Blood Harmane (1-methyl-9H-pyrido[3,4-b]indole) and Mercury in Essential Tremor: A Population-Based, Environmental Epidemiology Study in the Faroe Islands

Abstract

Background

Essential tremor (ET) is among the most prevalent neurological diseases. Its environmental determinants are poorly understood. Harmane (1-methyl-9H-pyrido[3,4-b]indole), a dietary tremor-producing neurotoxin, has been linked to ET in a few studies in New York and Madrid. Mercury, also a tremor-producing neurotoxin, has not been studied in ET. The Faroe Islands have been the focus of epidemiological investigations of numerous neurological disorders.

Objective

In this population-based, case-control study, we directly measured blood harmane concentrations [HA] and blood mercury concentrations [Hg] in ET cases and controls.

Methods

1,328 Faroese adults were screened; 26 ET cases were identified whose [HA] and [Hg] were compared to 197 controls.

Results

Although there were no statistically significant differences between diagnostic groups, median [HA] was 2.7x higher in definite ET (4.13 g⁻¹⁰/ml) and 1.5x higher in probable ET (2.28 g⁻¹⁰/ml) than controls (1.53 g⁻¹⁰/ml). Small sample size was a limitation. For definite ET vs. controls, p = 0.126. [Hg] were similar between groups.

Conclusions

We demonstrated marginally elevated [HA] in definite and probable ET. These data are similar to those previously published, and possibly extend etiological links between this neurotoxin and ET to a third locale. The study did not support a link between mercury and ET.

Introduction

Essential tremor (ET) is among the most common neurological diseases, with an estimated 7 million people affected in the US (i.e., 2.2% of the entire US population and 2.9% of the US population age 20 and older) [1]. Prevalence rises exponentially with age, and reaches >20% in advanced age groups [2,3]. The underlying causes of ET are unclear, but are thought to be both genetic [4–6] and environmental [7–9]. With numerous studies noting that fewer than one-half of patients report a positive family history of ET [10], and twin studies reporting monozygotic concordance of approximately 60% [11], the role for environmental factors is apparent.

Among the environmental factors most studied is harmane (1-methyl-9H-pyrido[3,4-b]indole), a dietary tremor-producing toxin whose blood [12–15] and brain levels [16] have been demonstrated to be elevated in ET in a small number of studies. These studies have been geographically-restricted to the US (New York) and Spain (Madrid) [12–15]. Given its high concentration in dietary animal protein [15], studies of harmane in populations with high animal protein consumption, such as the Faroe Islands [17], are of particular value.

Other neurotoxins are tremor-producing. Mercury is one of these, and it has long been linked with tremor [18–20]. Yet curiously, we are unaware of any studies directly assessing the links between mercury and ET.

The Faroe Islands, located in the North Atlantic Ocean, between Norway and Iceland, are inhabited by 49,121 individuals of whom 24,154 are age ≥40 (January 1, 2016). Due to the isolated geographic location and homogenous population, they have been the focus of epidemiological investigations of numerous neurological disorders [21–24]. The Faroese diet is high in marine food [17], and high level of exposure to methyl mercury through diet makes it a valuable location to study the effects of this neurotoxicant [25].

In this population-based, case-control study, we directly measured blood concentrations of harmane and mercury in ET cases and controls. The goal was to extend the etiological links between harmane and ET to a third locale and to take the first steps in exploring the etiological links between mercury and ET. Given the extraordinarily prevalence of ET in human populations, the identification of toxicant-risk factors would have considerable public health impact in terms of disease prevention.

Methods

Study Population and Sampling Frame

We used a two-phase, population-based study design [24,26,27]. All study procedures were approved by the local ethical review committee of the Faroe Islands and the Institutional Review Board at Yale University, with participation on a voluntary basis and signed informed consent obtained from each enrollee. As described [24,26,27], the study was undertaken between August 2016 and December 2017. The names and addresses of 4,798 individuals aged ≥40 were obtained from the Faroese Population Registry, which included 24,154 individuals aged ≥40 years living in the Islands. These individuals were selected based on six randomly-selected birthdates - 10^th, 16^th, 17^th, 18^th, 22th, and 30^th [24]. From this group of 4,798, all 1,155 individuals aged ≥70 years were selected into the screening group, while the remaining 1,845 were selected by random sampling using SPSS [24]. Thus, the screening group comprised 3,000 Faroese individuals aged ≥40 years (Figure 1 in Eliasen et al.[24]).

First Phase Screening

During the first screening phase (September 2016 - December 2016), these 3,000 individuals in the screening group received an invitation letter to participate in a study of lifestyle, diet and neurological conditions [24]. The invitation letter included a screening package (questionnaires, a request for hand-drawn spirals, a return stamped envelope). The questionnaires included demographic and clinical questions, and seven screening questions for tremor [24]. The letter included instructions on drawing two Archimedes spirals with each hand and four blank sheets of paper [24]. A total of 1,334 (44.5%) individuals returned the completed screening package (Figure 1 in Eliasen et al.[24]).

Tremor on each spiral was rated by a senior movement disorders neurologist (E.D.L) using an ordinal clinical rating scale (0 – 3.0). Based on data from the questionnaire and spiral scores, participants were stratified into four groups: (1) high likelihood of having ET, (2) intermediate likelihood of having ET, (3) low-intermediate likelihood of having ET and (4) low likelihood of having ET [24].

Second Phase

In the second phase, a subsample of 282 individuals who had returned the spirals and questionnaire was invited to participate in an in-person clinical evaluation at the Department of Occupational Medicine and Public Health or in their own homes [24]. All 85 individuals in group 1 were invited. Also, a randomly-selected subsample of individuals in group 2 (n = 97), group 3 (n = 32) and group 4 (n = 68) were invited. A total of 227 (80.5%) individuals accepted and completed the in-person clinical evaluation, while 54 declined participation and one had died. The in-person clinical evaluation was conducted by a trained nurse (E.H.E.) and included 1) anthropometric measures (body weight [kilograms] and height [cm]), 2) a questionnaire encompassing data on demographics, socioeconomic status (e.g., education, number of rooms in house), medical comorbidity (i.e., Cumulative Illness Rating Scale[28]), smoking habits, family history of ET or tremor, diet, ethanol consumption, fasting status, and hours since last food consumption, 3) a detailed videotaped tremor examination, and 4) phlebotomy for blood harmane and blood mercury determinations. The videotaped tremor examination included (1) an ET-specific metric comprised of one test for postural tremor and five for kinetic tremor (e.g., pouring, drinking) performed with each arm (12 tests total, total tremor score = 0 – 36), (2) the motor portion of the Unified Parkinson’s Disease Rating Scale (UPDRS) [29] excluding an assessment of rigidity, and (3) an assessment of dystonia.

ET Diagnoses

As described, ET diagnoses (definite, probable, possible) were assigned by E.D.L. by review of questionnaire data and the videotaped neurological examination and based on published diagnostic criteria (moderate or greater amplitude kinetic tremor during three or more activities or a head tremor in the absence of Parkinson’s disease or another known cause [e.g., medication-induced tremor]) [30]. These diagnostic criteria for ET were developed for a population-based genetic study and based on data from approximately 2,000 normal (non-diseased controls)[30] the criteria carefully specify the specific examination maneuvers during which tremor should be present and the severity of tremor that should be evident during these maneuvers to distinguish normal from ET. These criteria have been shown to be both reliable [31] and valid [32], and have been used by tremor investigators in the United States and internationally [33–44]. As described, a diagnosis of ET was assigned to 27 of 227 individuals [24].

Measurement of Blood Harmane Concentration

After phlebotomy, blood harmane was quantified, blinded to clinical information, by a well-established high performance liquid chromatography method used in our previous studies [12,14,16,45]. Concentrations were reported in g⁻¹⁰/ml.

Measurement of Whole Blood Mercury

Analyses were conducted in the laboratory at the Department of Environmental Medicine, University of Southern Denmark, blinded to demographic, clinical and diagnostic information. Total mercury was determined in a volume of 100 μL of whole blood on a Direct Mercury Analyzer system (DMA-80) from Milestone, Sorisole, Italy [46]. The concentration in whole blood was reported in ng/ml. Certified quality control samples from Seronorm Trace Elements (Sero, Norway) were included in all sample series, and the calculated imprecision was less than 5% [46]. The limit of detection was 0.15 ng/ml.

Final Sample

There were 26 ET cases and 196 controls with successful phlebotomy and hence, blood concentrations of harmane and mercury. None of the controls came from the same household as an ET case.

Statistical Analyses

Differences between the ET and non-ET group were tested with Student’s t-test (continuous variables) or chi square test or Fischer’s exact test (categorical variables). Mann-Whitney tests were used when continuous variables were not normally distributed. Neither blood harmane nor blood mercury concentrations were normally distributed (Kolmogorov Smirnov test); hence, non-parametric tests were used to compare blood levels across diagnostic groups, and median values were also reported. As several prior publications report log-transformed harmane concentrations, we also do so here. In our control sample, we assessed the correlation between blood levels and demographic and clinical variables using non-parametric tests (Spearman’s r, Mann-Whitney tests). For these analyses, log-tranformed values were used to minimize the effects of outliers. Statistical analyses were performed using SPSS version 26.0.

Study Power

The study was designed, through screening, to yield 108 ET and to compare these to a similar number of controls. For comparisons of blood harmane concentrations, this number was deemed sufficient, as similar sample sizes (n = 100 – 150 per group) had revealed significant differences in all prior studies [12–15]. There were no prior studies of mercury in ET on which to base sample size.

The screens yielded far fewer ET than expected, as the crude prevalence of ET was only 2.9% among persons age ≥40 [24]. Hence, the final number of ET cases for these analyses was 26, with the resultant power for the harmane analysis at 30%.

Results

The 26 ET cases were similar to the 197 controls in all demographic and clinical variables - age, gender, education, number of rooms in home, body mass index, cigarette smoking, ethanol consumption, Cumulative Illness Rating Scale score, years since most recent hospitalization, fasting status, and hours since last meal or liquid consumption (Table 1). As expected, family history of ET or tremor was more prevalent among ET cases than controls, and the total tremor score was higher in ET cases than controls (Table 1). Definite ET cases were older than controls (Table 1).

Table 1:

Clinical characteristics of 222 study participants by tremor diagnosis

	ET (all)	Definite ET	Probable ET	Possible ET	Control
N	26	3	12	11	196
Age in years	65.7 ± 11.7 [67.0] 1	77.7 ± 5.8 [81.0] 1 *	68.0 ± 11.9 [69.0] 1	59.8 ± 9.7 [58.0] 1	61.8 ± 12.2 [60.0]
Male gender	17 (65.4) 2	2 (66.7) 2	8 (66.7) 2	7 (63.6) 2	105 (53.6)
Education #
<High School	8 (33.3) 2	2 (100) 2	2 (16.7) 2	4 (40.0) 2	54 (28.0)
High School	1 (4.2)	0 (0.0)	1 (8.3)	0 (0.0)	7 (3.6)
Trade School	3 (12.5)	0 (0.0)	2 (16.7)	1 (10.0)	43 (22.3)
Associates	5 (20.8)	0 (0.0)	1 (8.3)	4 (40.0)	19 (9.8)
Bachelors	5 (20.8)	0 (0.0)	4 (33.1)	1 (10.0)	57 (29.5)
Masters	1 (4.2)	0 (0.0)	1 (8.3)	0 (0.0)	11 (5.7)
Doctorate	1 (4.2)	0 (0.0)	1 (8.3)	0 (0.0)	2 (1.0)
Number of rooms in home	7.4 ± 3.3 [6.0] 1	5.3 ± 1.2 [6.0] 1	7.6 ± 3.1 [6.5] 1	7.8 ± 3.8 [6.0] 1	7.8 ± 2.9 [8.0]
Body mass index in Kg/m2	28.0 ± 4.0 3	24.6 ± 4.3 3	29.4 ± 4.1 3	27.5 ± 3.4 3	28.4 ± 4.6
Current cigarette smoker	8 (30.8) 2	1 (33.3) 2	2 (16.7) 2	5 (45.5) 2	38 (19.4)
Number of cigarettes smoked per day (including smokers and non-smokers)	3.1 ± 6.2 [0.0] 1	2.3 ± 4.0 [0.0] 1	2.3 ± 6.0 [0.0] 1	4.1 ± 7.1 [0.0] 1	2.2 ± 5.6 [0.0]
Number of drinks of ethanol per month	10.5 ± 12.3 [4.7] 1	5.2 ± 5.2 [5.2] 1	13.0 ± 16.6 [6.0] 1	9.8 ± 9.9 [3.3] 1	20.1 ± 34.6 [8.4]
Cumulative Illness Rating Scale score	3.6 ± 2.5 [3.0] 1	6.7 ± 3.8 [5.0] 1	3.8 ± 2.3 [3.0] 1	2.4 ± 1.5 [2.0] 1	3.5 ± 3.1 [3.0]
Years since most recent hospitalization	8.2 ± 7.8 [5.8] 1	8.0 ± 2.8 [8.0] 1	6.1 ± 6.9 [3.0] 1	10.3 ± 9.2 [7.5] 1	11.8 ± 13.7 [5.0]
Fasting #	3 (12.5) 2	0 (0.0) 2	1 (9.1) 2	2 (18.2) 2	19 (9.8)
Hours since last liquid intake	2.9 ± 3.7 [1.8] 1	2.8 ± 0.4 [2.8] 1	2.6 ± 3.3 [1.5] 1	3.2 ± 4.4 [1.0] 1	2.1 ± 2.9 [1.0]
Hours since last meal	3.1 ± 3.6 [2.3] 1	2.8 ± 0.4 [2.8] 1	3.0 ± 3.2 [2.0] 1	3.2 ± 4.4 [1.0] 1	2.7 ± 3.3 [1.5]
Family history of ET or tremor	10 (38.5) 2 *	1 (33.3) 2	6 (50.0) 2 *	3 (27.3) 2	37 (18.9)
Total tremor score	16.8 ± 3.8 [16.6] 1 ***	21.8 ± 2.8 [20.3] 1 **	16.1 ± 4.4 [18.0] 1 ***	16.1 ± 2.3 [15.5] 1 ***	10.3 ± 3.3 [10.3]

ET = essential tremor.

Values are mean ± standard deviation [median] or number (percentage).

^#Some data are missing.

¹Mann-Whitney test compared to controls

²Chi square test compared to controls

³Student’s t test compared to controls.

^*p<0.05

^**p < 0.01

^***p < 0.001.

Higher blood harmane concentration was marginally correlated with older age and was correlated with higher body mass index but not with any other variables (Table 2). Higher blood mercury concentration was correlated with older age, male gender, lower education and higher body mass index (Table 2). Neither blood harmane nor mercury concentrations were correlated with fasting status or hours since last food or liquid consumption (Table 2).

Table 2:

Correlations between clinical characteristics and harmane and mercury concentrations in controls

	Harmane	Mercury
Age in years	r = 0.14, p = 0.053	r = 0.22, p = 0.002
Gender
Male	0.24 ± 0.39 [0.20]	1.11 ± 0.41 [1.11]
Female	0.25 ± 0.34 [0.18]	0.92 ± 0.40 [0.94]
	p = 0.87 1	p = 0.002 1
Education
<Bachelors	0.25 ± 0.36 [0.19]	1.06 ± 0.42 [1.09]
Bachelors or higher	0.25 ± 0.38 [0.19]	0.93 ± 0.38 [0.96]
	p = 0.82 1	p = 0.03 1
Number of rooms in home	r = 0.01, p = 0.86	r = 0.03, p = 0.66
Body mass index (Kg/m2)	r = 0.15, p = 0.035	r = 0.14, p = 0.049
Current cigarette smoker
Yes	0.30 ± 0.37 [0.28]	0.96 ± 0.42 [1.00]
No	0.23 ± 0.36 [0.18]	1.04 ± 0.41 [1.06]
	p = 0.35 1	p = 0.46 1
Number of cigarettes smoked per day (including smokers and non-smokers)	r = 0.03, p = 0.71	r = −0.06, p = 0.39
Number of drinks of ethanol per month	r = 0.74, p = 0.40	r = 0.04, p = 0.63
Cumulative Illness Rating Scale score	r = 0.09, p = 0.20	r = 0.10, p = 0.18
Years since most recent hospitalization	r = −0.01, p = 0.90	r = 0.047, p = 0.57
Fasting
Yes	0.22 ± 0.35 [0.24]	1.05 ± 0.44 [1.00]
No	0.25 ± 0.37 [0.19]	1.02 ± 0.41 [1.06]
	p = 0.96 1	p = 0.93 1
Hours since last liquid intake	r = 0.058, p = 0.42	r = 0.045, p = 0.53
Hours since last meal	r = 0.02, p = 0.80	r = 0.049, p = 0.50
Family history of ET or tremor
Yes	0.17 ± 0.31 [0.13]	0.92 ± 0.39 [0.91]
No	0.26 ± 0.38 [0.24]	1.05 ± 0.42 [1.08]
	p = 0.20 1	p = 0.07 1
Total tremor score	r = −0.05, p = 0.50	r = 0.02, p = 0.75

ET = essential tremor.

r = Spearman’s r.

Harmane (in g⁻¹⁰/ml) and mercury levels (in ng/ml) were log-transformed.

Values are r values or mean ± standard deviation [median].

¹Mann-Whitney test.

Although there were no statistically significant differences between diagnostic groups (Table 3), median blood harmane concentration was 2.7 times higher in definite ET (4.13 g⁻¹⁰/ml) and 1.5 times higher in probable ET (2.28 g⁻¹⁰/ml) than controls (1.53 g⁻¹⁰/ml). For the comparison of definite ET vs. controls, p = 0.126 (Table 3, Figure 1). Mercury levels were similar between groups (Table 3).

Table 3:

Harmane and mercury concentrations by tremor diagnosis

	ET (all)	Definite ET	Probable ET	Possible ET	Control
N	26	3	12	11	196
Harmane	1.74 [2.32 ± 2.05] p = 0.92	4.13 [4.86 ± 3.86] p = 0.126	2.28 [2.35 ± 1.82] p = 0.91	1.27 [1.60 ± 1.18] p = 0.28	1.53 [2.51 ± 2.29]
Log Harmane	0.24 [0.21 ± 0.40] p = 0.92	0.62 [0.58 ± 0.40] p = 0.126	0.36 [0.23 ± 0.40] p = 0.91	0.10 [0.09 ± 0.37] p = 0.28	0.19 [0.25 ± 0.37]
Mercury	10.11 [17.30 ± 17.78] p = 0.86	11.41 [10.49 ± 8.22] p = 0.64	11.30 [21.31 ± 20.77] p = 0.51	8.40 [14.80 ± 16.19] p = 0.86	11.36 [16.18 ± 17.41]

ET = essential tremor.

Values are median [mean ± standard deviation].

Harmane was in g⁻¹⁰/ml and mercury was in ng/ml.

All statistical tests are Mann-Whitney test compared to controls.

For comparison of harmane concentration in definite and probable ET (n = 15) vs. 196 controls, p = 0.44 (Mann-Whitney test); for comparison of mercury concentration in definite and probable ET (n = 15) vs. 196 controls, p = 0.70 (Mann-Whitney test).

Figure 1:

Median blood harmane concentrations in g⁻¹⁰/ml

We compared ET cases with a family history of ET or tremor to those without such history, and blood harmane concentrations were similar: 2.38 ± 2.60 g⁻¹⁰/ml [median = 1.85 g⁻¹⁰/ml] vs. 2.28 ± 1.72 g⁻¹⁰/ml [median = 1.71 g⁻¹⁰/ml], Mann Whitney p = 0.78). We compared ET cases with a family history of ET or tremor to those without such history and blood mercury concentrations were similar: 15.09 ± 11.80 [median = 9.96] vs. 18.69 ± 20.93 [median = 10.25], Mann Whitney p = 0.82).

ET cases and controls did not differ with respect to any demographic or clinical variables (Table 1); hence, there were no confounding variables to consider in adjusted models. Neither blood harmane nor mercury concentrations were correlated with fasting status or hours since food or liquid consumption; hence, there was no need to consider these confounding variables in additional analyses either.

Discussion

The contribution of environmental factors to disease etiology has been examined in numerous epidemiological studies of Parkinson’s disease, Alzheimer’s disease and amyotrophic lateral sclerosis [47–53]. Yet the study of toxins in ET lags far behind, with surprisingly few studies [7,8].

Although there were no statistically significant differences between diagnostic groups (Table 3), median blood harmane concentration was 2.7 times higher in definite ET (i.e., 4.13 g⁻¹⁰/ml) and 1.5 times higher in probable ET (2.28 g⁻¹⁰/ml) than controls (i.e., 1.53 g⁻¹⁰/ml). In a prior study in New York [15], harmane concentrations were similarly highest in definite ET, second highest in probable ET and similar in possible ET and controls (Table 4), although in that study, the numbers of definite and probable ET (38 and 53 respectively) were 5 to 10 times higher than the present study and the differences reached statistical significance. The current study, which is a fifth study of blood harmane concentrations in ET [12–15], suggests that the higher levels of harmane extend now to a third locale, the Faroe Islands, and as with a prior study[15], mainly affect those with more severe or more diagnostically definitive ET. In addition, one study demonstrated higher brain harmane concentrations in ET than controls [16].

Table 4:

Studies comparing harmane concentrations in ET cases and controls

Authors (location)	ET (all)	Definite ET	Probable ET	Possible ET	Control
Louis et al 2002 [12] (New York)	5.21 g−10/ml (n = 100) p = 0.005				2.28 g−10/ml (n = 100)
Louis et al 2005 [15] (New York)	Log harmane = 0.61 ± 0.67 g−10/ml (n = 106) p = 0.035	Log harmane = 0.74 ± 0.69 g−10/ml (n = 38)	Log harmane = 0.56 ± 0.67 g−10/ml (n = 53)	Log harmane = 0.48 ± 0.64 g−10/ml (n = 15)	Log harmane = 0.43 ± 0.72 g−10/ml (n = 161)
Louis et al 2008 [13] (New York)	2.61 g−10/ml (n = 150) p = 0.016				1.82 g−10/ml (n = 135)
Louis et al 2013 [14] (Madrid)	2.90 g−10/ml in familial ET (n = 62) 2.41 g−10/ml in sporadic ET (n = 68) p = 0.049 a				2.09 g−10/ml (n = 135)
Current study 2019 (Faroe Islands)	1.74 g−10/ml (n = 26) p = 0.92	4.13 g−10/ml (n = 3) p = 0.126	2.28 g−10/ml (n = 12) p = 0.91	1.27 g−10/ml (n = 11) p = 0.28	1.53 (n = 196)

Values are median harmane unless otherwise specified as log harmane.

^ain an adjusted logistic regression model comparing familial ET to controls.

p values show comparisons with controls.

The mechanism for a higher blood harmane concentration in ET is unclear and has been discussed in several or our earlier papers [12–15], although the data most lend support to the notion that it could be the result of a genetically-driven reduction in harmane metabolism. In prior analyses in a cohort in New York, we did not major detect dietary differences in animal protein consumption between ET cases and controls [15] and in the current cohort, we collected detailed data on animal protein consumption and food cooking practices (i.e., level of doneness of meat, which is associated with blood harmane concentrations), and similarly did not detect significant differences between ET cases and controls (data not presented), suggesting that the case-control difference we observed in harmane concentration are not driven by dietary differences.

The study did not support a link between mercury and ET. Action tremor has been a well-described outcome of mercury exposure in a broad and varied range of settings in numerous studies spanning many years [18–20], and further interest in mercury relates to the fact that it is toxic to the Purkinje cell population in the cerebellum [54,55], a population of neurons increasingly linked to the pathophysiology of ET [56–58]. With respect to ET, specifically, however, we are unaware of any study that has directly compared mercury exposures in ET cases vs. controls or that has quantified and compared blood mercury concentrations in ET cases vs. controls. The mercury levels we report are relatively high, and diet, namely the intake of whale meat contaminated with mercury, could account for this. Self-reported whale meat consumption was similar in the two groups (data not shown) as were blood mercury levels.

This study should be interpreted within the context of several limitations. First, despite the large screening sample, the number of ET cases was small (n = 27 and 26 with blood levels of harmane and mercury), thereby limiting the power to detect differences and increasing the possibility of Type II statistical error. Indeed, even robust differences (e.g., difference in harmane concentrations between definite ET and controls or probable ET vs. controls) did not achieve statistical significance. A larger sample of ET cases would achieve greater power, but would require the screening of several thousand more participants. The second issue is that we did not assess fasting blood harmane concentrations; fasting concentrations are less likely to reflect recent food consumption. However, prior data indicated that there was no association between time elapsed from last food or beverage ingestion and blood harmane concentration (Spearman’s r = 0.12, p = 0.34) or HA/HI ratio (Spearman’s r = 0.11, p = 0.41), indicating that blood harmane was not a function of time since last food consumption [59]. Our current analyses confirmed that this was the case (Table 2).

The study had a number of strengths. First, this is the first population-based study of harmane in ET, with all prior studies largely or exclusively sampling cases from clinics [12–14]. Second, it extends studies of this neurotoxin from only two locales [12–15] now to a third locale. Third, it is the first study to directly measure blood mercury concentrations in ET cases. Finally, all ET diagnoses were carefully assigned by a movement disorder neurologist with longstanding interest in ET based on a detailed clinical assessment using stringent, validated diagnostic criteria.

In summary, we demonstrated marginally elevated blood harmane concentrations in probable and definite ET cases. These data seem to extend the etiological links between this neurotoxin and ET to a third locale. The study did not support a link between mercury and ET.