Kinetic tremor is apparent in this drawing of an Archimedes spiral in a patient with essential tremor.
One possible schema of the heterogeneity of essential tremor. The heterogeneity in essential tremor may be organized by (1) disease etiology, (2) tissue-level changes that occur after the disease process is initiated and as it develops, and (3) clinical features that are the end product of these underlying pathological processes.
When quantified using different methods, Purkinje cell loss in essential tremor is significant compared with age-matched controls. Bielschowsky-stained section (original magnification × 50) of the cerebellar cortex in a 79-year-old individual with essential tremor showing normal Purkinje cellularity (right) vs a region with segmental loss of Purkinje cells (left).
Louis ED. Essential TremorsA Family of Neurodegenerative Disorders?. Arch Neurol. 2009;66(10):1202-1208. doi:10.1001/archneurol.2009.217
Essential tremor (ET) is the most common pathologic tremor in humans. The traditional view of ET, as a monosymptomatic condition, is being replaced by an appreciation of the spectrum of clinical features, with both motor and nonmotor elements. These features are not distributed homogeneously across patients. In addition, postmortem studies are now demonstrating distinct structural changes in ET. There is growing evidence that ET may be a family of diseases rather than a single entity. Furthermore, this aging-associated, progressive disorder is associated with neuronal loss and postmortem changes that occur in traditional neurodegenerative disorders.
Essential tremor (ET), the most common pathologic tremor in humans, is among the more ubiquitous neurologic diseases, with a prevalence (age ≥40 years) of 4.0%.1,2Its most recognizable clinical feature is an 8- to 12-Hz kinetic tremor of the arms (ie, tremor during voluntary movement) (Figure 1), which is often later accompanied by head and voice tremors. The condition is global, affecting human beings in a variety of settings, ranging from the remote Okapa subdistrict of Papua New Guinea to the urban Washington Heights–Inwood community in northern Manhattan, New York.1,2The traditional view of ET as a monosymptomatic condition characterized by only kinetic or postural tremor has been supplanted in recent years. First, the tremor phenomenology is broad and many patients have other motor manifestations3and nonmotor features, including cognitive and psychiatric.3,4Furthermore, this expanded constellation of clinical features is not distributed homogeneously across patients. For example, some patients develop head tremor while others do not. Some have dementia while others do not; some develop Parkinson disease (PD) while others do not.5Recent studies of the pathology of ET have demonstrated several distinctive structural changes, including neuronal loss.6- 10With evidence of clinical and pathologic heterogeneity emerging, several important questions arise. First, is ET a single disease entity or a family of diseases? Second, is ET neurodegenerative?
Neurology is replete with historical examples in which clinical-pathological investigations have resulted in the partitioning of disease entities and the refinement of classification. For example, many patients with upper motor neuron signs were clinically lumped together, yet in the 1800s, Charcot and others observed that among such patients who had sclerosis on postmortem examination, 1 subgroup had a pattern that could be characterized as amyotrophic lateral sclerosis (ALS), while a second subgroup had a pattern that could be characterized as multiple sclerosis (MS). We now know that ALS and MS are different diseases with distinct mechanisms, clinical features, and treatments. Likewise, in the mid-20th century, advances in neuropathology and careful clinical-pathological studies redefined the concept of parkinsonism as a family of disease entities rather than a single disease. Idiopathic PD could be distinguished from other entities like progressive supranuclear palsy and corticobasal ganglionic degeneration. It is conceivable that we are now arriving at such a primary juncture in our understanding of ET. Evidence of clinical heterogeneity and pathologic heterogeneity is emerging, raising the question as to whether it is possible to reformulate ET as a cluster of separable clinical-pathological entities, that is, a family of diseases—the essential tremors.
Although clinical phenotypic variability by itself is not an argument for separate diseases, its presence further opens that possibility. Clinically, the view of ET as a single neurologic sign no longer seems tenable. First, the tremor phenomenology itself is multifaceted. Although kinetic and postural tremors are the core features upon which diagnostic criteria are generally based (Table 1), intention tremor18and tremor at rest19may also occur in subsets of patients. The relative severity of different tremor types (kinetic more severe than postural),20the favored sites of anatomical involvement (arm more often than head, and head more often than jaw),21and the typical direction of spread over time (from arms to head rather than the converse)22are distinctive, adding a degree of subtlety and complexity to the recognition and diagnosis of a disorder that is often viewed as relatively ordinary and featureless. Indeed, perhaps owing to a lack of familiarity with these features, misdiagnosis is exceedingly common; 30% to 50% of patients diagnosed with ET do not have ET,23which may make this one of the most commonly misdiagnosed neurologic disorders. Moreover, kinetic tremor of the arms, though not the head, may be an adverse effect of many medications, further contributing to this diagnostic difficulty. The following medications may produce kinetic tremor: amiodarone, bronchodilators, cyclosporin, lithium, methylphenidate, phenylpropanolamine, procainamide, pseudoephedrine, selective serotonin reuptake inhibitors, steroids, theophylline, thyroxine, tricyclic antidepressants, and valproic acid. Second, there are motor features aside from tremor. In a large number of studies,18,24postural instability and mild to moderate gait ataxia, beyond that observed in normal aging, have been demonstrated in subgroups of patients with ET. In addition to these gait problems, eye motion abnormalities (impaired smooth pursuit initiation and pathologic suppression of the vestibulo-ocular reflex time constant) were recently described in a study of 17 ET cases.25Third, there is a growing appreciation of the existence of a variety of nonmotor features, including cognitive, psychiatric, and sensory. Cognitive features, especially mild problems with executive function and memory, were first reported in 200126and soon after by numerous other investigators.27Furthermore, a population-based study in Madrid28first demonstrated in 2006 that the odds of prevalent dementia were nearly twice as high among the subset of ET cases with older-onset disease when compared with age-matched controls. The same group later demonstrated5that the risk of developing incident dementia was nearly twice as high in the ET cases than controls. Also, a number of psychiatric correlates have been observed. The presence of specific personality traits29,30has been demonstrated. In one such study,29the observed personality profile was not related to functional disability or tremor severity, suggesting that it could be a primary disease feature rather than a response to disabling tremor.29Anxiety,31depressive symptoms,32and social phobia33have been shown to occur in patients with ET more often than in controls. Traditionally, these have been viewed as psychiatric responses to disabling tremor. Yet in a recent prospective study,4depressive symptoms preceded the onset of the tremor by several years (ie, the presence of baseline self-reported depression was associated with an increased risk of developing incident ET during follow-up). Sensory abnormalities, including olfactory deficits in some studies34and hearing loss in others,35have been reported in ET cases compared with age-matched controls, further drawing attention to the domain of nonmotor manifestations. Although a single disease may have a broad array of clinical manifestations (eg, Huntington disease), such clinical variety can also be an indication that one is dealing with a group of diseases (eg, parkinsonisms such as PD, progressive supranuclear palsy, and corticobasal ganglionic degeneration).
It is important to emphasize that these clinical features are heterogeneously distributed across patients with ET. Although some patterns are now becoming apparent (eg, patients with intention tremor have more gait difficulties and eye motion abnormalities25; patients with rest tremor generally have long-standing disease with severe kinetic tremor19; dementia is associated with older age of onset of ET5; and older age of onset and unilateral onset are associated with more rapid progression36,37), a clear separation of distinct clinical subtypes has yet to emerge.
The mainstays of therapy for ET are propranolol and primidone, though several promising new agents have been introduced in recent years (Table 2); surgical treatment (deep brain stimulation) is also highly effective.3One reoccurring feature of pharmacotherapeutic trials in ET is that the response to a particular medication is usually patchy, with approximately one-half of the patients showing some degree of tremor reduction and others none at all38; this phenomenon is frequently observed by practitioners as well. Among other possibilities (eg, differences in disease duration across treated patients), this heterogeneity of response could be a marker of different underlying disease mechanisms in subsets of patients, though this remains to be demonstrated.
Disease mechanisms in ET have been elusive. Despite its high prevalence, until recently few brains with ET were examined and little information was available about its pathology. An intensive effort was launched in 2003 to bank ET brains.6In contrast with previous studies, these brains were systematically examined to quantify cerebellar and other brainstem pathologies and were compared with control brains. These analyses, based on 33 ET brains, indicated that the structural pathologic changes appeared to be of 2 types.6,7,9,39Most commonly (75%), brains were characterized by clear cerebellar degenerative changes, including a 40% reduction in the number of Purkinje cells, a 6-fold increase in the number of torpedoes (ie, swellings of the Purkinje cell axon that likely represent a cellular response to injury), and Purkinje cell heterotopia and dendrite swellings. These brains did not have Lewy bodies. The remaining brains were characterized by Lewy bodies confined mainly to the locus ceruleus relative to total sparing of other brainstem structures.6The prevalence of Lewy bodies was significantly greater than that observed in similarly aged controls, indicating that they were not likely incidental.6Furthermore, this particular pattern of Lewy body involvement in ET has not been described in series of atypical Lewy body brains (ie, brains that do not follow the Braak staging scheme).40The ET brains with Lewy bodies did not have excessive torpedoes or Purkinje cell loss.6These 2 pathological patterns were labeled cerebellar ET and Lewy body variant of ET.6,41Other recent series of ET brains have also confirmed a heterogeneous pathology involving Purkinje cell loss in the cerebellum in some cases and changes, including cell loss, in the locus ceruleus in others.8These recent postmortem studies have helped identify degenerative structural alterations in the cerebellum and its connecting pathways in ET. How changes in the locus ceruleus could produce ET is less clear, though neurons of the locus ceruleus synapse with cerebellar Purkinje cell dendrites.42These projections are important for the normal development and maintenance of Purkinje cells.43Impaired activity in the locus ceruleus could result in a diminution of stimulatory output from that locus to the Purkinje cells.
On an etiologic level, ET is often considered to be genetic.44There are many examples of families in which the proband and multiple relatives have ET and in which the pattern of inheritance is consistent with an autosomal dominant model. In 1997, linkage to a region on chromosome 2p was demonstrated,45and in that same year, to chromosome 3q in other families.46A third study demonstrated linkage to a region on chromosome 6p in several families.47Aside from these 3 studies, others have failed to demonstrate linkage to these 3 regions, indicating that there will likely be more than 3 genes responsible for this disease.44,48It is important to note, however, that the genetic studies have not progressed further and the specific ET genes have not yet been identified.44- 47In the absence of a specific genotype for ET, there are as yet no specific genotype-phenotype correlations, though it appears that young-onset cases are generally familial (ie, likely to have a genetic susceptibility).49Aside from genetic factors, there is also a growing understanding that environmental factors likely contribute to the etiology of ET as well, indicating that there is further heterogeneity on an etiologic level. Several lines of evidence support the role of these factors. First, while commonly stated that 50% or more of ET cases have a genetic basis, the precise derivation of this estimate is unclear and its validity is also doubtful.50Indeed, some estimates are as low as 17%.50In twin studies,51,52concordance in monozygotic twins was only 60% in one study and 63% in another. Second, the well-known existence of intrafamilial differences in age of onset, location of tremor, and severity of tremor53also suggests that environmental (or perhaps other genetic) factors may be serving as modifiers of underlying susceptibility genotypes. In terms of environmental factors, a growing number of case-control studies54have implicated several specific toxins, namely β-carboline alkaloids (eg, harmine and harmane, a group of highly tremorogenic dietary chemicals) and lead; and further studies of these putative environmental toxins are needed.
Information presented in the previous section may be organized by etiologies, pathologies, and clinical features (Figure 2). Given the sheer prevalence and ubiquity of this condition, the historical tendency to divide disease entities as new knowledge arises, the appreciation of a broader variety of clinical features in several separate domains (eg, tremor, other motor, cognitive, and psychiatric), the observation that these are not uniformly present in patients with ET, the evidence that multiple genes will likely be responsible for this disease, and preliminary evidence of distinct pathologic patterns, it is likely that ET will turn out to be a family of diseases rather than a single disease entity. It is likely that this family of diseases, united by the presence of kinetic tremor, would be separable on the basis of etiologic, clinical, therapeutic response, and pathologic features, though further work is needed to explore these relationships.
A question implicitly raised in the above discussion is whether this disease (or these diseases) are neurodegenerative. The idea that ET could be neurodegenerative is not new. In 1948, Critchley and Greenfield wrote: “Although anatomical proof is as yet lacking, there are at least a number of clinical points to make question whether ‘essential tremor’ may not, at times any rate, represent an incomplete or a premature variant of one of the cerebellar atrophies.”55Although not further elaborated on by those authors, these clinical points include its insidious onset, association with advanced aging (ie, both incidence and prevalence increase with aging), gradual yet progressive nature, and the presence of cerebellar features (eg, intention tremor and ataxia) on examination.
Neurodegenerative diseases traditionally have been defined as diseases that begin insidiously, pursue a gradually progressive course that may continue for many years, and are characterized by the selective involvement of anatomically and physiologically related systems of neurons owing to intrinsic processes rather than an identifiable outside influence (eg, vascular, autoimmune). Cell loss is also considered by many to be a prominent feature of these diseases.56Furthermore, their occurrence often increases markedly with advancing age. What is the evidence that ET is neurodegenerative? The clinical points noted above are important. Essential tremor has an insidious onset53,57and then follows a gradual yet progressive clinical course.36,37There is a marked and continued rise in disease occurrence in advanced ages.1,2This clinical constellation is somewhat compelling; however, none of these features in isolation is specific to neurodegenerative diseases. On a tissue-based level, the evidence is more compelling. Selective involvement of an anatomically and physiologically related system of neurons, Purkinje cells, has been reported recently in ET cases both in our series (Figure 3)6,7and in the other large series.8When quantified using different methods, Purkinje cell loss is significant. There is an approximate 40% loss of these cells in ET cases compared with age-matched control brains,6,7which persists even when one adjusts for age and other confounding pathologies (eg, mild Alzheimer-type changes). Additional evidence that the Purkinje cells are diseased is that there are significantly more torpedoes in the ET brain, where their numbers are 6 times higher than expected for age and a preponderance of displaced (ie, heterotopic) Purkinje cells as well as Purkinje cells with dendritic swellings.6In contrast to these cases, the ET cases with normal cerebella have Lewy bodies,6,9lesions that have long been considered important in the pathogenesis of another neurodegenerative tremor disorder, namely PD. While none of these pathologic changes in the cerebellum are disease-specific (eg, other forms of cerebellar degeneration may be characterized by Purkinje cell loss and torpedo formation), this just indicates that the changes seen in the cerebellum in ET occur more broadly in the cerebellar degenerations. Aside from this structural-pathologic evidence suggesting a neurodegenerative process, other evidence suggests that ET is neurodegenerative. While many of these features in isolation are not specific to neurodegenerative diseases, the constellation of findings, all present in the same disease, is more compelling (Table 3). For example, there is a long-standing clinical association between ET and PD; indeed, having ET increases the risk of developing incident PD 4- to 5-fold.60Furthermore, having older-onset ET increases the risk of developing Alzheimer disease nearly 2-fold.5This association between ET and subsequent development of these neurodegenerative diseases suggests that ET may share pathogenic mechanisms with these disorders.
There is some evidence to suggest that ET is a family of diseases rather than a single entity. These disorders, perhaps better termed the essential tremors, are aging-associated, progressive, and associated with cell loss and other types of changes (Lewy body formation) that traditionally occur in neurodegenerative disorders. Future study is needed to continue to shape our evolving notion of the entity that we currently refer to as essential tremor.
Correspondence:Elan D. Louis, MD, MSc, Neurological Institute, Unit 198, 710 W 168th St, New York, NY 10032 (firstname.lastname@example.org).
Accepted for Publication:December 3, 2008.
Financial Disclosure:None reported.
Funding/Support:This work was supported by grant R01 NS42859 from the National Institutes of Health, Bethesda, Maryland; the Parkinson's Disease Foundation, New York, New York; the Arlene Bronstein Essential Tremor Research Fund (Columbia University); and the Claire O’Neil Essential Tremor Research Fund (Columbia University).
Additional Contributions:Blair Ford, MD, Steven Frucht, MD, and Vinita Sehgal, MD, read the manuscript and provided critical commentary.