Task-specific tremor (TST) is a specific type of action tremor occurs only or predominantly when an affected individual is performing or attempting to perform a specific task. The clinical entity remains heterogeneous. TST can be induced by active movements or by adopting a specific position simulating the task, and is usually non-progressive [1]. TSTs mostly involve the upper limbs, especially dominant limbs, during specific skilled tasks though sometimes the orolingual area (e.g., lip, chin) is affected [2]. TSTs have a mean frequency of 5–7 Hz (range 3–8 Hz), which may be accompanied by a jerky component in some cases [3]. To date, there are no known patients with TSTs involving the lower limbs.

The involved tasks are variable. For example, most commonly, TSTs are elicited while writing and referred to as “primary writing tremor” (PWT) [4]. TSTs in musicians (TSTM) occur mainly while playing an instrument, with cases reported in string instrumentalists [5–7] or flutists [8]. TSTs can also occur during other daily activities. For instance, finger tremors when playing carroms [9], lip tremors while drinking [10–13], chin tremors only while brushing teeth [14], finger tremors with the use of scissors [15], and wrist tremors during weightlifting [16–18]. Given its various clinical subtypes, limited case numbers, and diagnosis uncertainty, there are no accurate numbers for the prevalence and incidence of TST among general populations.

Despite limited reported cases, PWT and TSTM are the two most prevalent TST subtypes. Two case series with 21 and 56 patients with PWT, respectively [4, 19], reported the mean age of onset to be around 50 years of age (broad range: 16–72 years), with a male predominance (70%–95.2%), and up to 33%–44% of the patients reported a positive family history of PWT. These findings suggest that there may be a possible genetic susceptibility to PWT, in addition to environmental factors. However, no causative gene or mutation has been identified so far. In contrast to PWT cohorts, a case series of 23 musicians with TSTM reported the age of onset to be 44.6 ± 13.6 years, with equal gender distribution, and without a positive family history. Besides, TSTM was associated with a relatively long average duration of playing an instrument (35 years) prior to tremor onset [6]. The variable clinical features implied that different types of TST may not share an identical pathophysiology.

Some recent studies have alluded to the possibility that TST may be an early symptom before the onset of other parkinsonism features in patients with Parkinson’s disease (PD) [14, 20, 21]. A case series reported that three of the fve patients with PWT (onset age between 46 and 76 years), later developed PD (within 1–5 years of PWT onset). All three patients had reduced uptake in DaTscan contralateral to the tremor-affected side, and were refractory to propranolol/primidone, but responded to carbidopa-levodopa treatment [20]. However, the interval between TST onset and a diagnosis of PD has been reported to be even longer (average: 13.66 years) in another case series [21]. Currently, the relationship between TST and PD is unclear.

The pathophysiology of TST has been debated in the past decade. The clinical presentation of being focally distributed and task-specific, sometimes with abnormal posturing [4, 22], the presence of coactivation and overflow of muscular activity to adjacent muscles in electromyography [23], a better response to botulinum toxin therapy, suggests a possible correlation between TST and dystonic tremor syndromes (DTS), which included both dystonic tremor and tremor associated with dystonia [24]. The alleviation of symptoms by gestes antagonistes has been described as a clinical hallmark characteristic of dystonia but was only reported in one patient with TST in the previous literature [25]. On the other hand, many studies have reported considerable symptomatic relief of TST by ethanol or propranolol [26–28], and identified a genetic susceptibility, with one case reported with bilateral involvements [26], which points to a possible relationship between TST with ET. Therefore, whether TST is an isolated tremor syndrome, a tremor associated with task-specific dystonia, or a variant of ET, remains uncertain [29, 30]. In this review, we explored the current electrophysiological and functional neuroimaging studies of TST and discussed the possible pathophysiology of TST.

Functional magnetic resonance image (fMRI) techniques provide a non-invasive assessment of the structural, functional, and metabolic alterations of neurological disorders. Numerous imaging studies have been performed in patients with ET and DTS, but the studies on TST are sparse.

In an early fMRI study involving three patients with PWT, PWT was shown to be associated with increased activity of the cerebellum bilaterally, with a more pronounced area of activation on the side ipsilateral to the affected hand, along with bilateral activation of the parietal lobule with a more pronounced activation on the side contralateral to the affected hand [68]. Conversely, recent studies have shown opposite findings in the cerebellum. For example, Hirdesh Sahni et al. showed overactivations of the primary and supplementary motor areas and reduced activity in the cingulate motor area and the cerebellum in six patients with PWT [69]. Another recent study using voxel-based morphometry and diffusion tensor imaging (DTI) found that there was predominantly gray matter atrophy in the frontal lobe and the cerebellum, along with white matter changes in the frontal lobe and the cingulum in patients with PWT when compared with healthy subjects [70]. Lenka et al. further applied graph theory-based neural network analysis to fMRI to explore connectivity during the resting state of the functional brain [71]. In this study, the brain was modeled as a complex functional network with two measurements including “clustering coefficient”, which quantified the local connectivity as an index of network segregation; and “path length,” which quantified the global connectivity as an index of network integration. The results of this analysis demonstrated that patients with PWT had a significantly lower clustering coefficient and a higher path length in the bilateral medial cerebellum, right dorsolateral prefrontal cortex, and left posterior parietal cortex, suggesting significant disruptions of the small-world brain architecture in these regions.

To our knowledge, to date, there are no studies that directly compared patients with TST to patients with ET or DTS. However, numerous studies have discussed the structural, functional, and metabolic presentations between patients with ET and patients with tremors associated with dystonia. Through understanding the difference between ET and DTS in the MRI images may shed lights on the pathophysiology of TST. Findings from DTI studies suggest an increased mean diffusivity and a decreased fractional anisotropy of the cerebellum in patients with ET, indicating possible microstructural tissue damage and a loss of cellular integrity [72–74]. fMRI studies in ET patients have further clearly demonstrated abnormal cerebellar function and altered connectivity in the cerebello-thalamico-cortical circuitry [75]. Another recent MRI study demonstrated grey matter hypertrophy of the thalamus and motor cortex in the cerebello-thalamo-cortical circuit among patients with DTS [76]. The author concluded that deficient input from the cerebellum towards the thalamo-cortical circuit with hypertrophy of the thalamus, may play a key role in the generation of DTS. To compare patients with ET and DTS, a functional MRI during a grip-force task as a proxy of tremor-related cerebral activity showed similar reduction of functional connectivity in the cerebellum in both patients with ET and DTS [77]. Nevertheless, when the region of interest was outside the cerebellum, compared to patients with ET, those with DTS have more widespread areas of reduced functional connectivity in the cortical regions when the seed regions were placed either in cortical regions, such as the sensorimotor cortex and inferior parietal lobule or subcortical areas, such as globus pallidus interna. Another study using multi-modal imaging combining resting-state functional MRI and DTI showed reduced functional connectivity between the cerebellum and dentate nucleus bilaterally for the ET group but not the DTS group, compared to healthy subjects [78]. From the treatment response viewpoint, both ET and DTS improved after deep brain stimulation were significantly correlated to the stimulation of the dentato-rubro-cortical tract, while only DTS, but not ET, presented a significant additional correlation to the pallidothalamic tract [79]. These findings point towards a second pathophysiological mechanism involving the basal ganglia in patients with DTS. Taken together, connectivity dysfunction of both the cerebello-thalamo-cortical and the basal ganglia-thalamo-cortical networks may both be involved in driving the pathophysiology of DTS [80, 81] which was different from ET who presented mainly cerebello-thalamo-cortical connectivity impairment.

There is an ongoing debate about whether TST is a distinct disease entity, a variant form of ET, or a focal task-specific dystonia with dystonic tremor. Based on current evidence, it is reasonable to classify TST as a subtype of DTS, rather than a subtype of ET. Clinically, TST occurs when the patient performs a specific task, similar to patients with writer’s cramp who present with dystonic postures when they are writing. Moreover, TST usually affects the dominant hand only, unlike ET, which involves both sides bilaterally. On the contrary, the findings of electrophysiological studies suggest that TST showed normal spinal inhibitory circuits and motor cortical excitability, but a disinhibited brain stem inhibitory circuitry is evident from the reduced EBCC and R2BRrc. The loss of LICI modulation by PAS and reduced SICI are present in both TST and DTS patients. Nevertheless, patients with dystonia usually demonstrate other forms of hyperexcitability of the motor cortex, for example, a reduced CSP, or hyperexcitability of the spinal cord and a loss of reciprocal inhibition. Therefore, the overall electrophysiological characteristics of TST imply that the underlying pathophysiology is not entirely identical to dystonia.

Due to the variable methodologies used in fMRI studies and the sparsity of fMRI studies in patients with TST, it remains inconclusive whether TST is distinct from ET or DTS. Although cerebellar functional connectivity impairments were observed in PWT, it could also represent a fundamental abnormality for any tremor syndrome, since patients with ET and DTS also demonstrate a decreased connectivity in the cerebello-thalamo-cortical circuits. Whether the additional basal ganglion-thalamo-cortical circuits are involved, or whether a more widespread reduction in functional connectivity in the cortical regions occurs in the patients with TST is still uncertain. From the structural point of view, whether patients with TST presented thalamic hypertrophy, which implied dystonia characteristics, may be another clue to interpret the pathophysiology of TST in the future. All these aspects may be critical to distinguishing the underlying pathophysiology between TST, ET and DTS. Table 1 compares the different features, such as the clinical presentation, electrophysiological findings and fMRI results, between TST, ET, and DTS.

Although the association between TST and Parkinson disease (PD) is less depicted in the previous literature, especially in the electrophysiological assessment, however, a recent study of eight patients with TST who later developed into PD showed an optimal response to apomorphine but was refractory to other dopaminergic agents [82]. Therefore, TST responses to the apomorphine test might provide an early hint to indicate that TST may be full-blown to PD in the future. Figure 1 delineated the current position of TST in the tremor syndrome by integrating electrophysiological and fMRI findings, and indicated the knowledge gap that might help clinicians to better understand the pathophysiology of TST in the future.

There is still a lack of comprehensive and consistent understanding of TST due to limitations in the currently available studies. First, most studies have a small sample size, with a large intra-subject variability. Second, the inclusion criteria in each study are varied, and some studies conducted even before the development of tremor classification and the definitions for the patient groups are ambiguous and non-standardized in some studies. For example, dystonic tremor or tremor with dystonia may not be necessarily shared the same pathophysiology, although they both can be sorted in the same disease population as “dystonic tremor syndromes” in most of the studies. A significant portion of the studies were conducted before the development of tremor classification criteria [83]. Third, the different methodologies and paradigm designs used in each study, including both electrophysiological and neuroimage aspects, have led to inconclusive results. Fourth, most of the TST studies mentioned in this review focused on PWT patients, which might only represent a specific subtype of TST though still giving us an insight of the picture of the underlying pathophysiology. Moreover, most studies lack long-term follow-up. Thus, additional neurological signs that emerge over time may be left undetected (e.g., Parkinson’s disease), which may have led to unreported but critical findings, misinterpretations, or incorrect inferences.

Findings from the available electrophysiological and fMRI studies on patients with PWT suggest that TST may be an isolated tremor entity or a spectrum of DTS, rather than an ET variant. This is consistent with the latest consensus statement on tremor classification from the task force on tremors of the International Parkinson and Movement Disorder Society [83], in which TST has been separately classified as a specific action-induced tremor, different from DTS or ET. Regular follow-ups and comprehensive symptoms documentation with longitudinal electrophysiological and neuroimaging assessment are the keys to fully understanding the underlying pathophysiology of each individual patient with TST.