The Roles of Age and Aging in Essential Tremor: An Epidemiological Perspective

All humans are aging. Furthermore, with advances in life expectancy, the number of humans who are entering advanced age has been rising considerably [1, 2]. For example, in 2050, the population in the United States aged 65 and over is projected to be 83.7 million, which is almost double its estimated population of 43.1 million in 2012 [1]. As aging is associated with a variety of medical conditions, and the costs of these conditions are considerable [3-5], it behooves us to gain an understanding of the intersections between aging and human disease.

Movement disorders are common neurological disorders that are characterized by either an excess or a paucity of movements. Among the most common of these is essential tremor (ET) [6-8], a disease whose hallmark feature is action tremor of the arms and hands [9, 10]. As it is with many movement disorders, the prevalence of ET increases with age [6]. In this paper, I review the points of intersections between age and aging and ET, particularly as these relate to aspects of disease epidemiology.

In March 2018, the author used PubMed to search relevant literature using the term “essential tremor” with additional search terms (e.g., “age,” “aging”; Table 1). These terms were restricted to the title of the article. Duplicates were removed and all articles published in English language (n = 25) were reviewed (Table 1). Additional articles of relevance were identified through the reference list of these 25.

As I will be using a variety of related and intersecting terms, it is important to define these terms from the start (Table 2). “Age” refers to a length of time, whereas “aging” refers to a process. Children have an age and they are aging; hence, the term “advanced age” brings us into the sphere of older adults who are near the end of the human life cycle. “Age of onset” refers to the age at which a disease commences.

Population-based prevalence studies have consistently demonstrated that the prevalence of ET increases markedly with age [6] (Table 3). This is easily understandable given the fact that the incidence of ET increases with age (see further below) and mortality in ET is only modestly increased relative to that of age-matched controls (see further below).

Although results vary across studies, most report a crude prevalence of 4% or higher in persons age 60 and older, with some reporting values considerably higher (see Table 2 in [6]). With advanced age, the prevalence appears to increase in an exponential manner (Fig. 1) [6, 11, 12], with values in the oldest old (e.g., 95 and older) in excess of 20% in some studies [6]. Both the increase in prevalence with age as well as the exponential manner of that increase, are consistent with the view that the disease is neurodegenerative [13-17]; such increases in prevalence are similarly found in patients with Parkinson’s disease, Alzheimer’s disease and other degenerative diseases of the nervous system [18, 19].

There are 2 published studies of the incidence of ET [20, 21], and both demonstrated an increase in the incidence of disease with age (Table 3). One of these was a retrospective medical record review that utilized the records linkage system of the Rochester Epidemiology Project in the United States; it encompassed all age groups [20]. The other was a prospective, population-based study of individuals age 65 and older in Spain [21]. In the first study, the incidence per 100,000 increased consistently with each 10- or 20-year age stratum (e.g., 2.9 [ages 0–19], 6.3 [ages 20–39], 19.2 [ages 40–49], 46.8 [ages 50–59]), with the highest value in those age 80 and above (135.7 per 100,000) [20]. In the second study [21], the incidence per 100,000 increased in each 5-year age stratum (e.g., 568 [ages 65–69], 583 [ages 70–74], 754 [ages 75–79]) until the final stratum (age 85 and older), for which the sample size was small and the estimate unstable.

There is some evidence that age influences the clinical expression of ET. For example, head tremor may eventually develop in as many as 30–60% of ET patients, yet it is unclear why some individuals develop this type of tremor and others do not [22]. One possibility is that it is merely a function of advancing disease duration. Alternatively, age could be a primary factor. In a cross-sectional study, 363 ET patients underwent a detailed neurological examination that included an assessment of the presence or absence of head tremor [22]. Head tremor was present on examination in 38.6% of patients. Young patients, even with longer-duration tremor, rarely had head tremor: 2/27 (7.4%) patients < 40 years old with tremor duration ≥10 years had head tremor versus 121/283 (42.8%) older patients (> 60 years old) with tremor duration ≥10 years (p < 0.001) [22]. In a multivariate logistic regression analysis, head tremor was associated with age (p < 0.001), but it was not independently associated with tremor duration (p = 0.26). With the exception of 1 patient, head tremor did not begin before the age of 36 years. The data from this study raise the possibility that the appearance of head tremor in ET depends on a biological factor that is intrinsic to the patient (i.e., their age) and is not a clear consequence of advancing disease duration.

Another study, with a cross-sectional design, enrolled 55 community-based ET cases [23]. Disease duration was associated with the total tremor score, a measure of tremor severity (p = 0.02). Similarly, age was associated with the total tremor score (p = 0.025) [23]. In a linear regression analysis, duration (p = 0.02) and age (p = 0.02) were independently associated with the total tremor score. In the same study, results were similar in 79 ET cases enrolled from a tertiary referral center [23]. One interpretation of these data is that the reported increase in tremor severity in ET may be related not only to the inherent worsening of the disease pathophysiology with increasing duration but also with age and age-related processes such as neuronal attrition [23, 24]. This will be discussed further below.

Other studies have noted that older ET patients, when compared to age-matched controls, are more likely to have certain clinical deficits (e.g., deficits in gait and balance [25, 26]) than are younger ET patients when similarly compared with respective age-matched controls; however, it is not clear from these studies whether the authors adequately adjusted for the effects of disease duration.

Age of onset refers to the age at which an individual acquires a disease. The onset age of first symptoms or signs is often used as a close surrogate. Age of onset of symptoms can be difficult for patients to precisely recall, as the disease often has an insidious onset and initial presentation. This should be kept in mind when interpreting data that involve age of onset. In ET, age of onset has several important epidemiological and clinical correlates (Table 3). The first is that familial forms of the disease are more likely to have a younger age of onset [27]. In one study, investigators used a carefully characterized sample of 376 ET cases (61.7% familial and 38.3% sporadic) enrolled in a clinical-epidemiological study in order to contrast the age of onset distributions in familial ET versus sporadic ET [28]. The majority (71/82, 86.6%) of ET cases that appeared during childhood were familial rather than sporadic. Additionally, although the onset of ET occurred after age 40 in a majority of cases, the proportion was 125 out of 232 (53.9%) in familial ET and even more (118/144, 81.9%) in sporadic ET [28].

Second, the age of onset seems to run in ET families. In a family study, data from 26 probands and 52 relatives were analyzed [29]. The probands’ age of onset correlated significantly with their relatives’ age of onset (r = 0.50, p = 0.001), and in 57.7% of cases, the relative’s age of onset was within 10 years of the proband’s onset (i.e., a 20-year age range) [29].

Third, age of onset may be associated with the natural history of ET. As in several neurodegenerative disorders, in older onset cases, it seems that the disease may advance more rapidly [30-35]. For example, in one cross-sectional study, 60 ET cases from a community and 55 from a tertiary referral clinic underwent an interview and videotaped tremor examination [36]. The rate of progression was estimated based on the tremor severity and reported disease duration at the time of evaluation [36]. The authors reported an association between age of onset and rate of progression; cases with older age of onset (> 60 years) progressed more rapidly (p < 0.001) [36]. In another cross-sectional study, the investigators quantified the rate of tremor progression (i.e., tremor severity/duration) in 348 ET cases enrolled in an epidemiological study and characterized the relationship between age of tremor onset and rate of tremor progression in that sample [30]. Furthermore, they characterized the relationship between age of tremor onset, rate of tremor progression, and severity of underlying brain (i.e., cerebellar degenerative) changes in 9 ET cases from a brain repository [30]. The degeneration index was the number of abnormal Purkinje cell swellings (i.e., torpedoes) per section/Purkinje cell linear density [30]. In the epidemiological study, older age of tremor onset was associated with a faster rate of tremor progression (p < 0.001) [30]. In the brain repository, older age of tremor onset was associated with higher degeneration index (p = 0.037) [30].

Further data on age of onset and its associations with the natural history of ET relate to the literature on cognitive issues in ET. In several of these studies, cognitive impairment was found to be more a feature of older-onset ET cases than younger-onset cases [37-39]. However, whether the driving force in this difference was older age of onset or advanced age is not clear.

Finally, older thinking was that the penetrance of ET was considered complete by age 65 years. In other words, it was thought that by age 65, all individuals who were genetically predisposed to develop ET would manifest it (i.e., age of onset was always prior to 65) [40]. More recent data indicate that this is not likely to be the case. Studies have shown that relatives of ET cases are more likely than relatives of controls to exhibit mild, subclinical forms of tremor [41-43]. For example, in one study, the investigators enrolled more than 200 relatives of ET cases and a similar number of relatives of controls [41]. Relatives with ET were excluded from the analysis. In all age groups, including older relatives (i.e., ≥60 years of age), there was an increased prevalence of individuals with higher tremor scores among case relatives than control relatives, suggesting that even into advanced ages, subclinical ET seems to be present and penetrance still may not be complete [41].

The intersection points between ET and aging are numerous. One other example is that there is considerable evidence that ET is associated with several conditions of aging [44]. These include Parkinson’s disease, mild cognitive impairment (MCI) and dementia [44] (Table 3). First, it is well known from clinical practice settings that patients with ET often later develop Parkinson’s disease [45-47]. Indeed, family studies have shown an increased co-occurrence of the 2 diseases within the same families above that expected by chance alone [48, 49], and case-control studies have demonstrated an increased co-occurrence of the 2 disorders within the same individuals above that expected by chance alone, with increased odds being at 5 times [50]. A prospective analysis similarly indicated that patients with ET have a more than fourfold increased risk of developing incident Parkinson’s disease [51].

Second, several epidemiological studies have demonstrated an association between ET, and especially that which begins after age 65 years, and cognitive impairment and dementia [44, 52]. Although the clinical diagnosis in many of the ET cases with dementia was Alzheimer’s disease, in the absence of postmortem studies, the pathophysiological basis for the dementia in these cases remains unelucidated. A series of papers from the Neurological Disorders in Central Spain study examined the associations between ET and cognitive dysfunction [53, 54], prevalent MCI [38], and both prevalent and incident dementia [55, 56]. In Neurological Disorders in Central Spain, ET cases were ascertained from 3 communities in Spain and all were 65 years of age or older. Investigators used a population-based case-control study design and matched ET cases and controls based on age, gender, education, depressive symptoms and medications that could potentially affect cognition. The neuropsychological battery tested global cognitive performance, executive function, memory and pre-morbid intelligence. Participants were further asked whether they suffered from forgetfulness or depression. The investigators demonstrated that ET cases performed less well than controls on neuropsychological testing, particularly on tests of memory and executive abilities; furthermore, a larger proportion of cases reported forgetfulness [53, 54]. In addition, cross-sectional analyses revealed that ET cases whose tremor began after age 65 years had increased odds of MCI (adjusted OR 1.57, p = 0.03), whereas cases with tremor onset prior to age 65 years and controls were equally likely to have MCI (adjusted OR 0.73, p = 0.43) [38]. These same study subjects were then followed longitudinally for a mean of 3.2 years, and the rate of cognitive decline appeared to be faster in the ET cases with tremor onset after 65 years of age than controls [57]. The same study demonstrated an association between ET with onset after 65 years of age and prevalent dementia [55] (adjusted OR 1.70, p = 0.03) and incident dementia [56] over a mean follow-up duration of 3.2 years (relative risk [RR] 1.98, p = 0.01). A second prospective, population-based study of older individuals (study enrollment age ≥65 years but ET age of onset unknown) was conducted in northern Manhattan, New York [58] and it demonstrated similar results. In cross-sectional analyses, 25.0% of ET cases had prevalent dementia versus 9.2% of controls (adjusted OR 1.84, p = 0.01). In prospective analyses, 18.3% of ET cases versus 8.7% of controls developed incident dementia (adjusted hazard ratio [HR] 1.64, p = 0.055). Thus, in a second population-based study of elders, ET was associated with increased odds of prevalent dementia and increased risk of incident dementia [52]. Additional longitudinal data support an association between older onset ET cases and risk of dementia [39].

Older ET patients are less able to tolerate the side effects of pharmacotherapy than are younger patients [59]. Both the cognitive side effects of primidone and the cardiac effects of propranolol (i.e., its effects on heart rate) are more difficult for older patients to tolerate. Hence, the age of the patient limits tolerance to the front-line medications (i.e., it creates a narrower therapeutic window) [59].

The biological basis for the observed associations between ET and age/aging is not entirely clear. One possible explanation for the increase in disease incidence with age as well as the reported increase in tremor severity with age may be neuronal attrition [23, 24]. The pathophysiology of ET is not completely understood, although there is considerable evidence that it involves a mild form of cerebellar degeneration centered on and around the Purkinje cell population [17, 60-62]. It is well known that aging is associated with progressive Purkinje cell loss [63, 64]. It is possible that the rate at which new cases arise as well as the progression of symptoms in existing cases could be impacted upon by age-associated deficits in the Purkinje cell population. The observation that older age of onset is associated with more rapid progression may similarly be based in differences in the substrate of the older versus younger brain; in the former, disease-linked pathology develops on top of age-associated changes (e.g., neuronal loss), leaving less of a reserve to cushion the effects of the disease. The associations between ET and degenerative conditions of aging are not clear, although they have been discussed elsewhere [44]. In essence, the issue is whether Lewy body pathology and tau pathology is more prevalent in ET cases than controls [65, 66]. Alternatively, there is a sizable literature demonstrating that neurodegenerative diseases may be associated with one another, with the notion being that the development of one such disorder is a marker of a biological propensity/vulnerability for the development of others [44].

One final biological consideration should be discussed. In recent literature, the term “aging-related tremor” has been introduced to deal with some variability in clinical features of ET cases with different ages of onset [67] (Table 3). The proposal was to label individuals whose tremor began after age 70 as “aging-related tremor” rather than ET [67]. This was because of limited data that these individuals might show worse aging parameters and mortality than controls and ET cases whose tremor began prior to that age cut-off [67]. However, the notion of age-related tremor is problematic for several reasons [68]. First, when Parkinson’s disease begins after age 70 years, one does not refer to it as “age-related tremor, bradykinesia, and rigidity” and when Alzheimer’s disease begins after age 70 years, one does not refer to it as “age-related cognitive dysfunction” [68]. In a related point, in Parkinson’s disease, older age of onset is associated with poor survival, akin to what is seen with ET, yet one does not remove the diagnostic label “Parkinson’s disease” from cases whose onset is older [68]. Second, even if it were conceptually of merit, the cutoff point for such an age-related tremor is arbitrary [68]. Even the authors who suggested this term have defined “late-onset” differently in various papers (e.g., defining it as ≥46 years in another paper) [69]. Finally, postmortem studies of humans suggest that the same degenerative pathology is found in both younger onset and older onset ET cases to the same extent [70]. In summary, there is little current evidence to support the notion that age of onset is the basis for distinct and separable subtypes of ET or that cases whose disease begins after a certain age cutoff do not have ET [68].

Here, I reviewed the intersections between age, aging, and ET, particularly as they relate to disease epidemiology. More specifically, I focused on the related but separable entities of “age,” “aging,” “advanced age,” and “age of onset.” Overall, there were several important intersection points between age, aging, and ET. The prevalence and incidence of ET both increase with age, and age may independently drive several features of the natural history of ET, including the increase in tremor severity with time and the appearance of head tremor with time. Epidemiological studies were reviewed, which have shown that ET is associated with several conditions of advanced age, including Parkinson’s disease, MCI, and dementia. Aside from age, age of onset has several important clinical correlates, and these were discussed. For example, familial forms of ET are more likely to have younger ages of onset, and older ages of onset are associated with more rapid disease progression. The biological basis for the observed associations between ET and age/aging is not entirely clear. One possible explanation may be neuronal attrition (i.e., age-associated deficits in the Purkinje cell population). More work is needed, and further knowledge about the basis for these observations will enhance our understanding of this disease and inform our care of patients.

Dr. Elan D. Louis has received research support from the National Institutes of Health: NINDS #R01 NS094607 (principal investigator), NINDS #R01 NS085136 (principal investigator), NINDS #R01 NS073872 (principal investigator), NINDS #R01 NS085136 (principal investigator) and NINDS #R01 NS088257 (principal investigator). The author does not have any conflicts of interest to discloses.