J Neurol Neurosurg Psychiatry 85:759-769 doi:10.1136/jnnp-2013-305532
  • Movement disorders
  • Review

The treatment of dystonic tremor: a systematic review

  1. Anthony E Lang1,3
  1. 1Division of Neurology, Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada
  2. 2Department of Neurology, Istituto di Neurologia, Università Cattolica del Sacro Cuore, Roma, Italy
  3. 3Department of Neurology, The Edmond J. Safra Program in Parkinson's Disease, Toronto, Ontario, Canada
  1. Correspondence to Dr Alfonso Fasano, Movement Disorders Center, Toronto Western Hospital, 399 Bathurst Street, Toronto, Ontario, Canada M5T 2S8; alfonso.fasano{at}
  • Received 1 April 2013
  • Revised 16 September 2013
  • Accepted 26 September 2013
  • Published Online First 28 October 2013


Tremor is one of the clinical manifestations of dystonia; however, there are no specific therapeutic trials evaluating the efficacy of treatments for dystonic tremor (DT), tremor associated with dystonia or primary writing tremor (PWT). We systematically reviewed the literature available up to July 2013 on the treatment of these tremors and retrieved the data of 487 patients published in 43 papers detailing the effects of given interventions on tremor severity. Treatment outcome was highly variable, depending on the specific type of intervention and tremor distribution. No specifically designed studies were available for the treatment of tremor associated with dystonia. As for the other tremors, drug efficacy was generally disappointing and a moderate effect was only found with anticholinergics, tetrabenazine, clonazepam, β-blockers and primidone; levodopa was only efficacious on tremor due to dopa-responsive dystonia. The largest amount of data was available for botulinum toxin injections, which provided a marked improvement, particularly for the management of axial tremors (head or vocal cords). In refractory DTs, deep brain stimulation of several targets was attempted. Deep brain stimulation of globus pallidus internus, thalamus or subthalamic area led to a marked improvement of dystonic axial or appendicular tremors in most cases refractory to other treatments. Few other non-invasive treatments, for example, orthotic device in PWT, have been used with anecdotal success. In conclusion, considering the lack of good-quality studies, future randomised controlled trials are needed. In absence of evidence-based guidelines, we propose an algorithm for the treatment of DT based on currently available data.


Different forms of tremor can be associated with dystonia. Dystonic tremor (DT) is defined as a postural/kinetic tremor occurring in the body region affected by dystonia.1 Generally these are focal tremors with irregular amplitudes and variable frequencies (mostly below 7 Hz). In many patients with DT antagonistic gestures lead to a reduction in the tremor amplitude. Moreover, DT usually worsens when the patient voluntarily moves the affected body part against the major direction of pulling caused by dystonia (eg, a patient with right torticollis might present more DT when he turns the head towards left). Conversely, DT may diminish and even disappear when the body part is positioned where the dystonia wants to place it. Some patients exhibit focal tremors even without overt signs of dystonia. They have been included among DTs because some of them later develop dystonia.2 Tremor associated with dystonia (TAD) is another type of tremor, which is present in a body region not affected by dystonia, but dystonia is present elsewhere (eg, bilateral hand tremor in a patient with cervical dystonia). This is a relatively symmetric, postural and kinetic tremor usually showing higher frequencies than typical DT.3 Dystonia gene-associated tremor is analogous to TAD, but it is an isolated finding in a patient with a dystonic pedigree but not personally suffering from dystonia.4

The prevalence of DT or TAD is not known. Head tremor in the context of cervical dystonia is the most common form of DT, found in up to 68% of these patients in one survey.5 In one Brazilian cross-sectional study, it was estimated that approximately 20% of patients with dystonia have postural tremor of the upper limbs (ie, TAD).6 This proportion does not differ between idiopathic and acquired dystonia but seems to be more common in cervical dystonia than in other forms.7 DT is still a debated entity and different definitions have been proposed by clinicians.8–10 Virtually, every dystonic syndrome may present with DT.

Primary writing tremor (PWT) is a condition in which tremor predominantly or only occurs during writing.11 Although there is some resemblance to essential tremor (ET) (where tremor is present on action, on maintenance of a posture, and may affect hand writing), its focal task-specific nature, the lack of response to propranolol and a well-documented effect of central cholinergic drugs12 have suggested that PWT may be more closely related to focal dystonia than ET.13 ,14 Other task-specific DTs have been reported. A 5 Hz jaw tremor induced only when speaking or attempting to drink from a glass has been reported15 ,16 as well as tremors of the lips induced by smiling17 or by speaking18 or orolingual tremor in patients with embouchure dystonia.19

The pathophysiology of these tremor conditions is largely unknown but is likely related to the basal ganglia abnormality postulated for dystonia itself.1 Consequently, in spite of their profound impact on patients’ functioning and quality of life, there is a paucity of information about the treatment of the DT syndromes.20 The present systematic review is aimed at collecting the available evidence guiding clinicians in the treatment of DT, TAD and PWT.


We searched PubMed until July 2013 with the search terms “Dystonic Tremor”, “Tremor Associated with Dystonia” or “Primary Writing Tremor” (yielding 321, 0 and 50 papers, respectively) and with the search terms “dystonia” AND “tremor” AND “therapy” OR “treatment” (yielding 693 papers). Additional 476 papers, retrieved from the references quoted in the papers found in our literature search, were also evaluated. Given the lack of meta-analyses or randomised controlled trials, we included all the available studies. We only considered papers on human subjects and written in English; among them, we only considered those detailing the effect of a given intervention on tremor severity (a total of 43 papers). No study on TAD fulfilled the aforementioned research criteria. Dystonia was retrospectively classified using the information from the articles according to recently published criteria.21

The outcomes of papers not reporting rating scales but only descriptive results were arbitrary transformed to the percentage of improvement (ie, no improvement=0%, slight improvement=1–25%, moderate improvement=26–50%, marked improvement=51–75%, tremor disappearance=76–100%). When assessing the effects on a specific body site, we only considered cases with tremor exclusively involving one region (eg, only head). When assessing the effects of a specific intervention, we did not consider cases undergoing combined treatments (eg, drugs and surgery).

Due to the small and inhomogeneous sample, statistical analysis was performed by means of nonparametric tests (Mann–Whitney U Test and χ2 with Yates and Fisher corrections, as needed).


No study on the treatment of TAD was found, whereas a total of 43 studies enrolling 487 patients with DT or PWT were included in the analysis (table 1): 12 were prospective, all but two were not blinded or controlled and only one was randomised; assessment tool was detailed in only 19 studies. The vast majority of studies enrolled a small sample size (median of 2 cases, range: 1–44), genders were almost equally distributed (male to female ratio of 1.2 : 1), the majority of patients were adults (range: 1–76 years) and disease duration ranged from 1 to 60 years. Idiopathic dystonia was the most common underlying aetiology (in 35 studies), other causes were acquired (largely perinatal anoxia and head injury) and ‘dystonia-plus’ (dopa-responsive dystonia—DRD—and myoclonus-dystonia) (in 6 and 4 studies, respectively). The majority of patients had tremor only involving head/trunk (63.0%), followed by hands (25.3%), vocal cords (9.0%) and lower face (0.8%); remaining cases had a variable combination of body sites involved.

Table 1

The features of the 43 studies (enrolling 487 patients with DT or PWT) retrieved by the systematic review of the literature


Different dugs in variable combinations were used in 71 cases (13 studies, table 2); no controlled or randomised studies are available and detailed information is generally missing (eg, in many of these studies drug doses were not specified). Overall, many drugs have been used unsuccessfully in patients with DT. Data on carbamazepine, acetazolamide, baclofen, amitriptyline, imipramine, fluvoxamine and sulpiride are too scanty to draw any useful conclusions as these drugs were anecdotally employed in single cases.22 ,23

Table 2

Tremor distribution and outcome according to the treatment

Given their consolidated use in the treatment of dystonia,24–26 many DT/PWT patients have been treated with anticholinergics. Trihexyphenidyl 4 to 10 mg has been the most commonly used.27–29 Anticholinergics were found to have slight to moderate effectiveness in reducing tremor amplitude in a total of 16 patients with hand or head DT.27 ,30–33 However, in three studies, a total of five patients did not achieve any benefit from anticholinergic treatment.22 ,23 ,30 Anticholinergics may also provide benefit in patients with PWT.12

Many patients were also treated with drugs commonly used in the treatment of ET.34 A positive effect of propranolol at unknown doses has been described in two studies, in which a moderate to marked improvement of arm tremor was reported in nine patients with idiopathic focal dystonia.30 ,31 By contrast, three other studies did not show efficacy of propranolol in a total of 10 patients with head and jaw tremor.22 ,23 ,35 Another β-blocker, timolol, was found to improve head DT in one patient at the dose of 15 mg.35 Primidone failed in one patient with head tremor35 and improved PWT in three of four cases.31

Benzodiazepines (in particular clonazepam at the dose of 0.5 to 3 mg) have been used in patients with DTs involving the arms, jaw and head. Tremor outcome was quite variable, ranging from no effect (two patients)22 to slight–moderate improvement (six patients).22 ,23 ,27 ,35 The total abolishment of tremor was reported in three patients, so that the authors defined this tremor as “clonazepam-sensitive dystonic tremor”.36 Interestingly, clonazepam dramatically suppressed the tremor presented in 14 patients believed to be affected by a ‘kinetic predominant’ variant of ET.37 The diagnosis of ET in these cases was supported by the absence of cerebellar signs, a positive family history of tremor and alcohol responsiveness; however, the frequency of tremor (3.5–6 Hz) and its poor responsiveness to propranolol in contrast to the striking response to clonazepam might also suggest a diagnosis of “clonazepam-sensitive dystonic tremor” in these cases.

Although tetrabenazine has been successfully used in dystonia,25 ,38 its role in the management of DT/PWT has been poorly investigated to date. Tetrabenazine was found to improve a case of PWT31 and one of three cases with dystonic jaw tremor.23 Levodopa is the treatment of choice in patients with DRD and accordingly it was found to improve dystonic postures and the associated tremor.39 The limited published data on levodopa treatment in patients with DT not related to DRD did not demonstrate a reduction in tremor amplitude, but instead occasionally showed worsening.40 ,41

Riluzole 100 mg/day was associated with a slight to moderate improvement of dystonic head tremor in one study42; unfortunately, no other report confirms these findings.

Finally, when addressing the outcome of the drugs lumped according to the main mechanism of action (anticholinergic, antiepileptic or β-blocker), no significant differences emerged (figure 1A).

Figure 1

Improvement of dystonic tremor or primary writing tremor according to the involved body site and the specific pharmacological or surgical approach (either stimulation or ablation). Numbers under the bars represent the respective number of studies and enrolled patients (note that the sum of studies does not match the total because patients with more than one tremor site have been excluded). Gpi, globus pallidus internus.

Botulinum neurotoxin

Botulinum neurotoxin (BoNT) injections have been employed in a total of 330 patients (8 studies) with PWT,33 head tremor43–46 or tremulous spasmodic dysphonia.47 As a consequence, its effectiveness in the management of these tremor disorders is well documented.48 Task-specific jaw tremor has been found to improve after BoNT injections into the digastric and masseter muscles.16 Fewer studies on PWT are available and BoNT has been reported to be successful in some of them.33 No conclusion can be drawn about the efficacy of BoNT in patients with arm tremor not caused by PWT, as no case has been reported.


A surgical approach (either deep brain stimulation—DBS—or ablation of different brain nuclei) was employed in 37 patients (20 studies). Severe cases of DT in the setting of a generalised dystonia have been successfully treated with DBS of the globus pallidus internus (Gpi),49 although none of the larger DBS trials for dystonia have specifically taken into account the effect on tremor. Seven cases were treated with only GPi stimulation or ablation: two had dramatic improvements,50 ,51 four had moderate improvement,52 whereas one with cervical DT did not improve.53 Stimulation of the thalamic ventralis intermedius nucleus (Vim) or ablations within the same region improved tremor in all cases in which they were employed: in 11 cases as unique target54–58 and in seven cases combined with GPi DBS.58–61 In addition, stimulation of the posterior part of the ventrolateral thalamic nucleus (VLp) also led to the improvement of tremor (notably, without improving dystonia) in four patients with dystonic arm tremor.62

Bilateral stimulation of subthalamus has been reported to improve cervical dystonia, dystonic head and hands tremor in two patients.63 ,64 Others reported that DBS of the subthalamic white matter, including the zona incerta, remarkably improved proximal DT that had been refractory to Vim thalamotomy.65 Moreover, stimulation of posterior subthalamic area provided encouraging results in three patients with dystonic arm tremor.66 ,67

Thalamotomy68 as well as Vim DBS55 ,69 ,70 have been successfully performed in patients with PWT.

Finally, when considering the outcome of specific surgical targets (thalamus, globus pallidus, subthalamic area), no significant differences emerged (figure 1B).

Other treatments

Other treatments were employed in 65 cases (five studies, table 2). Transcutaneous electrical nerve stimulation (TENS) in patients with dystonic arm tremor,71 selective peripheral denervation in patients with dystonic head tremor72 or other approaches did not show great efficacy in most of the reported cases with few exceptions. For example, dystonic jaw tremor completely disappeared in a single case treated with dental splint, which probably acted as a sensory trick.73 Recently, TENS has been also used in patients with PWT, again with inconsistent results.74 In contrast, a significant improvement with a simple hand orthotic device has been demonstrated, leading the authors to suggest that this non-invasive risk-free therapy could be useful during the initial assessment, before any other invasive option is considered.75

Comparison of outcomes according to tremor distribution

Figure 2 and table 2 detail the tremor distribution and the outcome of the given treatment. Treatment outcomes were highly variable. Overall, BoNT and surgery were the most successful strategies, significantly more effective than drugs (−69.9±21.3% and 69.9±25.1% vs −42.2±33.2%, p=0.02 for both comparisons). With respect to tremor distribution, BoNT was the most useful strategy in the management of axial (head/trunk or vocal cord) tremor, although no statistically significant difference emerged. As to hand tremor, BoNT and surgery were equally effective (−74.5±17.0% and −74.7±23.7%), significantly more than other strategies, though the outcome of BoNT was only based on PWT data (figure 2A). When specifically analysing the outcome of the PWT group, BoNT and surgery were found to be more effective than drugs (−74.5±17.0% and 93.2±7.5% vs −30.3±26.0%, p=0.0005 and p=0.01, respectively). By contrast, no data were available on the efficacy of BoNT treatment for DT involving upper limbs (figure 2B).

Figure 2

Improvement of dystonic tremor (DT) or primary writing tremor (PWT) according to the involved body site and the given treatment: the outcomes were highly variable, depending on the specific type of intervention and tremor distribution. Numbers under the bars represent the respective number of studies and enrolled patients (note that the total of studies exceeds 43 because some of them adopted more than a single therapeutic approach). Botulinum neurotoxin (BoNT) and surgery are the most successful strategies, significantly more effective than drugs. With respect to tremor distribution, BoNT was the most useful strategy in the management of axial (head/trunk or vocal cord) tremor, although no statistically significant difference emerged. As for hand tremor as a whole group (PWT or tremor in the context of upper limb dystonia), BoNT and surgery were equally effective, significantly more than other strategies (A). When splitting the hand tremor group into PWT and DT involving upper limbs, drugs are less effective in PWT than other strategies, whereas no data are available for BoNT in DT (B).

Discussion and recommendations

There are very few studies specifically addressing the treatment of the DTs and, more importantly, almost all of them are retrospective and not randomised. Moreover, in many studies the tremor outcome was poorly documented as were the assessment tools, which were very rarely based on objective measurements (electromyography (EMG) or kinematics). These limitations account for the lack of evidence-based guidelines in the treatment of these tremor disorders.34 ,76

In addition, it is impossible to know whether the reviewed studies reliably distinguished TAD and DT. In order to clarify this issue, we also reviewed the literature on TAD but unfortunately we did not find any study specifically addressing its treatment. TAD is usually treated with the medications used for ET77 and believed to respond in a comparable fashion, although there is little available data to support this supposition. Consequently, the data on the effectiveness of β-blockers or primidone in DT are limited by these diagnostic difficulties. Indeed, sometimes it is very difficult to distinguish between different tremor conditions, particularly with writing tremor, and isolated head or voice tremor.77 Therefore, it is possible that some of the DT patients reported in the literature were actually affected by TAD or even ET. In order to limit this bias, we focused our recommendations on DT and PWT.

In this systematic review, we found that the outcome of all the possible interventions (drugs, BoNT, surgery and alternative approaches) is highly variable, depending on the specific type of intervention and tremor distribution. Figure 3 provides a possible algorithm for the treatment of DT/PWT, admittedly based on very limited evidence available to date. On the basis of tremor distribution, BoNT is the most useful strategy in the management of axial (head or vocal cord) tremors and has the largest amount of data in the literature confirming its efficacy. With respect to appendicular tremor, BoNT and surgery might be effective in PWT, whereas no data are available for DT of the upper limbs. Drugs are generally less efficacious but could be tried especially when tremor involves different body regions, thus allowing the use of BoNT in a stepwise approach (in order to further improve the regions not adequately controlled by drugs). Apart from tremor in DRD, which typically responds to levodopa, the first-line drugs should be anticholinergics, as their use has been described in large samples of patients.27 ,30–33 Second-line drugs have been less extensively investigated; trials might include tetrabenazine, due to the reported efficacy in dystonia25 and DT/PWT,23 ,31 and clonazepam, especially in the light of the description of “clonazepam-sensitive dystonic tremor”.36 ,37 Third-line drugs could be β-blockers, primidone and many others occasionally found to improve DT. Drug tolerability is an important issue for clinical decision-making as most of adult patients experience unwanted effects at lower doses than those needed for yielding a benefit. Even though the available literature did not specifically address the side effects produced by medications, the choice to rely on medications (eg, increasing the dose or adding another compound) or to adopt other approaches (ie, BoNT and/or surgery) should be based on these aspects as well.

Figure 3

A possible algorithm for the treatment of dystonic tremor (DT) or primary writing tremor (PWT) on the basis of the limited evidence available to date. Botulinum neurotoxin (BoNT) is the most useful strategy in the management of axial (head or vocal cord) tremors, whereas appendicular tremor—with the exception of PWT—should be firstly treated with drugs, thus allowing the use of BoNT in a stepwise approach (in order to further improve the regions not adequately improved). By contrast, BoNT might be first-choice therapy in PWT, followed by drugs or surgery. Surgery (DBS) should be considered only when the disability derived by tremor overcomes the risks of its invasiveness. Unilateral procedures can be tried in case of appendicular tremor, whereas bilateral surgery is indicated for head tremor. As for the target of DBS, Vim and Gpi are the most used ones, based on the predominance of tremor (Vim) or dystonic postures (Gpi). The combined approach Vim+Gpi might be considered in case of failure of the single target procedure. Other targets (VLp, STN and surrounding areas, ie, PSA or cZi) might be considered in very selected cases. cZi, caudal zona incerta; DBS, deep brain stimulation; Gpi, globus pallidus internus; PSA, posterior subthalamic area; STN, subthalamus; Vim, ventralis intermedius nucleus of thalamus; VLp, posterior part of the ventrolateral thalamus.

Finally, DBS could be considered only when the disability derived from tremor overcomes the risks of its invasiveness. For example, the disability caused by PWT can vary from mild to considerable, depending on the profession of the patient. The handicap that it produces in Western cultures has generally not been considered sufficient to merit the risks involved in stereotactic surgery. However, in Japan where calligraphy is an important occupation, neurosurgery has been successfully performed for this condition, as it can threaten the professional career of the patient.68 Taking into account the potential ameliorative effect on the other dystonic features, GPi should be viewed as the preferred target for either stimulation50 or ablation51; thalamic stimulation may be added in cases in which the benefit is insufficient,59–61 because it can also alleviate the tremor but can occasionally lead to worsening of dystonia itself.54 ,56–58 Stimulation of thalamic VLp is considered as less efficacious in dystonia and tremor treatment than GPi DBS.62 Noteworthy, a recent study challenged this assumption, concluding that DT patients with a mild dystonia should be considered for Vim DBS while the coexistence of severe DT and dystonia may be successfully controlled by combined GPi and Vim DBS.52 Subthalamic area might be an interesting target for stimulation, given the early experiences with lesions and the hypothesis that many fibres can be modulated in this ‘bottle-neck’ anatomical region78; however, the available data are too scanty to support its use beyond an experimental approach.

In conclusion, paradoxically the most important result of our systematic review is the demonstration of the lack of good-quality studies addressing the treatment of DT. For instance, although surgery seems to be a useful treatment for tremor refractory to drugs or BoNT, the outcomes of either DBS or ablations of different brain nuclei have been reported in only 37 patients (20 studies); even the recent GPi DBS trials79–81 did not focus on the outcome of tremor, although they were conducted according to the criteria of evidence-based medicine. Future randomised controlled trials are needed, particularly those with active comparators, thus allowing the comparison of different treatment approaches (eg, BoNT vs drugs or GPi vs thalamic DBS).


  • Contributors AF: conception and design, analysis and interpretation of data, drafting the article and revising it critically for important intellectual content; final approval of the version to be published. FB: analysis and interpretation of data, drafting the article, final approval of the version to be published. AEL: conception and design, revising the article critically for important intellectual content, final approval of the version to be published.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement We have the excel sheets where we collected the information retrieved from the literature.


Visit the full archive of podcasts for JNNP here >>

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.