Essential tremor plus rest tremor: current concepts and controversies

Demographics and clinical features of ET + RT have been extracted from 23 studies involving 3500 patients. A summary of these studies is provided in Table 1. Prevalence of ET + RT ranged from as low as 1.9% (Louis et al. 2015) to 72.9% of ET cases (Rajalingam et al. 2018). Importantly, the prevalence of RT in ET has reported to be dependent on the recruitment setting, being the lowest in population-based setting and highest in a brain bank study (Louis et al. 2015). In many (Rajalingam et al. 2018; Bellows and Jankovic 2021; Campbell et al. 2022; Prasad and Pal 2019; Peng et al. 2022; Gilmour et al. 2021; Steffen et al. 2020), but not all studies (Pandey and Bhattad 2021; Manorenj et al. 2019), ET + RT was among the commonest ET-plus subtypes.

In the vast majority of studies reporting this information, patients with ET + RT were older than ET (Louis et al. 2015, 2005; Rajalingam et al. 2018; Campbell et al. 2022; Peng et al. 2022; Louis and Jurewicz 2003; Djaldetti et al. 2008), with only one study reporting a similar age between ET + RT and ET (Louis et al. 2012). There are conflicting evidence of whether the latter result depends on longer disease duration of ET + RT with respect of ET without RT. Four studies found a longer disease duration in ET + RT than in ET (Cohen et al. 2003; Rajalingam et al. 2018; Prasad and Pal 2019; Djaldetti et al. 2008), whereas three did not (Louis and Jurewicz 2003; Louis et al. 2005, 2012). One study (Louis et al. 2015) found this relationship to be true only in one of four examined settings (ie, population-based, genetic study, study of environmental epidemiology, brain bank) and another study found this association only in young onset cases (< 40 years) but not in patients with older age at onset (Louis et al. 2021). In keeping with the latter finding, the majority of the studies (Louis et al. 2015, 2005; Peng et al. 2022; Louis and Jurewicz 2003; Martinelli et al. 1987), but not all (Rajalingam et al. 2018; Djaldetti et al. 2008; Louis et al. 2012), reported patients with ET + RT having a higher age at onset than ET. Many studies found ET + RT to be more severely affected than ET cases (Louis et al. 2015; Peng et al. 2022; Louis and Jurewicz 2003) and to have more likely cranial-cervical tremor (Cohen et al. 2003; Louis et al. 2015), although this finding was not confirmed in another study (Louis et al. 2005).

Although the majority of studies did not detail the onset of RT with regard to action tremor, it was commonly reported as a late feature of the condition. However, the only study which detailed these data reported 76.3% of cases having action tremor and RT since onset, 16.6% developing action tremor first and 4.1% developing RT first (Martinelli et al. 1987).

The majority of the clinical studies did not detail the distribution of RT in ET + RT and it has been reported to be either unilateral or bilateral (either symmetric or asymmetric) (Cohen et al. 2003; Koller and Rubino 1985). RT in ET + RT has been reported to be not re-emergent upon holding a posture (Koller and Rubino 1985; Papengut et al. 2013), to be significantly attenuated during walking (Uchida et al. 2011) and to be less common during standing (Louis et al. 2021; Nisticò et al. 2022) and when lying down as compared to PD (Nisticò et al. 2022). The other studies comparing ET + RT to ET (without RT) for features suggestive of PD found no differences between the two groups in terms of smell performances (Louis and Jurewicz 2003; Djaldetti et al. 2008), writing (Martinez-Hernandez and Louis 2014) and color discrimination (Louis et al. 2012). However, another study found REM sleep Behavior Disorder (RBD) in 11.6% of ET cases and RT to be more common in ET with RBD (100%) as compared to ET without RBD (34.2%) (Bugalho et al. 2021).

Interestingly, the available longitudinal studies showed that no patients with ET + RT developed PD during a follow-up of 24–64 months (Gilmour et al. 2021; Martinelli et al. 1987; Bugalho et al. 2021), but also that RT was an unstable feature, reverting from present to absent in more than one-third of cases (Iglesias-Hernandez et al. 2021).

Overall, the majority of the studies reported patients with ET + RT to have higher age at onset and a greater tremor severity than ET. Moreover, these patients missed clinical features suggestive of parkinsonism.

Six studies involving a total of 60 patients employed electrophysiology in ET + RT. Thus, RT in ET-plus cases has been shown to have a synchronous EMG pattern as opposed to the alternating pattern observed in PD (Fekete and Li 2013; Nisticò et al. 2011; Vescio et al. 2021), a finding that has been suggested to differentiate the two conditions at individual level. Conversely, tremor frequency measurement either at rest or across different activating conditions does not distinguish the two conditions (Cichaczewski et al. 2014). Interestingly, Nisticò et al. explored the integrity of the blink reflex recovery cycle (BRRC) in ET + RT and ET, showing that the former had an increased R2 component of the BRRC (Nisticò et al. 2012). Given that these patients had a normal DAT-SPECT, they suggested ET + RT to be dystonic in nature rather than being an ET subtype (Nisticò et al. 2012). On the other hand, another study reported patients with ET + RT to have electrophysiological features suggestive of parkinsonism (ie, slow spiral drawing speed and increased decrement of spiral drawing speed with radius) (Cohen et al. 2003). Overall, the electrophysiological features of ET + RT seem to be similar to those of ET. Yet, individual studies reported electrophysiological features consistent with dystonia or suggestive of parkinsonism.

The hypothesis that ET + RT might be sustained by a dopaminergic denervation has been tested in 10 studies using DAT-SPECT and involving a total of 192 patients. Thus, Lee et al. reported patients with ET + RT to have intermediate uptake binding values (suggesting a mild substantia nigra neuronal loss) between ET without RT and patients with PD (Lee et al. 1999). In these patients the mean disease duration was of about 18 years (but RT only developed as a late feature for a mean duration of approximately 3 years) as opposed to 16 years of patients with ET without RT (Lee et al. 1999). In the aforementioned study by Djaldetti et al. (Djaldetti et al. 2008), 25% of ET + RT cases (mean disease duration of about 15 years) had abnormal DAT-SPECT findings despite normal smell performances. A similar frequency was found in another study in which 22% of patients with ET + RT (mean disease duration of about 11 years) had evidence of dopaminergic denervation (You et al. 2013), whereas a much higher figure has been found in other three studies (Ceravolo et al. 2008; Verdal et al. 2013; Marsala et al. 2017). Thus, Ceravolo et al. reported 50% of patients with ET + RT (mean disease duration of about 10 years) to have abnormal DAT-SPECT (Ceravolo et al. 2008). Interestingly, 60% of these had a clinical evolution to PD, whereas the remaining 40% still presented with an isolated tremor syndrome at the follow-up of about 2 years (Ceravolo et al. 2008). de Verdal et al. reported 75.7% of patients with ET + RT (mean disease duration of 13 years) with an abnormal DAT-SPECT, of whom about the half had unilateral RT and the other half bilateral RT (Verdal et al. 2013). Finally, Marsala et al. found 88.2% of abnormal scans in patients with ET + RT with a mean disease duration of about 19 years (Marsala et al. 2017). At a mean follow-up of 5 years, these patients showed mild parkinsonian features, such as the reduction of pendular movements, micrography, and slight muscle rigidity (Marsala et al. 2017). In contrast to the above, Asenbaum et al. found normal DAT-SPECT findings in all 32 investigated patients with ET, of whom 18 (56.25%) had RT (Asenbaum et al. 1998). Of note, in this study patients with ET + RT had a shorter disease duration (of about 9.3 years) than ET patients without RT (11.5 years) (Asenbaum et al. 1998). Similar findings were obtained in three more studies (Fekete and Li 2013; Nisticò et al. 2012; Caligiuri et al. 2017) in which all patients with ET + RT had normal SPECT findings. Therefore, the percentage of abnormal DAT-SPECT findings grossly varied across studies and ranges from 0% to 88.2% of the recruited sample, with higher figures being reported in patients with longer disease duration.

Seven studies involving a total of 265 patients used different magnetic resonance imaging (MRI) techniques to explore the neuroanatomical correlates of ET + RT (Caligiuri et al. 2017; Nicoletti et al. 2015; Li et al. 2021, 2020; Novellino et al. 2016; Prasad et al. 2018; Cherubini et al. 2014). A connectivity-based study by Caliguri et al. found common alterations of the cerebellothalamo-cortical (CTC) network in ET + RT and ET cases without RT, with the former additionally showing reduced connectivity in a pathway connecting globus pallidus, caudate, and supplementary motor area (Caligiuri et al. 2017). In keeping with these results, in a task-based (ie, during continuous writing) functional MRI (fMRI) Nicoletti et al. found a reduced activation of various cortical areas and internal globus pallidus in ET + RT, beyond the shared reduced activation of the CTC network in both ET and ET + RT (Nicoletti et al. 2015). They additionally failed to evidence morphometric changes between the two groups (Nicoletti et al. 2015). As opposed to the above, Li et al. explored spontaneous brain activity in a resting-state fMRI in ET and ET + RT showing that two groups shared a reduced activity in the basal ganglia, but differed in terms of cerebellar and limbic system activity, respectively (Li et al. 2021). Novellino et al. explored cerebellar microstructure in patients with ET + RT as compared to ET cases without RT and healthy controls (HC) by Diffusion Tensor Imaging (DTI) (Novellino et al. 2016). Thus, whereas all patients as a group showed increased cerebellar gray matter (GM) diffusivity than HC, ET + RT showed intermediate values between ET without RT and HC, suggesting that cerebellar involvement does not seem to fully account for ET + RT, with additional networks being possibly involved (Novellino et al. 2016). In contrast with these results, however, Prasad et al. (Prasad et al. 2018) found no abnormalities in cerebellar GM between ET + RT and ET without RT, nor when comparing these groups to HC, in keeping with results obtained with voxel-based morphometry (Nicoletti et al. 2015). However, they found a consistent pattern suggestive of higher cerebellar white matter (WM) microstructural damage in ET + RT, suggesting this could be a severe variant of ET (Prasad et al. 2018). Of note and differently from Novellino (Novellino et al. 2016), Prasad et al. adopted a whole-brain approach and did not confirm basal ganglia involvement in ET + RT (Prasad et al. 2018). Two studies compared ET + RT with tremulous PD (Li et al. 2020; Cherubini et al. 2014). Thus, Li et al. using a resting-state fMRI showed a decreased activation of the default mode network, bilateral putamen and bilateral cerebellum in both groups, with patients with PD additionally showing alterations in the bilateral supplementary motor area and precentral gyrus (Li et al. 2020). Finally, Cherubini et al. used a support vector machine (SVM) approach combining morphometric GM and WM data with DTI-derived mean diffusivity and fractional anisotropy data to distinguish ET + RT from PD (Cherubini et al. 2014). They showed that SVM classification of individual patients was not accurate if based on a single predictor, whereas when all predictors were combined in a multimodal algorithm, SVM distinguished patients with ET + RT from PD cases with an accuracy of 100% (Cherubini et al. 2014). However, this study did not provide any information on the specific abnormalities observed in each group, therefore hampering any speculations on the involved circuitry in the generation of RT in either disorder. Overall, the seven studies exploiting MRI did not show consistent findings and only suggested that the abnormality responsible for RT might localize in a circuitry different from that causing action tremor.

Only 2 studies involving a total of 38 patients explored post-mortem findings in ET + RT. Thus, in a series of 20 cases with ET, of whom 6 had additional RT, Rajput et al. found no consistent pathology at a group level, nor differences between ET + RT and ET (Rajput et al. 2004). Similarly, Louis et al. (Louis et al. 2011) compared 9 cases with ET + RT with 9 age-matched cases with ET and found no differences between the two groups. Namely, in both there was found an increased number of torpedoes of the cerebellar cortex and lower Purkinje cells, as compared to normative control values, whereas neither demonstrated Lewy body-containing neurons, Lewy neurites or other relevant pathological changes in the basal ganglia were minimal, suggesting that ET + RT is not the expression of underlying Lewy body pathology (Louis et al. 2011). Overall, these two anatomopathological studies did not demonstrate PD pathology in ET + RT.

There is a body of works suggesting that RT can be a late manifestation of ET (Cohen et al. 2003; Louis et al. 2015, 2021; Rajalingam et al. 2018; Djaldetti et al. 2008), which echoes the initial suggestions by Critchley (Critchley 1949) and Marsden (Marsden et al. 1984). In keeping with this hypothesis, patients with ET + RT had longer disease duration and greater tremor severity than ET cases without RT. Supporting this view and excluding the possibility that these cases might have developed PD, there are also studies showing preserved dopaminergic functionality, as assessed by DAT-SPECT in the majority of these patients (Djaldetti et al. 2008; Fekete and Li 2013; Nisticò et al. 2012; Lee et al. 1999; You et al. 2013; Asenbaum et al. 1998; Caligiuri et al. 2017) as well as anatomopathological studies failing to demonstrate Lewy body pathology (Rajput et al. 2004; Louis et al. 2011). Moreover, the electrophysiological studies comparing ET + RT and PD cases further demonstrated tremor characteristics (i.e., synchronous pattern) in the former that bear resemblance to those of action tremor in ET and contrast to what is observed in parkinsonian tremor (Fekete and Li 2013; Nisticò et al. 2011; Vescio et al. 2021). The evidence that RT might arise from the involvement of a different circuitry of that sustaining action tremor (Caligiuri et al. 2017; Nicoletti et al. 2015; Li et al. 2021; Novellino et al. 2016) does not have a unique interpretations: following the account that RT develops as a late manifestation of ET, one might speculate that this phenomenon occurs as a function of the progression of the underlying disease, but alternative hypotheses are possible, as continued below.

Supporting an alternative hypothesis, there are evidence suggesting the RT might not reflect disease duration (and progression), but aging. Many studies reported patients with ET + RT to be older than ET cases without RT, despite a similar disease duration, and to have a higher age at onset (Louis et al. 2015, 2005, 2012, 2021; Louis and Jurewicz 2003). A proportion of these patients might therefore have tremor, including RT, as a manifestation of pathological aging (Deuschl et al. 2015), which would arguably be less dependent on genetic susceptibility (Martinelli et al. 1987). This hypothesis is intriguing further because it would recapitulate the bimodal distribution in terms of age at onset that has been classically observed in ET in general (ie, before the new tremor classification) (Haubenberger and Hallett 2018). This hypothesis has not yet been formally tested and might represent a topic for future investigation. In this context, however, it is worth noting that two studies attempted a factor analysis of the clinical features of ET (Louis et al. 2005; Louis 2009). The results from this type of approach are driven by the data and are hence not generated following a particular hypothesis (Erro et al. 2016). In both studies, RT emerged as an independent factor (Louis et al. 2005; Louis 2009), which confirms that these patients represent a different entity from ET. However, in one of the studies including age as a feature, it did not group with RT (Louis 2009), which is in contrast to the hypothesis raised above. However, it should be noted that disease duration was not included in the primary analysis, nor secondary analyses were performed to compare patients based on the identified factors to test this hypothesis. What remains from these data-driven approaches is that RT would seem to identify a group of patients different from ET. From a mechanistic standpoint, the difference between the two groups might stand in the involvement of not entirely overlapping networks, as suggested by some works using MRI (Caligiuri et al. 2017; Nicoletti et al. 2015; Li et al. 2021; Novellino et al. 2016). In such cases, the involvement of such circuitry would be therefore triggered by pathological aging rather than by the progression of the underlying disease itself.

Finally, there is also evidence that at least in a subset of patients with ET + RT the clinical diagnosis was erroneous. Nisticò et al. showed these patients having an abnormal BRRC in keeping with the diagnosis of dystonia (Nisticò et al. 2012). They subsequently suggested that this electrophysiological marker could be helpful to distinguish ET + RT from patients with PD (Nisticò et al. 2014), although the findings in the two groups largely overlapped and might not be that useful at individual level. However, the main conclusion that derives from these studies is that there might be a misdiagnosis of ET in rarer forms of tremor (Erro and Reich 2022), including dystonia. This evidence reiterates previous claims (Quinn et al. 2011) and is in line with recent findings showing that RT is common in dystonia (Erro et al. 2014). Interestingly, a recent study showed that the diagnostic allocation of tremulous patients is partly dependent on the specific expertise of the evaluators (Becktepe et al. 2021), a finding which might explain a proportion of the misdiagnoses.