Association of Essential Tremor With Dementia and Affective Disorders: A Meta-Analysis

Background The dementia and affective disorders are common non-motor features in patients with essential tremor (ET). However, the relationship of ET with cognitive impairments and affective disorders remains controversial. This meta-analysis aimed to analyze the association of ET with dementia and affective disorders. Methods Original studies published from January 1999 to October 2019 were systematically searched from the database of Medline (OvidSP), EMBASE (OvidSP), and the Cochrane Central Register of Controlled Trials. Pooled standard mean difference (SMD, random effect model), odds ratios (ORs), relative risk (RR), and 95% CI were calculated. Results Compared with the Non-ET group, patients with ET had significantly lower Mini-Mental State Examination (MMSE) score (SMD, −1.16; 95% CI, −1.75 to −0.58; p = 0.0001) and had significantly higher depressive and anxiety symptoms scale score (SMD, 0.55; 95% CI, 0.22–0.87; p = 0.0009). The OR for dementia and affective disorders in individuals with ET compared with individuals without ET was 2.49 (95% CI, 2.17–2.85, p < 0.00001). While there was no significant difference in Montreal Cognitive Assessment (MoCA) score between ET and Non-ET groups (SMD, −0.52; 95% CI, −0.16 to 0.13; p = 0.23), there was a significant difference in the risk of mortality between ET and Non-ET groups (RR = 4.69, 95% CI, 2.18–10.07). Conclusion The non-motor symptoms should not be neglected among patients with ET. However, the causal relationship between ET and dementia, depression, and anxiety is unclear.


INTRODUCTION
Essential tremor (ET) is one of the common movement disorders and is characterized by isolated tremor syndrome of the bilateral upper limbs for at least 3 years. The prevalence of ET is about 5% in the general population and over 20% in the elderly (1,2). ET is often familial, with a typically autosomal dominant pattern, and is most often observed in the elderly population over 65 years old. The main clinical manifestations of ET patients include motor features (intention tremor and ataxia) and non-motor features (cognitive impairments and affective disorders). ET often has a bilateral presentation but can also be asymmetric in nature and is rarely seen at rest. Tremor is exaggerated under stressful conditions, and 50-70% of ET cases can be ameliorated by consuming alcohol (3).
There is an increasing awareness that patients with ET may present non-motor features, such as cognitive impairments and mood disorders. In 2001, it was suggested that tremor was not the only manifestation of ET. Since then, a rising number of non-motor symptoms were recognized in patients with ET, such as not only cognitive impairment and affective disorders but also anxiety, personality changes, fatigue, hearing impairment, olfactory dysfunction, upper airway dysfunction, and sleep disturbances (4,5). Meanwhile, the cognitive deficits of ET were first noticed prior to thalamic deep-brain stimulation (DBS) for refractory ET (6). Cognitive deficits were observed in 69.2% of patients with ET, and 11.4% of ET had dementia. The cognitive deficits affected the domains of attention, executive function, and memory (7).
An increasing number of studies have shown the poor quality of life among patients with ET. Compared to physical impairments, psychological and cognitive deficits more likely affected patients with ET because of the psychological burden and negative influence on the recovery (4). Therefore, it is necessary to better understand the correlation of ET with cognitive impairment and affective disorders. This meta-analysis aimed to analyze the correlation of ET with dementia and affective disorders.

Search Strategy
This meta-analysis followed the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group reporting guideline (Supplementary File S1) (8), and the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines (Supplementary File S2). The literature search, title screening, abstract screening, final decision on eligibility after full-text review, methodological quality appraisal of included studies, and data extraction were independently performed by two investigators.
Original studies published from January 1999 to October 2019 were systematically searched from the database of Medline (OvidSP), EMBASE (OvidSP), and the Cochrane Central Register of Controlled Trials (CENTRAL, last issue). Searching terms were listed as follows: 1) Essential Tremor. 2) (essential tremor or Familial Tremor).mp. 3) 1 OR 2. 4) (observational or "case control" or cohort or "cross sectional" or "follow up" or risk factor * ).mp.

5) 3 AND 4.
Publication type and language were not restricted in the searches. Reference lists of all relevant articles were also searched for additional studies.

Inclusion and Exclusion Criteria
Identified articles were then further selected if they satisfied all of the following conditions: 1) Compared ET participants with healthy control.
2) The age of participants with more than 18 years. 3 Searched articles that met one of the following criteria were excluded: 1) Cannot find the original paper.
2) Data cannot be extracted.
3) Authors did not use mean (SD) report continuous outcomes or did not report event numbers for dementia or anxiety, depression, or did not report outcomes with OR, HR, RR together with 95% CI. 4) The participants with nervous and psychiatric system diseases seriously affecting cognitive function and emotion.

Study Selection and Data Extraction
In the initial screening, two authors (Xinjie Chen and Mingda Ai) independently assessed all of the abstracts retrieved from the search and excluded studies that met the exclusion criteria. Then full manuscripts of the potentially eligible studies were obtained and screened twice for inclusion. A third author (Xiaolei Liu) solved any disagreements based on the criteria of inclusion and exclusion. In addition, the following variables were extracted when available (shown in Table 1), i.e., (1) author; (2) year of published; (3) region; (4) age of participants; (5) educational status of participants; (6) age at onset for tremor; (7) duration of illness; (8) method used to assess cognitive study or depression or anxiety (i.e., structured diagnostic interview or coded diagnosis given by clinician or screening tool or validated algorithm); (9) study design; and (10) follow-up period.

Methodological Quality Appraisal
The Agency for Healthcare Research and Quality (AHRQ) was used to rate the methodological quality of cross-sectional studies. An item would be scored "0" if it was answered "NO" or "UNCLEAR" and scored "1" if it was answered "YES". Article quality was assessed as follows: low quality = 0-3; moderate quality = 4-7; and high quality = 8-11. The quality of studies was assessed using the Newcastle Ottawa Scale (NOS) generating a maximum of nine stars for each study, i.e., four stars for the selection of participants, two stars for the comparability of participants, and three stars for the assessment of outcomes. Quality was assigned according to the scores so that 7-9 stars indicated high quality, 4-6 stars for middle quality, and 0-3 stars for low quality (28).

Statistical Analysis
All statistical analyses were conducted using statistical software Cochrane Collaboration (RevMan5.3) and Stata15.0.
Results were reported as OR with corresponding 95% CI for dichotomous data. The mean difference (MD) or SMD was calculated for continuous data. Transformed into SMD when different scales were used for the same outcome domain. Heterogeneity between study results was assessed using a standard I2 test, with an I 2 >50% regarded as an indicator of substantial heterogeneity, then the random-effects model will be implemented (29). Publication bias was assessed via funnel plots and more formally with the Begg's test. Potential sources of heterogeneity across studies were explored by subgroup analyses.
The following variables were considered: (1) type of disease; (2) study characteristics; and (3) method used to assess cognitive study or depression or anxiety. Benito-León et al. (10) 2006 They had an action tremor of the head, limbs, or voice without any other recognizable cause; Second, the tremor had to be of gradual onset (i.e., slow and progressive) and present for at least 1 year or accompanied by a family history of the same disorder (at least one reportedly affected first degree relative). Third, on an Archimedes spiral, tremor severity had to be moderate or greater.
Louis et al. (11) 2007 Age >65 years; participants were diagnosed as having ET if they had an action tremor of the head or limbs without any other recognizable cause. The tremor had to be of gradual onset (i.e., slow and progressive) and either (i) present for at least 1 year or (ii) accompanied by a family history of the same disorder (at least one reportedly-affected first degree relative); dementia was diagnosed using criteria from the DSM-IV.
Louis et al. (12) 2007 Diagnostic criteria for ET, which were used both in participants who were examined and those whose medical records were reviewed, were closely modeled on those used in the Sicilian Study17and those recommended by Consensus from the Movement Disorders Society.
Kim et al. (13) 2009 All patients were diagnosed as having either definite or probable ET based on National Institutes of Health diagnostic criteria; Thawani et al.
2009 A total tremor score ≥5.5 or rated the handwritten sentence ≥2 (moderate or greater tremor, equivalent to a rating Inclusion criteria for patients with essential tremor were: (i) integrity of the nigrostriatal dopaminergic terminals, as evidenced by a normal dopamine transporter scan, to exclude parkinsonisms; (ii) no traumatic brain injury and past or current substance abuse, particularly alcohol; (iii) no dementia according to the DSM-IV; in particular, probable Alzheimer's disease was excluded according to the NINCDS-ADRDA criteria.
Benito-León et al. (19) 2011 Diagnostic criteria for ET (used in participants who were examined and in those whose medical records were reviewed) were similar those used in the Sicilian Study; depressive symptoms were assessed with the simple question "Do you suffer from depression?" Chandran et al.   Figure 1 shows the literature searching and study selection process. The initial search of databases retrieved 2,380 studies. After 756 duplicate records were removed, the remaining 80 studies with full text were read. Finally, 19 unique references were analyzed.

Study Characteristics
The basic characteristics of the included studies are summarized in Table 1

Quantitative Analysis of Dementia
Ten studies, i.e., 13,087 participants, were evaluated for MMSE. Heterogeneity across studies was high (I 2 = 98%), so the random-effects model was chosen to analyze the results. In subgroup analysis, ET patients had significantly lower MMSE scores than that of Non-ET groups (SMD, −1.16; 95% CI, −1.75 to −0.58; p = 0.0001; Figure 2A). After 2 studies were excluded for sensitivity analyses, the heterogeneity was lower (I 2 = 0%) but the results remained significant (Figure 2B). There was no evidence of publication bias (Begg's test, p = 0.474).

Quantitative Analysis of Affective Disorder
For depressive and anxiety symptoms (6 studies with 616 participants), the SMD was 0.55 (95% CI, 0.22-0.87; Figure 4A). Heterogeneity across studies was high (I 2 = 71%), so the randomeffect model was adopted. In subgroup analysis, heterogeneity was lower (I 2 = 0%) and patients with ET had significantly higher depressive and anxiety symptom scale score than that of Non-ET patients (p = 0.0009; Figure 4B). There was no evidence of publication bias (Begg's test, p = 0.707).

Qualitative Analysis of Dementia and Affective Disorder
Essential tremor was significantly associated with these two endpoints. Eleven studies that included 24,954 participants reported 4,170 events of dementia, affective disorders. Heterogeneity across studies was low to moderate (I 2 = 42%; Figure 5A), the fixed-effect model was implemented. Heterogeneity was lower (I 2 = 0%) in subgroup analysis and we found an increased risk of developing dementia, depression, and anxiety in the ET group compared with the Non-ET group (OR = 2.49, 95% CI, 2.17-2.85, p < 0.00001; Figure 5B). There was no evidence of publication bias (Begg's test, p = 0.062).

Quality of All Studies
For cross-sectional studies, the AHRQ scores varied from 5 to 7 ( Table 3). For those prospective studies, the NOS scores varied from 6 to 8 stars ( Table 4).

DISCUSSION
Essential tremor is characterized by a wide range of motor and non-motor symptoms. Patients with ET may exhibit dementia, depression, and anxiety. This meta-analysis indicated that there was a higher prevalence of dementia, depression, and anxiety in ET individuals not only in old age but also in the youth. The prevalence of ET among individuals aged between 18 and 60 was 1.60% (30). Recent studies have revealed that these nonmotor features occur similarly in younger and elder patients (31). A study that included 45 young patients with ET and 35 age-matched controls showed that compared with the control group, depressive and anxiety symptoms were more common in the ET group, while total MoCA scores were lower in the ET group (32). This meta-analysis showed that the age of ET patients with cognitive and affective impairment ranged from 27 and 84 years old. Therefore, non-motor symptoms seem to be disease-associated rather than age-associated manifestations. ET appears to be a risk factor for developing dementia, depression, and anxiety. A clinical study demonstrated that Chinese patients with ET showed a high prevalence of depression and anxiety, suggesting that depression and anxiety were risk factors for dementia (33). In a prospective population-based study, elderlyonset ET was associated with an increased risk of dementia (34). Our study showed that the OR for dementia, depression, and anxiety in individuals of the ET group was 2.49-fold than that of Non-ET group. Therefore, ET was speculated as an independent factor of dementia and affective disorders. Early screening and intervention for dementia, depression, and anxiety in patients with ET should be a crucial strategy to delay cognitive or affective impairment progression.
Our meta-analysis also revealed that the all-cause mortality rate in the ET group was higher than that of the control group. However, whether dementia, depression, or anxiety are risk factors for all-cause death in the ET group remains unclear. For Parkinson's disease (PD), several studies examined the extent to which cognitive impairment increased the risk of death (35)(36)(37)(38)(39)(40)(41)(42). Eight studies included in this meta-analysis had a follow-up of 3.2-3.8 years. Among them, only one reported the mortality rates in ET and Non-ET groups. To find out whether cognitive or affective dysfunctions are the risk factors of all-cause mortality in ET patients, future studies with more participants and longer follow-up are needed.
In the general population of non-demented individuals, the predictors of cognitive change include demographic, genetic, medical, subjective cognitive, and biologic factors (30,(43)(44)(45)(46)(47)(48)(49)(50). According to previous studies about PD, the predictors of cognitive impairment include a distinctive set of cognitive, neurologic, psychiatric, biologic, and genetic factors that may be disease-specific and factors related to the clinical severity of PD (51). Understanding the predictors of cognitive and emotional changes will be helpful for early screening, timely intervention, and improvement of quality of life with ET patients. In our study, we could not analyze the predictors of cognitive and affective disorders due to the lack of data from the multiplefactor analysis. In the future, prospective studies are needed to explore predictors.
The cognitive impairment of patients with ET mainly affected the executive function. It is evident that the cerebellum is responsible for executive function and emotion-related behaviors (52,53). Executive deficits in patients with ET have generally been attributed to the inefficient cerebellarcortical networks, particularly, those projecting to and from the prefrontal cortex (53)(54)(55)(56)(57)(58)(59). A possible link between Lewy body dementia, and/or PD and ET has been reported (60). One neuroimaging study has shown that the white matter and gray matter abnormalities significantly correlated with tremor severity, cognitive profile, and depression (61). Furthermore, a study reported the correlation between brain microstructural changes in the amygdala/left ventrolateral prefrontal cortex and the severity of depressive/anxiety symptoms among patients with ET (62). Therefore, dynamic assessment of dementia, depression, and anxiety and neuroimaging with ET patients will help to explain the mechanism of the occurrence and development of these diseases.
Several limitations of our study should be noted. Firstly, different evaluation tools were used in the studies to assess cognitive function, depression, and anxiety, which may cause high heterogeneity in this meta-analysis. Secondly, most of these studies were cross-sectional ones and patient information was incomplete. Since both outcomes and exposures are ascertained at the same time, the temporal relationship between the two might be unclear (63). Thirdly, our meta-analysis is based on observational studies, which compared ET patients and Non-ET controls. In future studies, uniform diagnosis and evaluation criteria are needed in prospective studies to better understand the association of ET with dementia, depression, and anxiety.
In conclusion, we found a higher prevalence of dementia, depression, and anxiety in the ET group compared with the Non-ET group. Therefore, non-motor symptoms should not be neglected among patients with ET. However, the causal relationship between ET and dementia, depression, and anxiety is unclear and should be further investigated.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
YS, XC, MA, XG, SD, MZ, CY, LW, JZ, and LZ: collected and analyzed data. TB and XL: designed and supervised the study. All authors contributed to the article and approved the submitted version.