaterally, and trait anxiety and QA in tracts from amygdala to motor cortex, sensorimotor cortex and parietal association area bilaterally (FDR ≤0.05).

Conclusion

Our efforts to map anxiety to brain networks in CD highlight the role of the amygdala in the pathophysiology of anxiety in CD.

Cervical dystonia (CD) is the most common adult‐onset, focal dystonia in clinical practice. While diagnosis is largely based on motor findings, ¹ psychiatric aspects of dystonia require greater recognition. The likelihood of a current or lifetime diagnosis of a psychiatric illness of any type in CD is as high as 91.4% compared to 35% of adults in the general population. ² Patients with idiopathic dystonia, including CD, show higher rates of psychiatric diagnoses and medication prescription. These may predate or follow a diagnosis of dystonia. ³ A systems‐biology approach shows that dystonia‐associated genes and heritability of psychiatric disorders enrich within the same co‐expression modules, thereby indicating that associated psychiatric symptoms could be intrinsic to dystonia pathophysiology. ⁴

Anxiety, an internalizing phenomenon, can be categorized based on combinations of fear and distress, but also rumination, obsessive cognitions and avoidance. The separation of anxiety into state and trait allows for studying these two broad aspects: an acute and emotional response to stressors (state anxiety) and a tendency to have general feelings of worry for current and future events, and phobias (trait anxiety). An example of state anxiety may be fear due to needles prior to botulinum toxin injections. Trait anxiety, on the other hand, may reflect relatively persistent generalized worry and fear in CD patients. The State–Trait Anxiety Inventory (STAI) instrument is a reliable tool to capture state and trait anxiety across a range of anxiety disorders. ⁵ Overall, anxiety is one of the most common psychiatric diagnoses in CD. ² It may precede motor symptoms and has been shown to be associated with an earlier onset of dystonia. ⁶ Improvement in motor severity and anxiety with botulinum toxin may poorly correlate. ⁷ These clinical features corroborate the possibility of anxiety being intrinsic to its pathophysiology.

Neuroimaging studies, using diffusion tensor imaging (DTI) and functional magnetic resonance imaging (fMRI) highlight the importance of the basal ganglia, cerebellum and the sensorimotor cortex in the pathophysiology of CD. ⁸ , ⁹ , ¹⁰ , ¹¹ Using causal brain lesions, positive connectivity to the cerebellum and negative connectivity to the somatosensory cortex has been noted to be a specific marker for CD. ¹² Vermian atrophy using voxel based morphometry has been associated with a tremulous form of CD. ¹³ As such, a number of brain regions and their networks have been implicated in the pathophysiology of CD. ¹⁴ , ¹⁵ These biologically connected networks may be key to understanding the association of CD and anxiety. Activation of the amygdala has been consistently reported in anxiety‐provoking situations. ¹⁶ It is hypothesized to be central to acquisition and expression of conditioned fear, ¹⁷ and accordingly, in the pathophysiology of anxiety disorders. As a key component of the limbic system, the amygdala evaluates and integrates surrounding sensory information, assigns “emotional dimensions” and regulates changes in short‐ and long‐term synaptic plasticity to guide motor response. As such, it is known to play a key role in the limbic‐motor interface. ¹⁸

Despite recent advances in the genetic relationship between dystonia and neuropsychiatric disorders and known limbic‐motor interactions in otherwise healthy individuals, our understanding of anxiety networks in CD remains inadequate. Through his study, we sought to test the hypothesis that cervical dystonia patients with anxiety will demonstrate involvement of the limbic system, specifically the amygdala. Characterization of anxiety and associated networks in cervical dystonia will additionally explore the interaction between anxiety and motor manifestations of dystonia.

Methods

Patient Population and Experimental Design

Participants with idiopathic CD were prospectively recruited from the Rush Movement disorders clinic. Assessment of CD severity was done (by AM and DS) using the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS), with subscales for motor severity, activities of daily living and pain on the day of botulinum toxin injections. The TWSTRS is a validated measure of CD. Severity of anxiety was determined using the STAI. The STAI is a reliable instrument to capture anxiety in anxious persons compared to non‐anxious controls across the range of anxiety disorders with demonstrated use in CD. ⁷ , ¹⁹ Demographic and clinical data, including age, age of onset of CD, duration of CD and associated medications, were collected from electronic medical record. All subjects underwent an MRI Brain without contrast. Details on the MRI protocol and analysis are included as Supporting Information (Supporting Information: MRI Protocol and analysis).

Deterministic fiber tracking methods may be used to track trajectories of a fiber pathway. ²⁰ The use of Quantitative anisotropy (QA) improves their use. ²⁰ An FDR threshold of 0.05 was used to select tracks. To estimate the false discovery rate, a total of 4000 randomized permutations were applied to the group label to obtain the null distribution of the track length. Correlational tractography was then used to correlate quantitative anisotropy with STAI‐S and STAI‐T subscales.

Other Statistical Analysis

Descriptive statistics and nonparametric Spearman correlations were run on demographic information and clinical scales to understand univariate central tendency and dispersion, as well as bivariate relationships among the clinical scales. In order to clarify the relationship between STAI scales and quantitative anisotropy, we implemented nonparametric Spearman partial correlations, which controlled for the effect of age, sex, generalized anxiety (GAD‐2), and dystonia severity (total TWSTRS score).

All statistical analyses were performed using SAS. This study was approved by the Rush University Institutional Review Board (#21032912‐IRB01) Informed patient consent was collected from all subjects.

Results

Data on 26 subjects were collected. The mean age of subjects was 58.4 years (SD 11.8) and 76.6% of the sample were female. The mean total TWSTRS score was 30.8 (SD. 11.2), with its mean severity being 15.31 (SD = 5.77), mean disability being 7.46 (SD = 4335), and mean pain being 8.13 (SD = 4.53). The mean STAI‐T was 38.08 (SD = 11.65) and STAI‐S was 37.42 (SD = 12.23) The mean age of onset of CD was 40.5 years. The mean duration of symptoms was 17 years (SD. 15.8). Among the subjects, 26.9% had significant trait anxiety (n = 7) and 42.3% had significant state anxiety (n = 11). ²¹ STAI‐S scores correlated well with STAI‐T scores. (rho = 0.78, P < 0.001) STAI‐T scores showed significant correlation with motor TWSTRS scores (rho = −0.39, P = 0.04). All subjects regularly received botulinum toxin to address CD (Table S1). There were no differences in TWSTRS total score nor STAI subscales among those on and without neuroactive medications (anti‐dystonia and anti‐anxiety medications).

Results of Correlational Tractography

State Anxiety (STAI‐S) (Fig. 1)

Figure 1.

State‐anxiety: correlational tractography analysis shows increased Quantitative Anisotropy from the amygdala to the thalamus/ pulvinar region on the right (bottom) and the left (top).

The connectometry analysis found tracts showing increased QA from the right (FDR ≤0.05) and left (FDR ≤0.05) amygdala to the thalamus/ pulvinar region.

The connectometry analysis did not find tracts showing significant decreased QA.

Trait Anxiety (STAI‐T) (Fig. 2)

Figure 2.

Trait anxiety: correlational tractography analysis shows increased Quantitative Anisotropy from the amygdala to the motor cortex, sensorimotor cortex and the parietal association area on the right (bottom) and the left (top).

The connectometry analysis found tracts showing increased QA from the right (FDR ≤0.05) and left (FDR ≤0.05) amygdala to the motor cortex and the parietal association area. A clear and robust correlation was found with a large number of significant tracts from the amygdala to the ipsilateral primary motor and sensorimotor cortices.

The connectometry analysis did not find tracts showing significant decreased QA.

Discussion

These data confirm the role of amygdala and associated tracts in CD and present a direct correlation of microstructural integrity of tracts with severity of anxiety.

In otherwise healthy subjects, white matter microstructure involving the amygdala and the cingulate cortex have been suggested to be pathogenic in treatment‐naïve generalized anxiety disorder. ²² The amygdala has been implicated in anxiety and emotional processing with structural integrity of the amygdala–ventromedial prefrontal cortex predicting trait anxiety. ²³ Reduced connections between the amygdala and prefrontal areas (associated with modulation of emotion) and enhanced connectivity with somatosensory areas is associated with greater trait anxiety. ²⁴ In our study on subjects with CD, state anxiety severity directly correlated with strength of projections from the amygdala to the thalamus/ pulvinar region. The thalamus, especially paraventricular nucleus, has been implicated as an integration and relay node to modulate a behavioral response to threat, anxiety and fear. ²⁵ Trait anxiety directly and robustly correlated with the strength of projections from the amygdala to the motor cortex, sensorimotor cortex and parietal association cortex (which processes somatic, visual, acoustic and vestibular sensory information and plays a role in spatial cognition and motor control). ²⁶ Our study adds novelty by confirming a central role of amygdala in anxiety in CD. Our analysis using a more accurate, robust measure, QA, highlights the networks associated with state and trait anxiety in CD. Finally, corroborated by the correlation between trait anxiety and motor dystonia severity scores, act as a bridge between imaging data and clinical care by explaining the anecdote of worsening dystonia motor severity in patients with untreated anxiety. It validates the notion that anxiety may be an intrinsic personality trait in cervical dystonia.

Imaging investigations in dystonia have traditionally focused on motor manifestations representing microstructural changes in the cortico‐basal ganglia‐cerebellar and cortico‐striato‐pallido‐thalamic networks, with recent report of involvement of the corticospinal tract and posterior parietal cortex. ¹⁰ , ²⁷ Our study adds to literature due to our focus on imaging correlates of anxiety severity in CD.

To further improve specificity given our relatively low sample size, we used topology‐informed pruning which reduces false‐discovery by automatically cleaning‐up noisy fibers. ²⁸ While a strength of our methodology, it is possible that this approach limits our sensitivity to capture small fiber pathways or branches. ²⁸ As an example, the amygdala is known to be connected to the basal ganglia circuit through projections to the ventral pallidum and ventral striatum, before these projections are relayed back to the cortex via the dorsomedial nucleus of the thalamus. ²⁹ It is likely that these small branches to the basal ganglia were not captured. A majority of our study subjects identified as female. Such female preponderance for dystonia is consistent with literature. ³⁰ A limitation of our study is the absence of a control group which limits a comparative assessment of circuitry associated with anxiety in otherwise healthy controls or other dystonias. As such, we are unable to comment on whether these findings are specific to CD. Future studies should validate our findings with a larger sample size and an appropriate control group.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. Writing of the first draft, B. Review and Critique.

A.M.: 1A, 1B, 1C, 2A, 2C, 3A, 3B

T.S.: 2A, 2B, 2C, 3B

D.G.: 2A, 2B, 2C, 3B

G.S.: 2A, 2C, 3B

G.G.: 1B, 1C, 3B

T.W.R.: 1B, 1C, 3B

D.S.: 1C, 3B

C.P.: 2B, 3B

M.Y.: 2B, 3B

C.C.: 1A, 1B, 2C, 3B

Disclosures

Ethical Compliance Statement: This study was approved by the Rush University Institutional Review Board (#21032912‐IRB01). An approved‐informed patient consent was collected from all included subjects. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding sources and Conflicts of Interest: The study received funding from the Dystonia Medical Research Foundation and an internal funding mechanism at Rush University Medical Center. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures from the Last 12 Months: The authors declare that there are no additional relevant disclosures to report.

Supporting information

Supplementary TABLE S1. Clinical and demographic data of all subjects.

Data S1. MRI Protocol and analysis.