Spectrum of Dystonia in Progressive Supranuclear Palsy: A Prospective Observational Study and Review of Literature

Introduction

Progressive Supranuclear Palsy (PSP) is a sporadic neurodegenerative disorder within the spectrum of parkinsonism, pathologically defined by the accumulation of 4-repeat tau [1]. Classically, it presents with an akinetic-rigid syndrome, early falls, supranuclear gaze palsy, and pseudobulbar features. The global prevalence is estimated at around 5–6.4 per 100,000 population, although this varies depending on the diagnostic criteria and healthcare settings [2]. The mean age at onset is approximately 63 years, with a median survival of nearly 7 years from symptom onset [3].

Although the clinical spectrum of PSP has broadened considerably, it is still often described in terms of its classical presentation, overlooking the diversity of its subtypes [4, 5]. The 2017 Movement Disorders Society (MDS) criteria classify PSP into distinct variants based on four functional domains: ocular motor dysfunction, postural instability, akinesia, and cognitive dysfunction [6]. These are further stratified into levels of diagnostic certainty, supported by clinical features and imaging markers [7]. While definitive diagnosis requires neuropathological confirmation, the use of these criteria has improved the recognition of non-classical PSP variants with greater sensitivity and specificity [8, 9, 10].

PSP involves multiple brain regions, particularly the subthalamic nucleus, globus pallidus, substantia nigra, and brainstem structures, with the distribution of pathology influencing the clinical phenotype. The histopathological hallmark is the presence of tau-positive globose neurofibrillary tangles, reflecting widespread 4R-tau pathology. Involvement of multiple neurotransmitter systems and both pre- and post-synaptic pathways contributes to the limited and often short-lived response to pharmacological therapies [11].

Neuroimaging markers such as the midbrain-to-pons ratio and the magnetic resonance parkinsonian index (MRPI) are useful in supporting the diagnosis of PSP and differentiating it from other parkinsonian disorders [12, 13].

Dystonia is a recognized feature of PSP, affecting both limb and axial musculature [14, 15]. Previous studies have described its prevalence and clinical characteristics in atypical parkinsonian syndromes [16, 17, 18, 19]. Despite this, dystonia remains under-emphasized in routine clinical descriptions, often overshadowed by more prominent features such as falls and gaze palsy. Moreover, much of the existing literature is based on retrospective analyses and small cohorts, with limited focus on variation across PSP subtypes or on associated clinical and imaging correlates. A systematic characterization of dystonia is clinically relevant because its phenomenology may provide valuable clues for recognizing PSP subtypes, particularly in patients with non-classical presentations. Furthermore, identifying subtype-specific patterns and clinical correlates of dystonia may facilitate earlier and more accurate diagnosis. This, in turn, may enable timely implementation of targeted symptomatic therapies, such as botulinum toxin treatment, ultimately improving patient care and functional outcomes.Thus, the present study was undertaken to systematically evaluate dystonia in PSP, including its prevalence, phenomenology, anatomical distribution, subtype associations, clinical and imaging correlates, and therapeutic implications.

Methods

This prospective study was conducted at our tertiary care teaching institute, Amrita Institute of Medical Sciences, Faridabad, India, [2024–2025]. The study was approved by the Institutional Ethics Committee. Patients attending our Movement disorder clinic were clinically examined and those diagnosed as possible or probable PSP based on the 2017 International Parkinson and MDS-PSP study group diagnostic criteria were included in the study [6]. Written informed consent was obtained from all patients for the publication of their videos. PSP patients were subclassified into variants such as PSP-RS (Richardson’s syndrome), PSP-P (parkinsonism), PSP-F (frontal), PSP-PGF (progressive gait freezing), PSP-CBS (cortico-basal syndrome), PSP-PLS (primary lateral sclerosis), PSP-C (cerebellar ataxia), PSP-SL (speech/language disorders), PSP-PI (postural instability), and PSP-OM (oculomotor dysfunction). The cases were classified based on the clinical assessment done by two movement disorder specialists [KG, SP]. Patients fulfilling the 2017 MDS diagnostic criteria for ‘probable’, ‘possible’, or ‘suggestive of’ PSP were included. Clinical phenotypes were assigned according to the corresponding MDS criteria. Demographic data such as age of onset, gender, disease duration, clinical characteristics, and PSP subtype were collected. Presence of dystonia, anatomical distribution, timing of dystonia, and characteristics across various PSP subtypes were recorded. MRI findings such as midbrain to pons ratio, magnetic resonance parkinsonism index (MRPI) and cortical atrophy patterns were recorded. The data was entered into an MS Excel sheet and statistically analysed.

In addition, a literature review was performed to evaluate the spectrum of dystonia in PSP. We searched PubMed on 8^th May 2026 and identified 271 articles on ‘Dystonia AND Progressive supranuclear palsy’, and 204 articles on ‘Dystonia AND Corticobasal syndrome’. The search query combined these terms using: articles containing “Dystonia” [Mesh] OR “Dystonia” [Title/Abstract]) AND (“Progressive supranuclear palsy” [Mesh] OR “Progressive supranuclear palsy” [Title/Abstract] OR “Corticobasal syndrome” [Mesh] OR “Corticobasal syndrome” [Title/Abstract] were included. After removal of duplicates, all relevant English-language articles were screened. Studies lacking details regarding dystonia, not focused on PSP or CBS, or published in languages other than English were excluded. A total of 29 articles were included in the detailed review (see Figure 1).

Figure 1

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics (IBM Corp., Armonk, NY). Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range [IQR]), depending on the distribution, assessed using the Shapiro-Wilk test. Comparisons between PSP patients with and without dystonia were conducted using Welch’s independent samples t-test to account for unequal variances and group sizes. Continuous variables across major PSP subtypes (PSP-RS, PSP-P, and PSP-CBS) were compared using one-way analysis of variance (ANOVA) for normally distributed data, and the Kruskal–Wallis test for non-normally distributed variables. Subtypes with very small sample sizes (n ≤ 4) were excluded from inferential statistical comparisons and are described descriptively. Categorical variables, including gender, subtype distribution, and presence of early dystonia (onset of dystonia <= 12 months), were analysed using the Chi-square test of independence or Fisher’s exact test where appropriate. All tests were two-tailed, and a p-value < 0.05 was considered statistically significant.

Results

Discussion

With the advent of newer diagnostic criteria, the clinical spectrum of PSP has expanded, allowing improved recognition of subtypes beyond the classical PSP-RS [6]. This classification is clinically relevant, as outcomes vary across PSP subtypes [20]. Subcortical variants such as PSP-P and PSP-PGF are associated with longer disease duration and slower progression compared to PSP-RS, which remains the most common and more aggressive phenotype [21]. While earlier literature has described PSP-CBS as relatively rare, nearly one-fourth of the cases in our prospective cohort were classified as PSP-CBS [9, 22, 23]. This higher frequency likely reflects systematic data collection and focused evaluation for dystonia and cortical features. As PSP-CBS is recognized as a distinct clinical phenotype within the 2017 MDS diagnostic criteria, its inclusion allowed evaluation of dystonia across the full spectrum of clinically recognized PSP variants. Although limb dystonia is a characteristic feature of PSP-CBS and may have contributed to the overall prevalence of dystonia, our subtype-specific analyses demonstrated distinct patterns of dystonia across the different PSP phenotypes. Within this evolving framework, dystonia emerges as an important yet under-recognized clinical feature of PSP.

We reviewed data from 29 studies evaluating dystonia in PSP. Most available literature consisted of case reports describing focal craniofacial manifestations, including facial, oromandibular, lingual, and laryngeal dystonia [24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36]. A few reports also described levodopa-induced or dopaminergic therapy-induced craniofacial and oromandibular dystonia [24, 29, 31, 33]. Across published studies, the reported prevalence of dystonia in PSP ranged from approximately 30% to 60% (see Table 4) [14, 15, 16, 17, 37, 38, 39, 40]. Much of the existing literature is retrospective and based on small cohorts, suggesting that dystonia may have been historically under-recognized and overshadowed by more prominent features such as vertical gaze palsy and postural instability [41]. Nevertheless, dystonia has clear clinical relevance, as it contributes to pain, impaired dexterity, and increased risk of falls, thus affecting the quality of life of the patient [42].

Dystonia in PSP shows considerable heterogeneity in its distribution. Some studies report a predominance of cranio-cervical involvement, particularly blepharospasm and cervical dystonia, while others describe a higher frequency of limb dystonia [15, 16, 39]. In our cohort, limb-only dystonia was the most frequent pattern, followed by cranio-cervical dystonia and combined involvement. This variation likely reflects differences in cohort composition as well as the prospective nature of our study, which allowed systematic detection.

A clear relationship between dystonia patterns and PSP subtypes was observed in our study. Dystonia was more frequently associated with PSP-RS and PSP-CBS, whereas patients without dystonia were more commonly classified as PSP-P and other subtypes. Limb dystonia was strongly associated with PSP-CBS, while cranio-cervical dystonia was predominant in PSP-RS. Apraxia of lid opening was also more common in PSP-RS, in keeping with the existing literature describing blepharospasm and ALO as characteristic features of this subtype [43, 44]. Importantly, dystonia appeared earlier in the disease course in PSP-CBS. Patients with PSP-CBS showed shorter latency to dystonia onset and a higher proportion of early dystonia. In several cases, dystonia was the presenting feature, most commonly in PSP-CBS. These findings suggest that focal limb dystonia may serve as an early clinical marker, even before overt cortical signs become apparent. Previous reports have similarly described asymmetric limb dystonia as an early manifestation in PSP-CBS and have documented cases in which dystonia preceded the development of ophthalmoplegia and gait impairment by several years [14, 38]. However, few studies have systematically evaluated the latency to dystonia onset across PSP subtypes. Our findings therefore extend the existing literature by demonstrating subtype-specific differences in the temporal evolution of dystonia.

The prominence of limb dystonia in PSP-CBS also reflects the overlap between PSP and Corticobasal Degeneration. Features such as asymmetric limb dystonia, apraxia, and alien limb phenomenon are traditionally associated with CBS but are increasingly recognized in PSP-CBS [45, 46, 47]. Current diagnostic criteria do not explicitly incorporate asymmetric limb dystonia into the classification of possible PSP-CBS [6]. As a result, some patients presenting with early asymmetric dystonia without cortical signs may be misclassified as PSP-RS. This highlights the importance of longitudinal follow-up, as such cases may represent evolving PSP-CBS [48, 49, 50]. Moreover, Rafal and Friedman observed that the severity of limb dystonia did not correlate with the severity of ophthalmoplegia in their patients, suggesting that dystonia may evolve independently of the classical oculomotor involvement [14].

Consistent with previous literature, asymmetric upper limb dystonia was the most frequent manifestation in PSP-CBS, while lower limb and cranio-cervical involvement were less common [23, 39, 51]. Earlier reports have suggested that dystonia is less frequent than rigidity in PSP-CBS; however, this was not observed in our cohort, where the majority of PSP-CBS patients exhibited dystonia [52, 53].

In contrast, PSP-P was less frequently associated with dystonia. Patients without dystonia were more commonly classified as PSP-P, which is consistent with previous literature describing this subtype as having fewer atypical features early in the disease course and often being mistaken for Parkinson’s disease (PD) [54, 55]. The potential evolution of PSP-P into PSP-RS remains debated, and it is unclear whether this reflects true disease progression or initial diagnostic misclassification [56, 57]. This distinction remains important for clinical interpretation and follow-up. Tremor, however, was most commonly observed in patients with PSP-P. In patients with PSP-CBS, distinguishing action tremor from cortical myoclonus may be challenging on clinical examination alone. As electrophysiological confirmation was not performed systematically, we cannot exclude the possibility that some cases clinically classified as action tremor represented cortical myoclonus. Future studies incorporating electrophysiological assessment are warranted to better characterize these hyperkinetic manifestations.

The pattern of dystonia also appeared to relate to the disease duration. Patients with combined limb and cranio-cervical involvement had longer disease duration compared to those with focal dystonia, suggesting that dystonia may evolve and generalize over time, reflecting wider neuroanatomical involvement as the disease progresses. However, disease duration did not differ significantly between patients with and without dystonia overall. Therefore, although progressive network involvement may contribute to changes in the distribution of dystonia, the cross-sectional design precludes conclusions regarding whether longer disease duration directly influences the development or progression of dystonia.

Neuroimaging findings supported these clinical observations. Patients with dystonia demonstrated greater brainstem involvement, as indicated by lower midbrain-to-pons ratios and higher MRPI values. This suggests more severe degeneration in brainstem structures.Although these imaging markers are well established for supporting the diagnosis of PSP, their relationship with dystonia has not been systematically evaluated. One possible explanation is that more extensive degeneration involving the midbrain, basal ganglia, and their interconnected brainstem networks disrupts motor control circuits, thereby predisposing to the development of dystonia. In PSP-CBS, cortical atrophy was also observed in the contralateral parietal region, corresponding to the side of limb dystonia. These findings support the role of distributed network involvement, including basal ganglia, brainstem, and cortical regions, in the pathogenesis of dystonia.The relationship between dystonia and dopaminergic therapy in PSP remains uncertain. Earlier reports have suggested that dystonia may be induced or exacerbated by levodopa in some cases, but this association has not been consistent with several case series describing this as a rare occurrence [14, 16, 17, 33]. In our cohort, no clear relationship was observed between dystonia and levodopa exposure, with several patients being levodopa-naïve at the time of onset. This supports the view that dystonia in PSP is more likely a manifestation of the underlying disease rather than a treatment-related effect, in contrast to PD [40].

The coexistence of parkinsonism and dystonia represents a clinical paradox. Parkinsonism is hypokinetic, whereas dystonia is hyperkinetic. This combination likely reflects dysfunction across interconnected basal ganglia, brainstem, and cortical circuits, rather than a simple dopaminergic deficit [58, 59]. Such network-level involvement explains why dystonia is frequently observed across atypical parkinsonian syndromes, including PSP [60, 61, 62].

Therapeutic management in PSP remains challenging due to widespread pathological involvement and clinical complexity [63]. In this context, targeted symptomatic therapies are important [64, 65, 66]. Botulinum toxin was used in one-third of our patients and resulted in meaningful clinical improvement. These findings support its role as an effective symptomatic treatment for dystonia, helping reduce pain and stiffness and improving the quality of life in selected patients [67].

Studies from South Asian cohorts have highlighted considerable phenotypic variability in PSP, with a broader spectrum of clinical features than traditionally described [68]. Our findings add to this observation, demonstrating a relatively higher proportion of PSP-CBS and a strong association between dystonia patterns and specific subtypes. This underscores the importance of careful phenotypic characterization and prospective evaluation. Taken together, these findings support the concept of a “dystonia spectrum” in PSP, encompassing cranio-cervical and limb manifestations, and reflecting underlying neuroanatomical involvement.

This study has certain limitations. The absence of pathological confirmation limits diagnostic certainty [69]. In addition, patients were assessed at a single time point, which precludes evaluation of the longitudinal evolution of dystonia and potential transitions between PSP subtypes. Furthermore, dystonia severity and its functional impact were not systematically assessed using validated dystonia-specific rating scales, limiting our ability to quantify its effect on disability and daily functioning. Future longitudinal and clinicopathological studies are needed to better understand the natural history of dystonia in PSP and its relationship with disease progression and underlying tau pathology.

Conclusion

Dystonia is a common and clinically significant feature in PSP, with distinct phenotypic patterns across subtypes. The differential distribution of limb and cranio-cervical dystonia, along with a trend toward earlier onset in PSP-CBS, highlights its potential value as a clinical marker. Importantly, targeted interventions such as botulinum toxin can provide meaningful symptomatic relief. Overall, our findings support a “dystonia spectrum” in PSP with distinct clinical and anatomical correlates. Future longitudinal and clinicopathological studies are warranted to better delineate its evolution and underlying disease mechanisms.

Ethics and Consent

We confirm that we have read the Journal’s position on issues involved in the ethical publication and affirm that this work is consistent with those guidelines. We also confirm that the patients/their family members have given written informed consent for the publication of their video.

Author Contributions

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.

Kartika Gulati (KG): 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B.

Sanjay Pandey (SP): 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B.