The impact of neuropsychiatric symptoms on patients’ lives cannot be overstated - an illustrative example is social phobia and social anxiety. Social phobia is common in cervical dystonia (CD), attributable to a sense of body deformation and anticipation of negative perception by others related to the disease. This perception is out of proportion to motor severity, leading to self-stigmatization: a self-conscious anxiety in public spaces and social situations, increasing isolation and depression [10, 11]. There seems to be an overall hypervigilance, an erroneous assignment of salience to unimportant stimuli that is common in CD with NMS. There is known network dysfunction in dystonia involving salience network structures, like the amygdala, and structures interacting with the salience network, like the cerebellum [6, 12–14]. In fact, the cerebellum modulates input related to sensorimotor, salience, and default mode networks, implicating it across nonmotor symptoms like pain and sensation perception, depression and anxiety [12]. Indeed, resting-state functional neuroimaging studies have found abnormal activity in salience networks as directed by the cerebellum, correlating with symptoms of depression [12]. Social phobia illustrates the challenge in identifying which NMS are primary to the disease and which are a result of other NMS, or of the motor symptom impact of dystonia. For instance, higher rates of disordered eating and substance use disorder are found in patients with dystonia, suggesting a primary etiology [5, 11, 15]. However, as shown above, a symptom like social phobia could lead to isolation and subsequent depression and maladaptive coping strategies. Anxiety and depression are also commonly reported in writer’s cramp, laryngeal dystonia, oromandibular dystonia (OMD), and blepharospasm, and are unrelated to motor symptom severity, even predating the onset of motor symptoms again suggesting a primary role in the disorder [16–21]. 32.3% of patients with isolated dystonia report a history of suicidal behavior, a rate significantly higher than that of the general population, highlighting the severity of impact of these symptoms on patients’ wellbeing [22].

Fatigue and excessive daytime sleepiness (EDS) are prevalent and significantly impact the QoL of patients with dystonia, influencing their perceptions of the disease and satisfaction with treatment [4, 10, 16, 20, 23–25]. Compared to healthy controls, patients with dystonia exhibit reduced daytime activity, as measured by self-report and accelerometry data [26]. Sleep disturbances are present in up to 70% of patients with CD [10]. Across various dystonia subtypes, patients experience higher rates of insomnia, poorer sleep quality, reduced sleep duration, and lower sleep efficiency, as well as greater EDS, compared to healthy controls [5, 26]. Dysregulation of circadian rhythm networks, particularly cerebellar hyperactivity, has been proposed as a potential contributing factor to these sleep-wake disturbances [7].

Pain is a prevalent feature of dystonia and has a significant impact on health-related QoL [5, 27]. Patients with blepharospasm frequently experience eye pain and photophobia [17]. Similarly in CD, pain commonly reported and relates to complex mechanisms, including increased muscle contraction affecting muscle spindles and their associated pain receptors, alterations in peripheral sensitivity due to neuropeptide activity, abnormal processing of pain pathways, and impaired inhibitory signalling [28]. A systematic review has demonstrated that botulinum toxin (BoNT) treatment can significantly alleviate pain, as evidenced by improvements in the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS) pain scores [28]. Although BoNT treatment generally provides effective pain control, pain may recur between injection cycles and its persistence is associated with lower patient satisfaction with treatment [29, 30]. Deep brain stimulation (DBS) has shown variable impact on pain outcomes in prior studies, with some research indicating no effect and other studies showing significant improvements in pain scores independent from motor symptom improvement, suggesting that pain may be a primary and distinct symptom of dystonia [28].

Altered cognitive function is one of the least understood NMS of dystonia, as clinical changes are subtle and nonprogressive, and are difficult to glean from common screening measures like the Montreal Cognitive Assessment (MoCA) or the Mini Mental Status Exam (MMSE) [31]. Social cognition has been a focal point of cognitive research in CD, with deficits in this domain having substantial implications for comorbid anxiety [32]. Patients with CD demonstrate a reduced ability to recognize emotional expressions in both auditory and visual tasks compared to controls in some studies [33, 34]. Additionally smaller studies have reported lower performance in spatial working memory, word retrieval, and working memory in individuals with CD [35].

Patients with myoclonus-dystonia and CD both demonstrate impairment in their explicit sense of agency, a cognitive function dependent upon one’s ability to predict their actions’ impact on the external world [36, 37]. This deficit may implicate the cerebellum, which plays a key role in predicting feedback from self-generated actions and may indirectly contribute to dysfunction in the sense of agency network [36–39]. Some small studies have found executive deficits and slowed cognitive speed in CD patients; specifically, deficits of inhibitory control seem to be a feature of CD, even when controlling for patients on benzodiazepines [40].

The purpose of this study is to provide a descriptive review of recent publications related to the treatment of NMS in dystonia published over the past 5 years, and to provide a comprehensive description of this literature. We aim to reveal the benefits and limitations of common dystonia treatment approaches on NMS, and to uncover areas requiring further study.

Surgical interventions may be utilized in medication and toxin-refractory cases of dystonia. Various targets were investigated over the past 5 years including DBS of the globus pallidus internus (GPi), subthalamic nucleus (STN), and ventral intermediate nucleus of the thalamus (Vim); ablation of pallidothalamic tracts by focused ultrasound and radiofrequency; and cervical spinal cord stimulation. Motor symptoms improved with many of these interventions, however, the response of NMS was variable. There were 13 publications in the past 5 years (six case series; one randomized controlled trial [RCT]; five prospective, non-controlled studies; and one prospective, crossover, double-blind study) that reported on the impact of surgical interventions on NMS as secondary outcomes using a variety of scales [Table 1].

Eleven publications (one RCT, eight prospective non-controlled studies, one prospective controlled study, and one retrospective case-control study) investigating the response of NMS to BoNT injection were reviewed, with most studies focusing on CD populations [Table 2].

Eight papers were published investigating physical therapy and psychotherapy approaches to treating dystonia and its NMS. Of these, three were RCTs, one was an open-label, proof-of-concept study (Table 3), and three were descriptive analyses based on interviews with patients and healthcare professionals.

In one RCT, 40 patients with CD received specialized physical therapy, while 32 patients followed a standard physical therapy protocol. Both groups showed significant improvements in pain and mood as measured by TWSTRS pain scale, PRNS, BAI, and BDI. However, no between-group differences were observed at 12-month follow up [70]. Another negative crossover RCT, involving 17 patients with CD, compared BoNT treatment alone, BoNT with kinesiotaping, and BoNT with sham taping. No differences were found between groups in NMS, as measured by the Dystonia Non-Motor Symptoms Questionnaire, HADS and PSQI [71]. Xu et al. (2024) conducted a proof-of-concept study on the effects of vibrotactile stimulation on pain management in 44 patients with CD. While modest improvements in pain were observed, the study lacked a control group, limiting the interpretation of results [72]. McCambridge et al. (2019) published a descriptive analysis of physical activity in 263 patients with dystonia, including 40% with CD. The majority of participants reported worsening of motor and NMS, such as pain and fatigue, with high-impact physical activities (e.g., running, jogging, and brisk walking), while low-impact exercise (e.g., light walking, general stretching, yoga, or pilates) were less likely to exacerbate symptoms [73].

Four papers focused on psychotherapeutic or behavioral interventions for managing NMS in dystonia. Wadon et al. (2021) investigated an internet-based cognitive behavioral therapy (CBT) program as compared to routine care in a small, randomized controlled study of 20 patients with CD over a 6-month period. Although participants in the intervention group showed trends toward improved anxiety and depression scores, and reported that the treatment was beneficial, no significant differences were observed between the two groups on measures including BDI, HAMD, HAMA, and the Generalized Anxiety Disorder-7 [74]. Detari et al. (2023) interviewed 14 healthcare professionals - coaches, physicians, researchers, and physical therapists - on their treatment approaches for managing NMS in musician’s dystonia. Strategies varied widely from addressing underlying perfectionism to motor retraining exercises [75]. O’Connor et al. (2023) surveyed 118 participants with CD, finding that illness perception and coping strategies accounted for 59% of the variance in anxiety, and 61% of the variance in depression and health-related QoL, highlighting the importance of these factors in the psychological adjustment to CD symptoms [76]. Zetterberg et al. (2024) interviewed patients with CD to explore their experiences with symptom management, identifying the use of coping strategies, acceptance of their condition, and adherence to BoNT treatments as key factors contributing to improved wellbeing. Common frustrations included the impact of stress on symptom severity, negative self-image leading to avoidant behaviors, and persistent pain and fatigue. These findings offer valuable insights into the patient experience and suggest areas of focus in multidisciplinary treatment strategies [77].

There were some notable risks of bias impacting this systematic review. The review protocol was not published, and was expanded after the initial PubMed review to include the other databases. Reporting bias remains a possibility in that non-significant results may have been overlooked in our selection and presentation of results. Within the PubMed search results, evidence selection bias is highly likely, in that studies with negative outcomes may not have been published. Within the primary studies, common biases included performance bias, given the lack of a control arm in most reports, and outcome reporting bias, especially as NMS were mostly secondary outcomes and non-significant findings were less likely to be included for publication. Figure 2 illustrates the presence of bias within the primary studies.

The overall quality of the studies reviewed was limited by small sample sizes and the absence of control groups. Additionally, the systematic analysis of the literature was limited by the use of only two keywords and future reviews should include specific NMS terms as keywords to capture a larger number of reports. Meta-analysis of reviewed data is precluded by the heterogeneous use of scales in NMS assessment, a well-documented challenge in dystonia research. The development and routine use of NMS-specific, disease-specific scales has been proposed to address this issue, similar to their application in Parkinson’s disease [8]. Several validated scales have been developed for distinct dystonia subgroups, enabling standardized outcome reporting. However, these scales generally focus on motor symptoms, limiting their capacity to comprehensively assess NMS outcomes. For example, the TWSTRS-Pain subscale used in CD populations may be less informative than other, more in-depth pain measures like the McGill Pain Questionnaire. Similarly, the TWSTRS-Psych may not capture the nuances of anxiety and depression as effectively as the Beck and Hamilton inventories, or the STAI. The growing recognition of NMS in dystonia is relatively recent, and the field of research remains in its early stages. However, progress has been made in the past 5 years, in this regard, including the development of a digital patient-centered outcome tool to measure NMS in CD, the Pain Scale for Adult-Onset Idiopathic Dystonia (PIDS) and its validation in CD, the Dystonia Pain Classification System (Dystonia-PCS), and the OMD Rating Scale (OMDRS), which includes NMS domains [80–83]. Future NMS scale development must strike a balance between specificity, depth, and ease of use in large study populations.

Most of the studies reviewed here primarily assessed motor symptoms, with NMS impact measured as secondary outcomes. This common theme underscores the need for high-quality research focusing on NMS improvement as a primary outcome in dystonia treatments. While still a subject of ongoing debate, there is a consensus that NMS in dystonia are distinct from motor symptoms and should be considered independent features of disease. Several studies support this view: da Silva-Júnior et al. (2022) found a high prevalence of NMS in patients with idiopathic isolated dystonia, without correlation to motor severity [3]. Li et al. (2020) reported that psychiatric symptoms in generalized dystonia did not correlate with motor severity, suggesting that NMS may be primary features of the disease [84]. Foley et al. (2017) concluded that cognitive symptoms and mood disorders in isolated dystonia occur independently of motor symptoms [85]. This view is further supported by studies identifying clinical subtypes of cervical dystonia based on NMS severity rather than motor symptom severity [86, 87].

Recent studies have identified distinct NMS-predominant subgroups within the cervical dystonia population. Wadon et al. (2022) and Costanzo et al. (2021) demonstrated two phenotypes: one motor-predominant, with fewer NMS, lower disability, and better QoL, and another NMS-predominant, with higher rates of depression, anxiety, sleep dysfunction, cognitive complaints, and pain [86, 87]. These subgroups did not differ in sex, motor severity, duration of dystonia, or the presence of tremor or alleviating maneuvers, though a potential age difference was noted, with the NMS-predominant group being older [86]. Further research is needed to verify these findings.

Regarding sex differences, while no significant differences between subgroups were observed in the aforementioned studies, other research has shown sex-based disparities in the prevalence of NMS in dystonia. Yang et al. (2021) found that women with blepharospasm had a higher frequency of NMS and were more likely to have multiple comorbid NMS [88]. Additionally, women with CD report worse health-related QoL compared to men [89]. The impact of sex on dystonia and NMS warrants further investigation, especially given the increased penetrance of CD in women, who are affected at least twice as often as men [90]. Notably, temporal discrimination threshold abnormalities, a hallmark feature of CD, are fully penetrant in unaffected first-degree female relatives, but only 40% penetrant in unaffected male relatives [89]. Further studies are needed to understand sex-based differences in disease susceptibility and phenotype.

In conclusion, NMS in dystonia significantly impact patient outcomes and require increased attention in both the clinic and research. An illustrative example of progress in this area is the work by Martino et al. (2023), who developed a care pathway for addressing mood symptoms in adult-onset isolated dystonia [91]. Further investigation is needed to refine dystonia subgroup phenotyping, explore sex differences, and standardize NMS rating scales across studies. These efforts would facilitate meta-analyses, improve the reproducibility of treatment outcomes, support the standardization of clinical practice, and improve patient outcomes.